The Science of

Learning Physics
Cognitive Strategies for Improving Instruction

11998_Master4.indd 1 2/11/2020 [Link] PM

 b2530   International Strategic Relations and China’s National Security: World at the Crossroads

This page intentionally left blank

b2530_FM.indd 6 01-Sep-16 [Link] AM

 The Science of
Learning Physics
Cognitive Strategies for Improving Instruction

José P. Mestre
University of Illinois at Urbana-Champaign, USA

Jennifer L. Docktor
University of Wisconsin-La Crosse, USA

World Scientific
NEW JERSEY • LONDON • SINGAPORE • BEIJING • SHANGHAI • HONG KONG • TAIPEI • CHENNAI • TOKYO

11998_Master4.indd 3 2/11/2020 [Link] PM

 Published by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

Library of Congress Cataloging-in-Publication Data

Names: Mestre, Jose P., author. | Docktor, Jennifer L., author.
Title: The science of learning physics : cognitive strategies for improving instruction /
José P. Mestre, University of Illinois at Urbana-Champaign, USA,
Jennifer L. Docktor, University of Wisconsin-La Crosse, USA.
Description: New Jersey : World Scientific, 2021. | Includes bibliographical references and index.
Identifiers: LCCN 2020042831 | ISBN 9789811226540 (hardcover) |
ISBN 9789811227769 (paperback) | ISBN 9789811226557 (ebook) |
ISBN 9789811226564 (mobi)
Subjects: LCSH: Physics--Study and teaching. | Cognition--Mathematical models.
Classification: LCC QC30 .M365 2020 | DDC 530.071/1--dc23
LC record available at [Link]

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Copyright © 2021 by World Scientific Publishing Co. Pte. Ltd.

All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy
is not required from the publisher.

For any available supplementary material, please visit

[Link]

Printed in Singapore

Lakshmi - 11998 - The4Science of Learning [Link] 1

11998_Master4.indd 12/10/2020
2/11/2020 [Link]
[Link] am
PM
 Contents

Chapter 1: Introduction 1
1.1. What is this book about? 1
1.2. Why is this book needed? 2
1.3. What does it mean to teach better? 4
1.4. What are some contributing factors leading to many
students finding physics difficult to learn? 5
1.5. What view undergirds this entire book? 6
1.6. How are the subsequent chapters of this book organized? 7
Reference 9

Chapter 2: The Formation of Concepts and How to Fix 11

Broken Ones
2.1. What does research on students’ concept formation
tell us? 11
2.1.1. A comparison of behaviorist and constructivist
views of learning 11
2.1.2. Conceptual difficulties in physics 14
2.1.3. Example: A study with balls and tracks 17
2.2. What are the implications of research on students’
concept formation for instruction? 22
2.3. Examples of teaching interventions based on learning
research 25
References 34

Chapter 3: How Learning Looks for a Novice: 37

Implications of Expert–Novice Research for
Physics Teaching and Learning
3.1. What does expert–novice research tell us? 37
3.1.1. Broad observations about novices and experts in
physics 38

11998_Master4.indd 5 2/11/2020 [Link] PM

 vi The Science of Learning Physics

3.1.2. Some important expert–novice differences 40

3.1.3. Differences in what novices and experts notice
when “doing” physics 41
3.1.4. Problem solving differences between novices and
experts in physics 45
3.1.5. Can instruction be designed to promote
expert-like behavior in physics? 46
[Link]. Short-term interventions 46
[Link]. Longer-term interventions 55
3.2. What are the implications of expert–novice research for
instruction? 63
3.3. Examples of teaching interventions based on learning
research 65
References 78

Chapter 4: From Manipulating Equations to a More 83

Conceptual Approach:
How to Improve Problem Solving
4.1. What does research on problem solving tell us? 83
4.1.1. Research on problem solving skills and
approaches 83
4.1.2. Problem representation is important 86
4.1.3. Problem solving frameworks 91
4.1.4. Assessment of problem solving 95
4.2. What are the implications of research on problem
solving for instruction? 100
4.3. Examples of teaching interventions based on learning
research 102
References 104

Chapter 5: Active Learning Strategies: 111

Engaging Students in their Own Learning
is the Key to Learning
5.1. What does the research tell us? 111

11998_Master4.indd 6 2/11/2020 [Link] PM

 Contents vii

5.1.1. Introduction and background 111

5.1.2. Reforming the “traditional” lecture classroom 116
5.1.3. Reforming the traditional laboratory or recitation
experience 120
5.1.4. Reforming the “out-of-class” experience 123
5.1.5. Barriers to active learning instruction 125
5.1.6. The use of Communities of Practice to encourage
reformed instruction 127
5.2. What are the implications of research on active
learning for instruction? 129
5.3. Examples of teaching interventions based on
learning research 131
References 133

Chapter 6: Students’ Perceptions of Learning and their 141

Study Habits
6.1. What does research on students’ perceptions of
learning and studying tell us? 141
6.1.1. Research on study strategies 142
6.1.2. What are the actual patterns of learning and
forgetting? 149
 
6.1.3. Students’ metacognition about studying and
learning 152
6.2. What are the implications of research into students’
view of learning and studying for instruction? 156
6.3. Examples of teaching interventions based on learning
research 158
References 162

Chapter 7: Testing in the Service of Learning: 169

The Testing Effect and How it Promotes
Long-Term Retention
7.1. What does research on the “testing effect” tell us? 169
7.1.1. What is the testing effect? 170

11998_Master4.indd 7 2/11/2020 [Link] PM

 viii The Science of Learning Physics

7.1.2. Experimental findings on the testing effect 171

7.1.3. Relevance of the testing effect for physics
instruction 173
7.1.4. Mechanisms leading to long-term retention
from the testing effect 176
7.2. What are the implications of the testing effect for
instruction? 178
7.3. Examples of teaching interventions based on learning
research 179
References 182

Chapter 8: Concluding Remarks 185

Reference 187

Index 189

About the Authors 201

11998_Master4.indd 8 2/11/2020 [Link] PM

 c hapter
1 Introduction

1.1. What is this book about?

This book is about teaching and learning physics. It is intended
for use by college-level instructors of physics, although high
school instructors might also find it very useful. It targets under-
graduate instruction with an emphasis on introductory physics,
which is where most students taking physics are enrolled.
However, the ideas in this book can also be used to improve
teaching and learning in both upper-division undergraduate
physics courses as well as graduate courses. Whether you are new
to teaching physics or a seasoned veteran, we present a variety
of ideas and strategies for you to consider. Some ideas might be
a small “tweak” to your existing practices whereas others require
more substantial revisions to instruction.
The discussions of student learning herein are based on
research evidence accumulated over decades from various
fields, including cognitive psychology, educational psychology,
the learning sciences, and discipline-based education research
including physics education research. Likewise, the teaching sug-
gestions are also based on research findings. As for any other
scientific endeavor, physics education research is an empirical
field where experiments are performed (in the messy environ-
ment of classrooms, or in the lab where student behaviors can be

11998_Master4.indd 1 2/11/2020 [Link] PM

 2 The Science of Learning Physics

explored under more controlled settings), data are analyzed and

conclusions are drawn. Evidence from such research is then used
to inform physics teaching and learning.

1.2. Why is this book needed?

There are three main reasons for why a book like this is needed:
1. Advances in knowledge about teaching and learning.
There is a substantial body of knowledge about how people
learn and how to make learning more efficient/effective.
Although applications of this knowledge are finding
their way into physics classrooms, much of the research
on student learning remains unknown by physicists, or
perhaps worse, unheeded by them. We aim to highlight
many salient findings about teaching and learning, discuss
their instructional implications, and provide teaching
examples of how to apply them. Currently there are no
resources like this for physics instructors; existing books
on physics teaching are either sorely outdated or narrower
in scope.
2. Physics is leading the way in reforming undergraduate
STEM (Science, Technology, Engineering and Math­
ematics) instruction. Anyone who has taught physics
quickly realizes that the concepts and problem solving
techniques that form the backbone of physics instruction
are difficult to learn by large numbers of students. Perhaps
because they like solving difficult problems, physicists have
viewed difficulties in learning physics as a challenging
problem to be solved, and for many years, physicists have
engaged in studying ways to make physics learning more
tractable to larger numbers of students. Many examples
exist of how this research has made its way to classrooms,

11998_Master4.indd 2 2/11/2020 [Link] PM

 Introduction 3

among the more notable ones is implementing active

learning in large classes using classroom polling technolo-
gies (with “clickers”).
3. The way Ph.D. training is structured. Ph.D. training
in just about every field except for education, where
pedagogy is emphasized, consists of learning the field
broadly through course work and by conducting research
deeply in a specific subfield and writing a dissertation.
Ph.D. programs produce experts in content knowledge
and neophytes in knowledge about teaching and learning.
With rare exceptions, the courses taken en route to a
Ph.D. are only in the discipline in question (in our case,
physics). The Ph.D.s in physics who eventually become
professors and teach physics are highly likely never to
have taken a course on how students learn, or on the use
of evidence-based instructional practices (EBIPs) to teach
physics. Like learning any new endeavor, learning to use
any of the many EBIPs available takes time and effort,
and professors at all levels have pressures on their time.
The result is that most take the path of least resistance in
teaching and do what was done unto them—they lecture
to students. Some people can learn well with the lecture
method (most of us who become Ph.D. physicists, for
example), probably not because they learn everything the
first pass through by listening to a lecture, but because
they are disciplined and work hard at learning the material
outside of lectures. As someone teaching physics you may
have spent countless hours solving problems and thinking
about the subject matter on your own or by learning
through interactions with fellow students or professors.
As we will discuss later, more students will learn more
physics with EBIP-based instruction, if they would only

11998_Master4.indd 3 2/11/2020 [Link] PM

 4 The Science of Learning Physics

be used. Two important conclusions to be drawn from

the structure of Ph.D. programs are: (1) Upon graduation,
the newly-minted Ph.D. is ready for additional “education”
on how students learn and how to implement EBIPs, and
(2) If the newly-minted Ph.D. teaches only using lectures,
fewer students will learn and many capable students will
drop out of science entirely (Tobias, 1990). In addition,
someone teaching in the K-12 realm or at a teaching-
focused institution might be expected to engage in
innovative teaching practices without much training on
how to do so.

1.3. What does it mean to teach better?

In higher education, it is a common practice to measure the
quality of teaching by using teaching evaluations, which are
almost entirely based on the opinions of those students receiving
the instruction. Even if you get observed by a colleague, their
evaluation might be focused on surface-level characteristics
such as whether your presentation of material was sufficiently
“polished.” If the goal of teaching is to have students learn, it is
strange that we do not use assessments of student learning as
the measure of quality teaching. If an instructor has a modest
degree of charisma, shows some degree of sympathy toward stu-
dents, and is clear in the lecture delivery, he/she is practically
guaranteed high teaching evaluations. However, sitting in a lec-
ture and enjoying “the show” is no different than going to the
movies and enjoying the show. Enjoying a lecture and being
able to follow it develops familiarity with the subject covered,
and often a false sense of competence, rather than actual com-
petence. In fact, we will present evidence later that indicates that
typical student study habits (e.g., reviewing notes and homework

11998_Master4.indd 4 2/11/2020 [Link] PM

 Introduction 5

solutions) do not provide evidence that she or he can actually

apply concepts to solve problems—attempting to solve problems
is the way of gaining competence in problem solving. One does
not train to run a marathon by watching videos of marathon
runners.

1.4. What are some contributing factors leading to

many students finding physics difficult to learn?
Physics instructors are very adept at explaining the concepts and
problem solving procedures needed to succeed in introductory
undergraduate physics courses. Therefore, it is often not obvious
to physics instructors why exactly learning physics is difficult for
many. We mention two important factors at play here. The first
is referred to as “the expert’s blind spot,” meaning that once we
know something at a deep level (physics in our case), it is diffi-
cult to put ourselves in the shoes of a novice to determine which
are the ideas that are not obvious and why. For example, when
faced with a physics problem, our students might tell us that
they do not even know where to start, whereas we have seen so
many similar exercises in the past, it has become a routine pro-
cedure. The decisions are so automatic that we might have dif-
ficulty explaining how we know what to do. There is in fact
considerable tacit knowledge that we possess as physics experts
that we take for granted but deploy in our reasoning that stu-
dents are not privy to unless made explicit.
The second factor is largely the answer to the question: Why
don’t students simply follow the procedures they are taught in
class given that they are presented clearly and illustrated with
worked out examples? Students have strong and deeply rooted
erroneous intuitions about how the physical world works—
something that has been known for over four decades through

11998_Master4.indd 5 2/11/2020 [Link] PM

 6 The Science of Learning Physics

research on “misconceptions.” Those intuitions surface in “knee-

jerk” fashion when students encounter a situation that they
seemingly can explain with their intuitions, and when this hap-
pens, what we had taught them in our classes fades to the back-
ground. We will discuss this in one of the upcoming chapters
in detail. In fact, we will illustrate how students who are able to
solve a rather complicated mechanics problem display a major
misunderstanding when asked what physicists would deem a
simple conceptual question about the physical situation of the
problem they just solved. Hence among many students, espe-
cially those who find learning physics difficult, there is a con-
stant tension between adopting physicists’ ways of thinking and
abandoning their experiential (erroneous) intuitive notions. Why?
Their intuitions were formed through many experiences and
they make sense to them. Young children believe heavy objects
fall faster than light object because they do in their worlds from
many observations (rocks fall faster than leaves). Newton’s Third
Law explaining action-reaction is obvious to physicists but try
convincing a student that when a 300-pound football linesman
tackles them running at full speed, the force that the linesman
exerts on them is equal to the force that s/he exerts on the
linesman. It is intuitively obvious to the student that they will
“feel” the hit more than the linesman will, and they translate that
to mean that the larger, heavier object exerts a bigger force on
the smaller, lighter object.

1.5. What view undergirds this entire book?

Throughout this book we will promote the view that you should
“teach the way you do research.” That is, do not work on hunches
of what might work well in teaching students difficult concepts
and problem solving, but rather, start with research on learning,

11998_Master4.indd 6 2/11/2020 [Link] PM

 Introduction 7

apply it to classroom instruction, collect data on its effective-

ness, and cyclically refine the process. If you must try teaching
hunches of what you believe would be effective teaching tech-
niques, evaluate them in terms of evidence of student learning.
Teach the way you do research implies that you should let evi-
dence guide you; you can experiment with your teaching but
base your approaches on current knowledge about best teaching
and learning practices and collect data that allows a determina-
tion of whether or not students learned better.
We recognize that instructors are not teaching in isolation—
there are situational factors that also impact decisions about
what to teach and how to teach it. Perhaps at your institution
there is an expectation to do things the way they have always
been done, or there is pressure to stay consistent across mul-
tiple sections of a course. There might be limitations on the class-
room enrollment size and format. Maybe you are afraid that your
teaching evaluation scores will go down if you try something
new in your classroom. Regardless of your particular situation,
another view that undergirds this book is the need for a cul-
tural shift within STEM and especially Physics Departments—we
should be supportive of faculty who want to try EBIPs in their
classroom, and not hindering progress toward enhanced student
learning.

1.6. How are the subsequent chapters of this book

organized?
We have taken the view that faculty members are busy people
who must manage their professional time between teaching,
service and scholarly research. To simplify things we have orga-
nized every chapter into three sections, answering the following
three questions: (1) What does the research tell us? Here, we will

11998_Master4.indd 7 2/11/2020 [Link] PM

 8 The Science of Learning Physics

discuss salient research about the chapter’s topic, (2) What are
the implications of the research for instruction? Here, we inter-
pret the research into a few simple messages about what we
should, and should not be doing, to help students learn in the
most effective manner possible, and (3) What are some specific
examples of applying the research base to teaching physics?
Here, we will provide instructional examples in specific contexts
that illustrate ways of structuring instruction that are consistent
with research on learning.
Chapter 2 starts with a review of conceptual development
and the negative impact of misconceptions in learning physics
concepts. Chapter 3 focuses on how learning looks for a novice
and implications of expert–novice research on physics teaching
and learning. Chapter 4 reviews research on the teaching of
problem solving. Chapter 5 reviews active learning strategies for
the physics classroom. This is a good place to start if you are
unfamiliar with EBIPs or are looking for some quick ideas to
enhance your instruction. Chapter 6 addresses student percep-
tions of learning and their study habits—you might be surprised
to learn that several of students’ study strategies (such as high-
lighting or rereading) are not supported by research. Chapter 7
examines the topic of assessment, particularly some recent
research on the value of testing for learning and retention.
Chapter 8 provides some reflections on this book.
The book can be used in (at least) two ways. As for any book,
it can be read “all the way through” cover-to-cover if the reader
has the time. That is recommended since each chapter builds on
prior chapters and examples build on prior examples. A second
way to use the book is as a teaching reference guide by selecting
a particular topic/chapter to read. A busy faculty member may
want some ideas and insights on how to structure testing in a

11998_Master4.indd 8 2/11/2020 [Link] PM

 Introduction 9

course, or on ways of helping novice students put aside their

naive ideas and adopt the concepts and procedures of physics,
which are quite unforgiving if misapplied.
No matter how you choose to use this book, we hope you
find it both informative and thought provoking. We have strived
for brevity while still providing enough details for the ideas pre-
sented to make sense.

Reference
Tobias, S. (1990). They’re not Dumb, They’re Different: Stalking the Second
Tier. Tucson, AZ: Research Corporation.

11998_Master4.indd 9 2/11/2020 [Link] PM

 b2530   International Strategic Relations and China’s National Security: World at the Crossroads

This page intentionally left blank

b2530_FM.indd 6 01-Sep-16 [Link] AM

 c hapter
2 The Formation of Concepts and
How to Fix Broken Ones

2.1. What does research on students’ concept

formation tell us?
When students take a physics class, they come with a range of
prior knowledge and experiences that can affect the way they
learn physics. To understand how concepts form and evolve, it
helps to begin by discussing the current view of learning, called
constructivism (Trowbridge et al., 2000; Von Glasersfeld, 1989).
Constructivism is useful due to its ability to explain a wide range
of human behavior, including concept formation and problem
solving. However, to appreciate constructivism, it is best to begin
with a brief overview of its predecessor, namely the behaviorist
view of learning (see Wikipedia entry on behaviorism for a good
summary).

2.1.1. A comparison of behaviorist and constructivist

views of learning
The behaviorist approach for teaching a complex task is to break
up the task into components, teach each one, and later teach
by combining the components to achieve the desired behavior.
Thus in the behaviorist view, achieving competence is manifested
by the individual exhibiting behavior leading to correct com-
pletion of a task; positive reinforcement is used to encourage

11

11998_Master4.indd 11 2/11/2020 [Link] PM

 12 The Science of Learning Physics

appropriate behavior and negative reinforcement is used to dis-

courage inappropriate behavior. For example, a dog receives a
treat for performing a trick (positive reinforcement); or a dog
pays more attention when learning a new trick to avoid its mas-
ter’s displeasure (negative reinforcement).
Behaviorism does not care about two important aspects of
human learning, the first being the cognitive mechanisms used
by the individual to learn a complex task. Knowing how an indi-
vidual reasons while completing a complex task would seem
important in terms of designing effective instructional strategies;
that is, performing sub-components and stringing them together
does not add up to the understanding of the entire complex pro-
cess. The second thing that behaviorism does not care about is
whether or not the complex task, or processes leading to its com-
pletion, makes sense to the individual or whether or not the indi-
vidual would be able to appropriately apply what was learned to
a related or similar task. These would seem like important con-
siderations because if a complex task makes no sense to an indi-
vidual, then they will impose their own sense making, which may
not be accurate from a scientific standpoint; further, if knowl-
edge learned under behaviorism is so closely bound to the
task in which it was learned, then that knowledge is not flex-
ible enough to be applied broadly. Behaviorism is useful for
designing certain types of training (e.g., assembling a device,
testing for and applying chemicals to a swimming pool) but not
for educating a student on the nuances of physics concepts.
The constructivist view of learning holds that individuals
actively construct the knowledge they possess from observa-
tions and experiences that they encounter (National Research
Council, 2000; Piaget, 1978; Vygotsky, 1978). On the surface this
view seems obvious, perhaps even trite, since at one level, it

11998_Master4.indd 12 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 13

says that everyone’s knowledge is their own, and built brick by

brick throughout life. However, there is an important implica-
tion in constructivism for instruction, namely that people are
not blank slates when learning new knowledge—the knowledge
they already possess impacts all future new knowledge that will
be learned. This means that when new knowledge conflicts with
the knowledge that a student already possesses, the student is
likely not to accept the new knowledge and overwrite resident
knowledge. Why is this? Constructing knowledge is a time-
intensive, effortful process and the student has a strong attach-
ment to the knowledge they have constructed since it “works” for
them in real life. For example, many young students (and some
college students) believe that heavy objects fall at a faster rate
than light objects. This knowledge is accurate in many situa-
tions: a piece of paper falls slower than a rock, a leaf falls slower
than an acorn, and a foam ball falls slower than a baseball when
both thrown up in the air. A student who has grown up making
these accurate observations cannot help but construct the con-
cept that heavy objects fall faster than light objects, especially
since it is impossible to turn off the effect of air resistance. When
they come to a physics class and the instructor states that all
objects fall at the same rate, this new knowledge handed down
from authority makes no sense to students, and so they often
reject it or perhaps construct it as an exception (e.g., objects fall
at the same rate only in physics class). When teaching, we often
make generalizations without the accompanying caveats, e.g.,
all objects fall at the same rate (in the absence of air resistance).
Similarly, most students enter an introductory physics class
believing that, in a collision between a large truck and a compact
car, the truck exerts a larger force on the car than the car does
on the truck. Students base this on the damage to both vehicles

11998_Master4.indd 13 2/11/2020 [Link] PM

 14 The Science of Learning Physics

after the collision. When told in physics class that Newton’s Third
Law states that when two bodies interact the forces they exert on
each other are equal and opposite, their intellect thinks “Are you
kidding me?!” since they equate force with degree of damage.
The challenge for instructors is not only teaching correct scien-
tific concepts, but also eradicating (if that is even possible) stu-
dents’ erroneous notions.

2.1.2. Conceptual difficulties in physics

And so we see the difficulties in teaching many physics concepts
to students who come to class with their own private under-
standing of how the physical world works. By the time students
reach our physics classes they have been constructing knowl-
edge for many years in order to organize their experiences and
observations so that they make sense and also so that they can
be used to make predictions. The corpus of private understand-
ings that students bring to physics class is often incomplete,
fragmented, and riddled with preconceptions. When preconcep-
tions are in conflict with scientific concepts they are commonly
called misconceptions (other names are naive theories, and alter-
nate conceptions). The interested reader may want to delve into
detailed studies of a classic acceleration misconception, in
which acceleration is treated by students as if it were velocity.
For example, students often describe the acceleration of a ball
thrown straight up in the air as diminishing as the ball ascends,
becoming zero at the top of the trajectory and increasing
again as the ball descends (Clement, 1981; Hestenes et al., 1992;
Trowbridge & McDermott, 1980, 1981). If asked to discuss the
forces on the same ball, students often state that a strong ini-
tial force was imparted on the ball but that this force diminishes
as the ball climbs, becoming 0 at the top, and then the force

11998_Master4.indd 14 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 15

of gravity “takes over” and increases so as to speed up the ball

as it descends. This is akin to the medieval theory of “impetus”
and has been found in several studies (McCloskey et al., 1980;
Clement, 1982; Halloun & Hestenes, 1985). More recent dis-
cussions of students’ conceptual difficulties can be found in
(McDermott & Redish, 1999), and in (Docktor & Mestre, 2014).
Given that misconceptions are deeply rooted and difficult to
dislodge, the question becomes: What can instructors do to help
students overcome misconceptions? There are two prominent
conflicting schools of thought on that issue. The first school of
thought on overcoming misconceptions holds that they need
to be supplanted with correct scientific concepts, and to do so,
requires a conflict/resolution approach (Posner et al., 1982; Strike
& Posner, 1992). Within this view the instructor would create dis-
satisfaction with the student’s current concept by pointing out,
for example, how it is inconsistent for explaining a wide range of
phenomena; at the same time, the new concept needs to be intel-
ligible or the student will not understand it. The instructor then
shows the plausibility of the scientific concept and its usefulness
in making predictions about physical phenomena. This approach
is an all-or-nothing replacement of the misconception, which
unfortunately seldom occurs. Studies show that students often
display correct understanding shortly after a scientific concept is
presented in class only to revert to their private understanding
weeks later (Clement, 1982; Halloun & Hestenes, 1985).
A more modern approach, termed “knowledge in pieces”
(diSessa, 1993; diSessa & Sherin, 1998; Hammer 1996a, 1996b,
2000; Smith et al., 1994) holds that students’ knowledge is made
up of many “pieces” (or mental “resources”) that are compiled in
real time to reason about physical situations, making this view
more dynamic. A misconception in this view is due to an error

11998_Master4.indd 15 2/11/2020 [Link] PM

 16 The Science of Learning Physics

in recalling and compiling (inappropriate) knowledge pieces to

yield an (incorrect) explanation. Within this view, the instructor
would help students use/compile appropriate knowledge pieces
in reasoning. That is, learning concepts within this view is man-
ifested by increased accuracy and consistency in coordinating
knowledge pieces to yield accurate explanations of scientific phe-
nomena. Note that eradicating misconceptions from memory is
not a goal in the knowledge-in-pieces view, especially since cog-
nitive science is not at a stage to determine whether it is even
possible to permanently and consciously forget something that
we previously learned. Thus, inappropriate knowledge pieces that
could be used in reasoning about physics concepts (e.g., heavy
objects fall faster than lighter objects) can co-exist with correct
knowledge pieces (e.g., all objects fall at the same rate in the
absence of air resistance) so long as they are judiciously coor-
dinated to explain phenomena. For example, explaining why a
piece of paper falls faster than a rock (in opposition to all objects
falling at the same rate) when held and dropped at arm’s length
is due to air resistance precluding the paper from falling faster,
but if the piece of paper is crumbled into a tight ball, the rate of
fall would be perceptibly the same for both objects.
The types of misconceptions mentioned above are the result
of students’ knowledge construction outside of physics classes.
There is also another, more insidious type of misconception
consisting of students applying a correct concept learned in
physics class to a context in which it does not apply, yielding
erroneous reasoning; this type of misconception is compiled
from knowledge pieces on-the-fly to reason about situations. In
fact, the detailed example that we are about to discuss is such
a case, and readers will likely be surprised at the bizarre stu-
dent behavior that can result. The example will also illustrate

11998_Master4.indd 16 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 17

how the knowledge-in-pieces view can be used to explain stu-

dent behavior. The example consists of university students
making judgments about realistic motion of two balls rolling
side-by-side along different tracks in a race. A surprisingly large
number of students taking introductory physics interpreted
a piece of physics knowledge incorrectly and coordinated it
with other knowledge pieces to form a strong misconception
about the results of the race between the two balls. In contrast,
physics-naive college students who had not taken introductory
physics coordinated “common-sense” knowledge beautifully to
reason correctly about the race.

2.1.3. Example: A study with balls and tracks

To unpack this odd behavior, we begin by summarizing the task
given to students, with additional details found in the published
study (Thaden-Koch et al., 2006). The researchers constructed
five computer animations of two balls released simultaneously
from the top of two tracks, as shown in Fig. 2.1. Figure 2.1 shows
a time-lapsed picture of the position of the balls in the five ani-
mations at equal time intervals (labeled sequentially with num-
bers). The four animations depicting unrealistic motion were not
constructed from the researchers’ whims, but rather a group of
students was given a picture of the two tracks with only Ball A
drawn in at equal time intervals in the eight positions shown and
students were asked to draw the position of Ball B for the eight
time intervals on the second track having the “V” valley. The
researchers selected the four most popular erroneous renditions
and constructed computer animations of the two balls moving in
a race. Note that in Fig. 2.1, animation “e” (the last one) depicts
realistic motion with Ball B winning the race by a considerable
margin. We will also point out that animation “c” is very peculiar

11998_Master4.indd 17 2/11/2020 [Link] PM

 18 The Science of Learning Physics

(a)

(b)

(c)

(d)

(e) real
animation

Fig. 2.1. Pictures depicting ball behavior in the animations. Reproduced from Physical
Review: Physics Education Research, 2006, 2 (#2) 020107 with permission from the
American Physical Society.

and unrealistic since it depicts Ball B speeding up on the uphill

portion of the “V”; however, this animation is a very popular
selection among students who had just finished an introductory
physics course—more on why later.
Two groups of students participated in the study, one from
an honors calculus-based introductory course (participation took
place near the end of the course), and one from an introduc-
tory educational psychology course; the educational psychology

11998_Master4.indd 18 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 19

students had taken no college physics. Students were brought

into an office individually and shown two sets of animations.
The first set consisted of the five from each group of five that
depicted correct, realistic motion of the ball(s). Students had con-
trol of playing the animations in whatever order, and however
many times they wished (the order of the animations was ran-
domized). After making many comparisons, each student eventu-
ally chose one of the five animations as the animation depicting
the real, correct motion.
Tables 2.1 and 2.2 show the results, with Table 2.1 showing the
selections of the physics students and Table 2.2 the selections of
the physics-naive students. At the end of each row is the number
of students selecting Ball B animations as realistic and at the
bottom of each column is the number of students selecting the
animation they believed to be realistic in the two-ball race. Note
that for both groups of students, animation “b” and the real ani-
mation (“e”) were the most popular in judging the motion of
Ball B moving alone on the “V” track. This is not surprising: The
real animation shows Ball B speeding up on the downhill por-
tion of the “V” and slowing down on the uphill portion. In anima-
tion “b” the ball slows down considerably on the uphill portion
as if it experienced a lot of rolling friction. Thus we can conclude
that generally, students are acting reasonably in attempting
to judge realistic motion of one ball moving along a “V” track.
Importantly, note that not a single student chose animation “c”
as realistic, which depicts Ball B speeding up on the uphill por-
tion of the “V”.
The surprise comes when looking at the two-ball data (the
columns in Tables 2.1 and 2.2). Here, we see that 35% of the
physics-naive group chose the realistic animation whereas
only 4% (1 student) chose the realistic animation among the

11998_Master4.indd 19 2/11/2020 [Link] PM

 20 The Science of Learning Physics

Table 2.1. Animation preferences for 1- and 2-ball tasks, 24 students from Introductory
Calculus-based Physics (Honors). Note that entries on the diagonal correspond to students
who answer consistently across the 1-ball and 2-ball animations. Reproduced from Physical
Review: Physics Education Research, 2006, 2 (#2) 020107 with permission from the American
Physical Society.

Number choosing 1-ball and 2-ball 2-ball 2-ball 2-ball 2-ball (1-ball)
2-ball animations as realistic (a) (b) (c) (d) (e) (real) Row Total
(percent)

1-ball (a) - - 1 1 - 2 (8%)

1-ball (b) - 3 7 - - 10 (42%)

1-ball (c) - - - - - 0 (0%)

1-ball (d) - - - - - 0 (0%)

1-ball (e) (real) - 1 7 3 1 12 (50%)

(2-ball) Col. Total (percent) 0(0%) 4(17%) 15(63%) 4(17%) 1(4%) 24 (100%)

Table 2.2. Animation preferences for 1- and 2-ball tasks, 26 students from an Educational
Psychology class. Note that entries on the diagonal correspond to students who answer con-
sistently across the 1-ball and 2-ball animations. Reproduced from Physical Review: Physics
Education Research, 2006, 2 (#2) 020107 with permission from the American Physical Society.

(1-ball)
Number choosing 1-ball and 2-ball 2-ball 2-ball 2-ball 2-ball
Row Total
2-ball animations as realistic (a) (b) (c) (d) (e) REAL
(percent)

1-ball a) - 3 - 1 2 7 (23%)

1-ball b) 2 3 - - 1 9 (35%)

1-ball c) - - - - - 0 (0%)

1-ball d) - - - 1 - 1 (4%)

1-ball e) (real) 2 2 - - 6 10 (38%)

(2-ball) Col. Total (percent) 4(15%) 11(42%) 0(0%) 2(8%) 9(35%) 26 (100%)

11998_Master4.indd 20 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 21

physics students. Even more bizarre, we see now that 63% of

the physics students chose animation “c” as realistic (the one
showing Ball B speeding up while going uphill), whereas not a
single physics-naive student chose animation “c” as realistic.
The researchers report that when viewing the single-ball ani-
mations for Ball B, a lot of students from both groups laughed
when viewing animation “c” and dismissed it outright as unreal-
istic—after all, who would think that a ball would speed up going
uphill without a cause. Yet, no physics student laughed when
viewing the two ball “c” animation, and the majority of them
selected it as realistic. What reasoning could have led to this
bizarre behavior?
During the experiment, students were asked to “think aloud”
while making their decisions about realistic motion—that is, they
were asked to verbalize the reasoning that led to their choice of
realistic motion, and the entire session was audio-recorded for
later analysis. Generally, the reasoning of both groups showed
correct “knowledge pieces” brought to bear on the situation. For
example, students from both groups expected Ball B to speed
up on the downhill and to slow down on the uphill. The bizarre
behavior of 63% of the physics students selecting animation “c”
in the two-ball race as realistic stemmed from the inappropriate
application of a physics concept learned in class, namely con-
servation of mechanical energy. Those students reasoned that
both balls were released from the same height, and returned to
the same height at the end of the race, and that by conservation
of energy they had to have the same speed at the end and thus
they had to tie at the end. Note that these “knowledge pieces”
as applied by the physics students are all correct except for the
conclusion that equal speeds at the end implies equal posi-
tions at the end. Interestingly, the physics students invoking this

11998_Master4.indd 21 2/11/2020 [Link] PM

 22 The Science of Learning Physics

argument were looking for a tie result at the end of the race, and
only two animations had tie results (c and d); but animation “d”
did not meet the criteria of Ball B speeding up on the downhill
and slowing down on the uphill. The researchers also reported
that for many physics students invoking this argument, looking
for a tie outcome precluded them from noticing Ball B speeding
up in the uphill portion in the two-ball race.
There are several take-away messages from this study. First,
when learning a physics concept for the first time, students
can misinterpret or overextend the implications of the concept,
leading to erroneous reasoning. Second, having a strong expec-
tation from an incorrect application of a concept can lead to
not noticing important features of a situation (in this case, a
ball speeding up while going uphill). Third, students coordinate
pieces of knowledge “on-the-fly” to reason about situations, as
the knowledge in pieces view asserts. Finally, and on the positive
side, the physics students in this study did try to apply physics
knowledge that they learned in class, but unfortunately, lacking
the expert’s perspective they were prone to overextend or mis-
­
apply the conservation of mechanical energy concept.

2.2. What are the implications of research on

students’ concept formation for instruction?
We offer the following implications of the research reviewed
above for instruction:
● Students come to our classes with preconceptions that
often interfere with the physics concepts that we try
to teach them. Students construct their own private
understandings of the physical world around them
through observations and experiences. The preconceptions
formed by this process are often misconceptions that

11998_Master4.indd 22 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 23

run counter to correct physics concepts. Because it takes

time and mental energy to construct concepts, students’
misconceptions are deeply rooted and resistant to change.
It is important for instructors to be on the lookout for
students’ misconceptions since they interfere with the
learning of physics concepts. More on available inventories
to identify misconceptions provided in the Teaching
Interventions section below.
● “Teaching by telling” will likely not be sufficient to help
students overcome misconceptions. Although the lecture
method of instruction is effective for some students,
learning via this method is rather passive and allows little
opportunity for students to construct knowledge. Methods
that offer more engagement are better suited for knowledge
construction. If a student possesses a misconception it
becomes even more important to engage the student in
reconstructing appropriate conceptual understanding.
When listening to an instructor discuss a concept that
conflicts with previously constructed knowledge, a student
is not likely to give up their own private understanding and
replace it with the instructor’s version.
● Nuanced understanding about the meaning and
appropriate application of a concept takes time and
multiple contextual experiences. The example provided
above with the two rolling balls in a race along two tracks
of different shapes shows that students often overextend or
misinterpret the applicability of a concept. When presenting
a new concept, instructors should discuss both, the contexts
in which the concept applies, as well as the more salient
contexts in which it does not apply to allow students to
build a rich contextual knowledge base of the concept’s
applicability.

11998_Master4.indd 23 2/11/2020 [Link] PM

 24 The Science of Learning Physics

● Some misconceptions are easier to overcome than others.

Misconceptions that have been held for a long period of
time and which work in predicting physical outcomes in
the world we live in (e.g., heavy objects fall faster than
light objects, as in a rock falls faster than a leaf) are harder
to reconstruct than others that are formed on-the-fly by
compiling correct with incorrect knowledge pieces. The
balls-on-tracks example above is an example of the latter
case where students compile in real time some correct
knowledge pieces (e.g., the balls will have the same speed
in the final shelf due to conservation of mechanical energy)
with incorrect knowledge pieces (e.g., having the same speed
at the end implies the balls will arrive at the same time).
It is much easier for students to “fix” incorrect knowledge
pieces than to abandon long-held misconceptions that have
observable verification in our messy “real world” (e.g., heavy
objects fall faster than light objects).
● Expect some students to exhibit appropriate under­
standing of a physics concept immediately after teaching
­
it but to revert to an associated misconception they
possess related to the concept weeks later. It is common to
have students exhibit understanding of physics conceptual
questions when tested shortly following instruction only
to have misconceptions on the concept return when tested
weeks later. From a constructivist viewpoint, recall that
students construct their own private understanding of the
concept from observations and the constructed concept
makes sense to the student in terms of making predictions
about the physical world. It is difficult for students to give
up their constructed knowledge after a short exposure to a
physicist’s presentation of the concept. From a knowledge-
in-pieces perspective, the student has numerous knowledge

11998_Master4.indd 24 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 25

pieces in memory, some correct (e.g., concepts taught in

physics class) and some incorrect (misconceptions from
observing the physical world), and which piece of knowledge
gets activated in memory is highly context dependent.
● Misconceptions are difficult to uproot but instructors
can take steps to help students construct appropriate
understanding of physics concepts. It is unlikely that
one can ever consciously erase one’s own memory of
a misconception once formed. However, with the help
of instructors, one can always learn to cope with it. By
pointing out the different contexts in which a concept holds
and the contexts in which it does not, by pointing out how
a students’ misconception might be valid under certain
conditions but not universally, and by helping students
compile appropriate knowledge pieces to reason about
situations, instructors can help students become aware of
common pitfalls and develop a conceptual knowledge base
suitable for doing physics.

2.3. Examples of teaching interventions based on

learning research
The research reviewed in this chapter, as well as our teaching
experiences, indicate that students are not blank slates when it
comes to having their own private “theories” about how the phys-
ical world works. It is therefore important to probe for students’
conceptual understanding during the course of instruction so
that we can identify misconceptions and help students restruc-
ture their knowledge to reason effectively about the physical
world.
● Implement active learning (more details on active
learning in Chapter 5). Active learning takes many forms,

11998_Master4.indd 25 2/11/2020 [Link] PM

 26 The Science of Learning Physics

but essentially, it consists of getting students actively

engaged in constructing knowledge (e.g., using conceptual
“clicker” questions in large classes where students discuss
and then answer them in small collaborative groups).
Although active learning is not guaranteed to perfect
students’ conceptual understanding, it does offer many
benefits. For example, it allows instructors to probe for
misconceptions. Further, it also allows students and
instructors to openly discuss how misconceptions result
in an inconsistent description of the physical world and
sets the stage for students reconstructing knowledge.
Although not causal, there is strong correlational evidence
that courses taught using active learning strategies result
in fewer misconceptions following instruction com

­
pared to more traditional methods of instruction (Hake,
1998; Freeman et al., 2014). Such evidence comes from
administering a conceptual inventory (designed to probe
for misconceptions using multiple-choice questions) prior
to, and following instruction and comparing the pre-
to-post gains in performance. There are about a dozen
commonly used conceptual inventories in introductory
physics, among the most popular being the Force Concept
Inventory (Hestenes et al., 1992) and the Force and Motion
Conceptual Evaluation (Thornton & Sokoloff, 1998). A quick
reference guide of other conceptual inventories in physics
is found on page 24 in (Docktor & Mestre, 2014).

We now provide two examples of question sequences that have

been used in large lectures by one of us (JPM) using clicker tech-
nologies both to identify specific misconceptions and to help stu-
dents build appropriate conceptual knowledge. The first example
uses a classroom demonstration where a spring scale has two

11998_Master4.indd 26 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 27

light strings attached at either end, one end attached to a sta-

tionary hook and the other going over a pulley and attached to
a known weight. Figure 2.3 shows the situation. We present the
demo to students without first hanging the weight, and demon-
strate that the harder we pull on the string at the right, the
higher the reading on the scale. Then we cover the scale with a
cloth, hang the known weight, and ask the following multiple
choice question for students to discuss in small groups (note
that the cloth is not shown in Fig. 2.3).
When answers are collected with the clickers, not surprisingly,
all students answer correctly—4 lbs. Next we change the context
slightly by covering the scale with the cloth again, removing the
left end from the hook, attaching another 4 lbs weight to that
end, and putting the string over a second pulley at the left end
so that both weights hang without moving with the veiled scale
in the middle. Figure 2.4 shows the new set up and the question
asked (note that the cloth over the scale is not shown in Fig. 2.4).
If you guessed that the majority now answer 8 lbs, you would
be correct. (If you are not a physicist reading this and are sur-
prised, the answer does not change from before.) Most students
now think that hanging an equal weight from both sides dou-
bles the force reading on the scale and the tension in the strings.
They reason that gravity is pulling with 4 lbs on each side, which
adds up to 8 lbs—visually this seems reasonable to them without
analyzing the situation with Newton’s Second Law. Note that they
reasoned perfectly fine in the first case, thinking that gravity
pulls with 4 lbs and thus the scale must read that amount since
the string attached to the weight is holding 4 lbs and thus pulls
the scale with that force since the weight remains in equilib-
rium. Yet, in the second case, the weight hanging on the right is
still in equilibrium, so as far as the scale is concerned, nothing

11998_Master4.indd 27 2/11/2020 [Link] PM

 28 The Science of Learning Physics

A block weighing 4 lbs is hung from a

string that goes over a pulley and is
then attached to a scale. The scale is
attached to a hook on the wall. What
does the scale read?
4 lbs (a) 0 lb (b) 4 lbs (c) 8 lbs

Fig. 2.3

What does the scale read

now?
(a) 0 lb (b) 4 lbs (c) 8 lbs

4 lbs 4 lbs
Fig. 2.4

has changed. This offers an opportunity for the instructor to

refine students’ understanding by helping them analyze the two
hanging weights using Newton’s Second Law, showing that pre-
viously the hook provided the 4 lbs force needed to keep the
scale and weight in equilibrium, and in the second case the 4 lbs
weight on the left serves that same purpose. One can also pose
the argument to students that, if this weren’t so, there would be
a major contradiction with Newton’s Second Law: If the scale

11998_Master4.indd 28 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 29

read 8 lbs, then the two strings to either side of the scale would
have 8 lbs of tension, so that now, for both hanging weights,
there would be an unbalanced force of 4 lbs acting vertically up
on the masses, and thus they both would magically have to accel-
erate upwards!
Another two-question set also serves to highlight a common
misconception and ways of addressing it. The first question pres-
ents a physical situation consisting of two blocks of different
masses in contact with each other on a slippery horizontal table
having different forces applied from each side, and asked for the
acceleration of the blocks, as shown in Fig. 2.5.
Students are invited to discuss the question with their neigh-
bors and to answer when done. Not surprisingly, following
instruction on Newton’s Second Law, very few students get this
question wrong after answers are collected with clickers. When
asked to volunteer their reasoning, students argue correctly
that the two blocks can be considered a “body” for analysis with
Newton’s Second Law, and then add the two forces vectorially to
arrive at a net force of 12 N toward the right, which when divided
by the total mass of 4 kg yields the 3 m/s2 acceleration.
The surprise and conflict comes when asked the follow-up
question shown in Fig. 2.6. Now misconceptions magically sur-
face, as students abandon Newton’s Second Law and resort to
knee-jerk seemingly-reasonable answers that are flat-out wrong.
Students choosing choice [(a) and (b)] argue that the force
on the left (right) applied to the 3 kg (1 kg) block just “goes
right through” the block. Those answering (c) argue that the
force in the middle between the blocks is the difference of the
forces acting at the two ends. Those answering 16 N argue that
the force in the middle is the sum of the two forces since the
middle is being squashed by the two end forces. What is crucially

11998_Master4.indd 29 2/11/2020 [Link] PM

 30 The Science of Learning Physics

F1=14 N
F2=2 N
3kg
1kg

In the situation shown, two horizontal external forces are applied to the two blocks as
shown. There is no friction. What is the acceleration of the blocks?
(a) 4 m/s2 (b) 8 m/s2 (c) 3 m/s2 (d) 3.67 m/s2

Fig. 2.5

F1=14 N
F2=2 N
3kg
1kg

In the situation shown, two horizontal external forces are applied to the two blocks as
shown. There is no friction. What is the force that the 1 kg block exerts on the 3 kg block,
Fon 3 by 1? Recall that a = 3 m/s2.
(a) 14 N (b) 2 N (c) 12 N (d) 16 N (e) 5 N

Fig. 2.6

important to note is that in the first question, students applied

Newton’s Second Law flawlessly to arrive at a correct answer.
And so the question becomes: Why did they so quickly abandon
Newton’s Second Law to analyze the second situation? Had they
identified the 3 kg block as the body of interest, drawn a free-
body diagram of the forces on it, and analyzed it the same way
they analyzed the two blocks in the first question, it would have
inevitably led them to the correct answer of 5 N given that they
know the acceleration of each block from the first question.
Note that in both of these examples, two slightly different
contexts were used, with the second context tempting students
to abandon the physics presented in class in favor of their own
knee-jerk reasoning. The second question in each set created con-
flict in students’ minds, which with guidance from the instructor

11998_Master4.indd 30 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 31

could be resolved with appropriate physics reasoning. One of

us who has used these examples in class likes to add humor to
these situations during class by asking students “Did I ever teach
you ‘Newton’s Difference of Two Forces Law’ that states that when
confronted with two forces, just add them or subtract them?” or
“Did I ever teach you ‘Newton’s the-force-goes-right-through-a-
body Law’?” Students laugh, but questions like these prove the
point that physics in unforgiving unless one plays by the rules.
Examples like these also serve to caution students not to fall
prey to their knee-jerk intuitions and to instead rely on analyses
using physics concepts taught in class.
● Implement formative assessment classroom techniques
to elicit students’ ideas. Formative assessment techniques
(i.e., assessment in the service of learning rather than for
assigning grades) include a host of different low-stakes
strategies for making students’ thinking more visible to
the instructor (Keeley, 2008). You’re probably familiar
with think-pair-share where a student first thinks about
a question on their own, talks to a partner, and then
shares ideas with the class. This general strategy could
be implemented with the conceptual clicker questions
addressed in the previous bullet. Other strategies include
using mini whiteboards during class, giving students time
during class to pause and write/summarize the lesson so
far or identify the muddiest point, concept mapping, or
focused listing where a student brainstorms everything they
already know about a particular topic at the beginning of a
new unit. If you teach a physics course for future teachers,
the book series Understanding Student Ideas in Science by
Page Keeley contains hundreds of formative assessment
probes on K-12 topics that can also be used with college

11998_Master4.indd 31 2/11/2020 [Link] PM

 32 The Science of Learning Physics

students. For example, when provided with a list of objects

and asked which ones will stick to a magnet, most pre-
service teachers will say that all metals stick to a magnet.
Even after conducting several investigations where they
observe that magnets do not stick to a piece of aluminum
foil or a copper pipe, students will often revert to their
original ideas. When asked whether they will be able to see
an apple in a completely dark room, many students will say
that their eyes will eventually adjust to see the outline of
the apple or an apple without color, even when there is no
light for their eyes to detect. These examples illustrate that
many of the misconceptions that start in young children
permeate into adulthood.
● Implement other instructional strategies specifically
designed to identify and address students’ misconcep­
tions. There are other specific strategies to help students
overcome common stubborn misconceptions. We discuss
briefly two of them here but the reader may wish to read
more in the Docktor and Mestre synthesis of physics
education research (2014). The Washington University’s
Physics Education Group’s Tutorials in Introductory Physics
(McDermott & Shaffer, 2002) offers a supplementary curric-
ulum for use in small group discussion sections, although
they can be adapted for use in lecture settings. The curric-
ulum consists of “modules” treating specific topics (e.g.,
velocity & acceleration, Newton’s Laws, geometric optics,
gas laws) and the modules consist of pretests, worksheets
and homework assignments designed to identify miscon-
ceptions and to help students overcome them. A second
approach for use in large lecture settings are the Interactive
Lecture Demonstrations (ILD) (Sokoloff & Thornton, 2004).
This curriculum combines lecture demonstrations with

11998_Master4.indd 32 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 33

conceptual development through a specific sequence of

instructional activities. The activities start by introducing
a lecture demonstration, then asking students to make a
prediction on the outcomes after discussing with peers,
then they see the actual demonstration (usually showing
erroneous reasoning by a large portion of students),
and then attempt to reach resolution on the concepts
underlying the lecture demonstration. Each stage in this
approach is guided by questions in hand-out worksheets.
Studies on the effectiveness of ILDs show significant
improvements in the understanding of basic physics
concepts as measured by the FMCE. Another approach is
the use of bridging analogies and anchoring intuitions to
lead students to appropriate conceptions (Clement, 1998).
For example, to address the common misconception that
passive objects like tables cannot exert a force, a lesson can
start with an anchoring example that triggers the correct
conception: when a hand pushes down on a spring, the
spring pushes back on the hand. Then a series of bridging
analogies are used and discussed: a book resting on a
piece of foam, a book resting on a long flexible board, and
finally a book resting on a rigid table. Demonstrations can
be used with lasers and mirrors to illustrate that even a
“rigid” table undergoes small deformations when an object
is placed on top of it, and eventually the lesson can include
a microscopic model of atoms in a solid having spring-like
bonds between them. These approaches, as well as others,
have commonalities, such as creating conflict between
students’ current understanding and the targeted physics
concept, and then guiding students through the resolution
of the conflict. They also offer opportunities for discussing
concepts in multiple contexts.

11998_Master4.indd 33 2/11/2020 [Link] PM

 34 The Science of Learning Physics

References
Clement, J.J. (1981). Solving problems with formulas: Some limitations.
Engineering Education, 72, 158–162.
Clement, J.J. (1982). Students’ preconceptions in introductory mechanics.
American Journal of Physics, 50, 66–71.
Clement, J.J. (1998). Expert novice similarities and instruction using
analogies. International Journal of Science Education, 20(10), 1271–
1286.
diSessa, A.A. (1993). Toward an epistemology of physics. Cognition and
Instruction, 10, 105–225.
diSessa, A.A. & Sherin, B.L., (1998). What changes in conceptual change?
International Journal of Science Education, 20(10), 1155–1191.
Docktor, J.L. & Mestre, J.P. (2014). Synthesis of discipline-based educa-
tion research in physics. Physical Review Special Topics – Physics Edu

­
cation Research, 10(2), 020119 (58 pages). doi: 10.1103/PhysRevSTPER.
­
10.020119
Freeman, S., Eddy, S.L., McDonough, M., Smith, M.K., Okoroafor, N., Jordt,
H. & Wenderoth, M.P. (2014). Active learning increases student perfor-
mance in science, engineering, and mathematics. Proceedings of
the National Academy of Sciences, 111(23), 8410–8415. doi: 10.1073/
pnas.1319030111
Hake, R.R. (1998). Interactive-engagement versus traditional methods: A
six-thousand student survey of mechanics test data for introductory
physics courses. American Journal of Physics, 66, 64–74.
Halloun, I.A. & Hestenes, D. (1985). The initial knowledge state of college
physics students. American Journal of Physics, 53, 1043–1055.
Hammer, D. (1996a). Misconceptions or p-prims: How might alternative
perspectives on cognitive structure influence instructional percep-
tions and intentions. Journal of the Learning Sciences, 5, 97–127.
Hammer, D. (1996b). More than misconceptions: Multiple perspectives
on student knowledge and reasoning, and an appropriate role for
education research. American Journal of Physics, 64(10), 1316–1325.

11998_Master4.indd 34 2/11/2020 [Link] PM

 The Formation of Concepts and How to Fix Broken Ones 35

Hammer, D. (2000). Student resources for learning introductory physics.

American Journal of Physics, 68, S52–S59.
Hestenes, D., Wells, M. & Swackhamer, G. (1992). Force concept inventory.
The Physics Teacher, 30, (March), 159–166.
Keeley, P.D. (2008). Science Formative Assessment: 75 Practical Strategies
for Linking Assessment, Instruction, and Learning. NSTA Press.
McCloskey, M., Caramazza, A. & Green, B. (1980). Curvilinear motion in
the absence of external forces: Naive beliefs about the motion of
objects. Science, 210, 1139–1141.
McDermott, L.C. & Shaffer, P.S. (2002). Tutorials in Introductory Physics.
Prentice-Hall, Upper Saddle River, NJ.
McDermott, L.C. & Redish, E.F.(1999). Resource letter: PER-1: Physics
education research. American Journal of Physics, 67, 755–767.
National Research Council (2000). How People Learn: Brain, Mind, Experi-
ence, and School: Expanded Edition. Washington, DC: The National
Academies Press. https:// [Link]/10.17226/9853
Piaget, J. (1978). Success and Understanding. Cambridge, MA: Harvard
University Press.
Posner, G., Strike, K., Hewson, P. & Gerzog, W. (1982). Accommodation of a
scientific conception: Toward a theory of conceptual change. Science
Education, 66, 211–227.
Smith, J., diSessa, A. & Roschelle, J. (1994) Misconceptions reconceived:
A constructivist analysis of knowledge in transition. Journal of the
Learning Sciences, 3, 115–163.
Sokoloff, D.R. & Thornton, R.K. (2004). Interactive Lecture Demonstrations,
Active Learning in Introductory Physics. John Wiley & Sons, Hoboken,
NJ.
Strike, K.A. & Posner, G.J. (1992). A revisionist theory of conceptual change.
In Philosophy of Science, Cognitive Psychology, and Educational Theory
­
and Practice. In R.A. Duschl and R.J. Hamilton (Eds.) pp. 147–176,
State University of New York Press, New York.
Thaden-Koch, T., Dufresne, R. & Mestre, J. (2006). Coordination of
knowledge in judging animated motion. Physical Review Special

11998_Master4.indd 35 2/11/2020 [Link] PM

 36 The Science of Learning Physics

Topics: Physics Education Research, 2(2), 020107 (11 pages). doi:

10.1103/PhysRevSTPER.2.020107
([Link]
Thornton, R.K. & Sokoloff, D.R. (1998). Assessing student learning of
Newton’s laws: The force and motion conceptual evaluation and
the evaluation of active learning laboratory and lecture curricula,
American Journal of Physics, 66, 338–352.
Trowbridge, L.W., Bybee, R.W. & Powell, J.C. (2000). Teaching Secondary
School Science: Strategies for Developing Scientific Literacy, 7th ed.
Prentice-Hall, Upper Saddle River, NJ.
Trowbridge, D.E. & McDermott, L.C. (1980). Investigation of student
understanding of the concept of velocity in one dimension. American
Journal of Physics, 48, 1020–1028.
Trowbridge, D.E. & McDermott, L.C. (1981). Investigation of student
understanding of the concept of acceleration in one dimension.
American Journal of Physics, 49, 242–253.
Von Glasersfeld, E. (1989). Cognition, construction of knowledge, and
teaching. Synthese, 80, 121–140.
Vygotsky, L.S. (1978). Mind in Society: The Development of Higher Psycho-
logical Processes. Cambridge, MA: The Harvard University Press.
(Originally published 1930, New York: Oxford University Press.)

11998_Master4.indd 36 2/11/2020 [Link] PM

 c hapter
3 How Learning Looks for a Novice:
Implications of Expert–Novice Research
for Physics Teaching and Learning

3.1. What does expert–novice research tell us?

To understand how students taking introductory physics behave
in terms of solving problems, it helps to review findings from
the study of expertise, not because our job as instructors is nec-
essarily to turn those students into experts, but because how
experts and novices organize knowledge in memory and deploy
it to solve problems is quite different, and provides a window
into novices’ minds. The first thing that becomes evident as one
begins to learn a new area is that one lacks perspective on what
is important. We have all had experiences where we have felt
like a novice—perhaps learning to play a musical instrument, or
learning a new sport or complicated board game. For example,
in learning to play golf for the first time, should one swing the
club like a baseball bat only downwards? Should one swing the
different clubs the same way? Should one keep the arms flex-
ible during a swing or keep one arm stiff? Should one twist
one’s hips when swinging or keep them facing the ball during
the entire swing? Does one get more power by swinging harder?
Answers to these questions can be surprising to a novice, and in
fact, preconceived answers to these questions can get in the way,

37

11998_Master4.indd 37 2/11/2020 [Link] PM

 38 The Science of Learning Physics

making progress in learning the game. For example, the answer

to the last question is no—power (length that the ball travels for
a given club) is about technique, not strength of swing. One of
us who’s a “journeyman” in golf can attest to witnessing scrawny
guys and gals hitting very long shots indeed due to their excel-
lent form, whereas muscly guys who swing hard (with poor tech-
nique) usually hit embarrassingly short shots. As one plays more
and more, one starts to gain perspective on what is important,
and then comes refinement. The same is true for learning
physics—we begin to notice the subtleties of “playing the physics
game,” the more we learn physics.

3.1.1. Broad observations about novices and experts in

physics
Lacking any perspective on what is important in physics,
a beginner treats every new concept or equation encoun-
tered in an introductory course as equally important (or unim-
portant). All concepts, equations and facts are thrown into
memory with little by way of organization other than perhaps
the order in which they are covered or the context in which
they are presented in the course. Solving problems is usually a
pattern-matching exercise for the novice, hunting for equations
whose variables match those given in the problem, and then
blindly plugging in values to generate an answer. We have all wit-
nessed this behavior and wonder why our students do not orga-
nize knowledge and use it to solve problems “our way,” or the
way we try to teach them, namely to start with an analysis of the
problem and decide what “big idea” applies, and then to move
towards specifics such as the equation(s) and procedure(s) that
go with the application of the big idea. In fact, physicists will
likely agree that in an introductory mechanics course there are

11998_Master4.indd 38 2/11/2020 [Link] PM

 How Learning Looks for a Novice 39

only four big ideas covered (Newton’s Second Law, work-energy

theorem and conservation of mechanical energy, impulse-
momentum theorem and conservation of momentum, and angular
impulse-angular momentum theorem and conservation of
angular momentum) and a few secondary ideas (e.g., kinematics,
the other two Newton Laws). Yet introductory physics students
are always complaining to us that there is so much to remember,
meaning the hundred-some-odd equations they encounter in
a typical course, which is largely a consequence of poorly orga-
nized physics knowledge in memory. When experts activate a
physics concept or principle, the particular concept/principle in
memory is “attached” to a procedure(s)/equation(s) for applying
it; this makes remembering the plethora of equations needed
in physics problem solving easy for the expert since their entire
memory store is organized efficiently (National Research Council,
2000). It is important to be patient with our novice students,
since it is not that they are trying to be obstinate—they are just
behaving in the way that all novices behave. After all, those who
study expertise state that it takes about 10,000 hours to become
an expert at something (Ericsson et al., 1993), and by the end
of one semester of introductory physics, our best students are
only about 140 hours into it (assuming that they are diligent
and spend ~10 hours/week during the entire 14-week semester—
granted an optimistic estimate).
Although the research literature typically classifies partic-
ipants as either “novices” or “experts,” it is important to note
that the development of expertise occurs across a continuum.
Dreyfus and Dreyfus (1986) describe five stages of skill acquisi-
tion. A novice is someone who is still learning the facts and rules
of a skill; their decision-making process is very deliberate and
tasks require full concentration. With more experience, a novice

11998_Master4.indd 39 2/11/2020 [Link] PM

 40 The Science of Learning Physics

progresses to the stage of advanced beginner, or someone who

bases their decisions on similar situations they have encoun-
tered. As the beginner starts to organize situation features into
a hierarchy of important elements and base their decisions on
the presence or absence of particular features, they are labeled
as competent. Someone who is proficient at a skill has an intui-
tive ability to recognize important features of a task and organize
information, but still consciously analyzes the available options.
Finally, an expert on tasks that are routine for them will make
automatic, intuitive decisions. Another proficiency scale cited in
(Chi, 2006) is novice, initiate, apprentice, journeyman, expert, and
master.

3.1.2. Some important expert–novice differences

There are a number of salient expert–novice differences in
physics. One of the earliest studies of expertise in physics (Chi et
al., 1981) was a problem categorization study where physics nov-
ices (college students who had completed a mechanics course)
and experts (physics graduate students) were given 24 mechanics
problems written on individual 3×5 cards and asked to sort the
problems into piles according to similarity of solution (without
solving them first). The piles made by novices and experts had
distinctly different features: Novices’ piles shared surface simi-
larities based on objects/configurations (e.g., blocks on inclined
plane problems were in one pile), and/or physics terms (e.g., fric-
tion problems were in a pile); experts’ piles shared the major
physics principle that would be used to solve the problems (e.g.,
Newton’s Second Law, conservation of energy). The categoriza-
tion criteria used by experts and novices were not solely along
the surface features or principles but fell largely along those
dimensions. That same study also asked novices and experts to

11998_Master4.indd 40 2/11/2020 [Link] PM

 How Learning Looks for a Novice 41

discuss the “basic approach” that would be used to solve a set

of problems. The surface attributes and statements in the prob-
lems cued tacit knowledge in experts that allowed them to access
appropriate principles and procedures that could be applied to
solve the problems; in contrast, novices used surface attributes
to look for equations that seemed appropriate (e.g., matching
quantities given in the problems to quantities in equations).
This equation-based approach, often called “means–ends anal-
ysis,” attempts to reduce the “distance” between the givens in
the problem and the goal state. It is described by Larkin et al.
(1980) as follows: “Means–ends analysis begins with the desired
quantity and looks for equations including that quantity. Then
it works backward, marking as desired any unbound quan-
tity needed to solve such an equation.” (See also articles by Van
Heuvelen and Maloney (1999) and by Sweller (2011).) You may
have also heard this approach referred to as “plug-and-chug”
problem solving (Walsh et al., 2007) or a search-based strategy
(Gick, 1986). Students might also use a “memory-based” approach
by mimicking the solution to a similar problem they have seen.
A major shortcoming of this method of solving problems com-
monly employed by novices in physics is that it requires substan-
tial mental resources on the part of the solver (termed “memory
load”), hence leaving few mental resources for conceptual
learning.

3.1.3. Differences in what novices and experts notice when

“doing” physics
These findings indicate that what is noticed in physics problems
and situations is highly dependent on expertise. In particular,
experts notice important attributes in a physics situation and
use tacit knowledge to interpret the situation; lacking extensive

11998_Master4.indd 41 2/11/2020 [Link] PM

 42 The Science of Learning Physics

knowledge, important features of situations go unnoticed by

novices. For example, a study using eye-tracking (technology
that tracks eye movements while an individual performs a task)
revealed marked differences between those who answered physics
questions correctly and those who didn’t (Madsen et al., 2012).
All problems used had a diagram, and a conceptual question was
asked while participants’ eyes were tracked as they looked at the
diagrams and answered the question. Those who answered the
questions correctly spent a larger proportion of time looking at
thematically-relevant portions of the diagrams, whereas those
who answered incorrectly spent considerably more time looking
at diagram features that contained salient surface-feature attri-
butes but that were not thematically relevant for answering the
question. For example, one problem had roller coaster cars trav-
eling along different frictionless paths and ending at the same
height from which they started; people who correctly answered
a question about the speed at the end focused their atten-
tion on just the initial and final heights of the cars (which were
“thematically relevant” for an energy approach) whereas people
who answered incorrectly focused more visual attention on the
paths that the cars traveled.
Another recent study using a phenomenon called “change
blindness” from the cognitive psychology subfield of visual cog-
nition provides additional strong evidence for how novices and
experts direct their attention. Change blindness explores what
people notice during a change in a situation—for example, a par-
ticipant might be shown a picture on a screen, then distracted
from looking at it for a moment, during which something is pur-
posely changed in the picture, and then one ascertains if the
participant noticed the change. Another paradigm uses per-
son-change events; for example, an unsuspecting participant

11998_Master4.indd 42 2/11/2020 [Link] PM

 How Learning Looks for a Novice 43

walks up to a counter and interacts briefly with a person behind

the counter, then the person behind the counter finds an excuse
to duck below the counter (e.g., supposedly looking for a paper
form the participant needs), and a different person pops up and
continues the interaction. In about half the instances, the unsus-
pecting participant does not notice that the person behind
the counter changed (if you want to be amused, search for
“change blindness person change events” on YouTube and wit-
ness such experiments). The physics study (Feil & Mestre, 2010)
brought unsuspecting experts (physics graduate students) and
novices (students who had finished a mechanics course) to an
office where an experimenter showed a participant a computer
screen with a physics situation and asked the participant to for-
mulate an explanation for the behavior of an object in the situ-
ation. After the participant was ready to offer their explanation,
the participant was purposely distracted for a moment and the
experimenter clicked on a hidden wireless mouse that changed
the situation on the screen slightly. Two different situations were
used, one where the change made did not alter the underlying
physics needed to explain the behavior of the situation, and one
where the change did impact the underlying physics. Figure 3.1
contains the original and changed diagrams for the two situa-
tions used.
Somewhat surprisingly, in Situation 1, none of the 31 experts
or 48 novices who participated in the study noticed the ramp-
shape change when they looked back at the screen and started
to explain how they would go about finding the maximum com-
pression of the spring. Note that the ramp-shape change did
not change the underlying physics needed for the explanation;
the explanation that experts had formulated, based on conser-
vation of energy, could be used to explain how the maximum

11998_Master4.indd 43 2/11/2020 [Link] PM

 44 The Science of Learning Physics

Situation 1: Top diagram was the original and bottom diagram was the changed diagram.

Situation 2: Top diagram was the original and bottom diagram was the changed diagram.

Fig. 3.1. Diagrams used in change blindness study.

compression of the spring could be found in both ramp situa-

tions, and hence when they internalized the situation in terms
of the underlying physics, the physics principle needed for
explaining the situation remained the same for both cases, which
did not lead to any conceptual inconsistency. This was not the

11998_Master4.indd 44 2/11/2020 [Link] PM

 How Learning Looks for a Novice 45

case for Situation 2, where adding a string that tied the block on
the table to the wall changed the behavior of the system dras-
tically. In this situation, 77% of the experts noticed the change,
with the vast majority of them showing surprise and stating that
they had “missed” the string attaching mass M to the wall when
they had initially viewed the diagram and formulated an expla-
nation (perhaps, it is also surprising that only 77% of the experts
noticed). In contrast, only 27% of the novices noticed the added
string. Thus experts noticed that the change made to the situa-
tion changed the physics at about three times the rate as nov-
ices. In fact, the behavior and explanations offered by novices
who noticed the string addition were indistinguishable from the
experts that noticed it. In summary, novices and experts gener-
ally notice and attend to different things in physics situations.

3.1.4. Problem solving differences between novices and

experts in physics
Another difference between experts and novices is how they
approach problem solving. A novice behavior that all physics
instructors have witnessed is that when solving problems, novices
hunt for equations that contain the variables/quantities given
in the problem. In contrast, experts begin by conducting basic-
level descriptions and qualitative analyses of the problem based
on the major concepts and principles needed for constructing a
solution (Reif & Heller, 1982). Once identified, the expert knows
what procedure and equation(s) are typically used to apply the
concept. These observations were also identified in (Chi et al.,
1981) a study cited earlier. In fact, one of us (JM) who often taught
discussion sections at the University of Illinois, during which
students would solve difficult, “capstone” problems in collabora-
tive groups, would typically ask the foursome in a group what

11998_Master4.indd 45 2/11/2020 [Link] PM

 46 The Science of Learning Physics

“big idea” they had applied to solve a problem they had just fin-
ished. Students would have the look of a “deer in the headlights”
for a few seconds, and then start reporting what equation(s) they
had used in solving the problem. It was only near the middle to
end of the course that students started catching on to what was
meant by the big idea being applied to solve problems.

3.1.5. Can instruction be designed to promote expert-like

behavior in physics?
[Link]. Short-term interventions
Expert–novice research in physics also indicates that some nov-
ices (those who learn physics easily) can behave like experts,
findings that are evident in the change blindness study described
above. This leads to the question: Can expert-like behavior in
physics be taught to novices so that they begin to exhibit such
behavior. Some studies suggest a qualified “yes.”
An early study (Dufresne et al., 1992) attempted to answer
the question of whether or not constraining novices, after they
had finished an introductory course and received a respectable
grade (B or better), to perform expert-like analyses of problems
prior to solving them resulted in manifestations of expert-like
behavior. As educational experiments go, this one was compli-
cated since it required that the participants practice an expert-
like problem solving approach long enough for it to “take hold.”
The experiment took place over 9 one-hour sessions spread over
about 3 weeks, and participants were students who had finished
an introductory mechanics course with a grade of B or better.
The expert-like approach used to train students draws from the
research reviewed earlier in this chapter, namely that experts
begin by identifying major principles/concepts needed to solve

11998_Master4.indd 46 2/11/2020 [Link] PM

 How Learning Looks for a Novice 47

problems, and then move toward the procedure(s) and equa-

tion(s) needed for solution. Thus a rudimentary menu-driven,
computer-based “hierarchical analysis tool” or HAT was con-
structed that would allow novices to analyze problems by first
identifying one of the four major principles covered in the intro-
ductory mechanics course needed for solving the problem (in the
first menu), and then to continue refining the analysis through
subsequent menus until at the end of the analysis one or more
equations were provided that could be used to solve the problem.

A small block of mass M slides along a track having both curved and horizontal sections as
shown. If the block is released from rest at height H, what is its speed when it is on the
horizontal section of the track? The track is frictionless.

Which principle applies to this part of the problem solution?

1. Newton’s Second Law or Kinematics
2. Angular Momentum
3. Linear Momentum
4. Work and Energy
Please enter your selection: [4]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

Describe the system in terms of its mechanical energy.

1. Conservative system (conservation of energy)
2. Non-conservative system
Please enter your selection: [1]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

(Continued)

11998_Master4.indd 47 2/11/2020 [Link] PM

 48 The Science of Learning Physics

Describe the changes in mechanical energy. Consider only the energy of one body at a
time at some initial and final state.
1. Change in kinetic energy
2. Change in potential energy
3. Change in potential and kinetic energy
Please enter your selection: [3]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

Describe the changes in kinetic energy.

1. Change in translational kinetic energy
2. Change in rotational kinetic energy
3. Change in translational and rotational kinetic energies
Please enter your selection: [1]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

Describe the boundary conditions.

1. No initial translational kinetic energy
2. No final translational kinetic energy
3. Initial and final translational kinetic energies
Please enter your selection: [1]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

Describe the changes in potential energy.

1. Changes in gravitational potential energy
2. Changes in spring potential energy
3. Changes in gravitational and spring potential energy
Please enter your selection: [1]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

Describe the boundary conditions.

1. No initial gravitational potential energy
2. No final gravitational potential energy
3. Initial and final gravitational potential energy
Please enter your selection: [2]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

(Continued)

11998_Master4.indd 48 2/11/2020 [Link] PM

 How Learning Looks for a Novice 49

Is there another body in the system which has not been examined?
1. Yes
2. No
Please enter your selection: [2]
(B)ackup (M)ain menu (G)lossary (L)ist selections (Q)uit

The Work-Kinetic Energy Theorem states that the work done on the system by all non-
conservative forces is equal to the change in the mechanical energy of the system:

Wnc= −E f Ei

According to your selections,

Wnc = 0 (Conservative system: Mechanical energy is conserved)

2
= =
E f (1/2)Mv and Ei Mgy

Please press any key to continue

***Work and Energy***

1. Problem Solved
2. Return to Main Menu to continue solution
3. Review previous solution screens
Please enter your selection: [1]

Fig. 3.2. Sample problem and HAT menus needed to analyze problem.

Figure 3.2 shows a problem and the sequence of menus neces-

sary to appropriately analyze the problem.
It is important to note that the menu-driven computer tool
was not “smart” nor did it provide feedback on the appropriate-
ness of the selections made to solve a problem. In fact, the tool
(hence its “tool” designation) did not know what problem the stu-
dent happened to be analyzing. It simply provided the oppor-
tunity to perform a conceptual analysis, which if correct, would
reward the student with an equation that could be used to solve
the problem. A user of the tool could (and often did) make incor-
rect menu selections during the analysis, in which case the tool

11998_Master4.indd 49 2/11/2020 [Link] PM

 50 The Science of Learning Physics

would move to the next logical menu based on the previous

menu selections. Hence, if the student made wrong choices, s/he
would take an erroneous path that would not result in a correct
equation at the end. The hope was that the student would, at
some point during an incorrect analysis, recognize that the anal-
ysis was not going well, and back up to a previous menu, or start
over with a new analysis (hence the options at the bottom of the
menus that allowed students to back up or begin afresh, or to
have a term defined in a glossary or to list all previous selections
during the analysis).
In order to evaluate whether or not the tool made a differ-
ence, a control group participated in order to compare the focal
treatment, where students analyzed/solved problems with the
expert-like computer tool, with students who solved problems
on their own without any aid. Thus the volunteers were divided
into two groups, one focal group (the group that used the tool)
and one control group. All students in the two groups solved the
same 25 problems over five, one-hour treatment sessions (the
other four sessions were used for testing).
To measure the impact of the treatments, two assessments
were used, the first was a problem categorization task, the second
a problem solving task. As reviewed earlier, novices tend to cate-
gorize problems according to surface attributes (e.g., the objects
in the problem) whereas experts cue on the problems’ deep struc-
ture (e.g., principles needed for solution) to categorize them.
The expert-like tool asked in the first menu for the major prin-
ciple needed to solve the problem, thus one would hypothe-
size that students using the tool for an extended period of time
would be more likely to categorize problems according to prin-
ciples compared to the other control treatment. Whether or
not the problem solving measure would result in differential

11998_Master4.indd 50 2/11/2020 [Link] PM

 How Learning Looks for a Novice 51

performance was an open question since all students prac-

ticed solving the same total number of problems in the two
treatments.
Results of a study with the HAT tool showed that the treat-
ment group attempted to categorize problems according to prin-
ciples more often than the control group, and performed better
on the problem solving assessment at the end of the study; both
results were statistically significant. However, the accuracy of the
students using the expert-like computer tool in terms of catego-
rizing problems was not better than the control group. Hence,
using the tool for 5 hours shifted students’ categorization cri-
teria towards using principles significantly more often than the
control group but both groups were equally accurate in catego-
rizing problems according to principles. This is not surprising
since the ability to accurately categorize problems according to
principles is a trait of expertise and, as reviewed above, it takes
a long time to develop expertise—in other words, 5 hours is not
enough to turn novices into expert categorizers, but 5 hours of
practicing an expert-like analysis of problems did have an impact
on novices’ tendency to use principles in categorizing problems.
A more recent study (Docktor et al., 2012) demonstrated that
an even shorter intervention, lasting less than one hour, can
impact novices’ categorization criteria toward one that is prin-
ciple-based. In that study, two groups of students enrolled in
an introductory algebra-based mechanics course volunteered
for the study two weeks prior to the end of the course (so that
they had covered most of the material in the course). Students
were divided into two groups consisting of a control and a focal
treatment. In both treatments, students were presented with
pairs of problems and asked “Would these problems be solved
similarly?” and students would use a mouse to click “yes” or

11998_Master4.indd 51 2/11/2020 [Link] PM

 52 The Science of Learning Physics

“no”; immediately upon answering yes or no, students were

either given feedback on the appropriateness of their answer
(‘‘(In)Correct. These problems would (NOT) be solved similarly.” ),
or asked to type into a text-box (free-form) the reasoning behind
their answer; if prompted for reasoning, the feedback screen
appeared immediately upon completing the typed explana-
tion. Prompting for the reasoning that students used was done
eight times during the one-hour experiment equally spaced
throughout the experiment. Students in the control group were
only told whether or not they were correct after answering yes/
no to each problem pair. In contrast, the focal treatment, in addi-
tion, provided more elaborate feedback to the students, which
included statements about the principle needed to solve the
two problems. The control treatment will be referred to as the
“sparse-feedback condition” while the focal treatment will be
referred to as the “elaborate feedback condition.” Figure 3.3
shows a sample item and the sequencing of screens as the exper-
iment progressed. A total of 32 problem pairs were used that cov-
ered the major topics in the course.
The results in this study parallel the categorization results
described in the previous study where students used the expert-
like tool to analyze problems. Figure 3.4 shows the propor-
tion of reasoning statements judged to be principle-based by
two physics-expert-graders as a function of time as the exper-
iment progressed (recall there were eight different occasions
during which participants were asked to offer their reasoning
equally spaced throughout the experiment). During the first
half-hour of the experiment, reasoning statements based on
physics principles hovered around 20–35%, with no signifi-
cant differences between the sparse-feedback condition and the
elaborate-feedback condition. However, there was a dramatic

11998_Master4.indd 52 2/11/2020 [Link] PM

 How Learning Looks for a Novice 53

Problem 1 Problem 2

Would these two problems be solved similarly?

Yes No

Problem 1 Problem 2

Please explain the reason for your choice by typing in the box below.

Press the F1 key to continue.

(Continued)

11998_Master4.indd 53 2/11/2020 [Link] PM

 54 The Science of Learning Physics

Problem 1 Problem 2

Correct! These problems would NOT be solved similarly.

Problem 1 includes force and acceleration, so it would be
solved using Newton’s Second Law. Problem 2 includes mass
and change in speed with no external force on the system, so
it would be solved using Conservation of Momentum.

Press the Space Bar to continue

Fig. 3.3. Sample item used in categorization study. Reproduced from Physical Review:
Physics Education Research, 2012, 8(2), 020102 with permission from the American
Physical Society.

1 SPARSE FEEDBACK ELABORATE FEEDBACK

0.9
PROPORTION OF STATEMENTS

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1-2 3-4 5-6 7-8
REASONING STATEMENT NUMBERS

Fig. 3.4. Reasoning statements judged to be principle based as the experiment pro-
gressed. Reproduced from Physical Review: Physics Education Research, 2012, 8(2),
020102 with permission from the American Physical Society.

11998_Master4.indd 54 2/11/2020 [Link] PM

 How Learning Looks for a Novice 55

shift during the second half-hour of the experiment, with stu-

dents in the sparse-feedback condition lowering their use of
principle-based reasoning to slightly more than 10% of the
time, whereas the students in the elaborate-feedback condition
employed principle-based reasoning more than 50% of the time.
Thus it appears that even after one hour students can be influ-
enced to use principle-based reasoning with feedback that men-
tions the principles used to solve problems. However, like in
the previous study, there is also bad news here: Even though
the students in the elaborate-feedback condition used princi-
ple-based reasoning significantly more often than students in the
sparse-feedback group, the fraction of correct use of principles
when using principle-based reasoning was equally poor for both
groups; both groups were correct when using principle-based
reasoning about 50% of the time. This and the previous study
suggest that it is not difficult to steer students away from cate-
gorizing problems based on variable names or problems’ surface
attributes and more toward principle-based criteria—a good first
step toward developing superior habits, but that it takes a much
longer, sustained effort to help students use principle-based cat-
egorization correctly. Again, the latter finding is to be expected
since if we were able to help students categorize problems accu-
rately according to principles, which is a trait of expert physicists,
with relative short interventions, then we would be able to turn
novices into experts in short order, which is contrary to research
into the development of expertise across all domains studied to
date (e.g., chess, science, sports, …).

[Link]. Longer-term interventions

Thus far the studies reviewed above indicate that it is relatively
easy to influence novice students’ behavior towards a more

11998_Master4.indd 55 2/11/2020 [Link] PM

 56 The Science of Learning Physics

productive way of viewing physics as being comprised of a few

major principles/concepts that can be applied to categorize and
solve a wide range of problems, but their execution fell short,
especially given the brief duration of the interventions. What
if one could structure a study to promote a problem solving
approach throughout an entire semester that allowed students
to practice and receive feedback on a regimen where problems
to be solved were first analyzed conceptually and then analyzed
procedurally—might one see better execution of expert-like behav-
iors then? There are three examples of such studies, with all of
them resulting in positive findings.
One study implemented “strategy writing” in an introductory
calculus-based mechanics course (Leonard et al., 1996). Students
were informed early in the semester that a strategy was a written
paragraph with no equations that discussed the problem’s solu-
tion at a high level and contained three components: (1) identifi-
cation of the major principle(s) being applied, (2) a justification
for why the principle(s) was applied to the particular problem
under consideration, and (3) a procedure for applying the princi-
ple(s) to solve the problem (i.e., the what, why and how of solving
a problem). Strategies were modeled for students in lecture—
every time a sample problem was solved in lecture, the instructor
began with a strategy, and then executed it to generate an answer.
Strategies were also modeled in the posted written solutions to the
weekly homework assignments (homework was neither collected
nor graded), as well as by the instructors teaching the discussion
sections whenever problems were solved. In order for students
to take strategy writing seriously, all exams (75% multiple-choice
and a “work-out” problem that was hand graded worth 25%) had
one problem requiring a strategy and a solution, with 13 points
assigned for the strategy and 12 for the executed solution.

11998_Master4.indd 56 2/11/2020 [Link] PM

 How Learning Looks for a Novice 57

Students, being very grade-conscious and wanting a lot of

specificity, asked for more details about what a “good” strategy
should look like, but the course instructors simply stated that
there were many ways to write a strategy, and that as long as
they discussed a coherent plan for solving a problem that con-
tained the three elements it was fine; students were also told
that a litmus test for a good strategy would be the following: If
a student is stuck and does not know how to solve a problem,
then handing them a good strategy would allow them to follow
it to solve the problem. Figure 3.5 contains a strategy for a dif-
ficult two-principle problem from the posted solution to one of
the homework sets. As one would expect, writing strategies was
a struggle for students near the beginning of the course, but for
most students, it steadily improved as the semester progressed.
Figure 3.6 shows a problem from the third midterm exam
and sample student strategies, some good and some not-so-good.
Note how the quality strategies contained the three components
listed earlier, and displayed a coherent understanding of how to
structure a problem’s solution. On the other hand, it is difficult
to “fake” a strategy—as can be seen, the below-average strate-
gies were largely a “laundry list” of concepts and physics terms
that lacked coherence and definitely did not pass the litmus test
given earlier. By the end of the semester, about a third of the stu-
dents could generate good strategies, about a third (those strug-
gling in the course) were not able to put together a coherent
strategy, and the middle third was somewhere in-between.
In addition to the work-out problem that was hand-graded in
the exams that required a strategy, two additional measures were
used to evaluate the semester-long impact of writing strategies.
One was a problem categorization task consisting of 5 multiple-
choice problems given as part of the final exam.

11998_Master4.indd 57 2/11/2020 [Link] PM

 58 The Science of Learning Physics

Problem:
A stick of mass M and length L is hanging from a ceiling
as shown to the right. A piece of putty of mass m is L/2
traveling horizontally when it strikes and adheres to the v
stick at its midpoint. What must have been the speed
of the putty, v, if the stick (with putty attached) swings m
and comes momentarily to rest in a horizontal position L/2
before starting to swing down again?
M
M=2kg

Strategy:
This problem consists of two parts. First, the mass and stick undergo a totally inelastic
collision. Angular momentum is conserved about the pivot during the collision since
there is no net external torque (about the pivot). Even though there are two external
forces present during the collision, these do not provide a net external torque: The
force at the pivot has no moment arm (i.e., r = 0) and so it does not provide a torque
about the pivot. The gravitational force acting on the center of mass of the stick points
vertically down and initially is parallel to r (i.e., angle between r and Fg is 0). Assuming
the collision is instantaneous, the gravitational force delivers no angular impulse to the
system during the collision. So, assuming there is no friction between the pivot and
the rod, there are no angular impulses delivered to the putty-stick system. Thus, for
the collision portion of the problem, equate the initial and the final angular momenta
of the putty-stick system. Note that only the putty contributes to the initial angular
momentum since the stick is at rest prior to the collision.

The second part of the problem consists of the putty-stick system rotating to a horizontal
position before momentarily stopping. During this portion of the problem, mechanical
energy is conserved. Only two forces act on the system during this “swinging” portion:
The gravitational force and the force at the pivot. The force at the pivot does no work
since the pivot point does not move and we have assumed that there is no friction
between the pivot and the stick. Only the gravitational force (a conservative force) does
work, resulting in mechanical energy being conserved. Thus, for the second portion
of the problem, equate the initial and the final mechanical energies of the system.
Assuming the collision was instantaneous, the initial orientation of the putty-stick
system is vertical. The center of mass of the system is at the midpoint of the stick. (We
need this to evaluate the potential energy of the system.) Applying the two concepts of
angular momentum and energy conservation will result in two equations that will allow
you to solve for v in terms of the “givens” in the problem.

Fig. 3.5. Sample problem and strategy from posted homework solutions.
Reproduced from American Journal of Physics, 64, 1495–1503 with permission.

11998_Master4.indd 58 2/11/2020 [Link] PM

 How Learning Looks for a Novice 59

Problem: L/2
A disk of mass, M = 2v kg, and radius, R = 0.4 m, has string wound
around it and is free to rotate about an axle through its center. A
block of mass, m = 1m kg, is attached to the end of the string and
the system is released from rest with no slack in the string. What is
L/2 R=0.4m
the speed of the block after it has fallen a distance, d = 0.5 m. Don’t
M
forget to provide both a strategy and a solution. M=2kg
Above Average Strategies:
A1: Use the Conservation of Energy since the only non-conservative force in the
system is the tension in the rope attached to the mass M and wound around
the disk, (assuming there is no friction between the axle and the disk, and the
mass M and the air), and the work done by the tension to the disk and the mass cancel m=1kg
each other out. First set up a coord. syst. so the potential energy of the system at the
start can be determined. There will be no kinetic energy at the start since it starts at
rest. Therefore the potential energy is all the initial energy. Now set the Initial energy
d=0.5m
equal to the final energy that is made up of the kinetic energy of the disk plus the mass
m and any potential energy left in the system with respect to the chosen coord. system.
A2: I would use conservation of mechanical energy to solve this problem. The mass m has some potential energy
while it is hanging there. When the block starts to accelerate downwards the potential energy is transformed into
rotational Kinetic energy of the disk and Kinetic energy of the falling mass. Equating the initial and final states; and
using the relationship between v and w the speed of m can be found. Mechanical energy is conserved even with the
non-conservative Tension force because the Tension force is internal to the system. (pulley, mass, rope)

A3: The linear velocity is related to the angular velocity by the factor of the disc’s radius. In this problem, the pivot is
frictionless so all energy is conserved. By equating the initial energy (potential of hanging mass) with the final energy
(kinetic of mass & kinetic of rotating disc), it is then possible, after changing the rotational velocity to its linear form,
to isolate and calculate a value for the final linear velocity.
A4: Apply Newton Second Law to find the net force acting on the block. This is equal to its mass times its accelera-
tion. This net force, crossed with the radius of the disk, provides a net torque. The net torque is the moment of inertia
of the disk times the angular acceleration. Solve both equations for the accelerations. They are related because the
acceleration is the cross product of the radius and the angular acceleration. Substitute what you solved for in the two
equations above, and then solve for Tension (T). Substitute this for T in the equation for acceleration and then solve
using math.

Below Average Strategies:

B1: Using d = 0.5 m, we could find θ and from there we could figure out the time using θ, a, and ωo. After we found
this, we could use that, along with the other given information to determine the angular speed. Once we know this,
we can relate the angular information to the block.

B2: Not only do you have to consider the mass of the block on the string but also the force of gravity on the block.
Rotational kinematics must be used with the radius, mass and gravity.

B3: In trying to find the speed of the block I would try to find angular momentum kinetic energy, use gravity, I would
also use rotational kinematics and Moment of Inertia around the center of mass for the disk.

B4: There will be a torque about the center of mass due to the weight of the block, m. The force pulling downward is
mg. The moment of inertia about the axle is 1/2 MR2. The moment of inertia multiplied by the angular acceleration.
By plugging these values into a kinematic expression, the angular speed can be calculated. Then, the angular speed
times the radius gives you the velocity of the block.

Fig. 3.6. Sample student strategies from mid-term exam.

(The strategies above are transcribed verbatim from students’ exam papers).
Reproduced from American Journal of Physics, 64, 1495–1503 with permission.

11998_Master4.indd 59 2/11/2020 [Link] PM

 60 The Science of Learning Physics

Below are five choices labeled a–e containing one or more major concepts studied in the
course. Questions 1–5 consist of five problems that you do not need to solve. Your job is
to decide which major concept(s) needs to be applied to solve each problem in the most
efficient manner and make the appropriate selection. Use the same set of five multiple
choices for all five questions, and you may use each choice, A–E, once, more than once,
or not at all.

Multiple Choices:
(A) Newton's Second Law
(B) Work-Energy Theorem or Conservation of Mechanical Energy
(C) Linear Momentum or Conservation of Linear Momentum
(D) Conservation of Linear Momentum followed by Conservation of Mechanical Energy
(E) Angular Momentum or Conservation of Angular Momentum

Problems:
1. A 2 kg uniform metal bar of length 1 m resting on a frictionless horizontal surface is
free to rotate about a pivot at one end. A 5 g bullet traveling perpendicular to the stick
hits and embeds itself into the stick 50 cm from the pivot. If the initial speed of the bullet
is 300 m/s, what is the angular speed of the stick immediately following the collision.
2. A mass M is connected to a string of length L to form a simple pendulum, with the
other end of the string attached to the ceiling. The pendulum is released from rest at
height L/2 from the lowest point of the pendulum's swing. What is the speed of the mass
at the lowest point in the swing? Consider the string to be massless.
3. A block of mass m is moving at speed v along a horizontal, frictionless surface. The
block undergoes a perfectly inelastic collision with a second block of mass M. The two
blocks proceed up a frictionless inclined plane and momentarily come to rest part way
up the plane. What maximum distance along the inclined plane do the two blocks travel?
4. A 1 kg stick of length 2 m is placed on a frictionless surface and is free to rotate about
a vertical pivot through one end. A 50 g lump of putty is attached 80 cm from the pivot.
What is the magnitude of the force between the stick and the putty when the angular
velocity of the system is 3 rad/s?
5. A mass M is connected to a string of length L to form a simple pendulum, with the
other end of the string attached to the ceiling. If the mass has speed v at the bottom
of the swing, what is the tension in the rope at that point? Consider the string to be
massless.

Fig. 3.7. Categorization task.

(The five multiple choice questions were administered as part of the final exam).
Reproduced from American Journal of Physics, 64, 1495–1503 with permission.

11998_Master4.indd 60 2/11/2020 [Link] PM

 How Learning Looks for a Novice 61

Each question asked students to read a problem and then to

select from among five multiple choices the major principle(s)
that should be applied to solve the problem in the most efficient
way possible (students were not required to solve those five prob-
lems). The five questions are listed in Fig. 3.7. Because the first
component of a strategy was identifying the principle(s) needed
to solve a problem, it was expected that the strategy writing class
would do better on these five problems compared to students in
the same course, a semester later, that was taught traditionally
without strategy writing. Averaging over the five questions, the
traditional class selected the correct principle 48% of the time
whereas the strategy writing class did so 70% of the time (more
details are provided in (Leonard et al., 1996).
The last measure used probed students’ memories about
the most important physics ideas covered in the course months
after taking it. Since there was likely to be substantial forget-
ting as time passed, the researchers decided only to probe stu-
dents who had gotten a very high grade, hence volunteers were
sought who had earned grades in the A range. Because the two
courses were taught in different semesters, students from the
traditional class had finished the course six months prior, while
the strategy writing students had finished the course eleven
months prior. Students were asked to name/write down the most
important physics ideas used to solve problems in mechanics.
Students generally identified seven principles, namely Newton’s
three laws, conservation of energy, momentum, and angular
momentum, and the work-energy theorem. Both groups identi-
fied Newton’s three laws with about the same frequency. However,
the strategy writing students mentioned the other four principles
beyond Newton’s Laws at a much higher rate than the tradition-
ally taught students, with 92% mentioning one principle beyond

11998_Master4.indd 61 2/11/2020 [Link] PM

 62 The Science of Learning Physics

Newton’s Laws compared to 36% for the traditional class, and

69% mentioning two principles beyond Newton’s Laws compared
to 9% for the traditional class. Note that this does not mean that
one group was better, or less able than the other group in solving
problems since students were not tested for problem solving
skills months after the course was over, nor that the strategy
students had a better conceptual understanding of the mate-
rial since that was also not tested. However, focusing on a reg-
imen that consistently forces students to consider principles in
problem solving does impact their ability to recall and name the
major principles covered in the course months after it was over.
A second study by Van Heuvelen (1991a, 1991b) used an
approach to teach mechanics called Overview, Case Study
Physics (OCS). The approach was used in a large lecture course
and required two passes through the material. In the first pass,
which lasted about one-third of the semester, students were
exposed to the entire course content qualitatively, actively dis-
cussing concepts, representing them in diagram form and con-
fronting erroneous preconceptions. Students also worked on
qualitative problems during lecture in groups, and at the end of
the first third of the course, they were tested with an exam con-
sisting of qualitative questions about the entire course content.
In the second two-thirds of the course, the same material is revis-
ited but now in mathematical form where they solve quantitative
problems applying the concepts they learned previously. Students
were also provided with a flow chart with a hierarchical represen-
tation of the concepts in the course (e.g., starting at the top with
dynamics and conserved motion) and naming the major concepts
both in words and equation form. Students were assigned com-
plicated, multi-concept problems to solve in groups where they
were required to integrate and apply the concepts learned in the

11998_Master4.indd 62 2/11/2020 [Link] PM

 How Learning Looks for a Novice 63

course and generate a numerical solution. In assessments given

at the end of the semester, the OCS students (in both algebra-
based and calculus-based mechanics) outperformed tradition-
ally taught classes in both conceptual questions and in problem
solving. As in the strategy writing study, the OCS approach also
resulted in better retention. One month after finishing a tradi-
tional mechanics course and eight months after finishing an
OCS course, students were given a qualitative, conceptual diag-
nostic test where the OCS students scored 20% higher than the
traditional students (despite the much longer period of time over
which OCS students could have “forgotten” the material).
The third investigation consisted of a series of classroom-
based studies conducted over a number of years by Gautreau
and Novemsky (1997). Those studies attempted to adopt/
replicate the OCS approach by a faculty member who had not
been involved in the development of the OCS method. OCS was
used in teaching various courses to both mainstream students
as well as minority students enrolled in an Educational
Opportunity Program (EOP) with end-of-course exam perfor-
mance compared to students taking the same course taught
traditionally. In all cases that compared the performance of OCS-
students with traditionally-taught students, the OCS students sig-
nificantly outperformed traditionally-taught students in both
conceptual and problem solving measures. In fact, EOP students
who had experienced the OCS approach also outperformed the
traditionally-taught mainstream students.

3.2. What are the implications of expert–novice

research for instruction?
We offer the following implications of the research reviewed
above for instruction:

11998_Master4.indd 63 2/11/2020 [Link] PM

 64 The Science of Learning Physics

● Developing expertise in any endeavor, including a

complex domain such as physics, takes considerable
time and effort. Because the development of expertise
takes a long time, we cannot expect beginning students to
possess a well-organized memory store of physics concepts
and problem solving procedures (although perhaps the
most advanced beginning students will be on their way to
developing them). Lacking an expert’s perspective, novices
approach problem solving in the best way they can, which
largely consists of matching the variables in a problem to
equations containing those variables. Students would not
try this method if it did not have some merit—a lot of times,
finding relevant equations and manipulating them will
work in finding a solution to a problem. Students also tend
to work by analogy, matching problems they are solving to
similar-looking problems they have solved before or worked
out problems in textbooks. However, matching problems
on surface attributes can lead to erroneous solutions since
different concepts are often used to solve similar-looking
problems.
● As instructors, it is important to be aware that students
notice different things in physics situations than we do.
Experts possess considerable tacit knowledge, much of
which is not exposed to students during our teaching.
Experts’ knowledge allows them to “see” subtle features in
physics situations that help develop a deep understanding
of the situation. Novices are not so privileged, and often
miss things in physics situations that we as experts
consider obvious (recall the expert’s blind spot discussed in
Chapter 1).
● Research studies show promise for helping beginning
students adopt traits of experts in problem solving and

11998_Master4.indd 64 2/11/2020 [Link] PM

 How Learning Looks for a Novice 65

conceptual development. Although progression toward

expertise is slow, it can be made more efficient. Since
beginning students are adept at algebraic manipulations of
equations, that is their go-to method for solving problems.
Focusing on conceptual aspects of problem solving, such
as beginning to solve a problem by considering what
concepts/principles apply, is not something beginning
physics students typically do. However, research indicates
that interventions can be developed to help students
elevate the role of concepts in problem solving and consider
what concepts/principles can be applied to solve problems.
The only caveat is that getting novices to consider concepts
is not equivalent to having them consider appropriate
concepts for a given problem—that skill takes much longer
to develop.

3.3. Examples of teaching interventions based on

learning research
The research reviewed in this chapter suggests that introduc-
tory students could benefit from three types of instructional
strategies that are typically absent in traditional instruction:
(1) Elevating the role of concepts/principles in problem solving,
(2) Organizing physics knowledge more efficiently in memory,
and (3) Making the tacit knowledge that experts use in problem
solving explicit during instruction. Below we provide some
instructional suggestions for implementing these strategies
during the normal course of instruction; with the suggestions
below, we are not advocating wholesale change in teaching
practices but rather the integration of some new practices and
activities in homework and in-class practices. If you decide to
implement some of the instructional suggestions below, it is

11998_Master4.indd 65 2/11/2020 [Link] PM

 66 The Science of Learning Physics

important that students take the task seriously, and unfortu-

nately the only way to do so is to count it as part of the students’
course grade. Note how the suggestions below are largely qualita-
tive and not quantitative.
Instructional strategies to elevate the role of concepts/
principles in problem solving:
● Implement problem categorization. Problem categoriza

­
tion is an activity that offers fruitful opportunities for
blending conceptual knowledge with problem solving.
Categorizing problems according to the major concept/
principle needed for solving them is less time-consuming
than actually solving problems, thus one can go through
many more practice items in the same time than it would
take to solve just a few problems. Categorization also
offers the opportunity to compare/contrast the usefulness
of seeking a problem’s underlying conceptual structure
for strategizing a solution, as opposed to simply focusing
on surface attributes such as objects or variables in the
problem. Unfortunately, this task needs to wait until several
major principles are covered in a course (likely mid-way
through a course and beyond) because it would be rather
pointless to give items that all require the same major
principle for solution.
Problem categorization tasks can take various forms, each with
its own nuance (Hardiman et al., 1989). For example, giving pairs
of problems and asking if they are solved similarly together
with providing a justification allows one to draw attention to
extracting the deep structure from the problems’ surface attri-
butes. The “answer” to a two-problem categorization task, as seen
previously, is binary: The two problems are either solved similarly
(i.e., by applying the same principle) or not. But the problem pair

11998_Master4.indd 66 2/11/2020 [Link] PM

 How Learning Looks for a Novice 67

The following pair of problems match on surface features but not deep structure
Two blocks of mass 10 kg and 8 kg are connected by a light compressed spring of force con-
stant 80 N/m and held at rest. The blocks are released and the 10 kg block is observed to
move at 2 m1sec. Find the velocity of the 8 kg block.

A 15 kg block and a 10 kg block are connected by a light compressed spring of force con-
stant 200 N/m and held at rest. The blocks are released and observed to move at 2 m/s and
3 m/s respectively in opposite directions. Find the distance the spring was compressed from
its equilibrium length.

The following pair of problems match on deep structure but not on surface features
Two blocks of mass 10 kg and 8 kg are connected by a light compressed spring of force con-
stant 80 N/m and held at rest. The blocks are released and the 10 kg block is observed to
move at 2 m1sec. Find the velocity of the 8 kg block.

A bullet of mass 10 g is fired into a target of mass 10 kg. The bullet and target then have a
common velocity of 1 m/s. How fast was the bullet moving just before it entered the target?

Fig. 3.8. Two-problem categorization task.

could also match on surface attributes, or not. If done as a class

activity, the instructor could draw attention to how one extracts
the deep structure from problems (valuable tacit knowledge) and
warn students against their tendency to think that problems that
look alike are solved similarly. Two examples of a two-problem
categorization task are shown in Fig. 3.8, where in the first pair
the problems match on surface features but not on deep struc-
ture (thus they are not solved similarly), and in the second
pair the problems match on deep structure but not on surface
features.
Another form of a problem categorization task is a three-
problem format, as shown in Fig. 3.9 (for many more items see
(Hardiman et al., 1989)). Here one gives a “model problem” and
two “comparison problems” and the student’s job is to deter-
mine which of the two comparison problems is solved sim-
ilarly to the model problem. One can be more methodical with

11998_Master4.indd 67 2/11/2020 [Link] PM

 68 The Science of Learning Physics

Model Problem: Two blocks of mass 10 kg and 8 kg are connected by a light compressed
spring of force constant 80 N/m and held at rest. The blocks are released and the 10 kg block
is observed to move at 2 m/s. Find the velocity of the 8 kg block.

S Comparison Problem: A 15 kg block and a 10 kg block are connected by a light compressed

spring of force constant 200 N/m and held at rest. The blocks are released and observed
o move at 2 m/s and 3 m/s respectively in opposite directions. Find the distance the spring
was compressed from its equilibrium length.

D Comparison Problem: A bullet of mass 10 g is fired into a target of mass 10 kg. The bullet
and target then have a common velocity of 1 m/s. How fast was the bullet moving just before
it entered the target?

SD Comparison Problem: Two blocks each of mass 20 kg are connected by a light, com-
pressed spring of force constant 70 N/m. After the blocks are released, it is observed that one
of the blocks has received an impulse of magnitude 150 N/sec. Find the speed of the other
block.

N Comparison Problem: A 10 kg mass with initialvelocity2 m1sec passes over a rough hori-
zontal surface with a coefficient of kinetic friction 0.1. Find the acceleration of the mass when
it is on the rough surface.

Fig. 3.9. To form a three-problem categorization item, use the Model Problem and
the following two Comparison Problem combinations: S-D, S-SD, N-D, N-SD. The task
for the student is to determine which Comparison Problem is solved most like the
Model Problem.

this format in drawing contrasts between surface attributes and

concepts/principles needed for solution. Clearly one (and only
one) of the two comparison problems would need to match
the model problem on deep structure (the same concept(s)/
principle(s) would be applied to solve both) for there to be a
unique correct answer, but problems in the pair could be of four
types: (1) No match (N), meaning that a comparison problem
does not match the model problem on either surface attri-
butes or deep structure, (2) Match on surface features only (S),
meaning that a comparison problem matches the model problem
only on surface attributes, (3) Match on deep structure only (D),

11998_Master4.indd 68 2/11/2020 [Link] PM

 How Learning Looks for a Novice 69

meaning that a comparison problem matches the model problem

only on the concept(s)/principle(s) needed for solution, and
(4) Match on both surface attributes and deep structure (SD).
Thus, the pair of comparison problems could have one of the
following four forms: S-D, S-SD, N-D and N-SD. For students who
exhibit standard novice behavior, the hardest type of item would
be S-D since they would be tempted to match on surface attri-
butes, and the easiest would be N-SD since matching on either
surface feature or deep structure would result in a correct
answer. This type of task could be assigned for homework, for
an in-class collaborative activity with the instructor providing
coaching, or in discussion section as a group activity that is then
discussed in class-wide format.
A third form would be a multiple choice format as discussed
earlier (see Fig. 3.7), where a problem would be given together
with multiple choices for the principle(s) needed to solve the
problem. The student would attempt to choose the correct prin-
ciple(s) needed for solving the problem. This format lends itself
well for web-based homework formats or for multiple choice
exams.
In summary, problem categorization is an efficient way to
expose students to the role that concepts and principles play in
solving problems, and in addition, offers the instructor opportu-
nities to discuss the procedures for applying the major concepts/
principles, thereby revealing valuable tacit knowledge.
● Implement strategy writing. The classroom-based study
described above (Leonard et al., 1996) where students
wrote strategies for solving problems before solving them
is an excellent task for both, blending all the crucial
aspects of problem solving, and for revealing to students
the tacit knowledge that experts use in solving problems.

11998_Master4.indd 69 2/11/2020 [Link] PM

 70 The Science of Learning Physics

As we have defined it, the three components of a strategy

consist of identifying the major principle(s) needed to
solve the problem, justifying why the principle applies, and
describing a procedure in words for applying the principle.
In short, a strategy is a prose description of how to go
about solving a problem. A couple of caveats are in order.
First, strategy writing is a high level task that students will
find difficult to do (at least initially), so modeling it for
students is important. Also, since it is so unlike the typical
tasks assigned in physics courses, be prepared for students
resisting and offering gratuitous commentary such as
“writing belongs in English class.” Motivating students is
important for them to invest in an activity, so telling them
the purpose of strategy writing is important.
Note what strategy writing accomplishes that not only is cru-
cial for problem solving training, but that also is not typically
done in traditional instruction. First, it is a top-down approach
for solving problems, starting with identifying the princi-
ple(s) that can be applied to solve a problem, as opposed to
starting with equation manipulation. The second component
of a strategy also reveals valuable tacit knowledge that typi-
cally remains hidden during instruction, namely deciding the
conditions under which a particular principle can be applied.
For example, during the course of instruction we might work
out a problem for students and state that we are applying con-
servation of mechanical energy and move directly to the equa-
tion that sets the mechanical energy in the initial state to the
mechanical energy in the final state. Doing so sweeps under the
rug how we as experts decided that conservation of mechanical
energy could be applied to the situation, namely the absence of
work done by non conservative forces. Checking for conditions of
‑
11998_Master4.indd 70 2/11/2020 [Link] PM
 How Learning Looks for a Novice 71

applicability is an expert trait that students would find useful to

learn early on. The third component, describing a procedure for
applying the principle(s) to solve the problem also elevates pro-
cedural knowledge in ways that are generalizable. If all begin-
ning students do is work at the equation-manipulation level, it is
hard for them to pull out the meaning underlying the equations
they use. Describing a procedure in words reveals the structure
of the solution and provides opportunities for generalizability.
For example, the procedure for applying conservation laws in
mechanics is exactly the same whether we do it for mechanical
energy, momentum or angular momentum; the only thing that
changes is whether the equation used has variables E, P or L.
One way to ease students into strategies for solving problems
is to provide them with a strategy (written by the instructor)
and have students execute it to solve the problem. This would
be a considerably easier task than writing actual strategies, and
it allows students to interpret a strategy and follow it as a high-
level recipe for solving a problem. Note that providing students
with strategies to execute allows instructors to assign difficult
problems that might otherwise be very time-consuming for stu-
dents. For example, two-principle problems, such as the problem
shown in Fig. 3.5, are known to be harder for students than
single-principle problems, but with a recipe to follow, students
are much more likely to make rapid progress toward a solution.
Another approach to strategy writing is to provide much
more scaffolding to help students focus on the elements of a
strategy and then as they become proficient the instructor can
gradually fade the supports (Docktor et al., 2015). One example
of how to provide scaffolding is to use the format shown in
Fig. 3.10, where a strategy is broken down into the principle,
justification and plan (all clearly labeled), and then the plan is

11998_Master4.indd 71 2/11/2020 [Link] PM

 72 The Science of Learning Physics

A skateboarder enters a curved ramp moving horizontally with a speed of 6.5 m/s, and
leaves the ramp moving vertically with a speed of 4.1 m/s. The skateboarder and the skate-
board have a combined mass of 55 kg. Find the height of the ramp, assuming no energy
loss to frictional forces.

Principle:
Conservation of energy: the total mechanical energy (sum of kinetic and potential energies) of an
isolated system is the same in the initial and final states.

Justification:
Mechanical energy is conserved if there are no non-conservative forces that do not work on the
system. The normal force exerted on the skateboarder is a non-conservative force, but the work
that the normal force does is 0 because its direction is always perpendicular to the displacement.
The gravitational force is conservative (it is already included in the potential energy term), and
we are ignoring non-conservative frictional forces. Therefore, mechanical energy is conserved.

Plan:
1. Draw a picture and assign symbols for quantities in the problem. Choose a coordinate system.
2. Write an equation for conservation of mechanical energy. Expand the equation to include the
initial and final kinetic and potential energy terms.
3. Solve for the height of the ramp. Substitute values to get an answer.

Two-Column Solution:
Plan Step Equation(s) used in step
1. Draw a picture and assign symbols vf = 4.1 m/s
vi = 6.5 m/s
for quantities in the problem.
Choose a coordinate system.

hf

y
hi = 0

m = 55 kg Mass of the skateboarder and

skateboard combined
vi = 6.5 m/s Initial speed of skateboarder
vf = 4.1 m/s Final speed of skateboarder
hi = 0 m Initial height of skateboarder

(Continued)

11998_Master4.indd 72 2/11/2020 [Link] PM

 How Learning Looks for a Novice 73

2. Write an equation for conservation ∆E = 0 ⇒ E i = E f

of mechanical energy. Expand this
KE i + PE i = KE f + PE f
equation to include the initial and
final kinetic and potential energy 1 1
mv 2 + mghi = mv f 2+ mghf
terms. 2 i 2

1 1
mv 2 + =
0 + 2
mv mghf
2 i 2 f
1 1
3. Solve for the height of the ramp. mghf − mv f 2
= mv 2
2 2 i
Substitute values to get an answer.
1
2
mv f 2 − 12 mv i2
hf =
mg
1 2
v
2 f
− 12 v i2
hf =
g
1 1
2
(4.1 m/s )2 − 2 (6.5 m/s)2
=
(9.8 m/s2 )
= 1.3 m

Fig. 3.10. Scaffolding for problem-solving strategies.

executed using a two-column solution. Using this format, stu-

dents would be provided with a partially-filled “worksheet” (as in
Fig. 3.10) and task students to work collaboratively to complete
the worksheet.
● Implement assessments that promote conceptual under­
standing. Chapter 2 provided an overview of active learning
strategies to promote conceptual understanding. Another
suggestion we offer called “finding errors” consists of a
task whereby a problem is given together with a solution
consisting of a brief conceptual analysis of the problem
followed by a solution. The conceptual analysis and solution
contain a conceptual error that results in an incorrect
solution to the problem. The error is subtle and draws on
typical student misconceptions. Figure 3.11 provides an

11998_Master4.indd 73 2/11/2020 [Link] PM

 74 The Science of Learning Physics

example. The error here, consisting of the common student

misconception (Feil & Mestre, 2010; McDermott et al., 1994;
Mestre, 2002) that the tension in the string is equal to the
weight of the hanging mass, has been observed by many of
us in teaching introductory mechanics.
We need to point out several caveats about this task. Because this
task draws on common misconceptions and/or student errors,

A block of mass 0.25 kg is connected to a cart of mass 0.75 kg

by a string and a frictionless pulley, as shown in the figure
below. There is no friction as the cart is rolling on the hori-
zontal track. What is the acceleration of the cart?

Solution:
We can use Newton’s 2nd Law to find the acceleration of the
cart. The net force on the cart will equal the cart's mass multiplied by its acceleration. The ten-
sion in the string is equal to the weight of the hanging block. The net force on the cart is equal
to the tension in the string because the force of gravity on the cart and the normal force on the
cart cancel each other out. The mass of the cart is given (0.75 kg), so we can find the accelera-
tion of the cart.

Fnet = mcart a

T = mcart a

T
a=
mcart
T = mblock g

mblock g
a=
mcart

(0.25 kg)(9.81 m/s2 )

a=
0.75 kg

a = 3.27 m/s2

Explain what is wrong with this solution and why it is incorrect.

Fig. 3.11. Identifying conceptual errors in solutions.

11998_Master4.indd 74 2/11/2020 [Link] PM

 How Learning Looks for a Novice 75

it is a high-level, difficult task. We used it in a high school class-

room physics study evaluating innovative teaching strategies
(Docktor et al., 2015) as an end-of-course assessment and found
that the performance was at floor-levels (students averaged 6%
in an assessment comprised of items like the one above). Also,
as far as we know this type of exercise has not been tried at the
college level, although we would predict that the performance
would be higher than that at the high school level. In developing
items like that in Fig. 3.11, care should be taken so that the error
in the solution is not outrageous and easily identifiable—the
error should be plausible so that it “blends in” to the solution.
Instructional strategies to help students develop a hierar-
chical structure of physics:
● Implement concept map diagrams or flow charts. The Van
Heuvelen (1991a, 1991b) classroom study reviewed earlier
used a hierarchical flow chart to reveal the structure of
the physics course. There is no single “correct” flow chart,
although there should be substantial overlap in the flow
charts generated by different instructors. Such a flow chart
can be generated and filled in as the course progresses in
order to help students organize their physics knowledge
efficiently. Typically the hierarchy will show the major
principles at the top, and under each principle how different
conditions lead to specific forms of the major principle
(e.g., under work and energy, would be such branches as
work done by conservative and non-conservative forces,
potential and kinetic energies, the work-energy theorem
and conservation of mechanical energy). As one looks
down the hierarchy, the entries become more specific, for
example, discussing procedures and equations for applying
the concepts.

11998_Master4.indd 75 2/11/2020 [Link] PM

 76 The Science of Learning Physics

As reviewed earlier, novices find it difficult to “see” the structure

of a discipline until they gain considerable experience; however,
that does not mean that novices cannot be assisted in building
a more expert-like structure in memory of the subject matter.
Textbooks do not help to reveal the hierarchy of physics, and stu-
dents tend to view content coverage in lectures and textbooks as
a linear journey through unrelated topics. Building a hierarchical
flow chart and using it to think about solutions to problems is
one way of helping novices begin to restructure their amorphous
memory store of physics ideas into a more coherent whole. How
to “count” such a flow chart in a course’s grading structure so
that students take it seriously is a bit trickier. One approach is
to ask students to take problems they have previously solved
and trace the path of the solution through the hierarchy. Such
an exercise would need to take place during the last half of the
course when a substantial portion of the flow chart has been
filled in or the exercise becomes rather trivial (i.e., all kinematic
problems follow the same path more or less).
There has been some research on implementing “link maps”
in introductory physics courses to illustrate the key concepts
from lecture and how they are related to each other (Lindstrøm
& Sharma, 2009, 2011). In those studies, the instructor provided
a map of concepts and modeled how to use it during optional
problem solving “map meetings.” Students who attended the map
meetings had higher retention, more positive feedback about the
course, and scored higher on exams. The strategy was particu-
larly effective for low-achieving students.
● Highlight connections in the course content. As an
instructor, you might already highlight for your students
how the content builds or is an extension of what they’ve
learned previously. For example, if students in an electricity

11998_Master4.indd 76 2/11/2020 [Link] PM

 How Learning Looks for a Novice 77

and magnetism class are starting to learn about electric

forces between charged particles, you might remind them
how they previously solved force problems when they
learned about Newton’s laws of motion in mechanics
(breaking force vectors into components, analyzing the sum
of the forces along each coordinate axis, etc.). Or, instead of
pointing out these connections, you might elicit students’
existing knowledge on a topic and help them make their
own connections.
Instructional strategies for making tacit knowledge explicit:

● Justify problem solving decisions. As an expert on routine

-
physics problems or “exercises” in your view, it can be easy
to go into autopilot mode when writing down a problem
solution. It is important to remember that many of your
decisions have become automatic and you do not even
have to think about what comes next, but for a novice
this is a common struggle. It is valuable for beginning
students to see every procedure unpacked and explained.
When modeling how to solve an example problem for
your students, start from the very beginning and explicitly
discuss the justifications for each step. In addition, practice
steps that will benefit students—before writing equations,
start by representing the information given in the problem
(such as with a picture, diagram, and assigning symbols
for quantities). Then explain how you use that given
information to decide on an approach with the basic physics
concepts and principles. Highlight the conditions under
which those concepts and principles are applicable (see
strategy writing above). Show all steps in writing specific
equations and arithmetic to reach an answer, even if it
seems trivial. Model how you evaluate whether the answer

11998_Master4.indd 77 2/11/2020 [Link] PM

 78 The Science of Learning Physics

makes sense. Recall that a novice has to practice making

deliberate decisions before they can become an advanced
beginner or competent problem solver.
● Implement principle identification exercises. Experts
tacitly consider which principle(s) apply to problems
and then apply it(them) in equation form to solve them.
Novices tend to skip the principle identification stage and
go straight to equation manipulation. Exercises where
students are presented with problems, one at a time,
and asked to identify the principle needed for solution
(without actually solving the problem) would make this
type of tacit knowledge visible for students. Note that this
is the first step in strategy writing, however, the additional
two steps in strategy writing (justifying the principle and
describing a procedure for solving the problem) are rather
time consuming. Having only principle identification would
allow many problems to be considered in a short time span
and offer the instructor opportunities for discussing with
students how to extract the needed principle from the
problem’s surface attributes.

In Chapter 4 we will highlight additional instructional strategies

appropriate for teaching problem solving.

References
Chi, M.T.H. (2006). Two approaches to the study of experts’ characteristics.
In K.A. Ericsson, N. Charness, R.R. Hoffman & P.J. Feltovich (Eds.), The
Cambridge Handbook of Expertise and Expert Performance, pp. 21–30.
New York, NY: Cambridge University Press.
Chi, M.T.H., Feltovich, P.J. & Glaser, R. (1981). Categorization and represen-
tation of physics problems by experts and novices. Cognitive Science,
5, 121–152.

11998_Master4.indd 78 2/11/2020 [Link] PM

 How Learning Looks for a Novice 79

Docktor, J.L, Mestre, J.P. & Ross, B.H. (2012). Impact of a short intervention
on novices’ categorization criteria. Physical Review Special Topics –
Physics Education Research, 8(020102), 1–11. https:// [Link]/10.1103/
PhysRevSTPER.8.020102
Docktor, J.L., Strand, N.E., Mestre, J.P. & Ross, B.H. (2015). Conceptual
problem solving in high school physics. Physical Review Special
Topics – Physics Education Research, 11(020106), 1–13. https:// doi.
org/10.1103/PhysRevSTPER.11.020106
Dreyfus, H.L. & Dreyfus, S.E. (1986). Mind Over Machine: The Power of
Human Intuition And Expertise In An Era Of The Computer. New York:
Free Press.
Dufresne, R., Gerace, W.J., Hardiman, P.T. & Mestre, J.P. (1992). Constraining
novices to perform expert-like problem analyses: Effects on schema
acquisition. Journal of the Learning Sciences, 2, 307–331. https:// doi.
org/10.1207/s15327809jls0203_3
Ericsson, K.A., Krampe, [Link]. & Tesch-Romer, C. (1993). The role of
deliberate practice in the acquisition of expert performance. Psycho-
logical Review, 100, 393–394.
Feil, A. & Mestre, J.P. (2010). Change blindness as a means of studying
expertise in physics. Journal of the Learning Sciences, 19(4), 480–505.
Gautreau, R. & Novemsky, L. (1997). Concepts first—A small group
approach to physics learning. American Journal of Physics, 65, 418–428.
Gick, M.L. (1986). Problem-solving strategies. Educational Psychologist,
21(1&2), 99–120. https:// [Link]/10.1080/00461520.1986.9653026
Hardiman, P.T., Dufresne, R. & Mestre, J.P. (1989). The relation between
problem categorization and problem solving among novices and
experts. Memory & Cognition, 17, 627–638.
Larkin, J.H., McDermott, J., Simon, D.P. & Simon, H.A. (1980). Models of
competence in solving physics problems. Cognitive Science, 4, 317–345.
Leonard, W.J., Dufresne, R.J. & Mestre, J.P. (1996). Using qualitative problem-
solving strategies to highlight the role of conceptual knowledge in
solving problems. American Journal of Physics, 64(12), 1495–1503.

11998_Master4.indd 79 2/11/2020 [Link] PM

 80 The Science of Learning Physics

Lindstrøm, C. & Sharma, M.D. (2009). Link maps and map meetings:
Scaffolding student learning. Physical Review Special Topics –
Physics Education Research, 5(010102), 1–11. https:// [Link]/10.1103/
PhysRevSTPER.5.010102
Lindstrøm, C. & Sharma, M.D. (2011). Teaching physics novices at univer-
sity: A case for stronger scaffolding. Physical Review Special Topics –
Physics Education Research, 7(010109), 1–14. https:// [Link]/10.1103/
PhysRevSTPER.7.010109
Madsen, A.M., Larson, A.M., Loschky, L.C. & Rebello, N.S. (2012). Differ-
ences in visual attention between those who correctly and incorrectly
answer physics problems. Physical Review Special Topics – Physics
Education Research, 8(010122), 1–13. https:// [Link]/10.1103/PhysRev
STPER.8.010122
McDermott, L.C., Shaffer, P.S. & Sommers, M.D. (1994). Research as a guide
for teaching introductory mechanics: An illustration in the context of
the Atwood’s Machine. American Journal of Physics, 62, 46–55.
Mestre, J. (2002). Probing adults’ conceptual understanding and transfer
of learning via problem posing. Journal of Applied Developmental
Psychology, 23, 9–50.
National Research Council (2000). How People Learn: Brain, Mind, Experi-
ence, and School: Expanded Edition. Washington, DC: The National
Academies Press. https:// [Link]/10.17226/9853
Reif., F. & Heller, J.I. (1982). Knowledge structure and problem solving
in physics. Educational Psychologist, 17(2), 102–127. https:// [Link]/
10.1080/00461528209529248
Sweller, J. (2011). Cognitive load theory. In J. P. Mestre & B. H. Ross (Eds.),
The Psychology of Learning and Motivation: Cognition in Education,
Vol. 55, pp. 37–76. San Diego, CA, US: Elsevier Academic Press. http://
[Link]/10.1016/B978-0-12-387691-1.00002-8
Van Heuvelen, A. (1991a). Learning to think like a physicist: A review of
research-based instructional strategies. American Journal of Physics,
59(10), 891–897. https:// [Link]/10.1119/1.16667

11998_Master4.indd 80 2/11/2020 [Link] PM

 How Learning Looks for a Novice 81

Van Heuvelen, A. (1991b). Overview, case study physics. American Journal

of Physics, 59(10), 898–907. https:// [Link]/10.1119/1.16668
Van Heuvelen, A. & Maloney. D. (1999). Playing physics jeopardy.
American Journal of Physics, 67(3), 252–256. https:// [Link]/10.1119/
1.19233
Walsh, L.N., Howard, R.G. & Bowe, B. (2007). Phenomenographic study
of students’ problem solving approaches in physics. Physical Review
Special Topics – Physics Education Research, 3(020108), 1–12. https://
[Link]/10.1103/PhysRevSTPER.3.020108

11998_Master4.indd 81 2/11/2020 [Link] PM

 b2530   International Strategic Relations and China’s National Security: World at the Crossroads

This page intentionally left blank

b2530_FM.indd 6 01-Sep-16 [Link] AM

 c hapter
4 From Manipulating Equations to
a More Conceptual Approach:
How to Improve Problem Solving

4.1. What does research on problem solving tell us?

We have all witnessed that, when solving physics problems, begin-
ning physics students focus on searching for equations and plug-
ging in numbers to obtain answers. This approach works some
of the time, but when problems contain some subtlety, stu-
dents become stumped and do not know how to get started.
One of the primary goals of a physics class is to help students
become proficient at solving problems, thus, we as instruc-
tors need to help students acquire productive skills and behav-
iors. Since problem solving permeates all of physics instruction,
we have already discussed several problem solving activi-
ties in the previous two chapters. In this chapter we focus spe-
cifically on problem solving and review student approaches to
solving problems and some evidence-based instructional prac-
tices (EBIPs) for helping students adopt a more holistic problem
solving approach.

4.1.1. Research on problem solving skills and approaches

Early research into problem solving in physics focused on dif-
ferences between experienced problem solvers or “experts” and

83

11998_Master4.indd 83 2/11/2020 [Link] PM

 84 The Science of Learning Physics

inexperienced solvers or “novices.” As we reviewed in Chapter 3,

experts classify problems according to the concepts and princi-
ples used to solve them, whereas novices focus on surface fea-
tures, like the objects in the problem (springs, pulleys, inclined
planes, etc.), or the quantities given in the problem, like velocity
and acceleration (Chi et al., 1981; Docktor et al., 2012). This
implies that there are major differences in the way the knowl-
edge of problems is organized in memory for experts and nov-
ices (Eylon & Reif, 1984; Reif & Heller, 1982; Zajchowski & Martin,
1993). As a problem solver becomes more experienced, knowl-
edge can become “chunked” together with concepts and proce-
dures for solving problems about a particular topic. This reduces
cognitive load to free up working memory space (Sweller, 1988).
When faced with a problem, students will use a variety of
approaches and strategies. Sometimes these approaches are
effective in the short term to get an answer but are ineffective
for long term learning and retention. For example, students
might immediately look for equations that contain the quan-
tities given in the problem and start plugging in numbers, an
approach called “plug and chug,” or they might use a similarity-
based approach where they find a previously solved problem that
looks similar to the new problem and “pattern match” the solu-
tion (Gick, 1986; Tuminaro & Redish, 2007; Walsh et al., 2007).
They might use “means–ends analysis” to examine what they
are given and what they are trying to find and search for ways
to reduce the gap between the two (Gick, 1986). A more scien-
tific approach taken by experienced problem solvers involves per-
forming an initial qualitative analysis of the problem based on
concepts and principles and then plan out the solution proce-
dure prior to writing down equations (Larkin et al., 1980; Walsh
et al., 2007). The planning process might involve breaking up

11998_Master4.indd 84 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 85

the problem into relevant goals and sub-goals, a process called

problem decomposition (Gick, 1986). Expert-like approaches have
also been referred to as schema-driven strategies since the initial
qualitative representations of the question might cue the appli-
cation of a particular schema containing concepts and principles
along with appropriate procedures for applying those principles
(Gick, 1986). A schema in psychology refers to a mental construct
that helps the user organize experiences and situations.
Adams and Weiman (2015) identified forty-four subskills
involved in problem solving, broken into three different catego-
ries. Students need to have specific kinds of knowledge, such as
basic mathematical knowledge for addition/subtraction/multi-
plication/division, along with reading comprehension and some
real-world knowledge to make sense of situations. They need to
be able to engage in several different kinds of processes, such as
visualizing the problem, planning ways to get an answer, judging
the usefulness of information, adaptability and checking calcula-
tions. There are also several beliefs, expectations, and motivation
factors that can influence problem solving ability. For example,
students may have an inappropriate attitude that emphasizes
memorization over deeply understanding the material, or beliefs
about the kind of effort required to make sense of physics (Elby,
2001; Redish et al., 1998).
Others have highlighted the importance of metacogni-
tive skills throughout the problem solving process, where the
term metacognition refers to thinking about your own thinking.
For example, successful solvers periodically monitor their prog-
ress toward a solution with questions such as, “Does this pro-
cess make sense or is the solution getting too messy? Am I still
making progress toward the goal of the problem?” They also look
back at the end to evaluate their solution for reasonableness, for

11998_Master4.indd 85 2/11/2020 [Link] PM

 86 The Science of Learning Physics

example, checking the magnitude of the numerical answer, units

of the final answer, or using limiting cases in an expression to
check for consistency.
When viewing worked example problems, self-explanations
can be a useful metacognitive tool where students try to make
sense of the information by explaining what is going on in the
solution to themselves based on what they already know (Chi
et al., 1989). This is akin to keeping a mental dialogue while
reading, stopping to periodically ask whether what you are
reading makes sense or if you have questions about the material.
Frequent opportunities to receive feedback on their learning can
help students identify their personal strengths and weaknesses
and prompt them to reflect on how they will prepare for future
assignments and assessments. Mason and Singh (2010) found
that giving students opportunities for peer reflections helped
students learn more effective problem solving strategies. During
the reflections, small groups of students met to discuss home-
work problem solutions and identify characteristics that made
some solutions superior to others. In the next sections, we review
additional strategies you can use as an instructor to promote
productive problem-solving skills and behaviors in your courses.

4.1.2. Problem representation is important

There are many different ways to write a physics problem, and dif-
ferent formats can significantly impact the solution approaches
and strategies used by students. The term representation gener-
ally refers to the way information is described or depicted, such
as with a model or picture. Examples of representations relevant
for physics problems include text or verbal representations, dia-
grammatic, mathematical/symbolic, and graphical (Meltzer, 2005).
Although his study focused on multiple choice questions, Meltzer

11998_Master4.indd 86 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 87

(2005) observed that students gave inconsistent answers to dif-

ferent representations of the same question. Kohl and Finkelstein
(2005) and De Cock (2012) found similar results for homework and
quiz problems presented in those four different representations.
Follow-up studies focused on how experts and novices
use representations while solving problems. Kohl et al. (2007)
observed that for challenging problems, students who drew a
free-body diagram outperformed students who did not draw
a diagram. There was a small subset of students who did not
include a diagram yet had a correct solution, so it is possible
that they were very comfortable with the material and kept track
of information mentally rather than writing it down on paper.
When the problem was too easy or straightforward, it was pos-
sible for students to plug-and-chug to a correct answer without
including a diagram. Kohl and Finkelstein (2008) observed that
both experts and novices used similar representations when
solving problems with familiar contexts, such as drawing a free-
body diagram (likely because students had been instructed in
this manner). Where they found differences was in particular
stages of the problem solving process, such as analysis and explo-
ration. Experts spent more of their time using representations to
identify specific goals or sub-goals for the problem, whereas nov-
ices lacked a clear purpose for their representation use.
As you have likely observed in your own courses, students
perform significantly worse on questions that are purely sym-
bolic compared to ones where they are given numbers (Torigoe
& Gladding, 2011). Other features that can make a problem
more difficult include unfamiliar contexts or atypical situations,
excess information or missing information, two or more prin-
ciples required for a solution, a choice of possible principles,
abstract principles, mathematical complexity (such as vector

11998_Master4.indd 87 2/11/2020 [Link] PM

 88 The Science of Learning Physics

components, lengthy algebra, calculus or simultaneous equa-

tions), and vague statements such as no explicit target variable
(Heller & Hollabaugh, 1992). The question of whether providing
a supportive diagram in the problem statement helps or hinders
students has resulted in mixed findings. Maries and Singh (2018)
found that students performed worse on some problems where
they were provided a diagram, particularly for electrostatics
contexts. A study with eye-tracking data showed that students
spent less time reading the text of a problem when a diagram
was included with the problem statement, but providing a dia-
gram did not necessarily lead to a higher proportion of correct
responses (Susac et al., 2019).
Since standard problems presented in textbooks are typi-
cally well-defined, researchers have developed a host of alternate
types of problems to promote conceptual understanding and
physics reasoning in more realistic contexts (Hsu et al., 2004). As
reviewed in previous chapters, categorization tasks can be used to
help students practice identifying relevant concepts and princi-
ples for a problem without actually going through the process of
solving the problem (Dufresne et al., 1992; Hardiman et al., 1989).
As described in Chapter 3, students could also be presented with
a solution that contains a conceptual error and be asked to iden-
tify the error, a task called finding errors (Docktor et al., 2015).
Additional examples include context-rich problems (Heller &
Hollabaugh, 1992), problem posing (Mestre, 2002), jeopardy prob-
lems (Van Heuvelen & Maloney, 1999), ranking tasks (O’Kuma
et al., 2000), synthesis problems (Ding et al., 2011) and real-world
problems. Each of these alternate problem types encourage stu-
dents to make decisions and explicitly consider the conceptual
aspects of physics relevant for problem solving, as we will now
describe.

11998_Master4.indd 88 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 89

In problem posing, students are presented with a situation

(such as a diagram of two blocks connected by a string) along
with a concept scenario and they must generate a problem state-
ment that matches both the situation and concepts (Mestre,
2002). This notion of generating a problem statement is sim-
ilar to jeopardy problems that present students with a part
of a problem (such as a diagram or equation) and ask them to
devise a situation that could match with the presented informa-
tion, essentially working backward to write the question (Van
Heuvelen & Maloney, 1999).
Ranking tasks are conceptual exercises that provide a list of
several variations of a physical situation, oftentimes between
four to eight different diagrams or pictures and ask students
to rank them according to some criteria (O’Kuma et al., 2000).
Synthesis problems combine two physics concepts that are sep-
arated in a typical course timeline (Ding et al., 2011). Since most
end-of-chapter problems in a textbook typically only address a
narrow range of topics and even sometimes label problems with
a section header indicating the key topics and equations, this
helps draw students’ attention away from equations and toward
selecting relevant concepts when solving problems. When stu-
dents struggled to solve synthesis problems, Ding et al. (2011)
found that providing a series of conceptual questions helped to
guide students toward a solution.
Real-world problems use data and measurements from real-
life situations to construct physics questions (see the MIT site
https: / /[Link]/2019/making-it-real-mit-engineering-
class-0513). The data could include pictures of videos of situ-
ations. These problems have some similarities to the instruc-
tional approach called Problem-Based Learning or PBL (Duch et
al., 2001). In PBL, students work together in collaborative groups

11998_Master4.indd 89 2/11/2020 [Link] PM

 90 The Science of Learning Physics

to analyze, research, and solve complex real-world problems and

communicate their findings. Both of these approaches help to
motivate students by illustrating the relevance of physics in
everyday life situations.
Below we have provided some examples of different types of
physics questions.

Traditional textbook-style problem with numbers

A 200-kg block sits on a ramp that is inclined at θ =37.0
degrees above the horizontal. The coefficients of static and F
kinetic friction between the block and the ramp are µs =0.60
and µk =0.40, respectively.
(a) What horizontal force F is required to keep the block ɽ = 37°
stationary?
(b) If the horizontal force is removed, what is the block’s
speed after it slides 3.0 m down the ramp?

Traditional textbook-style problem without numbers

A block of mass M sits on a ramp that is inclined at θ degrees
above the horizontal. The coefficients of static and kinetic fric- F
tion between the block and the ramp are µs and µk respectively.
(a) Write an expression for the horizontal force F required
to keep the block stationary. ɽ
(b) If the horizontal force is removed, what is the block’s
speed after it slides a distance d down the ramp?

Context Rich Problem

You have a summer job with a professional moving company. You and another mover are
unloading a 200-kg wooden crate containing a piano from a truck using a 5-m long steel
ramp that is inclined 37 degrees above the horizontal. The cell phone of the other mover
rings and they ask you to hold the piano in place during the call. What horizontal force must
you apply to keep the piano stationary? You get tired after a few minutes, let go of the piano,
and quickly jump out of the way. If the crate is 2 meters wide, what is the piano’s speed at the
bottom of the ramp? You look in a book giving the properties of materials and find that the
coefficient of kinetic friction for wood on steel is 0.40 and the coefficient of static friction
is 0.60.
(Continued)

11998_Master4.indd 90 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 91

Problem Posing
Pose a problem that could be solved by applying Newton’s
Second Law by adding a statement or question to the situation
below to turn it into a “textbook-like” problem.

A block of mass M is situated on a inclined plane of angle θ. The

ɽ
coefficients of static and kinetic friction between the block and
the ramp are µs =0.60 and µk =0.40, respectively.

Jeopardy Problem
Invent a word problem that is consistent with the following equation:
(200 kg)(9.8 m/s2)sin(37°) - (0.40)(200 kg)(9.8 m/s2)cos(37°) = (200 kg) ax

4.1.3. Problem solving frameworks

One way to think about teaching and learning physics problem
solving is a theory called cognitive apprenticeship (Brown et al.,
1989; Collins et al., 1987). Like traditional apprenticeship models
for learning a trade skill, a novice works closely under the direc-
tion of a master to observe and practice the necessary skills of
the trade within the situation or context in which it will take
place. When this is applied to the classroom, the key pieces of
cognitive apprenticeship involve the instructor modeling how to
perform a skill with explanation while the apprentice (students)
attentively observe, then coaching the student(s) and offering
feedback as they attempt the skills on their own, and providing
scaffolding to support the specific behavior targeted, and finally
fading support structures as the student becomes more profi-
cient at the skill on their own.
When it comes to problem solving instruction, the modeling
phase is particularly important—every interaction students have
with problem solving sends a message about our expectations.
The way you model how to solve example problems during class,
the way the textbook structures example problems, and the way

11998_Master4.indd 91 2/11/2020 [Link] PM

 92 The Science of Learning Physics

homework solutions are written, all influence students’ behaviors

related to problem solving. If you want to encourage productive
behaviors such as “showing your work,” you will need to care-
fully consider the way problem solving is communicated in all
aspects of your course. The coaching and scaffolding pieces are
also important, such as the grading schemes you use on assess-
ments and the feedback you provide to students on homework,
quizzes, or exams. Giving points for things like drawing a dia-
gram or explaining reasoning steps can help to reinforce your
expectations for written problem solutions—we know from expe-
rience that if you do not grade something, students will not do it
since in their minds it does not “count.” If you are using online
homework systems or computer coaches, you want to ensure that
they emphasize the problem solving process and not just the
final numerical answer.
Problem solving frameworks or templates can be a useful way
to model expectations for students when formatting problem
solutions (Hsu et al., 2004). For example, if you want students to
start by drawing a picture of the situation and assigning sym-
bols for quantities in the question, you could provide them with
a worksheet designating space for that step. If you expect stu-
dents to solve equations in symbols prior to plugging in num-
bers, you will need to consistently model that process for them.
Several frameworks have been developed for physics instruction
to guide students’ use of an organized problem solving strategy
and break down complex processes used implicitly by experi-
enced solvers (Burkholder et al., 2020; Heller et al., 1992; Heller
& Heller, 2000; Van Heuvelen, 1991a, 1991b; Wright & Williams,
1986). As described below, many of these were derived from
research in mathematics.

11998_Master4.indd 92 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 93

One of the first people to document key steps in the problem

solving process was the mathematician Pólya in his book How
to Solve it (Pólya, 1945). His first step was Understanding the
Problem including summarizing known and unknown informa-
tion, introducing suitable notation, and drawing a figure (if rel-
evant to the problem). In step two of Devising a Plan, the solver
uses their knowledge to plan how to connect the given informa-
tion to the desired goal in the problem before they engage in
Carrying out the Plan. The final step is Looking Back to examine
whether the result makes sense.
Wright and Williams (1986) identified a four-step WISE
strategy for solving physics problems to place more emphasis
on physics and less emphasis on “number crunching” in their
courses:
● What’s happening? (identify physical principles, draw a
sketch or diagram, and identify knowns and unknowns)
● Isolate the unknown (select an equation, solve symbolically,
search systematically for other equations as needed)
● Substitute (plug in numbers and units)
● Evaluate (check sign, magnitude, and units of the answer)

Although initially they were met with some resistance from stu-
dents about perceived additional writing, ultimately they found
that it improved communication between the instructor and stu-
dents, increased the accuracy of solutions, and promoted orga-
nization of students’ solutions. These results are similar to those
found with the process of strategy-writing described in Chapter 3
where students first write down the principle they are planning
to apply to the problem and a justification for why that principle
is applicable, then they write out a procedure for solving the
problem (Leonard et al., 1996). The procedure could be formatted

11998_Master4.indd 93 2/11/2020 [Link] PM

 94 The Science of Learning Physics

as a two-column solution with the plan steps on the left column

and the execution of those steps in the right column, also shown
in Chapter 3 (Docktor et al., 2015).
Since the first step is critical and fairly involved for solving
physics problems, frameworks sometimes subdivide this into
more than one step. Heller and Reif (1984) suggested that effec-
tive problem solvers first generate a “basic description” that sum-
marizes information about the situation in symbolic, pictorial
and verbal forms prior to producing a “theoretical description”
that contains diagrams specific to physics concepts and princi-
ples. This is reflected in the five-step strategy developed by the
Physics Education Research Group at the University of Minnesota
(Heller & Heller, 2000; Heller & Hollabaugh, 1992; Heller et al.,
1992). This five-step strategy differs from other strategies in that
it explicitly asks students to write a solution plan starting from
the unknown quantity(ies).
1. Focus the problem: draw a picture and write down given
information, rephrase question(s) you are trying to answer,
general physics approach (key concepts/principles)
2. Describe the physics: draw physics-specific diagrams
and define quantities, identify target quantity(ies), write
quantitative relationships
3. Plan the solution: start with an equation which has the
target quantity(ies) and identify other unknowns, write
additional equations until you have sufficient equations to
solve for all the unknowns, then write a plan for solving
the equations
4. Execute the plan: follow the steps to solve for all unknown
quantities you identified in your plan, calculate a numerical
value for the target quantity(ies) as applicable, and check
units

11998_Master4.indd 94 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 95

5. Evaluate the answer: check whether the answer is properly

stated, unreasonable, and complete
Van Heuvelen (1991b) used a structured problem solving approach
as part of the Overview, Case Study Physics curriculum to
encourage multiple representations of information. Students
develop a pictorial representation and identify the given infor-
mation prior to constructing a physical representation, such as
a motion diagram or force diagram. Then they carry out a solu-
tion and evaluate the reasonableness of the answer. The frame-
work can be tailored specifically for particular topics such as
kinematics or forces.
An example of implementing the Minnesota strategy (Fig. 4.1)
and case study physics strategy (Fig. 4.2) are presented in the
frameworks below for the context rich piano moving problem
presented in the previous section:

You have a summer job with a professional moving company.

You and another mover are unloading a 200-kg wooden crate
containing a piano from a truck using a 5-m long steel ramp
that is inclined 37 degrees above the horizontal. The cell phone
of the other mover rings and they ask you to hold the piano in
place during the call. What horizontal force must you apply to
keep the piano stationary? You get tired after a few minutes, let
go of the piano, and quickly jump out of the way. If the crate
is 2 meters wide, what is the piano’s speed at the bottom of the
ramp? You look in a book giving the properties of materials and
find that the coefficient of kinetic friction for wood on steel is
0.40 and the coefficient of static friction is 0.60.

4.1.4. Assessment of problem solving

If you have graded students’ problem solutions on quizzes or
exams, a standard procedure is to set the total points possible for

11998_Master4.indd 95 2/11/2020 [Link] PM

 96 The Science of Learning Physics

(Continued)

11998_Master4.indd 96 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 97

Fig. 4.1. Minnesota five-step problem solving strategy.

11998_Master4.indd 97 2/11/2020 [Link] PM

 98 The Science of Learning Physics

Fig. 4.2. Van Heuvelen strategy from case study physics.

11998_Master4.indd 98 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 99

the problem and then assign points to particular components of

the solution (or deduct points for particular errors). This allows
the instructor some flexibility in weighing problem features and
tailoring a grading scheme to different topics or assessment sit-
uations. One thing to be cautious of when grading is to stay true
to your goals as an instructor. Even if you want students to show
their reasoning, research has indicated that instructors are hesi-
tant to deduct points from a sparse solution that might be cor-
rect, and they tend to make assumptions about students’ thought
processes when there are gaps in written work (Henderson et al.,
2004). If you are using a particular framework or template, points
should be awarded for each aspect of the framework to reinforce
the importance to including each step. Some samples of tem-
plates and point designations are provided in (Burkholder et al.,
2020).
Although there currently are no multiple choice diagnostic
assessments for problem solving like there are for concepts, some
researchers have focused on developing rubrics for assessing
problem solving skills. Docktor et al. (2016) describe the devel-
opment and validation of the Minnesota Assessment of Problem
Solving (MAPS) rubric designed to apply to any problem type
or topic. The rubric defines criteria to attain a score in five dif-
ferent areas: organizing problem information into a Useful
Description, selecting appropriate concepts and principles in the
Physics Approach, the Specific Application of Physics principles
to the conditions in the problem, using appropriate Mathematical
Procedures, and following an organized procedure or Logical
Progression. Each of these criteria are scored on a scale from
0 to 5 or designated as not applicable to the problem or solver.
Using a rubric like this can help instructors identify students’
strengths and weaknesses related to problem solving skills and

11998_Master4.indd 99 2/11/2020 [Link] PM

 100 The Science of Learning Physics

target instruction accordingly. The rubric was built on previous

rubrics such as the one used in (Heller et al., 1992) for use with
context rich problems and cooperative group problem solving.
Hull et al. (2013) criticize that most problem solving rubrics do
not go far enough in valuing the conceptual aspects involved in
mathematical procedures, and should be revised to better assess
the blending of conceptual and formal mathematical reasoning
during those procedures. A copy of the MAPS rubric is provided
in Fig. 4.3.

4.2. What are the implications of research on problem

solving for instruction?
We offer the following implications of the research reviewed
above for instruction:
● Problem solving needs to be taught explicitly. Students
use a variety of strategies to solve physics problems,
and while some of these are effective in the short term
they are ineffective for long-term retention and transfer.
Instructors need to explicitly model a process for students
that emphasizes principles and concepts over equation
manipulation, highlighting all of the decisions that have
become automatic for solving routine physics exercises.
● If you do not assess it, students will not do it. If you
expect students to produce carefully written problem
solutions and explain their reasoning steps, you will need
to design assessment structures in your course to reward
those behaviors. It is very easy for instructors to assume
that a student with a sparse solution who obtained the
correct result “knew what they were doing” and should
receive full credit, compared to a student who showed
more steps but made a conceptual error along the way

11998_Master4.indd 100 2/11/2020 [Link] PM

11998_Master4.indd 101

5 4 3 2 1 0 NA(Problem) NA(Solver)

The description The description Parts of the Most of the The entire The solution A description is A description is
is useful, is useful but description description is not description is does not include not necessary for not necessary for
USEFUL appropriate, and contains minor are not useful, useful, missing, not useful and/or a description and this problem. this solver.
DESCRIPTION complete. omissions or missing, and/or and/or contains contains errors. it is necessary (i.e., it is given
errors. contain errors. errors. for this problem in the problem
/solver. statement)

The physics The physics Some concepts Most of the All of the chosen The solution does An explicit An explicit
approach is approach and principles physics approach concepts and not indicate an physics approach physics approach
appropriate and contains minor of the physics is missing and/or principles are approach, and it is not necessary is not necessary
PHYSICS
complete. omissions or approach are inappropriate. inappropriate. is necessary for for this problem. for this solver.
APPROACH
errors. missing and/or this problem/ (i.e., it is given in
inappropriate. solver. the problem)

The specific The specific Parts of Most of The entire specific The solution does Specific Specific
application application of the specific the specific application is not indicate an application of application of
SPECIFIC
of physics is physics contains application application inappropriate application of physics is not physics is not
APPLICATION OF
appropriate and minor omissions of physics are of physics is and/or contains physics and it is necessary for this necessary for this
PHYSICS
complete. or errors. missing and/or missing and/or errors. necessary. problem. solver.
contain errors. contains errors.

The Appropriate Parts of the Most of the All mathematical There is no Mathematical Mathematical
mathematical mathematical mathematical mathematical procedures are evidence of procedures are procedures are
MATHEMATICAL procedures are procedures are procedures are procedures are inappropriate mathematical not necessary for not necessary for
PROCEDURES appropriate and used with minor missing and/or missing and/or and/or contain procedures, this problem or this solver.
complete. omissions or contain errors. contain errors. errors. and they are are very simple.
errors. necessary.

The entire The solution is Parts of the Most of the The entire There is no Logical Logical
problem solution clear and focused solution solution parts solution is unclear, evidence progression is not progression is not
LOGICAL is clear, focused, with minor are unclear, are unclear, unfocused, and/or of logical necessary for this necessary for this
PROGRESSION and logically inconsistencies unfocused, unfocused, inconsistent. progression, and problem. solver.
connected. and/or and/or it is necessary. (i.e., one-step)
inconsistent. inconsistent.
2/11/2020 [Link] PM

Fig. 4.3. Minnesota Assessment of Problem Solving (MAPS) rubric.

 102 The Science of Learning Physics

(Henderson et al., 2004). It has been shown that students

can solve a complex physics problem flawlessly but have a
flawed conceptual understanding of it (Morphew & Mestre,
2018). It is important to use consistent scoring criteria and
check assumptions.
● Students have attitudes, beliefs and expectations that
impact performance in physics. Problem solving in physics
is a complex process that can require students to stay moti-
vated and persistent. Sometimes beliefs about learning
physics can impact performance, like a belief that physics is
a set of disconnected facts and equations to be memorized
rather than a network of connected concepts. Students
might have attitudes toward learning physics that favor
getting an answer rather than understanding the appro-
priate procedures and methods for reaching that answer.
Instructors should be cognizant of students’ expectations
and how they can impact learning.

4.3. Examples of teaching interventions based on

learning research
The research reviewed in this chapter, as well as our teaching
experiences, indicate that the following strategies can be useful
for promoting skilled problem solving in your courses.
● Implement a problem solving framework to guide
students’ problem solutions. Providing a framework or
template for students can help students who just “don’t
know where to start” and can promote a more organized
solution process. Most of these frameworks begin with
drawing a sketch or picture of the situation and clearly
assigning symbols for quantities. Some frameworks
have students explicitly plan out their solutions before

11998_Master4.indd 102 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 103

carrying out the equation manipulation and mathematical

calculations. In this chapter, we reviewed at least four
different strategies and frameworks you could use. Actually
providing students with a worksheet template can help to
scaffold this process for them until they naturally follow
the procedure on their own.
● Carefully select problems for your course(s). There are
many different formats and styles for writing physics
problems. You will want to choose problems that discourage
common student behaviors like equation hunting and
chug-and-plug problem solving. So-called “traditional” end-
of-chapter problems include artificial situations like blocks
and pulleys on inclined planes and may have multiple
parts but this is not necessarily optimal for your particular
students. In writing problems you will have to decide things
like whether to include a diagram or not, numbers or no
numbers, symbols for quantities provided in the question
or not, and the difficulty of the mathematics required. There
are several research-based ways to write alternate formats
for problems which were reviewed in this chapter, including
context-rich problems, problem posing, jeopardy problems,
ranking task exercises, synthesis problems and real-world
problems.
● Model and assess productive problem solving behaviors.
When implementing an EBIP, such as a problem solving
framework, it is also important to model procedures for
students when solving example problems on the board
or in virtual worked examples. Explicitly verbalize the
decisions you are making and the conditions under which
a particular physics principle is applicable. Help students
connect math skills with physics reasoning and minimize
the use of nonfundamental formulae to discourage the

11998_Master4.indd 103 2/11/2020 [Link] PM

 104 The Science of Learning Physics

memorization of situation-specific equations (Zajchowski

& Martin, 1993). You will also need to model appropriate
behaviors in written solutions to homework, quizzes,
or exams as applicable. Solutions provided by textbook
companies as instructor solution manuals are typically
very sparse and missing steps—they are better suited for
instructors to view, not students.
Along with modeling productive problem solving
behaviors, it is also important to reinforce those behaviors
in the way you assess students’ solutions. Assign points
to each step of the process or of the written template so
students are rewarded for the process and not just the
final result. You could also implement research-based
instructional tools such as intelligent tutors, computer
coaches, or online homework systems to help students
practice problem solving skills—providing those systems
are aligned with the practices you are trying to encourage
in your course.
● Encourage the development of metacognitive skills for
problem solving. We want students to be reflective about
their learning when solving problems and how they might
apply what they have learned to future problems. When
students view worked example problems, encourage them
to use self-explanations to make sense of the procedures.
After they solve a problem, encourage students to evaluate
the reasonableness of their answer. Provide students
with frequent opportunities to receive feedback on their
performance and reflect on their learning.

References
Adams, W.K. & Wieman, C.E. (2015). Analyzing the many skills involved in
solving complex physics problems. American Journal of Physics, 83(5),
459–467. [Link]

11998_Master4.indd 104 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 105

Brown, J.S., Collins, A. & Duguid, P. (1989). Situated cognition and the
culture of learning. Educational researcher, 18(1), 32–42.
Burkholder, E.W., Miles, J.K., Layden, T.J., Wang, K.D., Fritz, A.V. &
Wieman, C.E. (2020). Template for teaching and assessment of
problem solving in introductory physics. Physical Review Physics
Education Research, 16(010123), 1–20. https:// [Link]/10.1103/
PhysRevPhysEducRes.16.010123
Collins, A., Brown, J.S. & Newman, S.E. (1987). Cognitive apprenticeship:
Teaching the craft of reading, writing and mathematics (Technical
Report no. 403). BBN Laboratories, Cambridge, MA. Centre for the
Study of Reading, University of Illinois, January, 1987.
Chi, M.T.H., Bassok, M., Lewis, M.W., Reimann, P. & Glaser, R. (1989). Self-
explanations: How students study and use examples in learning to
solve problems. Cognitive Science, 13(2), 145–182.
Chi, M.T.H., Feltovich, P.J. & Glaser, R. (1981). Categorization and represen-
tation of physics problems by experts and novices. Cognitive Science,
5, 121–152.
De Cock, M. (2012). Representation use and strategy choice in
physics problem solving. Physical Review Special Topics – Physics
Education Research, 8(020117), 1–15. http:// [Link]/10.1103/Phys
RevSTPER.8.020117
Ding, L., Reay, N., Lee, A. & Bao, L. (2011). Exploring the role of concep-
tual scaffolding in solving synthesis problems. Physical Review Special
Topics – Physics Education Research, 7(020109).
Docktor, J.L., Dornfeld, J., Frodermann, E., Heller, K., Hsu, L., Jackson,
K.A., Mason, A., Ryan, Q.X. & Yang, J. (2016). Assessing student
written problem solutions: A problem solving rubric with applica-
tion to introductory physics. Physical Review – Physics Education
Research, 12(010130), 1–18. http:// [Link]/10.1103/PhysRevPhys
EducRes.12.010130
Docktor, J.L, Mestre, J.P. & Ross, B.H. (2012). Impact of a short intervention
on novices’ categorization criteria. Physical Review-Special Topics:
Physics Education Research, 8(020102), 1–11. https:// [Link]/10.1103/
PhysRevSTPER.8.020102

11998_Master4.indd 105 2/11/2020 [Link] PM

 106 The Science of Learning Physics

Docktor, J.L., Strand, N.E., Mestre, J.P. & Ross, B.H. (2015). Conceptual
problem solving in high school physics. Physical Review Special
Topics – Physics Education Research, 11(020106), 1–13. https:// doi.
org/10.1103/PhysRevSTPER.11.020106
Duch, B.J., Groh, S.E. & Allen, D.E. (2001). The Power of Problem-based
Learning (1st Ed.) Sterling, Virginia: Stylus Publishing, LLC.
Dufresne, R., Gerace, W.J., Hardiman, P.T. & Mestre, J.P. (1992). Constraining
novices to perform expert-like problem analyses: Effects on schema
acquisition. Journal of the Learning Sciences, 2, 307–331. https:// doi.
org/10.1207/s15327809jls0203_3
Elby, A. (2001). Helping physics students learn how to learn. American
Journal of Physics, 69(7), S54–S64. https:// [Link]/10.1119/1.1377283
Eylon, B. & Reif, F. (1984). Effects of knowledge organization on task
performance. Cognition and Instruction, 1(1), 5–44. https:// [Link].
org/stable/3233519
Gick, M.L. (1986). Problem-solving strategies. Educational Psychologist,
21(1&2), 99–120. https:// [Link]/10.1080/00461520.1986.9653026
Hardiman, P.T., Dufresne, R. & Mestre, J.P. (1989). The relation between
problem categorization and problem solving among novices and
experts. Memory & Cognition, 17, 627–638.
Heller, J.I. & Reif, F. (1984). Prescribing effective human problem-solving
processes: Problem description in physics. Cognition and Instruction,
1(2), 177–216. [Link]
Heller, K. & Heller, P. (2000). Competent Problem Solver – Calculus Version.
New York: McGraw-Hill.
Heller, P. & Hollabaugh, M. (1992). Teaching problem solving through
cooperative grouping. Part 2: Designing problems and structuring
groups. American Journal of Physics, 60(7), 637–644. https:// doi.
org/10.1119/1.17118
Heller, P., Keith, R. & Anderson, S. (1992). Teaching problem solving
through cooperative grouping. Part 1: Group versus individual
problem solving. American Journal of Physics, 60(7), 627–636. http://
[Link]/10.1119/1.17117

11998_Master4.indd 106 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 107

Henderson, C., Yerushalmi, E., Kuo, V., Heller, P. & Heller, K. (2004). Grading
student problem solutions: The challenge of sending a consistent
message. American Journal of Physics, 72(2), 164–169. https:// doi.
org/10.1119/1.1634963
Hull, M.M., Kuo, E., Gupta, A. & Elby, A. (2013). Problem solving rubrics
revisited: Attending to blending of informal conceptual and formal
mathematical reasoning. Physical Review Special Topics – Physics
Education Research, 9(010105).
Hsu, L., Brewe, E., Foster, T.M. & Harper, K.A. (2004). Resource letter RPS-1:
Research in problem solving. American Journal of Physics, 72(9),
1147–1156. [Link]
Kohl, P.B. & Finkelstein, N.D. (2005). Student representational competence
and self-assessment when solving physics problems. Physical Review
Special Topics – Physics Education Research, 1(010104).
Kohl, P.B. & Finkelstein, N.D. (2008). Patterns of multiple representation
use by experts and novices during physics problem solving. Physical
Review Special Topics – Physics Education Research, 4(010111), 1–10.
Kohl, P.B. Rosengrant, D. & Finkelstein, N.D. (2007). Strongly and weakly
directed approaches to teaching multiple representation use in
physics. Physical Review Special Topics – Physics Education Research,
3(010108).
Larkin, J.H., McDermott, J., Simon, D.P. & Simon, H.A. (1980). Models of
competence in solving physics problems. Cognitive Science, 4, 317–345.
Leonard, W.J., Dufresne, R.J. & Mestre, J.P. (1996). Using qualitative problem-
solving strategies to highlight the role of conceptual knowledge in
solving problems. American Journal of Physics, 64(12), 1495–1503.
Maries, A. & Singh, S. (2018). Case of two electrostatics problems: Can
providing a diagram adversely impact physics tsudents’ problem
solving performance? Physical Review Physics Education Research,
14(010114).
Mason, A. & Singh, C. (2010. Helping students learn effective problem
solving strategies by reflecting with peers. American Journal of Physics,
78(7), 748–754.

11998_Master4.indd 107 2/11/2020 [Link] PM

 108 The Science of Learning Physics

Meltzer, D.E. (2005). Relation between students’ problem-solving perfor-

mance and representational format. American Journal of Physics,
73(5), 463–478.
Mestre, J.P. (2002). Probing adults’ conceptual understanding and
transfer of learning via problem posing. Journal of Applied
Developmental Psychology, 23(1), 9–50. https:// [Link]/10.1016/
S0193-3973(01)00101-0
Morphew, J.W. & Mestre, J.P. (2018). Exploring the connection between
problem solving and conceptual understanding in physics. Revista de
Enseñanza de la Física, 30(2), 75–85.
O’Kuma, T.L., Maloney, S.P. & Hieggelke, C.J. (2000). Ranking Task Exercises
in Physics (Prentice-Hall, Upper Saddle River, NJ).
Polya, G. (1945). How to Solve it. Princeton, NJ: Princeton University Press.
Redish, E.F., Saul, J.M. & Steinberg, R.N. (1998). Student expectations
in introductory physics. American Journal of Physics, 66(3), 212–224.
[Link]
Reif, F. & Heller, J.I. (1982). Knowledge structure and problem solving
in physics. Educational Psychologist, 17(2), 102–127. https:// [Link]/
10.1080/00461528209529248
Susac, A., Bubic, A., PLaninic, M., Movre, M. & Palmovic, M. (2019). Role of
diagrams in problem solving: An evaluation of eye-tracking parame-
ters as a measure of visual attention. Physical Review Special Topics –
Physics Education Research, 15(013101).
Sweller, J. (1988). Cognitive load during problem solving: Effects
on learning. Cognitive Science, 12, 257–285. https:// [Link]/10.1207/
s15516709cog1202_4
Torigoe, E.T. & Gladding, G.E. (2011). Connecting symbolic difficulties
with failure in physics. American Journal of Physics, 79(1), 133–140.
Tuminaro, J. & Redish, E.F. (2007). Elements of a cognitive model of
physics problem solving: Epistemic games. Physical Review Special
Topics – Physics Education Research, 3(020101), 1–22. https:// [Link]/
10.1103/PhysRevSTPER.3.020101

11998_Master4.indd 108 2/11/2020 [Link] PM

 From Manipulating Equations to a More Conceptual Approach 109

Van Heuvelen, A. (1991a). Learning to think like a physicist: A review of

research-based instructional strategies. American Journal of Physics,
59(10), 891–897.
Van Heuvelen, A. (1991b). Overview, case study physics. American Journal
of Physics, 59(10), 898–907. https:// [Link]/10.1119/1.16668
Van Heuvelen, A. & Maloney. D. (1999). Playing physics jeopardy. American
Journal of Physics, 67(3), 252–256. https:// [Link]/10.1119/1.19233
Walsh, L.N., Howard, R.G. & Bowe, B. (2007). Phenomenographic study
of students’ problem solving approaches in physics. Physical Review
Special Topics – Physics Education Research, 3(020108), 1–12. https://
[Link]/10.1103/PhysRevSTPER.3.020108
Wright, D.S. & Williams, C.D. (1986). A WISE strategy for introductory
physics. The Physics Teacher, 24, 211–216. https:// [Link]/10.1119/
1.2341986
Zajchowski, R. & Martin, J. (1993). Differences in the problem solving
of stronger and weaker novices in physics: Knowledge, strategies, or
knowledge structure? Journal of Research in Science Teaching, 30(5),
459– 470. [Link]

11998_Master4.indd 109 2/11/2020 [Link] PM

 b2530   International Strategic Relations and China’s National Security: World at the Crossroads

This page intentionally left blank

b2530_FM.indd 6 01-Sep-16 [Link] AM

 c hapter
5 Active Learning Strategies:
Engaging Students in their Own
Learning is the Key to Learning

5.1. What does the research tell us?

If you teach physics or another area of STEM, there is a good
chance that you have heard some buzz about “active learning”
instructional strategies. Maybe you have even tried some inno-
vative teaching methods in your classes with varied degrees of
success. Despite a growing body of literature in support of active
learning, the phenomenon has not spread as widely as might be
expected (Freeman et al., 2014; Henderson & Dancy, 2007, 2009).
In this chapter, we explore not only what active learning looks
like in a physics classroom but also examine the challenges asso-
ciated with implementation.

5.1.1. Introduction and background

Active instructional approaches have their roots in construc-
tivist learning theory, as discussed in Chapter 2—when we
encounter new information, we try to make sense of it based
on our existing ideas and experiences (NRC, 2000). Learning is
an active process of constructing knowledge, either cognitively
engaging with material or being physically active or both. As a
result, the extent to which students can effectively construct
new knowledge depends on the nature of their prior knowledge;

111

11998_Master4.indd 111 2/11/2020 [Link] PM

 112 The Science of Learning Physics

prior knowledge could be accurate but insufficient, or could be

inappropriate (sometimes called misconceptions or alterna-
tive conceptions). For example, recall from Chapters 1 and 2
that students might have an erroneous idea that heavier objects
fall faster than lighter ones or that in a collision between two
objects the bigger one exerts a larger force on the smaller one.
The “teaching by telling” lecture-based instructional approach
has been shown to be ineffective for addressing misconceptions
and facilitating long-term knowledge construction. This does not
mean you should never lecture—it just should not be the only
tool in your toolbox. There may be some situations for which lec-
turing is the most efficient way to convey information, however
being an effective educator typically means blending a variety
of strategies and methods to fit your context (NRC, 2000, 2012b,
2013; Schwartz & Bransford, 1998).
Constructivist epistemology also emphasizes the importance
of learners discovering or figuring things out for themselves
(Bruner, 1961). This does not mean students should simply be
left to their own devices to flounder with material; the instructor
still plays an important role in structuring learning activities and
serving as a facilitator or coach to guide students toward appro-
priate scientific thinking. This is consistent with the idea that
learning and cognition are sociocultural by nature; students can
make meaning from their observations, and interactions with
others (Bandura, 1986; Vygotsky, 1978).
As we examine active learning strategies in this chapter,
it will become evident that several of them incorporate meta-
cognition and overlap with formative assessment techniques.
Metacognition is the notion of getting students thinking about
their own thinking, such as reflecting on their reasoning pro-
cesses or what they have learned. This can be as straightforward

11998_Master4.indd 112 2/11/2020 [Link] PM

 Active Learning Strategies 113

as a student periodically questioning themselves: Does this infor-

mation make sense to me? How does this relate to what I already
know? What questions do I have about this material? or it could
be more explicitly structured into course activities, such as an
instructor prompting students to reflect: What were the key
physics principles and procedures we used to solve this problem?
Metacognitive strategies that help students monitor their own
learning have been shown to boost learning and transfer (NRC,
2000). In order to teach with metacognition in mind, instruc-
tors might implement classroom strategies to make students’
thinking visible to themselves and the instructor. These strate-
gies are oftentimes referred to as formative assessment class-
room techniques (FACTs) (Keeley, 2008). The term formative
broadly means that assessment is happening during the pro-
cess of instruction to help instructors identify students’ strengths
and weaknesses and inform the next steps for the teaching and
learning process. It is assessment for learning, not just assess-
ment of learning (Chappuis, 2015).
What exactly is “active learning” and what does it look like in
a physics classroom? The “traditional” lecture-based physics class
is usually described in the following way: the instructor does
most of the talking while deriving physics definitions and the-
ories, perhaps also illustrating how to solve quantitative prob-
lems. Students passively copy down notes, and are expected
to apply the information to algorithmic problem solving exer-
cises that appear at the end of the textbook chapter (Dancy &
Henderson, 2007; Docktor & Mestre, 2014; Hake, 1998). Students
might also attend laboratory sessions where they follow step by
step instructions to confirm or verify their experimental con-
sistency with physics principles. Almost anything that deviates
from this “traditional” view of instruction is colloquially referred

11998_Master4.indd 113 2/11/2020 [Link] PM

 114 The Science of Learning Physics

to as active learning or interactive engagement (IE) techniques

(Hake, 1998).
The entire field of Physics Education Research arose due to
a general dissatisfaction with student learning in traditional
physics classrooms, when instructors observed how students
could get a correct numerical answer to a problem, but they
could not explain the concepts (Cummings, 2011; Meltzer & Otero,
2015; Morphew & Mestre, 2018), or displayed conceptual under-
standing that did not conform with physics principles (i.e., mis-
conceptions). Some researchers suggest that in order to earn the
label of an “active learning” strategy, the instructional approach
needs to be grounded in research on student learning (Meltzer
& Thornton, 2012). Therefore, these practices are also referred
to as evidence-based instructional practices (EBIPs) or research-
based instructional strategies (RBIS) (Dancy & Henderson, 2007).
Non-traditional or alternative instruction includes significant
student–student discourse focusing on students’ ideas, and stu-
dents being physically active (interacting with equipment and
materials). Other disciplines might refer to these as student-centered
instructional approaches because students are afforded more con-
trol over their learning (NRC, 2012b). A summary of key elements
is provided below, see (Meltzer & Thornton, 2012; Mestre 2001).
Characteristics of active-learning instruction could include:
● Providing opportunities for students to share their ideas
and reasoning, either individually or with small groups of
peers
● Encouraging qualitative reasoning based on physics
concepts (which is discussed in Chapter 2)
● Encouraging construction and sense-making of physics
knowledge; for example, students are prompted to figure
things out for themselves

11998_Master4.indd 114 2/11/2020 [Link] PM

 Active Learning Strategies 115

● Providing opportunities for students to engage in the

process of “doing science”
● Providing opportunities for students to apply their
knowledge flexibly across multiple contexts (that is, to
practice “transferring” their knowledge to novel contexts)
● Incorporating formative assessment techniques to monitor
student learning and inform instruction
● Helping students organize content knowledge according to
some hierarchy (which is discussed in Chapter 3)
● Teaching metacognitive strategies to students; for example,
students reflect on their own reasoning and problem
solving practices
Teaching using active learning strategies means changing
common classroom “norms” to a culture where students are
willing to share their ideas even if those ideas end up being
incorrect (NRC, 2000). Students need to be willing to work
together as a team, take risks, and revise their thinking when
new information arises. Promoting these community-based atti-
tudes toward learning can prove challenging, especially for stu-
dents who have extensive experience in more individualistic,
competitive learning environments.
At the university level, physics courses can take a variety of
formats. Some institutions have “lecture” that is taught sepa-
rately from a laboratory portion of a course, whereas other insti-
tutions combine these two into an integrated lab-lecture format
called a studio format (Wilson, 1994), or Workshop Physics (Laws,
2004). Some institutions have a weekly recitation or discussion
session that is separate from lecture and lab whereas others
do not. Learning can also take place outside of formal class-
room environments as students complete homework, meet with
study groups or tutors, interact with online course management

11998_Master4.indd 115 2/11/2020 [Link] PM

 116 The Science of Learning Physics

systems or technology-enhanced resources (computer simula-

tions, etc.) As a result, research-based instructional strategies and
research-based curricula developed in physics tend to target a
particular classroom environment. Some of the most prevalent
RBIS are highlighted in the sections below, however a more com-
prehensive list of strategies and assessment tools is provided in
the resource PhysPort, https: / / [Link]/

5.1.2. Reforming the “traditional” lecture classroom

For instructors looking to “get their feet wet” with active learning,
most choose to make small changes to traditional lectures to
provide opportunities for students to pause and reflect on their
learning during class and/or share their reasoning by discussing
physics concepts with peers during class (Henderson & Dancy,
2009). One example is the formative assessment technique called
think-pair-share (Keeley, 2008; Lyman, 1981). Students are pre-
sented a question by the instructor and are given time to think
about it on their own first, then they pair up with another stu-
dent to discuss their thinking (see sample problems at the end
of Chapter 3 for questions that work well in this context). Some
pairs might be selected to share their ideas with the class as a
whole. This general strategy was more formally introduced
to the physics community in Eric Mazur’s 1997 book on Peer
Instruction, although it was already in use without that label, see
(Dufresne et al., 1996). In early versions of Peer Instruction, stu-
dents complete a pre-class reading and answer questions about
the reading using a method similar to Just in Time Teaching
(Novak et al., 1999). The instructor uses those responses to
design challenging multiple choice questions called ConcepTests
to present during the whole-class lecture period. These questions
typically focus on physics concepts and some of the options

11998_Master4.indd 116 2/11/2020 [Link] PM

 Active Learning Strategies 117

may include “distractor” choices based on common student dif-

ficulties. The class goes something like this: After 7–10 minutes
of lecture, the instructor stops and presents a conceptual ques-
tion. Students first reflect on the question and commit to an indi-
vidual answer (either electronically or using flash cards) and
the instructor reviews and/or displays the response distribution.
Then students are given 1–2 minutes to discuss their ideas with
peer(s) and revote on an answer choice. Based on the responses,
the instructor decides whether more explanation is needed or
whether to move on to the next concept. This process typically
takes 5–8 minutes, so by repeating the steps a one-hour lecture
is broken into approximately four 15-minute chunks of time. To
optimize interactions, Mazur suggests that the initial propor-
tion of students answering correctly should be between 35% and
70% (Crouch & Mazur, 2001; Redish, 2003), however other studies
have shown that peer discussion enhances understanding even
when no one in the peer group knows the correct answer (Smith
et al., 2009); nothing is better at getting students’ attention than
when the entire class misses a seemingly easy conceptual ques-
tion, thereby creating an opportunity for learning. Conceptual
understanding is also reinforced in other aspects of the course
by including conceptual multiple choice questions on exams.
One advantage of the Peer Instruction method is that the
instructor can see an “anonymous” (or not so anonymous) view
of what the whole class thinks, rather than getting a limited view
of students’ understanding based on verbal replies to a question
from a few students. The class can then be better tailored to stu-
dents’ needs. An advantage for students is that they are more
engaged during class and get immediate feedback on their level
of understanding of course material. There is also support for the
method from hundreds of research publications from a variety

11998_Master4.indd 117 2/11/2020 [Link] PM

 118 The Science of Learning Physics

of STEM disciplines. Most of these show higher gains on diag-

nostic concept inventories (Crouch & Mazur, 2001; Freeman et al.,
2014) or positive effects on students’ attitudes toward learning
physics (Zhang et al., 2017), however the effects on quantitative
problem solving are less clear (McDaniel et al., 2016). Some chal-
lenges with the method include time management decisions on
the part of the instructor. With less time spent on traditional lec-
ture, the instructor has to decide whether to only discuss a por-
tion of material in class and expect students to learn the rest on
their own through readings, recitation sessions, and homework;
or to decrease the number of topics covered in a course (Crouch
& Mazur, 2001). One approach developed at the University of
Illinois and now commercially available and used in many uni-
versities across the country (see [Link]) “out-
sources” presentation of basic information/concepts to 15-minute
web-based, interactive “pre-lectures” that students view prior
to coming to class, which allows the instructor to refine under-
standing rather than present new material; this method does
not reduce the amount of material covered in a typical intro-
ductory course. Peer Instruction places higher emphasis on con-
cepts rather than quantitative problem solving, so students must
learn those skills during other aspects of the course (such as in
recitation sections). In addition, students vary in their reactions
to active learning methods and some may be resistant to non-
traditional methods (we discuss barriers to reformed instruction
at the end of this chapter).
Another example of a lecture-based instructional strategy
with similar features is Interactive Lecture Demonstrations or
ILDs (Sokoloff & Thornton, 1997, 2004). Instead of presenting
students with a conceptual based multiple choice question, stu-
dents are shown the equipment for a demonstration and asked

11998_Master4.indd 118 2/11/2020 [Link] PM

 Active Learning Strategies 119

to make a prediction (such as predicting the shape of a force

vs time graph). After making an individual prediction, stu-
dents engage in a discussion with neighboring students and the
instructor elicits responses from the class. Students record their
final predictions and the instructor carries out the demonstra-
tion with measurements. Then the instructor facilitates a class
discussion about the results, and how to extend those results to
different physical situations based on the same concepts. This
method also shares similarities with the formative assessment
technique called P-O-E for predict, observe, and explain (Keeley,
2008). Similar to other RBIS, Interactive Lecture Demonstrations
have shown higher gains on concept inventories than tradi-
tional instruction (Sokoloff & Thornton, 1997, 2004). Another
study found that having students predict and discuss demonstra-
tions had a greater effect on performance on a free response test
compared to students who just observed the demonstration and
heard the teacher’s explanation (Crouch et al., 2004).
Other general strategies exist for making lectures more
interactive, referred to as formative assessment classroom tech-
niques or FACTs (Keeley, 2008). One of the most common exam-
ples is the think-pair-share strategy discussed above. Additional
examples include: Pausing during class for students to reflect
on their learning and write something down (such as a mud-
diest point, point of most significance in the lesson, or general
minute paper), creating concept maps, KWL chart (students write
down what they Know, Want to know, and what they’ve Learned),
or incorporating mini whiteboards as collaboration tools for stu-
dents. Students could also collaborate on solving physics prob-
lems during lecture (this is discussed further in the next section
with regard to recitation and discussion sessions).

11998_Master4.indd 119 2/11/2020 [Link] PM

 120 The Science of Learning Physics

5.1.3. Reforming the traditional laboratory or recitation

experience
Even though laboratories are somewhat active in that students
interact with equipment and each other, traditional laboratory
classes are referred to as “cookbook-style” because they provide
step-by-step instructions for students to follow. Efforts to reform
the physics laboratory experience to be more engaging typically
focus on the lab activities and curricula themselves, reforming
instructions to incorporate more decision making on the part
of students and encouraging them to figure things out with
less guidance. One challenge with implementing new laboratory
activities is that additional training might be required for fac-
ulty to learn about the RBIS or the technology associated with
the curriculum. Although computer-based tools for data collec-
tion and analysis have become quite common in physics class-
rooms, it is possible that a particular institution might have to
make a financial investment to maintain equipment and software
licenses. For institutions with undergraduate and/or graduate
teaching assistants, training might also be necessary for them
to carry out instruction as intended by curriculum developers.
Some example laboratory curricula are highlighted below.
The Investigative Science Learning Environment (ISLE) encour-
ages students to engage in a cycle of learning that is mod-
eled after the way physicists create knowledge (Etkina & Van
Heuvelen, 2001, 2006). Students encounter an interesting physical
phenomenon, gather data and propose multiple explanations,
then experimentally test those explanations. Finally, they apply
their ideas to real-world problems. ISLE is somewhat unique in
its explicit focus on fostering the development of scientific rea-
soning abilities. The learning system has been adapted into mul-
tiple textbooks (see Van Heuvelen & Etkina, 2006) and a set of

11998_Master4.indd 120 2/11/2020 [Link] PM

 Active Learning Strategies 121

video-based experiments further described at http: / /islephysics.

net/pt3
The RealTime Physics (RTP) laboratories include a mixture of
conceptual activities and laboratory experiments (Sokoloff et al.,
2004). Investigations typically incorporate technology tools for
data collection and analysis, such as sensors that dynamically
display graphs of data in “real time.” The research-based activi-
ties are designed to address common preconceptions students
might have about physical situations, encourage students to con-
struct their own models of phenomena, and encourage group
work and discussion. Each laboratory includes a set of pre-lab
questions and a post-lab assignment to reinforce concepts and
skills. Workshop Physics is an activity-based curriculum designed
for introductory calculus-based physics courses (Laws, 2004),
although it shares some of these same characteristics with RTP
labs. Students make predictions, conduct qualitative observations
and quantitative experiments, construct explanations, and use
computer-based tools to collect and analyze data and/or develop
mathematical models of phenomena. The use of video-based
methods for data collection and analysis are also becoming
increasingly common in laboratory curricula, as are interactive
simulations such as Physlets (Christian & Belloni, 2003) or PhET
(https:// [Link]/ ).
For physics courses that include a recitation or discussion
session, the emphasis is typically on problem solving or an exten-
sion of the lecture but with a smaller group of students. One
traditional view of the recitation was that it provided oppor-
tunities for students to ask questions about homework, or for
the TA/instructor to solve additional example problems for stu-
dents. Efforts to make recitation more “active” typically incor-
porate small groups and alternative types of problems and/or

11998_Master4.indd 121 2/11/2020 [Link] PM

 122 The Science of Learning Physics

conceptual worksheets. A discussion-based strategy developed at

the University of Minnesota is Cooperative Group Problem Solving
in conjunction with the use of context-rich problems (Heller
et al., 1992; Heller & Hollabaugh, 1992) where it was reported that
the quality of problem solutions produced by cooperative groups
of students was significantly better than the quality of solutions
from individuals in the group on matched problems (even the
strongest student in the group). Problem solving instructional
methods are reviewed in more detail in Chapter 4.
Another well-known strategy for recitation sessions is referred
to generally as tutorials, which are described by the developers as
guided-inquiry worksheets where students confront and resolve
conceptual difficulties (McDermott & Schaffer, 2002). Questions
in a tutorial workbook guide students to reason about concepts
and apply their knowledge to real-world situations. They also
obtain practice moving flexibly between multiple representations
of information (equations, graphs, diagrams, verbal descriptions,
etc.) Students complete pre-tests, homework assignments, and
post-tests in addition to the worksheets. Several adaptations of
tutorial curricula have been created, see PhysPort (https: / / www.
[Link]/[Link]) for a comprehensive list.
There have been several efforts in the field of physics to re
­
design classroom spaces and courses to combine lectures and lab-
oratory experiences. Class periods tend to be longer (2 to 2.5 hours)
and the classrooms are designed to promote cooperative group
work while students work on a variety of tasks, such as hands-on
activities, laboratories, conceptual questions, simulations, or
problem solving. An early implementation of this is referred to
as a Studio classroom (Wilson, 1994). The Workshop Physics curric-
ulum has been designed for the integrated lab-lecture classroom,
however the authors claim it can also be adapted to other settings

11998_Master4.indd 122 2/11/2020 [Link] PM

 Active Learning Strategies 123

(Laws, 2004). SCALE-UP (Student-Centered Active Learning for

Undergraduate Programs) is a similarly integrated learning en-
vironment which incorporates a variety of PER-based learning
activities but for a larger class of students (Beichner, 2008; Gaffney
et al., 2008). A key challenge with integrated classrooms is that
it could require substantial restructuring of the physical space
and time blocks of physics courses, which presents a financial
and/or logistical challenge. This is cited as a common barrier to
reformed instruction, as discussed below.

5.1.4. Reforming the “out-of-class” experience

Not all learning takes place during structured class time, and
most courses include an expectation that students will complete
at least some of the required course activities at other times such
as reading the textbook, solving homework questions, or inter-
acting with online resources. The advantage of these resources
is that students can be engaged in learning activities at a time
that might be more convenient for their schedule, like nights
and weekends. However, the challenge with “out-of-class” activ-
ities is that students might not have access to help resources
or personalized tutoring, and it is impossible to monitor all of
their interactions. Out of class experiences could actually rein-
force undesirable learning behaviors, like over-reliance on peers
or web searches for information. Although online homework and
tutoring systems are becoming increasingly sophisticated, many
still emphasize getting the correct numerical result and may not
provide much feedback on a student’s problem solving process.
Computer coaches will be reviewed further along with problem
solving in Chapter 4. Additional general strategies are reviewed
below.

11998_Master4.indd 123 2/11/2020 [Link] PM

 124 The Science of Learning Physics

One set of evidence-based practices is focused on gathering

student responses that the instructor can use to plan instruction
and respond to students’ needs. For example, the strategy Just-in-
time Teaching (JiTT) includes pre-class questions that students
answer online prior to class, and then instructors adapt and
fashion their instruction based on those responses (Novak et al.,
1999). Some learning management systems include communica-
tion tools like discussion posts so instructors and students can
ask and respond to questions. Students could also keep a reflec-
tive journal about what they learned and what questions they
have about the material, which the instructor and/or teaching
assistants can summarize and address with the whole class.
Another set of practices involve moving most of the basic
content delivery from face-to-face classes to be online content
viewed before class, sometimes referred to as a flipped class-
room (see [Link] for an example). In that situa-
tion, more in-class time is devoted to collaborative activities like
conceptual questions, or to refining students’ understanding and
developing problem solving skills. For example, students might
complete online multimedia learning pre-lectures in conjunction
with a course (Stelzer et al., 2010). These pre-lectures substitute
for material that is traditionally communicated in lecture, and/
or they could include additional things, like worked examples or
quizzes for students to check their understanding; interestingly,
students learn significantly more from interactive pre-lectures
than from textbook presentations of the same material (Stelzer
et al., 2009). Another advantage of video-based materials or ani-
mations is that students can periodically pause and rewind to
revisit material while viewing it, letting them control the pace
of instruction, and they can also revisit material again at a later
time. Web-based modules like this with narrated animations have

11998_Master4.indd 124 2/11/2020 [Link] PM

 Active Learning Strategies 125

been shown to provide some increases in students’ exam per-

formance, but also improve students’ attitudes toward a course
(Sadaghiani, 2011; Stelzer et al., 2010).

5.1.5. Barriers to active learning instruction

According to a large-scale web survey of physics faculty teaching
in the United States, nearly all faculty are familiar with one or
more RBIS (or EBIPs) and several have implemented one at some
point (Henderson & Dancy, 2007, 2009; Henderson et al., 2012).
Approximately half indicate they are currently using at least
one RBIS. Oftentimes the RBIS is substantially modified during
implementation, which could impact the outcomes of instruc-
tion; nearly a third of adopters end up abandoning the RBIS alto-
gether. When faculty discontinue using a particular RBIS, some
of the reasons listed were that it did not seem to work (they did
not see knowledge gains among students), it took up too much
in-class time, students had a negative reaction to the approach, it
was too cost-prohibitive to continue, and/or it was not supported
by their department (Henderson & Dancy, 2009). Faculty might
also be abandoning one approach in favor of trying out a dif-
ferent evidence-based instructional practice.
When a subset of faculty were interviewed and asked to elab-
orate on barriers to implementation, nearly all of them men-
tioned issues related to departmental norms and expectations
of content coverage, lack of time to learn about and implement
new approaches, and challenges with student attitudes or resis-
tance toward aspects of the course (Henderson & Dancy, 2007).
A smaller number cited structural constraints like the class size,
room layout, and time structure of the course. In a separate anal-
ysis of this survey data, the use of innovative teaching strate-
gies was found to be independent of the faculty member’s age,

11998_Master4.indd 125 2/11/2020 [Link] PM

 126 The Science of Learning Physics

type of institution, research productivity, and percentage of job

related to teaching (Henderson et al., 2012). Although faculty
might hold perceptions that it is more challenging to implement
innovative teaching at Ph.D.-granting institutions and it could
potentially hurt their research productivity, this was not sup-
ported by research findings.
Students are also victims of their own perceptions—in fact,
Chapter 6 discusses how students’ perception of what works
best for them in terms of learning run counter to measured out-
comes. A recent study found that in active learning classes, stu-
dents feel like they learned less, when in fact, they learned more
than in a traditional style lecture class (Deslauriers et al., 2019).
This could lead to lower rates of student participation, and even
lower evaluations at the end of a course (Gaffney & Gaffney,
2016). This underscores the need for faculty to explicitly talk
about why the course is structured in a particular way and the
values of active learning with their students. It is important to
acknowledge that it might not always feel like they are learning
more, but productively struggling with material leads to greater
learning gains. A major exception to these findings is the intro-
ductory physics sequence for science and engineering majors at
the University of Illinois, where the implementation of web-based
pre-lectures, use of class time largely devoted to Peer Instruction,
and a reduction of lecture time by 50 minutes per week, resulted
in significantly improved student attitudes, better exam scores,
and perceptions by students that the course was easier and that
the lecture portion was more valuable when compared to when
the course was taught traditionally (Stelzer et al., 2010).

11998_Master4.indd 126 2/11/2020 [Link] PM

 Active Learning Strategies 127

5.1.6. The use of Communities of Practice to encourage

reformed instruction
Recent work investigating the use of EBIPs at the University
of Illinois suggests a structure in which EBIPs can spread and
thrive, which is in stark contrast to the barriers discussed above.
The multi-year process leading to the spread of EBIPs at Illinois
is documented in three publications (Herman et al., 2018;
Ma et al., 2018; Mestre et al., 2019). The catalyst for the reform
was a grant from the National Science Foundation’s Widening
Implementation and Demonstration of Evidence-Based Reforms
(WIDER) program. The WIDER project is not intended to pro-
mote the discovery of new knowledge but rather to explore adop-
tion strategies for evidence-based reforms. Thirteen science and
engineering departments across two colleges participated in the
project. Each reform was centered around a team of faculty from
the department consisting of faculty who agreed to work col-
laboratively on identifying and adopting an EBIP that met the
context and need for an introductory course, to research its effec-
tiveness once implemented, and to use the EBIP long-term by
having different team members teach the introductory course
over the years. Crucially, one WIDER principal investigator was
embedded in each team and served as a mentor/resource to help
the team identify and adopt an appropriate EBIP; the principal
investigators possessed substantial pedagogical expertise that
they shared with the departmental teams as needed.
A Communities of Practice (CoP) model (Wenger et al., 2002)
was used in the reform teams, with team members and mentor
working collaboratively toward the common goal of adopting
an EBIP to reform a large introductory course. The EBIP
selected by each CoP was not prescribed but rather emerged
from lengthy discussions over many weekly meetings held by

11998_Master4.indd 127 2/11/2020 [Link] PM

 128 The Science of Learning Physics

the CoP, and was jointly “owned” by the CoP members and the
department (department chair buy-in was required). Four EBIPs
were predominantly adopted, namely flipped classrooms with
pre-lectures, use of active learning with clickers, problem-based
learning, and context-rich collaborative problem solving. In addi-
tion, a “teach the way you do research” mantra permeated the
project, not only because research is something that profes-
sors at a research university understand well, but also to adver-
tise to participating faculty that research into the effectiveness
of a reform was the only way to determine its efficacy; further,
this mantra served to dispense with proposed reforms based on
individual team member’s “hunches” on what worked best—the
“evidence” in Evidence-Based Reforms dictated that any reform
adopted had to have evidence that it worked. Various measures
used in evaluating student performance indicated improved
performance and higher student satisfaction following the
reforms.
By the end of the 4-year project, the EBIPs adopted in the
large “gateway” introductory science and engineering courses
were impacting over 17,000 students per year. Additional research
conducted by the WIDER principal investigators yielded some
very interesting findings. In weekly meetings held by the prin-
cipal investigators to discuss progress of the CoPs in which they
were embedded, it was becoming clear that certain EBIPs were
spreading across departments independent of the team’s efforts.
To study the spread of innovations, social network analyses were
performed on survey data from all CoP faculty and it was found
that the embedded mentors played a crucial role in the spread
of innovations; the mentors often discussed what they learned
in their weekly meetings about other reforms with the CoPs in
which they were embedded, thereby facilitating dissemination.

11998_Master4.indd 128 2/11/2020 [Link] PM

 Active Learning Strategies 129

There were several other indicators of both structural and cul-

tural change:
● Faculty in the CoPs began publishing about their educa-
tional innovations. Over 50 peer reviewed publications
emerged from 64 unique authors, 38 of which had never
published education-related work. This was accomplished
while these science and engineering faculty members main-
tained their own active research programs in their respec-
tive disciplines.
● CoP faculty submitted 21 external STEM education grant
proposals (including WIDER itself), totaling over $22 million,
with 8 being funded for a total of $6.0 million. There were
46 unique principal investigators in these proposals with 33
being STEM faculty (29 had never submitted an education-
related proposal) and 13 being education faculty. Thus
STEM faculty not only enjoyed their forays into educational
reform, but enjoyed their collaborations with education
faculty as well.
● Of the eleven CoPs, eight continued to operate after the
WIDER project was over, suggesting that for the STEM
faculty involved their education-related collaborations were
treated similarly to their disciplinary collaborations.

5.2. What are the implications of research on active

learning for instruction?
We offer the following implications of the research reviewed
above for instruction:
● Teaching by telling is ineffective for long term learning
-
and transfer. The research reviewed in this chapter
suggests that learning is not a spectator sport, as eloquently
discussed in the National Research Council’s publication,

11998_Master4.indd 129 2/11/2020 [Link] PM

 130 The Science of Learning Physics

How People Learn (2000). Passively listening to a lecture,

no matter how polished, may provide the listener with an
illusion of understanding but does little in terms of long
term retention and ability to transfer the ideas covered
in novel contexts. Instructors need to design learning
experiences to be more “active” from both a cognitive and
physical sense—traditional passive lectures result in lower
student performance and higher failure rates (Freeman
et al., 2014).
● Traditional labs are not enough. Some instructors might
be thinking that students already participate in “active
learning” instruction by doing hands-on laboratory activi-
ties. Research has shown that just being physically active
with equipment does not ensure learning—students might
be able to follow step-by-step instructions and fill in the
blanks for “traditional” lab activities, but walk away with
little understanding about what they did in an experiment
and why, and about how to design experiments. Curricula
need to make certain that students are cognitively engaged
and thinking deeply about the material. Ideally, EBIPs
should permeate all aspects of a course (lecture, laborato-
ries, recitation/discussion sessions, assessments, etc.) and
mesh together seamlessly to enhance student learning.
● Assessment is critical. At the very beginning of this
book, we made a case for the importance of approaching
classroom instruction in the same way we approach
research—by systematically conducting investigations. If
you are planning to implement one or more EBIPs, also
devote some time to planning out how you will evaluate
the effectiveness of instructional reforms. The Physics
Education Research community has a wide selection of
diagnostic tests and resources for assessing instruction.

11998_Master4.indd 130 2/11/2020 [Link] PM

 Active Learning Strategies 131

● Context is important. Implementing evidence-based in-

structional practices takes significant effort and buy-in
from instructors, and institutional barriers can quickly
derail efforts. These barriers can include anything from a
lack of support from administrators and/or colleagues, lack
of facilities or financial resources and student resistance to
reformed instruction.

5.3. Examples of teaching interventions based on

learning research
Physics students could benefit from increased opportunities to
discuss physics concepts with their peers during class, conduct
meaningful laboratory experiments, and engage in practices to
monitor their own learning. Some suggested instructional strate-
gies are provided below.
● Implement peer discussions about conceptually chal­
lenging questions. Traditional physics instruction tends to
focus on quantitative problem solving with little attention
to students’ qualitative understanding of concepts. Many
EBIPs include opportunities for student–student discourse
focused on students’ ideas at multiple points in instruction.
Earlier sections highlighted specific examples of this, such
as Peer Instruction (Mazur, 1997) and Interactive Lecture
Demonstrations (Sokoloff & Thornton, 1997, 2004). Also,
simply having students talk to each other does not comprise
an RBIS; having students discuss meaningful questions/
problems that are within their reach is crucial to lasting
learning, as well as instructor guidance. Chapter 2 included
some examples of “clicker” questions to promote peer dis-
cussions and conceptual understanding.

11998_Master4.indd 131 2/11/2020 [Link] PM

 132 The Science of Learning Physics

● Implement structured cooperative group work. In con-

junction with the first bullet, it is important to carefully
structure group activities to promote interactivity and
peer learning (Johnson et al., 1998). Whether students
are discussing conceptual questions, conducting a labo-
ratory experiment, or solving problems, there are class-
room structures that can facilitate positive interactions.
For example, it is important to design activities that blend
positive interdependence (we sink or swim together) with
individual accountability, foster social skills like communi-
cation and conflict management, and rotate roles within a
group. Other research studies have found better results in
problem solving with mixed ability groups and pairing up
2–3 women in a group instead of two men and one woman
(Heller & Hollabaugh, 1992).
● Structure laboratory experiences to reflect the processes
of “doing science”. Although laboratory experiments have
long been a part of physics instruction, the notion of
engaging students in more authentic scientific practices
started in the K-12 realm and is beginning to make its way
into postsecondary education. Students benefit from prac-
ticing scientific skills like developing and using models,
planning and carrying out investigations, analyzing and
interpreting data, and constructing explanations (NRC,
2012a). This chapter reviewed several laboratory curricula
that break the traditional format of “cookbook labs” and
provide increased opportunities for students to make deci-
sions and figure things out for themselves.
● Implement formative assessment classroom techniques
to monitor student learning and inform instruction.
Research indicates how prior knowledge plays a key role in
the process of learning (NRC, 2000). Instructors need to be
aware of students’ pre-existing ideas and design lessons to

11998_Master4.indd 132 2/11/2020 [Link] PM

 Active Learning Strategies 133

elicit and build on scientifically correct ideas, and to help

remedy stubborn misconceptions. This chapter explored
a variety of strategies for making students’ learning more
visible before, during, and after instruction using formative
assessment classroom techniques (Keeley, 2008). Some of
these techniques are enhanced by using technology like
classroom response systems or online assessment tools.
More strategies for using assessment in the service of
learning are provided in Chapter 7.
● Implement diagnostic assessments to evaluate students’
knowledge, skills, and attitudes. If you are currently using
or planning to use reformed instructional strategies, it is
helpful to collect evidence of student learning and attitudes
to better evaluate the effectiveness of the implementation
(and/or track multiple reform efforts). The Physics Education
Research community has developed a host of assessment
resources and tools for evaluating students’ conceptual
knowledge, attitudes and beliefs, scientific reasoning skills,
and problem solving skills (see https: / / [Link]/
[Link]). As Chapter 6 will discuss, effective learn-
ing strategies are counterintuitive for many students who
have erroneous notions of what works best for them.

References
Bandura, A. (1986). Social Foundations of Thought and Action: A Social
Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall.
Beichner, R.J. (2008). The SCALE-UP project: A student-centered active
learning environment for undergraduate programs. Paper presented at
the National Research Council’s Workshop on Evidence on Promising
Practices in Undergraduate Science, Technology, Engineering, and
Mathematics Education, Washington, DC. http:// [Link]-
[Link]/DBASSE/BOSE/DBASSE_071087

11998_Master4.indd 133 2/11/2020 [Link] PM

 134 The Science of Learning Physics

Bruner, J. S. (1961). The act of discovery. Harvard Educational Review,

31(1), 21–32.
Chappuis, J. (2015). Seven Strategies of Assessment for Learning, 2nd Ed.
New York: Pearson Education.
Christian, W. & Belloni, M. (2003). Physlet Physics: Interactive Illustrations,
Explorations, and Problems for Introductory Physics. Upper Saddle
River, NJ: Prentice-Hall.
Crouch, C.H., Fagan, A.P., Callan, J.P. & Mazur, E. (2004). Classroom
demonstrations: Learning tools or entertainment? American Journal
of Physics, 72(6), 835–838. https:// [Link]/10.1119/1.1707018
Crouch, C.H. & Mazur, E. (2001). Peer Instruction: Ten years of experience
and results. American Journal of Physics, 69(9), 970–977. https:// doi.
org/10.1119/1.1374249
Cummings, K. (2011). A developmental history of Physics Education
Research. Commissioned paper by the National Academies Board on
Science Education on Status, Contributions, and Future Directions of
Discipline-Based Education Research (DBER), Washington, DC. [http://
[Link]/DBASSE/BOSE/DBASSE_071087]
Dancy, M. & Henderson, C. (2007). Framework for articulating instruc-
tional practices and conceptions. Physical Review Special Topics –
Physics Education Research, 3(010103). https:// [Link]/10.1103/Phys
RevSTPER.3.010103
Deslauriers, L., McCarty, L.S., Miller, K., Callaghan, K. & Kestin, G. (2019).
Measuring actual learning versus feeling of learning in response
to being actively engaged in the classroom. Proceedings of the
National Academy of Sciences of the United States of America, 116(39),
19251–19257. https:// [Link]/10.1073/pnas.1821936116
Docktor, J.L. & Mestre, J.P. (2014). Synthesis of discipline-based
education research in physics. Physical Review Special Topics –
Physics Education Research, 10(020119). https:// [Link]/10.1103/
PhysRevSTPER.10.020119
Dufresne, R.J., Gerace, W.J., Leonard, W.J., Mestre, J.P. & Wenk, L. (1996).
Classtalk: A classroom communication system for active learning,

11998_Master4.indd 134 2/11/2020 [Link] PM

 Active Learning Strategies 135

Journal of Computing in Higher Education, 7(2), 3–47. http:// [Link].

org/10.1007/BF02948592
Etkina, E. & Van Heuvelen, A. (2001). Investigative Science Learning
Environment: Using the processes of science and cognitive strategies
to learn physics. Proceedings of the 2001 Physics Education Research
Conference. Rochester, NY, pp. 17–21.
Etkina, E. & Van Heuvelen, A. (2006). The Physics Active Learning Guide.
San Francisco, CA: Addison Wesley.
Freeman, S., Eddy, S.L., McDonough, M., Smith, M.K., Okoroafor, N., Jordt,
H. & Wenderoth, M.P. (2014). Active learning increases student perfor-
mance in science, engineering, and mathematics. Proceedings of the
National Academy of Sciences of the United States of America, 111(23),
8410–8415. [Link]
Gaffney, J.D.H. & Gaffney, A.L.H. (2016). Student satisfaction in interactive

­­
engagement-based physics classes. Physical Review Physics Educa-
tion Research, 12(020125). https:// [Link]/10.1103/PhysRevPhys
EducRes.12.020125
Gaffney, J.D.H., Richards, E., Bridget Kustusch, M. Ding, L. & Beichner,
R.J. (2008). Scaling up educational reform. Journal of College Science
Teaching, 37(5), 48–53.
Hake, R.R. (1998). Interactive-engagement versus traditional methods:
A six-thousand student survey of mechanics test data for introduc-
tory physics courses. American Journal of Physics, 66(1), 64–74. http://
[Link]/10.1119/1.18809
Heller, P., Keith, R. & Anderson, S. (1992). Teaching problem solving
through cooperative grouping. Part 1:Group versus individual
problem solving. American Journal of Physics, 60(7), 627–636. https://
[Link]/10.1119/1.17117
Heller, P. & Hollabaugh, M. (1992). Teaching problem solving through
cooperative grouping. Part 2: Designing problems and structuring
groups. American Journal Physics, 60(7), 637–644. https:// [Link]/
10.1119/1.17118
Henderson, C. & Dancy, M.H. (2007). Barriers to the use of research-
based instructional strategies: The influence of both individual

11998_Master4.indd 135 2/11/2020 [Link] PM

 136 The Science of Learning Physics

and situational characteristics. Physical Review Special Topics –

Physics Education Research, 3(020102). https:// [Link]/10.1103/Phys
RevSTPER.3.020102
Henderson, C. & Dancy, M.H. (2009). Impact of physics education research
on the teaching of introductory quantitative physics in the United
States. Physical Review Special Topics – Physics Education Research,
5(020107). [Link]
Henderson, C., Dancy, M. & Niewiadomska-Bugaj, M. (2012). Use of
research-based instructional strategies in physics: Where do faculty
leave the innovation-decision process? Physical Review Special Topics
– Physics Education Research, 8(020104). https:// [Link]/10.1103/Phys
RevSTPER.8.020104
Herman, G.L., Greene, J.C., Hahn, L.D., Mestre, J.P., West, M. & Tomkin, J. H.
(2018). Changing the teaching culture in introductory STEM courses
at a large research university. Journal of College Science Teaching,
47(6), 32–38.
Johnson, D.W., Johnson, R.T. & Holubec, E.J. (1998). Cooperation in the
Classroom 7th Ed. Edina, MN: Interaction Book Company.
Keeley, P.D. (2008). Science Formative Assessment: 75 Practical Strategies
for Linking Assessment, Instruction, and Learning. NSTA Press.
Laws, P. (2004). Workshop Physics Activity Guide Vol. 1-4. New York: Wiley
and Sons.
Lyman, F. (1981). The responsive classroom discussion: The inclusion of
all students. In A.S. Anderson (Ed.), Mainstreaming Digest, pp. 109–113.
College Park, MD: University of Maryland College of Education.
Ma, S., Herman, G.L., West, M., Tomkin, J. & Mestre, J. (2019) Study-
ing STEM Faculty communities of practice through social network
analysis. The Journal of Higher Education, 90(5), 773–799. doi: 10.1080/
00221546.2018.1557100
Mazur, E. (1997). Peer Instruction: A User’s Manual. Upper Saddle River,
NJ: Prentice Hall.
McDaniel, M.A., Stoen, S.M., Frey, R.F., Markow, Z.E., Hynes, K.M., Zhao, J.
& Cahill, M.J. (2016). Dissociative conceptual and quantitative problem

11998_Master4.indd 136 2/11/2020 [Link] PM

 Active Learning Strategies 137

solving outcomes across interactive engagement and traditional

format introductory physics. Physical Review Physics Education
Research, 12(020141).
McDermott, L. Schaffer, P. & the Physics Education Group at the Univer-
sity of Washington (2002). Tutorials in Introductory Physics. Upper
Saddle River, NJ: Prentice Hall.
Meltzer, D.E. & Otero, V.K. (2015). A brief history of physics education in
the United States. American Journal of Physics, 83(5), 447–458. https://
[Link]/10.1119/1.4902397
Meltzer, D.E. & Thornton, R.K. (2012). Resource Letter ALIP-1: Active-
learning instruction in Physics. American Journal of Physics, 80(6),
478–496. [Link]
Mestre, J.P. (2001). Implications of research on learning for the educa

­
tion of prospective science and physics teachers. Physics Education,
36(44), 44–51. [Link]
36/1/308/pdf
Mestre, J.P., Herman, G.L., Tomkin, J.H. & West, M. (2019). Keep your friends
close and your colleagues nearby: The hidden ties that improve STEM
education. Change: The Magazine of Higher Learning, 51(1), 42–49.
Morphew, J.W. & Mestre, J.P. (2018). Exploring the connection between
problem solving and conceptual understanding in physics. Revista de
Enseñanza de la Física, 30(2), 75–85.
National Research Council (2000). How People Learn: Brain, Mind, Experi-
ence, and School: Expanded Edition. Washington, DC: The National
Academies Press. https:// [Link]/10.17226/9853
National Research Council (2012a). A Framework for K-12 Science
Education: Practices, Crosscutting Concepts, and Core Ideas. Washington,
DC: The National Academies Press. https:// [Link]/10.17226/13165
National Research Council (2012b). Discipline-Based Education Research:
Understanding and Improving Learning in Undergraduate Science and
Engineering. Washington, DC: The National Academies Press. https://
[Link]/10.17226/13362

11998_Master4.indd 137 2/11/2020 [Link] PM

 138 The Science of Learning Physics

National Research Council (2013). Adapting to a Changing World –

Challenges and Opportunities in Undergraduate Physics Education.
Washington, DC: The National Academies Press.
Novak, G.M., Patterson, E.T., Gavrin, A. D. & Christian, W. (1999). Just-in-
time Teaching: Blending Active Learning with Web Technology. Upper
Saddle River, NJ: Prentice Hall.
Redish, E.F. (2003). Teaching Physics with the Physics Suite. New York:
Wiley.
Sadaghiani, H.R. (2011). Using multimedia learning modules in a hybrid-
online course in electricity and magnetism. Physical Review
Special Topics – Physics Education Research, 7(010102). https:// doi.
org/10.1103/PhysRevSTPER.7.010102
Schwartz, D.L. & Bransford, J.D. (1998) A Time For Telling. Cognition and
Instruction, 16(4), 475–5223. doi: 10.1207/s1532690xci1604_4
Smith, M.K., Wood, W.B., Adams, W.K., Wieman, C., Knight, J.K., Guild, N. &
Su, T.T. (2009). Why peer discussion improves student performance on
in-class concept questions. Science, 323(5910), 122–124. http:// [Link].
org/10.1126/science.1165919
Sokoloff, D.R. & Thornton, R.K. (1997). Using interactive lecture demonstra-
tions to create an active learning environment. The Physics Teacher,
35(6), 340–347.
Sokoloff, D.R. & Thornton R.K. (2004). Interactive Lecture Demonstrations:
Active Learning In Introductory Physics. New York: John Wiley & Sons.
Sokoloff, D.R., Thornton, R.K. & Laws, P.W. (2004). RealTime Physics: Active
Learning Laboratories (2nd ed.) Module 1: Mechanics, Module 2: Heat
and Thermodynamics, Module 3: Electric Circuits, Module 4: Light and
Optics. Hoboken, NJ: John Wiley & Sons.
Stelzer, T., Brookes, D.T., Gladding, G. & Mestre, J.P. (2010). Impact of
multimedia learning modules on an introductory course on electricity
and magnetism. American Journal of Physics, 78(7), 755–759. https://
[Link]/10.1119/1.3369920
Stelzer, T., Gladding, G., Mestre, J.P. & Brookes, D.T. (2009). Comparing
the efficacy of multimedia modules with traditional textbooks for

11998_Master4.indd 138 2/11/2020 [Link] PM

 Active Learning Strategies 139

learning introductory physics content. American Journal of Physics,

77(2), 184–190.
Van Heuvelen, A., and Etkina, E. (2006). The Physics Active Learning Guide.
San Francisco, CA: Pearson, Addison Wesley.
Vygotsky, L.S. (1978). Mind in Society: The Development of the Higher
Psychological Processes. Cambridge, MA: The Harvard University Press.
(Originally published 1930, New York: Oxford University Press.)
Wenger, E., McDermott, R. & Snyder, W. M. (2002). Cultivating Communities
of Practice. Cambridge, MA: Harvard Business Press.
Wilson, J. (1994). The CUPLE physics studio. Physics Teacher, 32, 518–523.
[Link]
Zhang, P., Ding, L. & Mazur, E. (2017). Peer Instruction in introductory
physics: A method to bring about positive changes in students’ attitudes
and beliefs. Physical Review Physics Education Research, 13(010104),
1–9. [Link]

11998_Master4.indd 139 2/11/2020 [Link] PM

 b2530   International Strategic Relations and China’s National Security: World at the Crossroads

This page intentionally left blank

b2530_FM.indd 6 01-Sep-16 [Link] AM

 c hapter
6 Students’ Perceptions of Learning
and their Study Habits

6.1. What does research on students’ perceptions of

learning and studying tell us?
We as instructors have a direct influence on students during
those times that we are in contact with them, such as during
class and office hours. We also have some indirect influence on
students. For example, the tests we administer and the home-
work we assign send messages to students about what we think
is important in a course, and students interpret and act on those
messages since they want to do well in our courses. However,
most of the learning that students do takes place outside of
our direct influence. Students study on their own, take tests on
their own and (mostly) do homework on their own. During those
types of “learning episodes” students make judgments and deci-
sions about how to optimize learning. However, as the research
reviewed in this chapter will point out, those judgments and
decisions quite often run counter to research evidence on opti-
mizing learning in ways that result in long-term retention and
in ability to apply the learning to novel contexts (i.e., transfer).
The research we will review in this chapter largely comes from
cognitive and educational psychology and was not carried out
within a physics context, or a science context for that matter.
But the findings below have generalizability and carry important

141

11998_Master4.indd 141 2/11/2020 [Link] PM

 142 The Science of Learning Physics

implications for how to structure our courses to help students

optimize their learning. We will discuss those implications as well
as suggest ways of using those implications to students’ benefit.

6.1.1. Research on study strategies

We begin with a discussion of study habits used by students. The
findings that we will review are very nicely summarized in two
reports (Dunlosky et al., 2013; Pashler et al., 2007). These reports
reviewed and synthesized hundreds of studies in order to iden-
tify both students’ study habits and instructional techniques
that held promise for student learning. What will become clear is
that quite often students’ perception of effective study strategies
are not borne out by research findings, while other study habits
that are deemed ineffectual by students are in fact quite effec-
tive. There are also some “mixed” findings in that certain study
habits preferred by students are actually very effective but with
important caveats that render them mediocre at best.
Let us start by dispensing with some ineffective study tech-
niques. A very popular technique among students, but among
the least effective, is highlighting/underlining text. Many of us
who have visited our campus’ bookstore and browsed some of
the used textbooks for sale have noticed that most have high-
lighting, ranging from sparse to dense highlighting. Highlighting/
underlining is a “feel good” strategy that helps students build
familiarity with the material but does little to develop compe-
tence with the material. In physics, being able to apply concepts
in multiple contexts is key to success, and highlighting does little
to develop the needed skills. Two other techniques earning low
effectiveness for similar reasons are rereading and summarizing.
As with highlighting, both are good for helping students develop
familiarity with the material, which unfortunately students

11998_Master4.indd 142 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 143

confuse with competence (more on this in the next chapter on

assessment). When students study, they need to make some deci-
sion about when they have studied enough, and these three tech-
niques build considerable familiarity with the material leading
to shallow learning, and unfortunately, to students judging that
they are well-prepared; when students who use these techniques
walk into an exam, everything will look familiar, but that does
not translate into being able to apply concepts to solve problems.
Note a common feature of the three ineffective study strate-
gies described in the previous paragraph: They are easy to imple-
ment and require minimal effort. The more successful learning
strategies require more effort on the part of the learner—
physical effort in the case of sports, mental efforts in the case
of academic subjects. Robert Bjork, a psychologist at UCLA who
studies learning and memory, refers to creating “desirable diffi-
culties” in order to promote lasting learning that transfers (Bjork,
1994; Bjork & Bjork, 2014; Schmidt & Bjork, 1992). He argues
that adding a certain type of difficulty to tasks forces learners
to notice and store away in memory subtleties that would oth-
erwise have been missed. The term “desirable” is crucial in the
descriptor since simply adding difficulty to a task accomplishes
little other than frustrating the learner—for example, giving stu-
dents a very difficult physics problem to solve right after cov-
ering a major topic but before the student has had a chance to
“explore the space” and develop competencies. A desirable diffi-
culty is making a task harder when the learner possesses all the
tools needed to make progress. Study strategies that have been
shown to be effective contain desirable difficulties, although
some evidence exists that desirable difficulties differentially ben-
efit high-ability students more than low-ability students (Wenzel
& Reinhard, 2019). The notion of desirable difficulties is similar

11998_Master4.indd 143 2/11/2020 [Link] PM

 144 The Science of Learning Physics

to productive struggle in mathematics instruction (Hiebert &

Grouws, 2007). Rather than immediately rushing to help students
when they are unsure about how to proceed, giving them some
time to wrestle with a problem on their own while providing
minimal scaffolding can promote the development of grit and a
growth mindset.
One successful study strategy is distributed practice, or spaced
practice. Perhaps the best example of distributed practice is to
space studying for an exam over several sessions, rather than
“massed practice” which consists of doing all the studying at
once just before the exam. Note that massed practice, otherwise
known as cramming, seems to be the method preferred by stu-
dents when preparing for an exam. It should be pointed out that
cramming is not without its virtues, otherwise students would
have abandoned the practice long ago. Massed practice is very
valuable for retaining information for regurgitation right after
the cramming session—at the time of the exam. Students often
feel rewarded for their efforts following an “all-nighter” after
noting a respectable grade on a test. However, studies consis-
tently reveal that massed practice is very ineffective for learning
that lasts (Dunlosky et al., 2013; Pashler et al., 2007; Schmidt &
Bjork, 1992). Studies comparing groups of students, where one
group distributes study over time and one group masses study
(equating for total study time) find that the massed practice
group slightly outperforms the distributed practice group on a
test administered immediately after cramming, but on a delayed
test administered one week later, the pattern reverses, with those
using distributed practice significantly outperforming those
who cram. It seems that this finding should be plastered all over
university dorms, academic buildings and dining commons!
Granted, cramming strategies may not be chosen willingly by

11998_Master4.indd 144 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 145

students, but may be a default strategy. Students are pressed for

time, with deadlines looming in every course, and so studying at
the last minute is one way of coping with lack of organizational
skills. However, taking the “distributed practice medicine” seems
very advisable since in the long run, it is a time-saver for such
things as studying later for a cumulative final exam, for retaining
and transferring knowledge for later use in one’s major, and for
reaping the most benefit from a very expensive college educa-
tion. What makes adoption of this strategy difficult for students
is that they have empirical evidence that cramming works (their
performance in an exam immediately after cramming is respect-
able), although that evidence is incomplete and flawed.
Another effective study strategy with desirable difficulty is
interleaved practice. Interleaved practice consists of varying prac-
tice across different types of items or activities. In stark con-
trast is blocked practice, consisting of practicing the same type
of item or activity. To use a sports analogy, for someone learning
to play tennis, after being introduced to different strokes and
playing tennis for some period of time (e.g., weeks), the player
would show more improvement in their overall game, if in prac-
tice sessions, they varied the strokes they practiced; for example,
mixing forehands, backhands and serves in every practice ses-
sion. Studies have shown that there is more improvement with
this type of practice than spending the same practice time
blocking strokes, for example, practicing only backhands in one
session, only serves in another, and so on. In physics, as in just
about every academic discipline, instruction is geared around
blocked practice. The chapters in a textbook are arranged by
topics, with exercises and problems at the back of each chapter
dealing with the topic of that chapter. Homework assignments
are also structured as blocked practice, with each homework set

11998_Master4.indd 145 2/11/2020 [Link] PM

 146 The Science of Learning Physics

containing problems from the last chapter or lectures covered. A

student opening a homework assignment therefore knows that
the particular assignment will deal with energy or momentum
concepts, for example. Further, although exam content is partially
interleaved, dealing with the topics covered in the last three or so
weeks of a course, typical semester exams are blocked, covering a
small number of topics. With rare exceptions, the only time when
students are forced to perform interleaved practice is during a
cumulative final exam. The reader can likely predict the instruc-
tional strategies that we will recommend later in this chapter,
which would blend interleaved practice throughout a course,
adding desirable difficulties for students and some extra work for
the instructor in implementing this structural change. In short,
interleaved practice allows students to differentiate which con-
cepts apply to which contexts.
A form of interleaved practice is also recommended in the
report by Pashler et al. (2007) namely to interleave worked-out
examples with problem solving exercises. There is a substantial
literature on the use of worked examples as a means of teaching
problem solving ((Sweller & Cooper, 1985; Ward & Sweller, 1990;
Pashler et al., 2007) and references therein). Generally the dual
benefits of studying worked examples are that they exhibit cor-
rect paths to solutions, as opposed to students taking incor-
rect paths and thereby reinforcing an incorrect approach, and
that studying worked examples take less time than solving prob-
lems on one’s own—for example, a student can likely study three
worked examples in the same time as it might take them to solve
a single problem on their own. Student engagement in studying
worked examples is a confounding factor, however—reading
a worked example with all the steps shown is typically easy to
follow but the mental processing needed to follow a solution is

11998_Master4.indd 146 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 147

considerably less than that needed to internalize the solution

so that the learner can reproduce it. Therefore adding desirable
difficulties in studying worked examples, such as leaving out
steps or asking deep questions along the way, helps the student
to internalize the solution better than passive reading of the
worked example. Providing worked examples at the right level is
also important since a worked solution for too difficult a problem
may not be understood by a student, and a worked solution to
too easy a problem will bore the student and serve little purpose.
One virtue of interleaving worked examples with problem solving
exercises is that a student may become stuck when attempting to
solve a problem, then by reading an isomorphic worked example,
realize what was hampering progress in the problem that they
were attempting to solve.
Two additional effective study strategies discussed in the
Pashler et al. report are worthy of mention, although for reasons
we will discuss, are better categorized as effective instructional
strategies. The first is asking and answering deep questions. This
would be a great study strategy if only beginning students, who
are novices in physics, were able to ask and answer deep ques-
tions on their own. Previous research has shown, for example,
that when provided with a physics scenario students are inept at
posing meaningful problems from the scenario (Mestre, 2002).
Therefore the “asking” part of asking and answering deep ques-
tions needs to be provided by instructors. The first two com-
ponents of strategy writing as discussed in Chapter 3, namely
identifying the major principle(s) needed to solve a problem
and providing a justification for why the principle(s) apply in
that particular instance, are examples of asking/answering deep
questions.

11998_Master4.indd 147 2/11/2020 [Link] PM

 148 The Science of Learning Physics

The second instructional strategy is to combine abstract and

concrete representations of concepts. This is common practice in
teaching physics, although perhaps not done extensively enough.
When instructors and textbooks introduce a new abstract con-
cept, it is followed by illustrations of ways of applying the con-
cept to solve problems. Leaving out concrete examples would
render the concept a mere abstraction with no anchor to the
physical world. Equally important is to provide enough concrete
cases where the concept can be applied, as well as cases where it
does not apply in order to promote transfer (i.e., the flexible and
appropriate application of the concept, see, for example, (Mestre,
2005)). Combining abstraction with concreteness, together with
appropriate and inappropriate contexts in which to apply a con-
cept, helps to build and refine a students’ knowledge base in
memory. For example, suppose that after introducing conser-
vation of mechanical energy, an instructor presents a single
sample problem to find the speed of a block at the bottom of a
curved ramp. Recall from Chapter 3 that novices tend to focus
on surface attributes, not on deep structure. If this is all the stu-
dent sees, it is highly likely that s/he will assume that conserva-
tion of mechanical energy is applicable only to blocks on curved
ramps. Many of us have noticed that after covering conservation
of mechanical energy, giving students a problem where a block
is released from the top of a straight, frictionless ramp results in
them applying Newton’s Second Law since that situation is asso-
ciated with that concept, as opposed to the easier conservation of
mechanical energy.

11998_Master4.indd 148 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 149

6.1.2. What are the actual patterns of learning and

forgetting?
Psychologists have been working to develop a model of learning
and memory for at least one hundred years, but many of those
models were based on remembering nonsense syllables or
words and did not take place in the context of a classroom
(Ebbinghaus, 1913). In the past, the term “learning curve” typi-
cally referred to the observation that with a repetitive task, the
time to complete the task decreased according to a power law
function (Newell & Rosenbloom, 1980). Most graphical represen-
tations of classroom learning are based on either an exponen-
tial rise in learning that occurs as a result of instruction or an
“S-curve” modeled as a logistic sigmoid function (Mitchell, 1997).
In an S-curve model (see Fig. 6.1), acquisition of knowledge starts
with small steps, then becomes larger steps, and reduces to suc-
cessfully smaller ones as learning reaches some maximum or
limit on a topic. After the learning reaches a maximum then with

Assessment (Homework/Quiz/Exam)

Lecture/lab
Post-instruc�on
instruc�on begins

Pre-instruc�on

Fig. 6.1. A sample model of learning and forgetting for a course topic.

11998_Master4.indd 149 2/11/2020 [Link] PM

 150 The Science of Learning Physics

time a forgetting curve takes over as an exponential decay or

power law decay to some new value. A sample smooth graph
of this process is presented (Fig. 6.1), however actual student
learning might be much more stepwise. Time scales can vary
widely from hours to days and even years depending on the
context.
Heckler and Sayre (2010) examined students’ responses to
questions about electricity and magnetism at multiple time
points throughout an introductory physics course. They found
an increase in student performance once lecture and laboratory
instruction began on a particular topic, and it peaked approxi-
mately one week later when online homework was due. Most stu-
dents completed their online homework within two days of the
deadline and received immediate feedback on their responses.
After the homework deadline, performance on the questions
decreased rapidly within a matter of a few days after the deadline
(Sayre & Heckler, 2009). The shape of the learning and forgetting
curves varied by topic; sometimes there was a sharp exponential
or stepwise increase just prior to the assessment and sometimes
performance did not exhibit such a rapid decline after the assess-
ment. A study looking at Newton’s Third Law saw an increase
in learning during the unit on forces that declined during the
energy unit, but rebounded again once students learned about
momentum and the interactions that happen during collisions
(Sayre et al., 2012).
A study comparing an activity-based physics course to tradi-
tionally taught course observed different decay patterns in the
forgetting curve (Franklin et al., 2014). The researchers found
that both groups of students saw a nearly identical increase in
learning with instruction, however, the students in the tradition-
ally taught course saw a rapid decline in performance after the

11998_Master4.indd 150 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 151

exam, returning back to their pre-instruction levels after three

weeks. The students in the activity-based course maintained their
level of performance or even exceeded it as they learned new
topics. This helps to explain why courses reformed to include
active learning often exhibit higher overall pre-post learning
gains on a concept inventory test even though the short-term
learning might be similar to traditionally taught courses.
In addition to learning and forgetting, studies have also
found that the phenomenon of interference can impact learning
of a topic. For example, immediately after learning about elec-
tric force students can correctly determine the direction of a
force on a charged particle on an electric field, but incorrectly
apply these ideas to particles in magnetic fields (Scaife & Heckler,
2011). After learning about magnetic fields they correct their
ideas, but then incorrectly apply the force direction from a mag-
netic field to electric fields. Another study found that after stu-
dents learned about electric potential, this information interfered
with their ability to correctly answer questions about the vector
nature of electric fields (Sayre & Heckler, 2009). These examples
illustrate the complex nature of learning and patterns of learning
and decay in a course.
How can we structure a course to address the rapid decline in
memory that typically happens after an assessment (a so-called
brain dump or mind wipe)? The research reviewed in this section
suggests that research-based instructional strategies or instruc-
tion that promotes active learning (see Chapter 5) can improve
long-term retention of information. The use of categorization
tasks or concept mapping can help students see the “big picture”
of topics in a physics course and the conditions under which
those concepts and principles are applicable. In the problem
solving chapter (Chapter 4) we described the idea of synthesis

11998_Master4.indd 151 2/11/2020 [Link] PM

 152 The Science of Learning Physics

problems that combine the use of multiple principles separated

in the timescale of a course, however scaffolding was needed to
assist students with solving those problems. Other strategies
to improve retention include distributed practice rather than
massed practice/cramming and frequent testing for students
to receive feedback on their performance (see Chapter 7). In
addition, designing assessments that require students to apply
information from previous chapters or units can reinforce the
importance of retaining this information.

6.1.3. Students’ metacognition about studying and learning

Thus far we have discussed effective study (and instructional)
strategies emerging from hundreds of cognitive and educational
psychology studies outside of physics. We have not discussed the
decisions students make about studying. When students study,
they make internal decisions that impact learning. For example,
how does a student decide when they have studied enough,
or learned enough, to put studying aside? These types of deci-
sions fall under the broad category of metacognition, which
loosely defined means thinking about, and regulating, one’s
own thought processes. We next discuss students’ metacognition
about studying and learning, and as we will see, students often
fool themselves into thinking that they have studied or learned
enough only to do poorly in an exam they prepared for.
In studying, students must decide what to study, how to
study, and how long to study. Consciously or subconsciously stu-
dents perform metacognitive monitoring in order to decide what
they need to study, and when they can stop studying. Such moni-
toring evaluates one’s current knowledge against some set of cri-
teria imposed by the learner and the context. It is important to
note that these processes are largely subjective—for example,

11998_Master4.indd 152 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 153

we discussed earlier in this chapter how students often mistake

familiarity or fluency with the material with competence with
the material, and so if familiarity/fluency is an important cri-
terion to be satisfied, a student may stop studying long before
she or he evaluates competence with the material (e.g., by taking
an old test under exam conditions). Unfortunately, studies have
shown that the fluency with which an item is retrieved from
memory impacts learners’ metacognitive judgments of learning
(Ackerman & Zalmanov, 2012; Koriat & Ma’ayan, 2005). For
example, individuals predicting their ability to recall answers
from a general knowledge test will often judge those items they
answered fastest as most accurate whereas accuracy of recall
is higher for items that take longer to answer (Benjamin et al.,
1998). This is particularly relevant for physics courses where
common misconceptions are often fluently retrieved. A more
subtle distinction between two types of fluency should also be
mentioned. Fluency with which a problem solution is generated
by the student is desirable and aligned with the student’s study
goals. On the other hand, fluency in reading a worked solution
is misleading since the student is good at reading (and likely
following the solution), but it is not clear that the student would
be able to generate the solution on their own later.
In order to make a sound decision on when to stop studying,
the student’s estimate of their knowledge and skills must
closely match their actual level of knowledge and skills (Ariel
et al., 2009; Metcalf & Kornell, 2005). However, for many stu-
dents the study techniques that they employ are not conducive
to reach the mastery they desire. For example, students errone-
ously believe that blocked and massed practices are more benefi-
cial than interleaved and distributed practices (Schmidt & Bjork,
1992). Another strong erroneous belief by students is that adding

11998_Master4.indd 153 2/11/2020 [Link] PM

 154 The Science of Learning Physics

more study sessions is more effective at preparing for an exam

than taking practice tests similar to those they are preparing to
take (Kornell & Son, 2009), even though using a test to judge
learning is an objective and more accurate measure of learning
(see Chapter 7 for additional research on this). One would think
that exam performance would be the best indicator of whether
or not a student is using successful study habits, but unfortu-
nately research indicates that students tend to ignore this evi-
dence and keep using the same study strategies throughout a
semester regardless of whether or not exam performance sug-
gests otherwise (Blasiman et al., 2017).
In studying the accuracy of students’ metacognitive mon-
itoring, a very common paradigm is to ask a learner to make
judgments about their learning at various points in the learning
process, and two types of judgments used are judgments of
learning (JOL) and retrospective confidence judgments (RCJ)
(Dunlosky & Thiede, 2013). JOLs are made after studying but
before the learner is tested on the learned material, while RCJs
are made after the learner is tested on the material. For example,
a JOL might consist of asking students to rate on a scale of
1–10 how well-prepared they think they are in an exam they are
about to take (or what grade they think they will earn), and an
RCJ might consist of asking students to rate on a scale of 1–10
how well they think they performed immediately after taking
the exam (or what grade they think they received on the exam
that they just completed). Findings reveal that accurate JOLs and
RCJs in a task are directly related to the learner’s ability to accu-
rately perform that task (Ehrlinger et al., 2008; Handel & Dresel,
2018; Handel & Fritzsche, 2016; Kelemen et al., 2007; Kruger &
Dunning, 1999; Rebello, 2012; Schneider, 2002); another way of
putting it is that when students’ self-awareness of their level of

11998_Master4.indd 154 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 155

learning is accurate, they tend to do well on the task they had

prepared for. What is perhaps most insidious is the finding that
low-performing students are less accurate in their metacogni-
tive judgments as measured by JOLs and RCJs; low-performers
not only think they are better prepared when walking into a
test than they actually are, but also think they performed above
average even after taking the test. In short, low-performers are
grossly overconfident in their preparation and fail to recognize
that they did poorly on an exam even after taking it. In contrast,
high-performers are often slightly under-confident as measured
by JOLs when walking into an exam and think they did slightly
worse than their actual performance as measured by RCJs
(Dunning et al., 2003; Griffin et al., 2009; Schlosser et al., 2013;
Shake & Schulley, 2014). This has led to low-performing students
being portrayed as suffering from a “double-curse,” as eloquently
described in the following quote:
“In many significant social and intellectual domains, the skills
necessary to recognize competence are extremely close if not
identical to those needed to produce competent responses.
… Thus, incompetent individuals suffer a double curse: Their
deficits cause them to make errors and also prevent them from
gaining insight into their errors. Several studies have now
shown that incompetent individuals (i.e., those performing
poorly relative to their peers) fail to show much insight into just
how deficient their performance is (Kruger & Dunning, 1999).
… College students scoring in the bottom 25% on a course exam
walked out of the exam room thinking that they outperformed a
majority of their peers (Dunning, Johnson, Ehrlinger, & Kruger,
2003). Compared with good students, poor students less success-
fully identify which specific questions they have gotten right on
an exam and which they have gotten wrong (Sinkavich, 1995).”
(Dunning, Johnson, Ehrlinger & Kruger, 2003, pp. 73–74)

11998_Master4.indd 155 2/11/2020 [Link] PM

 156 The Science of Learning Physics

Most of us have experienced students coming to our offices and

lamenting poor performance in an exam despite being “A” stu-
dents and spending considerable time studying for the exam;
when we probe their study strategies, we often find the pattern
described above (i.e., study strategies that build familiarity, not
competence, and poor judgments of learning), which does little
to build the competence needed to apply physics knowledge and
procedures to solve problems. This is a difficult problem to solve
since evidence also suggests that providing underperforming
students with feedback showing them that their judgments of
learning are highly inaccurate compared to actual exam per-
formance does little to help better calibrate future judgments
(Morphew, 2019).

6.2. What are the implications of research into students’

view of learning and studying for instruction?
We offer the following implications of the research reviewed
above for instruction:
● Students’ perceptions of good study strategies run
counter to research evidence. Common study strategies
favored by students to prepare for exams, such as
highlighting or underlining text, re-reading notes and
course material, and summarizing, build familiarity and
fluency with the material to be learned, not competence.
These study strategies require minimal effort and yield
minimal payoff. In a complex domain like physics learning
concepts and how they are applied to solve problems are
crucial to success, but study strategies that build familiarity
with concepts do little to help students learn what concepts
to apply, when to apply them, and procedures for applying
them across a variety of contexts.

11998_Master4.indd 156 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 157

● Effective study strategies contain “desirable difficulties”

that require more effort and deeper processing of
the to be learned material on the part of the student.
-
-
Effective study strategies, such as spaced/distributed
practice (in contrast to massed practice or cramming),
interleaved practice (in contrast to blocked practice), and
asking/answering deep questions, offer more promise for
learning that lasts. Unfortunately, these strategies are not
commonly used by students in studying or by instructors
in structuring their courses.
● Students who do poorly on tests do little to change their
study habits. Research indicates that even when faced with
evidence that their study habits are ineffective for exam
preparation, poorly performing students keep the same
study habits but simply add more study time in an effort
to remedy the situation. More study time using ineffective
study strategies does little to improve performance. If
students knew more effective learning strategies they may
be more likely to adopt them.
● Students use subjective judgments in deciding when to
stop studying for an exam, when objective judgments
would serve them better. Students tend to judge readiness
for an exam when they reach a certain level of familiarity
with the material, instead of competence with the material.
More subjective measures of exam preparation, such as
taking an old exam under exam conditions, are seldom
used by students, when such measures are more accurate
measures of preparation, and in addition, would help
students target their studying by focusing on topics and
problem types that they get wrong.
● Judgments of learning by poorly performing students
are grossly overestimated. Poorly performing students

11998_Master4.indd 157 2/11/2020 [Link] PM

 158 The Science of Learning Physics

overestimate the grade they will earn on an exam both

before, and after taking the exam, and providing feedback
on their miscalibration does not remedy the situation.
Researchers posit that the skills needed to make accurate
judgments of learning are the same as those needed
to perform well on exams, making the development of
intervention strategies to help underperforming students
challenging.

6.3. Examples of teaching interventions based on

learning research
Research on students’ study habits and judgments of learning
suggest that we can both, help students change some of their
study habits to improve learning, and coordinate instruction in
support of better study habits. The most difficult nut to crack
is the finding that poorly performing students have an inflated
view of their subject matter understanding, and further that
providing them evidence of their miscalibration does little to
improve it (Miller & Geraci, 2011a, 2011b; Morphew, 2019). Never
­
theless, there are some steps that can be taken to curb ineffective
student study behaviors and to improve instruction based on the
findings reviewed in this chapter.
● Summarize for students early in a course the salient re­
search findings on effective studying. Making students
aware of both effective (e.g., distributed practice, objec-
tive measures of learning like taking practice tests) and
ineffective (e.g., cramming, only reviewing course notes
and previous homework before an exam) study habits
could help them curb bad habits. Although students will
very likely be hesitant to change study habits that work
well in most of their other courses, telling them early on

11998_Master4.indd 158 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 159

that physics is unforgiving if not learned at a deep level

(e.g., understanding concepts and problem-solving proce-
dures and how to apply them across contexts) will “read
into the record” what is needed to do well in the course.
After the first exam’s grades are returned to students and
many come to see us to ask how they can do better, we can
refer back to the summary we provided to them of effec-
tive study habits and analyze the study habits they used
to study for the exam, as well as offer alternatives to their
study habits. Another strategy to support this reflection is
using a written exam wrapper which prompts students to
explicitly write down how they studied for the most recent
exam and what they will do differently to prepare for the
next exam (Ambrose et al., 2010). This also helps to place
the onus of learning on the student rather than blaming
the instructor for writing an exam that was “unfair” or too
hard in their view.
● Make old exams available for students to gauge their
preparation for exams. Point out to students that taking
old exams under regular exam conditions provides an un-
biased, objective method to gauge their exam preparation,
and also helps them target topics and problem types that
gave them difficulties for targeted studying. Students who
have come to the authors’ offices in the past asking for
help to do better on exams often state that they studied by
working out old exams. However, upon closer questioning,
we often find that they did not work the exam under
normal exam conditions (timed, with only a calculator and
an equation sheet, if allowed). Working out old exam prob-
lems casually with resources (course notes, web) and then
looking up the answer to each problem immediately after
working it out gives students a false sense of “knowing.”

11998_Master4.indd 159 2/11/2020 [Link] PM

 160 The Science of Learning Physics

To get a realistic assessment of their knowledge and an

accurate feel for the upcoming exam, students need to take
practice exams under the same conditions as the real exam,
and then grade the exam after taking it. Students can then
skip the topics/problem types they got correct and focus on
the topics/problem types they got wrong for further study.
As they are figuring out how to work out a problem that
they got wrong on a practice test, it is important to stress
to students that they need to identify/verbalize what exactly
caused the difficulty: Was it a conceptual error, and if so,
what was it and how can it be avoided in the future? Was
it a procedural error? Was it applying the wrong concept to
the problem’s context, and if so, what are the contexts under
which that concept applies and does not apply? These types
of questions force students to reflect on their learning and
add desirable difficulties to their exam preparation.
● Implement effective learning practices in organizing
courses. In preparing homework assignments and midterm
exams in our courses, try to incorporate interleaved prac-
tice by having some cumulative problems in assignments
and tests throughout the entire course. This serves several
purposes: To convey to students that physics knowledge is
cumulative and inter-related; to keep students’ knowledge
in the course current; and to convey to students that to
learn physics well one has to see both the big picture as
well as how concepts and procedures build on each other
and are related to one another. In addition, interleaving
some isomorphic worked examples within homework as-
signments would help students who are stuck identify the
nature of their difficulty and allow them to make progress.
Since asking/answering deep questions is also an effective
learning strategy, and since it is difficult for students to

11998_Master4.indd 160 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 161

pose deep questions on their own because their knowledge

and perspective are not well developed, provide students
with deep questions they can pose for themselves during
study sessions. For example, using a modified form of the
strategy writing activity discussed in Chapter 3, suggest
to students that, after solving a problem, they should ask
themselves to name the major principle/concept they ap-
plied, the reason why it applied to that context, and to state
in words the procedure used to apply the principle/concept;
this high level task can also be required for some subset of
homework problems. As discussed earlier, questions such as
these are difficult for students to answer and thus engender
useful reflection on problem solving practices.
● Help students improve their judgments of learning by
using objective measures. As discussed in the research
reviewed above, students, especially poorly performing
students, are likely to base their judgments of how well pre-
pared they are for an exam on familiarity with the subject
matter rather than on competence with the subject matter.
Attempting to remedy this problem is exacerbated by the
fact that poorly performing students think they performed
above average even after taking an exam. This disconnect
from reality is not amenable to quick fixes. One recommen-
dation we would offer is for instructors to construct and
have available a short 4–5 problem representative “mini-
test” of an upcoming test and offer it as a diagnostic tool
to students (in our office or at a predetermined place/time)
using realistic exam conditions. Have students predict how
many questions they expect to get correct prior to taking
the mini-test, and immediately following. Then grade the
mini-test and observe possible discrepancies between stu-
dents pre- and post-predictions and the actual grade on the

11998_Master4.indd 161 2/11/2020 [Link] PM

 162 The Science of Learning Physics

mini-test. We are likely to find the same result as in the pre-

viously reviewed research—students who perform poorly in
the mini-test will have predicted that they would do above
average in both of their predictions. Showing underper-
forming students first-hand that their predictions about
their current state of knowing are inaccurate, and that they
behave the same as students nationwide, opens the oppor-
tunity for a dialog on how to curb bad study habits and
how to obtain more objective measures of preparation.
We should keep in mind that students are very hesitant to
abandon study methods that work in other courses since
they will view it as risky behavior, and in addition, they
do not know what study habits to adopt instead. However,
seeing that what they are doing is not helping them suc-
ceed in the course allows instructors to offer concrete sug-
gestions for adopting more effective study strategies.

References
Ackerman, R. & Zalmanov, H. (2012). The persistence of the fluency-
confidence association in problem solving. Psychonomic Bulletin &
Review, 19, 1187–1192. doi: 10.3758/s13423-012-0305-z
Ambrose, S.A., Lovett, M., Bridges, M.W., DiPietro, M. & Norman, M.K.
(2010). How Learning Works: Seven Research-Based Principles for Smart
Teaching. San Francisco, CA: Jossey-Bass.
Ariel, R., Dunlosky, J. & Bailey, H. (2009). Agenda-based regulation of
study-time allocation: When agendas override item-based monitoring.
­
Journal of Experimental Psychology: General, 138, 432–447. doi:
10.1037/a0015928
Benjamin, A.S., Bjork, R.A. & Schwartz, B.L. (1998). The mismeasure of
memory: When retrieval fluency is misleading as a metacognitive
index. Journal of Experimental Psychology: General, 127, 55–68. doi:
10.1037/0096-3445.127.1.55

11998_Master4.indd 162 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 163

Bjork, R.A. (1994). Memory and metamemory considerations in the

training of human beings. In J. Metcalfe and A. Shimamura (Eds.),
Metacognition: Knowing about Knowing, pp. 185–205. Cambridge, MA:
MIT Press.
Bjork, E.L. & Bjork, R.A. (2014). Making things hard on yourself, but
in a good way: Creating desirable difficulties to enhance learning.
In M. A. Gernsbacher and J. Pomerantz (Eds.), Psychology and the
Real World: Essays Illustrating Fundamental Contributions to Society
(2nd edition), pp. 59–68. New York: Worth.
Blasiman, R.N., Dunlosky, J. & Rawson, K.A. (2017). The what, how much,
and when of study strategies: Comparing intended versus actual study
behaviour. Memory, 25, 784–792. doi: 10.1080/09658211.2016.1221974
Dunlosky, J., Rawson, K.A., Marsh, E.J., Nathan, M.J. & Willingham,
D.T. (2013). Improving students’ learning with effective learning
techniques: Promising directions from cognitive and educational
psychology. Psychological Science in the Public Interest, 14, 4–58.
Dunlosky, J. & Thiede, K.W. (2013). Metamemory. In The Oxford Hand
­
book of Psychology, edited by D. Reisberg, pp. 283–298. doi: 10.1093/
oxfordhb/9780195376746.013.0019
Dunning, D., Johnson, K., Ehrlinger, J. & Kruger, J. (2003). Why people fail
to recognize their own incompetence. Current Directions in Psycholog-
ical Science, 12, 83–86. doi: 10.1111/1467-8721.01235
Ebbinghaus, H. (1913). Memory: A Contribution to Experimental Psychology.
New York: Columbia University.
Ehrlinger, J., Johnson, K., Banner, M., Dunning, D. & Kruger, J. (2008).
Why the unskilled are unaware: Further explorations of (absent) self-
insight among the incompetent. Organizational Behavior and Human
Decision Processes, 105, 98–121. doi: 10.1016/[Link].2007.05.002
Franklin, S.V., Sayre, E.C. & Clark, J.W. (2014). Traditionally taught students
learn; actively engaged students remember. American Journal of
Physics, 82(8), 798–801. https:// [Link]/10.1119/1.4890508

11998_Master4.indd 163 2/11/2020 [Link] PM

 164 The Science of Learning Physics

Griffin, T.D., Jee, B.D. & Wiley, J. (2009). The effects of domain knowledge
on metacomprehension accuracy. Memory & Cognition, 37, 1001–1013.
doi: 10.1037/0003-066X.59.1.29
Händel, M. & Dresel, M. (2018). Confidence in performance judgment
accuracy: The unskilled and unaware effect revisited. Metacognition
and Learning, 13, 265–285. doi: 10.1007/s11409-018-9185-6
Händel, M. & Fritzsche, E.S. (2016). Unskilled but subjectively aware:
metacognitive monitoring ability and respective awareness in low-
performing students. Memory & Cognition, 44, 229–241. doi: 10.3758/
s13421-015-0552-0
Heckler, A.F. & Sayre, E.C. (2010). What happens between pre- and
post-tests: Multiple measures of student understanding during an
introductory physics course. American Journal of Physics, 78(7),
768–777. [Link]
Hiebert, J. & Grouws, D.A. (2007). The effect of classroom mathematics
teaching on students’ learning. In Lester, F.K. (Ed.) Second Handbook of
Research on Mathematics Teaching and Learnin, pp. 317–404. Informa-
tion Age Publishing Inc.
Kelemen, W.L., Winningham, R.G. & Weaver, C.A. III. (2007). Repeated
testing sessions and scholastic aptitude in college students’ metacog-
nitive accuracy. European Journal of Cognitive Psychology, 19, 689–717.
doi: 10.1080/09541440701326170
Koriat, A. & Ma’ayan, H. (2005). The effects of encoding fluency and
retrieval fluency on judgments of learning. Journal of Memory and
Language, 52, 478-492. doi: 10.1016/[Link].2005.01.001
Kornell, N. & Son, L.K. (2009) Learners’ choices and beliefs about
self-testing. Memory, 17, 493–501. doi: 10.1080/09658210902832915
Kruger, J. & Dunning, D. (1999). Unskilled and unaware of it: How difficul-
ties in recognizing one’s own incompetence lead to inflated self-
assessments. Journal of Personality and Social Psychology, 77, 1121–
1134. doi: 10.1037/0022-3514.77.6.1121

11998_Master4.indd 164 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 165

Mestre, J.P. (Ed.) (2005). Transfer of Learning from a Modern Multidisci-

plinary Perspective (393 pages). Greenwich, CT: Information Age
Publishing.
Mestre, J.P. (2002). Probing adults’ conceptual understanding and transfer
of learning via problem posing. Journal of Applied Developmental
Psychology, 23, 9–50.
Metcalfe, J. & Kornell, N. (2005). A region of proximal learning model of
study time allocation. Journal of Memory and Language, 52, 463–477.
doi: 10.1016/[Link].2004.12.001
Miller, T.M. & Geraci, L. (2011a). Unskilled but aware: Reinterpreting
overconfidence in low-performing students. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 37, 502–506. doi: 10.1037/
a0021802
Miller, T.M. & Geraci, L. (2011b). Training metacognition in the classroom:
The influence of incentives and feedback on exam predictions. Metacog-
nition and Learning, 6, 303–314. doi: 10.1007/s11409-011-9083-7
Mitchell, T. (1997). Machine Learning. New York: McGraw Hill.
Morphew, J.W. (2019). Metacognitive Calibration in Introductory Physics
Courses: Predictors and Interventions (Doctoral dissertation). Univer-
sity of Illinois at Urbana-Champaign, Champaign, IL.
Newell, A. & Rosenbloom, P.S. (1981). Mechanisms of skill acquisition and
the law of practice. In J.R. (Ed.) Cognitive Skills and Their Acquisition,
pp. 1–55. Hillsdale, NJ: Lawrence Erlbaum Associates.
Pashler, H., Bain, P.M., Bottge, B.A., Graesser, A., Koedinger, K., McDaniel,
M. & Metcalfe, J. (2007). Organizing Instruction and Study to Improve
Student Learning (52 pp). US Department of Education report, NCER
2007-2004.
Rebello, N.S. (2012). How accurately can students estimate their perfor-
mance on an exam and how does this relate to their actual perfor-
mance on the exam? In N.S. Rebello, P.V. Engelhardt & C. Singh
(Eds.), AIP Conference Proceedings, Vol. 1413, pp. 315–318. AIP. doi:
10.1063/1.3680058

11998_Master4.indd 165 2/11/2020 [Link] PM

 166 The Science of Learning Physics

Sayre, E.C., Franklin, S.V., Dyrnek, S., Clark, J. & Sun, Y. (2012). Learning,
retention, and forgetting of Newton’s third law throughout university
physics. Physical Review Special Topics – Physics Education Research,
8(010116). [Link]
Sayre, E.C. & Heckler, A.F. (2009). Peaks and decays of student knowledge
in an introductory E&M course. Physical Review Special Topics –
Physics Education Research, 5(013101), 1–5. https:// [Link]/
prper/abstract/10.1103/PhysRevSTPER.5.013101
Scaife, T. & Heckler, A. (2011). Interference between electric and mag

­
netic concepts in introductory physics. Physical Review Special Topics
Physics Education Research, 7(010104). http:// [Link]/10.1103/Phys
RevSTPER.7.010104
Schlosser, T., Dunning, D., Johnson, K.L. & Kruger, J. (2013). How unaware
are the unskilled? Empirical tests of the “signal extraction” counter

­
explanation for the Dunning-Kruger effect in self-evaluation of perfor-
mance. Journal of Economic Psychology, 39, 85–100. doi: 10.1016/j.
joep.2013.07.004
Schmidt, R.A. & Bjork, R.A. (1992). New conceptualizations of practice:
Common principles in three paradigms suggest new concepts for
training. Psychological Science, 3, 207–218. doi: 10.1111/j.1467-9280.
1992.tb00029.x
Schneider, W. (2002). Memory development in childhood. In U. Goswami
(Ed.), Blackwell Handbook of Childhood Cognitive Development, pp.
236–256. Malden, MA: Blackwell.
Shake, M.C. & Shulley, L.J. (2014). Differences between functional and
subjective overconfidence in postdiction judgments of test perfor-
mance. Electronic Journal of Research in Educational Psychology, 12,
263–282. doi: 10.14204/ejrep.33.14005
Sinkavich, F.J. (1995). Performance and metamemory: Do students know
what they don’t know? Journal of Instructional Psychology, 22(1),
77–87.

11998_Master4.indd 166 2/11/2020 [Link] PM

 Students’ Perceptions of Learning and their Study Habits 167

Sweller, J. & Cooper, G.A. (1985). The use of worked examples as a substi-
tute for problem solving in learning algebra. Cognition and Instruc-
tion, 2, 59–89.
Ward, M. & Sweller, J. (1990). Structuring effective worked examples.
Cognition and Instruction, 7, 1–39.
Wenzel, K. & Reinhard, M-A. (2019). Relatively unintelligent individuals
do not benefit from intentionally hindered learning: The role of desir

­
able difficulties. Intelligence, 77, 101405. doi: 10.1016/[Link].2019.
101405

11998_Master4.indd 167 2/11/2020 [Link] PM

 b2530   International Strategic Relations and China’s National Security: World at the Crossroads

This page intentionally left blank

b2530_FM.indd 6 01-Sep-16 [Link] AM

 c hapter
7 Testing in the Service of Learning:
The Testing Effect and How it Promotes
Long-Term Retention

7.1. What does research on the “testing effect” tell us?

In Chapter 6 we discussed effective and ineffective learning strat-
egies. It turns out that testing is an effective learning strategy
for long-term retention of information. Because testing is typ-
ically not used by students and faculty as a learning method,
and because of an abundance of recent research pointing to the
usefulness of testing for learning, we devote this short chapter
to testing as a learning tool. Testing is often considered a dirty
word among both students and faculty. Non-physics majors fear
physics tests likely more than tests in other subjects. Generating
several multiple-choice tests in large introductory courses
is not among the favorite activity of faculty given how time-
consuming it is, and open-response tests may be easier to make
up but time-consuming to grade. The types of tests alluded to in
the previous sentences are summative tests intended to evaluate
students’ knowledge to assign grades. Another type of testing,
called formative assessment, is not used to generate grades but
rather to help shape instruction to maximize student learning
(use of clicker questions in large classes, as discussed in previous
chapters, is an example of formative assessment). Over the last

169

11998_Master4.indd 169 2/11/2020 [Link] PM

 170 The Science of Learning Physics

two decades, an increasing number of studies from the learning

sciences indicate that testing (of the formative type) helps stu-
dents retain their learning over longer periods of time—an effect
called the testing effect.

7.1.1. What is the testing effect?

The testing effect can be summarized as follows (Dunlosky et al.,
2013; Pashler et al., 2007; Roediger & Kapricke, 2006): When com-
paring studying material to being tested on the material, those
spending their time being tested (after an initial study period)
show greater long-term retention of the material than those
spending an equal amount of time just studying the mate-
rial. The effect is robust—perhaps most surprising, the testing
effect holds even when performance on the tests is lack luster
and even if no feedback is given on performance immediately
following the tests. The qualifier “long-term retention” is crucial
in the previous description. As the research reviewed below will
demonstrate, repeated studying of material yields better perfor-
mance on summative tests administered immediately following
the study compared to equivalent time spent on testing; how-
ever, when tested one or more weeks later, the trend reverses,
with those who spent relatively more time being tested signifi-
cantly outperforming those who spent time in repeated studying.
Perhaps this is why so many students rely on cramming as the
preferred method of studying for high-stakes tests since it yields
positive (but ephemeral) short-term payoffs. As instructors we
want our students to retain the physics knowledge we teach
them and testing is a tool we can use to promote long-term
retention.

11998_Master4.indd 170 2/11/2020 [Link] PM

 Testing in the Service of Learning 171

7.1.2. Experimental findings on the testing effect

A typical study of the testing effect by Karpicke and Roediger
(2007) went as follows. Three groups of students were tasked to
learn 40 unrelated words. During study, the words were presented
on a screen at the rate of 3 seconds per word. During test, stu-
dents were given 2 minutes (note that this is the same amount
of time spent studying) to recall as many words as they could.
The three groups were in the conditions STST, SSST and STTT,
where S means a study trial and a T means a test trial (recalling
as many words as possible). Each S or T trial was administered
sequentially and words were randomized in each study trial. All
three groups did the same four-trial set 5 times. Thus the experi-
ment lasted approximately 40 minutes (plus a little “set-up” time
between trials). Participants were tested immediately following
each trial, and all participants returned after one week to take
a final recall test. Results showed that the percentage of words
recalled initially rose steeply across the first three four-trial sets
and then reached an asymptote in the 78–88% range by the last
set. The STST group did best with about 88% recall at the end of
the fifth set, followed by the STTT group at about 81%, followed
by the SSST group at about 78%. Thus all three conditions did
about the same at final testing, but with a trend favoring testing.
When tested one week later, the STST group did best with about
68% of the words recalled, followed by the STTT group at 63%
with the SSST group recalling the fewest words at 55%. This is
impressive given that the SSST groups had studied the list of
words 15 total times versus only 5 times for the STTT group.
Thus the group that engaged in the most study time did worst
one week following the experiment; in other words, testing trials
generated at least as much or more learning as did study trials.
These types of studies have been reproduced by others with

11998_Master4.indd 171 2/11/2020 [Link] PM

 172 The Science of Learning Physics

slightly varied conditions, all showing that long-term retention

is better for groups that practice memory retrieval with testing
rather than with studying. However, cramming (more study ses-
sions compared to testing sessions) consistently wins out for
tests conducted immediately following study.
A similar study by Wheeler, Ewers, and Buonanno (2003)
also involving memorizing 40 unrelated words and comparing
studying the list five times (SSSSS), versus one time with four
back-to-back recall tests (STTTT) resulted in the typical cross-over
pattern shown in Fig. 7.1. Studying the list five times has a large
advantage if testing is done immediately after study, but one
week later the single study followed by four recall tests condi-
tion does better. It is worth emphasizing that in the SSSSS condi-
tion participants were exposed to the entire list of 40 words five
iguretimes, whereas
1 Graph in thethis,
(Alex adjusts STTTT
then condition participants were exposed
creates screenshot)
to the 40 words only once and then re-exposed to only those

SSSSS
0.6
STTTT
Fraction of Words Recalled

0.5

0.4

0.3

0.2

0.1

0.0
5 minutes 1 week
Final Assessment Interval

Fig. 7.1. Recall of 40 words after all study, or study followed by four tests
after two different delay periods.

11998_Master4.indd 172 2/11/2020 [Link] PM

 Testing in the Service of Learning 173

words they could be recalled during the four tests! Roediger and
Karpicke (2006) describe these types of finding as follows:
Repeatedly studying material is beneficial for tests given soon
after learning, but on delayed criterial tests with retention
intervals measured in days or weeks, prior testing can produce
greater performance than prior studying (pp. 189).

Similar results have been obtained in studies using “paired

associates,” which are pairs of words to be learned. Word pairs
take a variety of forms in different experiments; for example,
English–English pairs such as chair-table or chair-horse, nonsense
words paired with English words, and even Swahili–English pairs
in some experiments.

7.1.3. Relevance of the testing effect for physics

instruction
The attentive reader is likely now thinking, “How is this relevant
to physics?” After all, conceptual learning and problem solving
in physics is a whole lot more complicated than learning list of
words. To be fair to lab-based cognitive psychology experiments,
researchers choose tasks that can be learned in single 1–2 hour
sessions to make experiments manageable. In order to conduct
learning experiments in subject areas, especially one as com-
plex as physics where learning takes considerable time, would
require longitudinal studies that are much more difficult to
conduct. Nevertheless, a smaller number of studies show sim-
ilar findings when using more complex reading materials (see
examples described in (Roediger & Karpicke, 2006)) and in class-
room studies where learning conditions are difficult to control
(McDaniel et al., 2012; McDermott et al., 2014). A recent study
also demonstrated a testing effect within engineering disciplines
(Morphew et al., 2019).

11998_Master4.indd 173 2/11/2020 [Link] PM

 174 The Science of Learning Physics

As an example of a study that used more academically rele-

vant materials, Roediger and Karpicke (2006b) had college par-
ticipants read scientific passages in two experiments. In one
experiment two groups were compared where one group studied
the passages twice (Study–Study) and one group studied the pas-
sages once and then in a free-recall test wrote down all the main
ideas they recalled from the passages (Study–Test). Three addi-
tional testing sessions were held 5 minutes, two days or one week
following learning during which participants again free-recalled
what they could remember from the passages. Figure 7.2 shows
the findings from the experiment. As in the word recall studies
reviewed above, the Study–Study group had the advantage in the
test administered 5 minutes after learning. However, two days
and one week later, the pattern reversed, with the Study–Test
group doing significantly better than the Study–Study group. A
second experiment compared three different groups who read
scientific passages, with group SSSS studying the passages in
four successive periods, group SSST studying the passages for
three successive periods followed by a recall test, and group STTT
studying once followed by three successive free-recall testing
periods. Final recall tests were then administered 5 minutes and
one week later. Figure 7.3 shows the findings. It is striking how
the SSSS group’s ability to recall passage details dropped pre-
cipitously after one week, whereas the groups that were tested,
although performing slightly lower that the SSSS group 5 min-
utes after learning, retained considerably more details one week
later.
Another finding is worthy of mention in studies of learning
that compared study to testing. When participants are asked
to predict how well they would remember what they had
studied, those in repeated study conditions thought they would

11998_Master4.indd 174 2/11/2020 [Link] PM

 Figure 3 Graph (Alex adjusts this, then creates screenshot)

Testing in the Service of Learning 175

100%

Percentage of Main Ideas Recalled

80

60

40

20

0
5 minutes 2 days 1 week
ure 2 Graph (Alex adjusts this, then creates screenshot)
Final Assessment Interval

Fig. 7.2. Studying scientific text folowed test after varying delays.

100%
Percentage of Main Ideas Recalled

80

60

40 SSSS SSST
STTT

20

0
5 minutes 1 week
Final Assessment Interval

Fig. 7.3. Studying scientific text followed by free recall of main ideas after
two different delays.

remember more than those in testing conditions (Dunlosky &

Nelson, 1992; Roediger & Karpicke, 2006), perhaps due to their
experience that cramming yields short term learning gains
even though it is ineffective for long-term retention compared

11998_Master4.indd 175 2/11/2020 [Link] PM

 176 The Science of Learning Physics

to testing; or perhaps this perception is due to repeated study

yielding more fluency than testing (see Chapter 6). Thus, one
cannot always trust one’s intuitions for what works best.

7.1.4. Mechanisms leading to long-term retention from

the testing effect
What mechanisms lead to better long-term retention with a com-
bination of study and testing rather than just repeated studying?
Several processes are likely at work. A hint comes from addi-
tional findings that tests requiring generation/production of
material (i.e., effortful retrieval from memory) result in larger
testing effect gains than tests requiring only recognition of the
material, as in multiple choice tests (see Roediger & Karpicke,
2006 and references therein). Thus it seems that more effortful
learning strategies result in better long-term retention. By prac-
ticing retrieval of information from memory during testing, stu-
dents engage in deep processing of the material and strengthen
memory pathways to the information (Kornell et al., 2009). It is
also thought that “test-potentiated learning,” that is, benefiting
from targeted study following a test, is another mechanism for
the testing effect (Bjork & Storm, 2011; Wissman et al., 2011).
Interestingly, when provided with correctness feedback immedi-
ately following testing, testing effect learning gains have been
found for items answered incorrectly during testing (Butler et al.,
2008; Richland et al., 2008) as well as for related information that
was untested (Carpenter et al., 2012; Little et al., 2012).
We have already discussed in Chapter 6 that popular yet inef-
fective learning strategies are those that provide students with
a feeling of familiarity or fluency with the to-be-tested material.
Repeated study is excellent for providing students with a feeling
of fluency, but as discussed in Chapter 6, fluency with material

11998_Master4.indd 176 2/11/2020 [Link] PM

 Testing in the Service of Learning 177

does not imply competence with the material. From a student’s

standpoint, learning by repeated studying seems like a good
choice—after all, it requires less effort and builds familiarity
quickly. Testing oneself, in contrast, is effortful and generates less
familiarity but (counterintuitively) results in better long-term
retention of knowledge by building more connections/pathways
to the information in memory. Test-enhanced learning fits Bjork’s
definition of a desirable difficulty discussed in Chapter 6, but it
is a challenge to get students to adopt a “test yourself ” strategy
when studying, or to get instructors to use frequent testing as a
learning strategy.
Many instructors make available to students old exams previ-
ously used in the course so that they may test themselves prior
to taking the actual exam. Although this is an excellent idea,
our experience is that students tend not to avail themselves of
this opportunity (recall that taking tests is more effortful than
re-studying material, and re-study generates more familiarity).
Further, those students who do use old exams to prepare for
an upcoming exam often use them inappropriately. In conver-
sations we have had with students in our courses who lament
they did not do well on an exam, we often ask them if they used
the “practice tests” made available to them. Those students typ-
ically answer “yes” but when we probe more deeply how they
used those practice tests they reveal that they used them casu-
ally and not under realistic exam scenarios. For example, they
solve the problems with notes and textbook at their sides, and
look up answers immediately after solving each problem. This
does not measure performance under pressure in real exam cir-
cumstances; however, it builds familiarity with the material that
is interpreted by students as competence with the material.

11998_Master4.indd 177 2/11/2020 [Link] PM

 178 The Science of Learning Physics

7.2. What are the implications of the testing effect

for instruction?
We offer the following implications of the research reviewed
above for instruction:
● Formative assessments (i.e., assessments in the service
of learning) are a powerful tool both for learning and for
retaining what is learned. Formative assessments afford
several positive influences on learning. First, there is no
pressure to do well on formative assessments so students
can focus on evaluating what they know and don’t know.
Further, students actively participate in retrieving and using
knowledge, thereby building and reinforcing retrieval links
to the knowledge in memory. Formative assessments also
makes students’ thinking visible to the instructor, allowing
instructors to spend relatively more time helping students
grasp difficult ideas, moving them closer to coaches of
learning rather than dispensers of information.
● Tests that require generation/production of knowledge
stored in memory promote lasting learning more than
tests that only require recognition. Tests that require
generation of material require significantly more effort
than those requiring only recognition. The added effort is
a way of injecting desirable difficulties into the learning
process, as discussed in Chapter 6. Students are likely to
resist the expenditure of effort in learning in favor of easier
but less effective learning approaches, such as studying and
re-studying, which largely build familiarity with the material
rather than competence with the material.
● Combining testing with other successful learning strate­
gies (as discussed in Chapter 6) would enhance the testing
effect. For example, when using formative assessment in a

11998_Master4.indd 178 2/11/2020 [Link] PM

 Testing in the Service of Learning 179

course the instructor could test both the material currently

being taught as well as previous material, thus combining
testing with interleaved practice (as opposed to blocked
practice).

7.3. Examples of teaching interventions based on

learning research
Research on the impact of the testing effect on long-term reten-
tion of learning suggests a number of study and instructional
strategies.
● Communicate with students the powerful impact of
the testing effect. Since the testing effect is counter

­
intuitive, communicating to students the value of testing
over repeated studying is valuable. Students will find it
hard to believe that intensive studying is not as effective
for learning and long term retention compared to a com-
bination of studying and testing, but providing them with
research findings will likely help them incorporate testing
in their learning strategies.
● Assess students’ ability to link concepts to conditions
for applicability. Problem solving is crucial in physics
and solving problems consists of applying concepts
and procedures. However, before applying a concept,
experienced problem solvers evaluate whether or not
a concept is applicable to the specific situation under
consideration; this type of knowledge is often referred to
as “tacit knowledge” since experts use it without overtly
making such knowledge visible during problem solving.
Students in introductory courses often do not know when
concepts can be applied to situations. Therefore, a very
useful type of assessment that requires “deep processing”

11998_Master4.indd 179 2/11/2020 [Link] PM

 180 The Science of Learning Physics

is to provide students with a concept and ask them to state

the conditions under which the concept can be applied,
as well as conditions under which it cannot be applied
(since this is a common error they commit). For example,
Newton’s second law is applicable to find the acceleration of
an object when the net force on the object is constant; an
object sliding down a frictionless curved ramp is not easily
solved with Newton’s Second Law due to the changing
nature of the net force; conservation of mechanical energy
can be applied only in cases where non-conservative
forces do no work. One way to structure an assessment
is to have a small group of students work collaboratively
on analyzing a problem and determining whether or not
a particular concept can be applied, and the underlying
reason. A subsequent class-wide discussion would bring
to light this type of tacit knowledge. Another activity
that has been previously discussed that would also link
concepts to conditions of applicability is strategy writing.
The first two elements of strategy writing consist of
identifying the concept(s) that applies to a problem and
stating why the particular concept(s) applies to that specific
situation, so one could shorten the strategy writing activity
to one where problems are considered in view of the
concept that could be applied to solve it and the reason for
why it applies.
● Use assessments that help students identify their own
misconceptions. Knowing that certain problems tempt
students to apply incorrect intuitive notions, we can use
those types of problems to refine students’ understanding.
One example is to give the problem below and ask students
to generate a short series of procedural steps needed to
solve it.

11998_Master4.indd 180 2/11/2020 [Link] PM

 Testing in the Service of Learning 181

First note that this assessment requires students to recall

and apply important physics ideas from memory. This assess-
ment would also help identify many students’ erroneous belief
that the tension in the string is equal to mg. The intent here is
to help students realize that application of physics concepts
and procedures, in this case applying Newton’s Second Law to
both masses individually yielding the equations T = Ma and
T − mg =ma−, is the appropriate approach which should
make obvious that the tension is not mg. Since it has also been
observed by physics education researchers that students can
often write down correct procedures yet have underlying miscon-
ceptions, even if students write down the correct equations, the
instructor can ask “Explain why the tension in the string is, or is
not, equal to mg,” hopefully leading to the obvious answer that
if the tension were mg, the net force on m would be 0 and the
mass would just hang there without moving. At some point later
in the course, instructors could check to see if students are less
likely to resort to incorrect intuitions by revisiting these situa-
tions under an Atwood Machine scenario.

M Two masses are attached to a frictionless,

massless pulley as shown. The table is fric-
tionless. Write a short series of procedural
steps that, if carried out, would allow you
m to find the acceleration of the blocks. Treat
the string as massless.

● Provide students with previous tests administered in

your courses for them to practice on. This was a bullet
listed in the previous chapter but it is worth repeating here.

11998_Master4.indd 181 2/11/2020 [Link] PM

 182 The Science of Learning Physics

It is important to ask students to take old practice tests

under normal exam circumstances since the impression
created by taking an old exam casually is that it is easier
than the real test (as students so often tell us, despite the
fact that exam averages remain fairly uniform year after
year). Again, students tend to take the easier approach to
exam preparation, namely studying and re-studying since
this approach requires less effort and builds familiarity
with the material. One analogy we like to use with students
is to tell them that one does not train to run a marathon
by watching videos of other people running marathons.
Taking practice tests will let students know the actual
state of their preparation so they can then take steps to
remedy gaps in knowledge (making use of test-potentiated
learning), as opposed to studying material they already
know well.

References
Bjork, E.L. & Storm, B.C. (2011). Retrieval experience as a modifier
of future encoding: Another test effect. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 37, 1113–1124.
Butler, A.C., Karpicke, J.D. & Roediger, H.L. (2008). Correcting a metacog-
nitive error: Feedback increases retention of low-confidence correct
response. Journal of Experimental Psychology: Learning, Memory, and
Cognition, 34, 918–928.
Carpenter, S.K., Cepeda, N.J., Rohrer, D., Kang, S.H.K. & Pashler, H. (2012).
Using spacing to enhance diverse forms of learning: Review of recent
research and implications for instruction. Educational Psychology
Review, 24, 369–378.
Dunlosky, J. & Nelson, T.O. (1992). Importance of the kind of cue for
judgments of learning (JOL) and the delayed-JOL effect. Memory &
Cognition, 20, 374–380.

11998_Master4.indd 182 2/11/2020 [Link] PM

 Testing in the Service of Learning 183

Dunlosky, J., Rawson, K.A., Marsh, E.J., Nathan, M.J. & Willingham,
D.T. (2013). Improving students’ learning with effective learning
techniques: Promising directions from cognitive and educational
psychology. Psychological Science in the Public Interest, 14, 4–58.
Karpicke, J.D. & Roediger, H.L. (2007). Repeated retrieval during learning
is the key to long-term retention. Journal of Memory and Language,
57, 151–162.
Kornell, N., Hays, M.J. & Bjork, R.A. (2009). Unsuccessful retrieval attempts
enhance subsequent learning. Journal of Experimental Psychology:
Learning, Memory, and Cognition, 35, 989–998.
Little, J.L., Bjork, E.L., Bjork, R.A. & Angello, G. (2012). Multiple-choice
tests exonerated, at least of some charges: Fostering test-induced
learning and avoiding test-induced forgetting. Psychological Science,
23, 1337–1344.
McDaniel, M.A., Wildman, K.M. & Anderson, L.L. (2012). Using quizzes to
enhance summative-assessment performance in a web-based class:
An experimental study. Journal of Applied Research in Memory and
Cognition, 1, 18–26.
McDermott, K.B., Agarwal, P.K., D’Antonio, L., Roediger, H.L. & McDaniel,
M.A. (2014). Both multiple-choice and short-answer quizzes enhance
later exam performance in middle and high school classes. Journal of
Experimental Psychology: Applied, 20, 3–21.
Morphew, J.W., Silva, M., Hermann, G. & West, M. (2019). Frequent
mastery testing with second chance exams leads to enhanced student
learning in undergraduate engineering. Applied Cognitive Psychology,
34, 168–181. doi: 10.1002/acp.3605
Pashler, H., Bain, P.M., Bottge, B.A., Graesser, A., Koedinger, K., McDaniel,
M. & Metcalfe, J. (2007). Organizing Instruction and Study to Improve
Student Learning (52 pp). US Department of Education report, NCER
2007-2004.
Richland, L.E., Kao, L.S. & Kornell, N. (2008). Can successful tests enhance
learning. In Proceedings of the Twenty-Eighth Annual Conference of the

11998_Master4.indd 183 18/11/2020 [Link] AM

 184 The Science of Learning Physics

Cognitive Science Society, p. 2338. Cognitive Science Society, Austin,

TX.
Roediger, H.L. & Karpicke, J. (2006a). The Power of testing memory: Basic
research and implications for educational practice. Perspectives on
Psychological Science, 1, 181–210.
Roediger, H.L., III & Karpicke, J.D. (2006b). Test enhanced learning: Taking
memory tests improves long-term retention. Psychological Science,
17, 249–255.
Wheeler, M.A., Ewers, M. & Buonanno, J.F. (2003). Different rates, of forget-
ting following study versus test trials. Memory, 11, 571–580.
Wissman, K.T., Rawson, K.A. & Pyc, M.A. (2011). The interim test effect:
Testing prior material can facilitate the learning of new material.
Psychonomic Bulletin & Review, 18, 1140–1147.

11998_Master4.indd 184 2/11/2020 [Link] PM

 c hapter
8 Concluding Remarks

As you can tell from this book and the synthesis we wrote a few
years ago (Docktor & Mestre, 2014), there is extensive research on
the teaching and learning of physics from the Physics Education
Research community as well as the learning sciences. Although
some of this research is being used in the service of teaching
and learning in physics classrooms, we still have a long way to go
before EBIPs are more the norm than the exception.
We have highlighted key research findings about teaching
and learning physics, however our review does not span the
entirety of Physics Education Research. One area of emerging
research in the field that we did not address is how non-cognitive
aspects play a role in learning, such as motivation, expectations
and beliefs about learning physics, mindset, stereotype threat,
and self-efficacy. Many of these non-cognitive aspects of learning
are also related to the under-representation of women and other
minorities in STEM fields, particularly issues of self-efficacy and
societal stereotypes and biases (for example, the notion that
physics is for “brilliant men”). For those interested, a good place
to start is the Physical Review focused collection on gender in
physics: https:/ /[Link]/prper/collections/gender-in-
physics . Hopefully in the coming years we will see advances
in this area with concrete recommendations for instructors to
promote a climate of diversity and inclusion in their classes.

185

11998_Master4.indd 185 2/11/2020 [Link] PM

 186 The Science of Learning Physics

We have focused on college-level intro physics, although

many of the findings and teaching suggestions discussed can
be used in high school physics and in upper-division college
courses. In fact, there is also a research literature on teaching
and learning in upper division physics (see the focused collec-
tion on PER in upper-division physics courses at https: / / jour-
[Link]/prper/collections/upper-division-physics-courses).
Another important hole that needs to be filled is how to
help our physics graduate students, who will become future
professors, learn about the contents of this book. Since grad-
uate courses for prospective physics Ph.D.s are virtually non-
existent, the physics community found a “patch” remedy by
having summer week-long programs for young faculty to learn
about physics education teaching resources (see the Workshop
for New Physics and Astronomy Faculty, https: / / [Link]/
Conferences/newfaculty/[Link]). It is sad that immediately
upon graduating with a Ph.D. in physics, young faculty need to
be trained on EBIPs, as opposed to learning about EBIPs during
the course of their graduate education.
It is also important to have a supportive structure in a
physics department’s administration to nurture and encourage
faculty to use EBIPs in instruction. As we discussed in this book,
many physics professors know about EBIPs and have even tried
some in their teaching, but they often abandon EBIPs citing lack
of support from the administration and time constraints. The
STEM institution-wide response we discussed in Chapter 5 indi-
cates the type of major accomplishments that can be achieved
when teams of like-minded faculty work within a supportive
administration. There is a story (whether or not it is apocry-
phal is unclear) of a department head coming to a young faculty
member’s lecture to observe him/her to provide feedback and to

11998_Master4.indd 186 2/11/2020 [Link] PM

 Concluding Remarks 187

supply information to the faculty member’s tenure portfolio, and

upon finding that the instructor had broken students into small
groups to work collaboratively, he yells back to the instructor
“I’ll come back some other time when you are teaching.”
Funny or not, it is also important to educate senior faculty and
administrators on the value of EBIPs in instruction. Physical
Review also has a focused collection of papers on preparing and
supporting university physics educators: https: / /[Link].
org/prper/collections/preparing-and-supporting-university-
physics-educators
We hope you found the contents of this book useful and
informative, and came away with a newfound excitement to try
out a variety of instructional innovations in your own classroom!

Reference
Docktor, J.L. & Mestre, J.P. (2014). Synthesis of discipline-based
education research in physics. Physical Review Special Topics –
Physics Education Research, 10(020119). https:// [Link]/10.1103/
PhysRevSTPER.10.020119

11998_Master4.indd 187 2/11/2020 [Link] PM

 b2530   International Strategic Relations and China’s National Security: World at the Crossroads

This page intentionally left blank

b2530_FM.indd 6 01-Sep-16 [Link] AM

 Index

A asking/answering deep questions 157,

160
abandoning intuitive notions 6 asking deep questions 147
abstract and concrete representations assessing instruction 130
148 assessing problem solving skills 99
abstract concept 148 assessment in the service of learning
acceleration 14, 29, 30, 32, 180 31
accuracy of solutions 93 assessment of learning 113
acquisition of knowledge 149 assessments 8, 50, 63, 73, 75, 92, 95,
active learning 8, 25, 26, 111, 113, 114, 100, 130, 133, 150, 151, 160, 180,
116, 118, 126, 128, 129, 151 181
active learning classes 126 assessments of student learning 4
active learning strategies 73, 111, attention 42
112, 115 attitudes 102, 125, 133
activity-based physics 150, 151 authentic scientific practices 132
adaptability 85
adopting physicists’ ways of thinking B
6
adoption strategies 127 balls and tracks 17
advanced beginner 40 balls-on-tracks 24
algebraic manipulations 65 barriers to active learning 125
algorithmic problem solving 113 barriers to implementation 125
alternate conceptions 14 basic description 94
amorphous memory store 76 basic mathematical knowledge 85
analogy 64 basic physics concepts 33
analysis 38 beginning students 64
analyzing problems 87 behaviorism 11, 12
anchoring example 33 beliefs 85, 102, 133
anchoring intuitions 33 beliefs about learning 185
angular impulse-angular momentum big idea 38
theorem 39 blank slates 13, 25
angular momentum 61, 71 blocked practice 145, 153, 157, 179
animations 19, 21, 124 brain dump 151
applying concepts 5, 143, 179 bridging analogies 33
applying knowledge flexibly 115
C
apprentice 40, 91
appropriate understanding 24 carrying out investigations 132

189

11998_Master4.indd 189 2/11/2020 [Link] PM

 190 The Science of Learning Physics

carrying out the plan 93 computer-based tools 120

categorization criteria 40 computer coaches 92, 104, 123
categorization tasks 88, 151 ConcepTests 116
categorizing problems 51, 55, 66 concept formation 11, 22
change blindness 42, 46 concept inventories 26, 118, 151
checking calculations 85 concept mapping 31, 151
checking the magnitude 86 concept maps 75, 119
chunking knowledge 84 concepts 2, 5, 6, 8, 11, 33, 38, 39, 45,
class discussion 119 62, 64, 65, 68, 69, 77, 84, 85, 88,
classroom demonstration 26 89, 100, 114, 116, 118, 119, 151,
classroom instruction 7 159, 179, 181
classroom learning 149 concept scenario 89
classroom response systems 133 concept(s)/principle(s) 66, 69
classroom spaces 122 conceptual analysis 49
class-wide discussions 180 conceptual approach 83
clicker questions 131, 169 conceptual development 8, 33, 65
clickers 27, 29, 31, 128 conceptual difficulties 14, 15, 122
clicker technologies 26 conceptual error 160
climate of diversity 185 conceptual knowledge 25, 26, 66, 133
coach 112 conceptual learning 173
coaches of learning 178 conceptually challenging questions
coaching 69, 91, 92 131
cognition 112 conceptual questions 63, 117, 122,
cognitive apprenticeship 91 124, 132
cognitive load 84 conceptual understanding 23, 25, 26,
cognitive mechanisms 12 73, 88, 117, 131
cognitive psychology 42, 141, 152 conceptual worksheets 122
cognitive psychology experiments concrete examples 148
173 conditions for applying concepts 180
collaborative activities 69, 124 conditions for applying principles 70
collaborative groups 45, 89 conditions of applicability 70
collaborative problem solving 128 conducting investigations 130
collect data 7 conflict 13, 14, 30, 33
common-sense knowledge 17 conflict management 132
communicate with students 179 conflict/resolution 15
communication tools 124 connections 76
Communities of Practice (CoP) 127 connections/pathways to memory
competence 4, 40, 142, 153, 155–157, 177
161, 177, 178 conservation laws 71
compiling knowledge pieces 25 conservation of angular momentum
complex real-world problems 90 39
computer animations 17 conservation of energy 21, 40, 43, 61

11998_Master4.indd 190 2/11/2020 [Link] PM

 Index 191

conservation of mechanical energy department chair 128

22, 39, 70, 75, 148, 180 describe the physics 94
conservation of momentum 39 designing experiments 130
constructing appropriate desirable difficulties 143, 146, 147, 157,
understanding 25 160, 177, 178
constructing concepts 23, 24 Devising a Plan 93
constructing explanations 132 diagnostic assessments 99, 133
constructing knowledge 13, 24, 26, 111 diagnostic tests 63, 130
constructivism 11, 13, 24 diagnostic tool 161
constructivist epistemology 112 disciplinary collaborations 129
constructivist learning theory 111 disconnected facts 102
constructivist views of learning 11, 12 discussion section 69
content coverage 76, 125 discussion session 121
context 38 dispensers of information 178
context-rich problems 88, 100, 103, 122 distributed practice 144, 152, 153, 158
contextual experiences 23 doing science 115
contextual knowledge 23 double curse 155
control group 50, 51
cookbook-style 120 E
Cooperative Group Problem Solving Educational Opportunity Program 63
100, 122 educational psychology 141, 152
coordinating knowledge 16 education faculty 129
coordinating knowledge pieces 22 education-related collaborations 129
course grade 66 effective learning practices 160
course management systems 115 effective studying 158
cramming 144, 157, 158, 170, 172, 175 effective study strategies 142, 157, 162
cultural shift 7 efficacy of reforms 128
cumulative exams 145, 146 effortful learning strategies 176
effortful retrieval 176
D
elaborate feedback condition 52
decline in memory 151 electric field 151
decline in performance 150 electric forces 77, 151
deeper processing 157 electricity and magnetism 150
deeply rooted misconceptions 23 electric potential 151
deep processing 176, 179 embedded mentors 128
deep questions 147 energy 42, 150
deep structure 50, 66–68, 148 engagement 23
deep understanding 64 engaging students 111
delayed testing 144, 173 engineering courses 128
demonstrations 119 engineering departments 127
departmental norms 125 engineering disciplines 173
departmental support 125 engineering majors 126

11998_Master4.indd 191 2/11/2020 [Link] PM

 192 The Science of Learning Physics

enhanced student learning 7 exploring problems 87

equal and opposite forces 14 exponential decay 150
equation-based approach 41 exponential rise 149
equation hunting 103 eye movements 42
equation manipulation 70, 78, 100 eye-tracking 42, 88
equations 38, 41, 46, 65, 102
equilibrium 27, 28 F
eradicating misconceptions 14, 16 FACTs 119
erroneous intuitions 5 fading 91
erroneous notions 14 familiarity 4, 142, 153, 156, 157, 161,
erroneous reasoning 16, 22 177, 178, 182
evaluate the answer 95 feedback 91, 92
evaluating solutions 85 feel good strategy 142
evaluation of classroom practices 7 feeling of familiarity 176
evidence-based instructional practices financial resources 131
(EBIPs) 3, 7, 83, 114, 125, 127, finding errors 88
130, 185–187 flipped classrooms 124, 128
evidence-based reforms 127, 128 flow charts 75
exam performance 125, 154 fluency 153, 156, 176
example problems 91 focus the problem 94
exam preparation 157, 159, 160, 182 force 13, 14, 33
exams 76, 92, 95, 104, 177 Force and Motion Conceptual
exam scores 126 Evaluation 26
exam wrapper 159 Force Concept Inventory 26
executed solution 56 force diagram 95
execute the plan 94 force of gravity 14
expectations 85, 102, 185 force problems 77
experienced solvers 92 forces 29, 30, 95, 150
experiment 7 force vectors 77
expertise 37, 39, 41, 64, 65 forgetting curve 150
expert-like analyses 46, 51 formative 113
expert-like approaches 85 formative assessments 31, 113, 115,
expert-like behavior 46 116, 119, 132, 133, 169, 178
expert-like computer tool 50, 51 formative assessment techniques 112
expert-like problem solving 46 formulating explanations 43
expert–novice differences 40 framework 102
expert–novice research 8, 37, 46, 63 free-body diagram 30, 87
experts 3, 38–43, 45, 46, 50, 55, 64, free-recall testing 174
69–71, 77, 78, 83, 87 frequent testing 152, 177
expert’s blind spot 5, 64
G
experts’ knowledge 64
expert’s perspective 22, 64 gas laws 32

11998_Master4.indd 192 2/11/2020 [Link] PM

 Index 193

gauging exam preparation 159 hunting for equations 45

gender in physics 185
general knowledge test 153 I
generation/production 176 identifying principles 70
generation/production of knowledge illusion of understanding 130
178 immediate feedback 150
geometric optics 32 impetus 15
grading 99 implement EBIPs 4
grading structure 76 implications of research on problem
graduate courses 186 solving 100
graduate education 186 important physics ideas 61
greater performance 173 improved communication 93
group activity 69 impulse-momentum theorem 39
group work 121 in-class practices 65
growth mindset 144 incompetent individuals 155
incorrect analysis 50
H
incorrect intuitions 181
hands-on activities 122 incorrect intuitive notions 180
heavy objects fall faster 6, 13, 16, 24, individual accountability 132
112 ineffective learning strategies 176
hierarchical analysis tool 47 ineffective study techniques 142
hierarchical flow chart 75, 76 initiate 40
hierarchical representation of concepts innovative teaching strategies 75
62 institutional barriers 131
hierarchical structure of physics 75 instructional activities 33
hierarchy 75 instructional implications 2
hierarchy of physics 76 instructional reforms 130
high-ability students 143 instructional strategies 32
higher retention 76 instructional suggestions 65
highlighting 8, 142, 156 instructional techniques 142
highlighting/underlining text 142 instructor guidance 131
high-performers 155 integrated classrooms 123
high school physics 186 integrated lab-lecture classroom 122
high-stakes tests 170 intelligent tutors 104
holistic problem solving 83 interactive engagement 114
homework 4, 65, 69, 92, 104, 141 Interactive Lecture Demonstrations
homework assignments 32, 56 32, 118, 119, 131
homework solutions 92 interactive pre-lectures 124
How People Learn 130 interference 151
how students learn 4 interleaved practice 145, 146, 153, 157,
human behavior 11 160, 179
hunches 6 internal decisions 152

11998_Master4.indd 193 2/11/2020 [Link] PM

 194 The Science of Learning Physics

interpreting data 132 lack of time 125

interventions 65 lasting learning 143, 178
introductory course 127 learning 13, 112, 113, 152, 156, 173,
intuitions 6, 176 174, 178, 179
Investigative Science Learning learning and forgetting curves 150
Environment (ISLE) 120 learning curve 149
isomorphic worked examples 147, 160 learning episodes 141
learning experiments 173
J learning gains 126, 176
learning is not a spectator sport 129
jeopardy problems 88, 89, 103
learning research 65
journeyman 40
learning sciences 170, 185
judging 85
learning strategies 160, 169, 177
judgments about learning 141, 154
learning word lists 173
judgments about realistic motion 17
lecture-based instruction 112
judgments of learning (JOL) 154,
lecture-based physics class 113
156–158, 161
lecture demonstrations 32, 33
justifications 77
lecture method 23
justifying principles 56
lectures 3, 4, 62, 122
justifying problem-solving decisions
limiting cases 86
77
link maps 76
Just-in-Time Teaching (JiTT) 116, 124
Logical Progression 99
K logistic sigmoid function 149
longer-term interventions 55
kinematics 39, 95 longitudinal studies 173
knee-jerk intuitions 31 long-term learning 84, 129
knowledge 13 long-term retention 100, 130, 141,
knowledge construction 13, 16, 23 151, 169, 170, 172, 175–177, 179
knowledge in memory 178 Looking Back 93
knowledge in pieces 15, 17, 22, 24 low-ability students 143
knowledge organization 37, 115 low-achieving students 76
knowledge organization in memory low-performers 155
39 low-performing students 155
knowledge pieces 16, 17, 21, 24
KWL chart 119 M

L magnetic fields 151

major principles 47, 56, 62, 75
laboratory activities 130 making predictions 14, 33
laboratory experiences 120, 122 making students’ thinking visible 113,
laboratory experiments 132 178
lack of facilities 131 management systems 124
lack of support 131, 186 manipulating equations 83

11998_Master4.indd 194 4/11/2020 [Link] AM

 Index 195

map of concepts 76 monitoring problem solving processes

massed practice 144, 153, 157 85
massed practice/cramming 152 motion diagram 95
master 40 motivating students 70
mathematical complexity 87 motivation 85, 185
Mathematical Procedures 99 multi-concept problems 62
means–ends analysis 41, 84 multiple-choice exams 69
measured outcomes 126 multiple-choice questions 26, 27, 116
mechanical energy 71 multiple-choice tests 169, 176
memorization 85 multiple contexts 115, 142
memory 38, 39, 65, 84, 148, 149, 176, multiple representations 122
178, 181
memory load 41
N
memory pathways 176 naive theories 14
memory retrieval 172 negative reaction 125
mental dialogue 86 negative reinforcement 12
mental resources 41 neophytes 3
metacognition 85, 86, 112, 152 net force 180
metacognitive judgments 153, 155 network 102
metacognitive monitoring 152, 154 new concept 23
metacognitive skills 85, 104 Newton’s Laws 32
metacognitive strategies 113, 115 Newton’s laws of motion 77
microscopic model 33 Newton’s Second Law 27–30, 39, 40,
midterm exam 57 148, 180, 181
mindset 185 Newton’s Third Law 6, 14, 150
mind wipe 151 Newton’s three laws 61
mini-test 161 non-cognitive aspects of learning 185
Minnesota Assessment of Problem non conservative forces 70
‑
Solving (MAPS) 99 nonsense syllables 149
minority students 63 novices 5, 37–43, 45, 46, 50, 55, 64,
miscalibration 158 76–78, 84, 87, 91, 147, 148
misconceptions 6, 8, 14–17, 22–26, 29, novices’ categorization criteria 51
32, 33, 73, 74, 112, 114, 133, 153, nuanced understanding 23
180 number crunching 93
misinterpreting concepts 23
O
misinterpreting implications 22
mixed ability groups 132 objective judgments 157
model appropriate behaviors 104 objective measures of learning 158
modeling 70, 77, 91 objective measures of preparation
modeling expectations 92 162
model of learning 149 objects fall at the same rate 13, 16
momentum 61, 71, 150 online homework 104, 123, 150

11998_Master4.indd 195 4/11/2020 [Link] AM

 196 The Science of Learning Physics

online homework systems 92 physics scenario 147

open-response tests 169 physics students 19, 21, 22
optimizing learning 141 Physlets 121
organizational skills 145 pictorial representation 95
organizing physics knowledge 65 planning 85
out-of-class experience 123 plan the solution 94
overcoming misconceptions 15 plug-and-chug 41, 84, 87
overconfidence 155 poorly performing students 157, 161
overextending implications 22 posing meaningful problems 147
Overview 95 positive feedback 76
Overview, Case Study Physics 62 positive interdependence 132
positive reinforcement 11
P potential and kinetic energies 75
pattern matching 38, 84 power law decay 150
patterns of forgetting 149 power law function 149
patterns of learning 149 practice exams 160
pedagogical expertise 127 practice tests 154, 182
peer discussions 131 pre-class questions 124
Peer Instruction 116–118, 126, 131 preconceptions 14, 22, 62, 121
peer learning 132 predictions 119
peer reflections 86 pre-lectures 118, 124
peer reviewed publications 129 pre-post learning gains 151
perceptions 126, 156 pretests 32
perceptions of learning 8, 141 principal investigators 128, 129
performance under pressure 177 principle-based categorization 55
personalized tutoring 123 principle-based criteria 55
perspective 38 principle-based reasoning 55
PhET 121 principle identification 78
physical representation 95 principle identification exercises 78
physical situations 119 principles 39–41, 45, 50, 51, 61, 62, 65,
physical world 14, 24 68, 69, 77, 84, 85, 88, 100, 151
Physics Approach 99 principles in categorizing problems
physics concepts 23 51
physics department’s administration prior knowledge 11, 111, 112
186 private “theories” 25
Physics Education Research 114, 185 private understanding 14, 15, 22, 23
physics education teaching resources problem-based learning 89, 128
186 problem categorization 40, 51, 66, 69
physics graduate students 43, 186 problem categorization task 50, 57
physics-naive students 19, 21 problem decomposition 85
physics principles 40, 44 problem posing 88, 89, 103
physics reasoning 31, 88 problem representation 86

11998_Master4.indd 196 2/11/2020 [Link] PM

 Index 197

problem solutions 102, 122 R

problem solving 2, 5, 6, 8, 11, 41, 45,
62–64, 66, 69, 76, 83, 85, 91, 92, ranking task exercises 103
ranking tasks 88, 89
100, 121–123, 132, 146, 173, 179
reading comprehension 85
problem solving assessment 51
realistic motion 21
problem solving differences 45
RealTime Physics (RTP) 121
problem solving difficulties 87
real-world knowledge 85
problem solving frameworks 91, 92,
real-world problems 88, 103
102, 103
recall tests 172, 174
problem solving instruction 91
recitation 121
problem-solving procedures 159
recitation experience 120
problem solving process 87, 92 reconstructing knowledge 26
problem solving skills 99, 104, 124, refining students’ understanding 28,
133 124, 180
problem solving strategies 86 reflection on problem solving 161
problem solving task 50 reflective journal 124
problems’ surface attributes 55 reformed instruction 127
procedural error 160 repeated studying 170, 174, 176, 177,
procedural knowledge 71 179
procedures 38, 41, 181 repetitive task 149
procedures for solving problems 56 representation of problems 87
productively struggling 126 rereading 8, 142
productive problem solving behaviors re-reading notes 156
103 research 8
productive problem-solving skills 86 research-based instructional strategies
productive struggle 144 (RBIS) 114, 119, 120, 125
proficient 40 research evidence 1, 156
promoting expert-like behavior 46 research on learning 6
proposals 129 research on problem solving 83
research productivity 126
Q research university 128
restructuring physical space 123
qualitative analyses 45, 84 re-studying 178, 182
qualitative observations 121 re-studying material 177
qualitative problems 62 retaining information 144
qualitative questions 62 retaining knowledge 145
qualitative reasoning 114 retention 84, 178
qualitative understanding 131 retrieval links 178
quantitative experiments 121 retrieval of information 176
quantitative problems 62 retrieving and using knowledge 178
quantitative problem solving 118, 131 retrospective confidence judgments
quizzes 92, 95, 104 (RCJ) 154

11998_Master4.indd 197 2/11/2020 [Link] PM

 198 The Science of Learning Physics

reviewing course notes 158 spread of innovations 128

reviewing notes 4 STEM 2, 7, 185
rewarding students 104 STEM faculty 129
role of concepts in problem solving 65 stereotype threat 185
role of concepts/principles in problem strategizing a solution 66
solving 65, 66 strategy 56, 57, 69–71
rolling balls 23 strategy writing 56, 61, 69, 70, 77, 78,
rubric 99, 100 93, 147, 161, 180
strategy writing class 61
S strategy writing students 61
scaffolding 71, 91, 92, 144, 152 strong expectations 22
schema 85 structural and cultural change 129
schema-driven strategies 85 structural change 146
science and engineering faculty 129 structural constraints 125
scientific concepts 14 structured cooperative group work
scientific passages 174 132
scientific reasoning skills 133 structured laboratory experiences
scientific thinking 112 132
S-curve 149 student attitudes 125, 126
search-based strategy 41 Student-Centered Active Learning
selecting appropriate concepts 99 for Undergraduate Programs
selection of problems 103 (SCALE-UP) 123
self-efficacy 185 student-centered instructional
self-explanations 86, 104 approaches 114
senior faculty 187 student perceptions 126
sense making 6, 12–14, 114 student resistance 131
shallow learning 143 students’ attitudes 118
short term learning 175 students’ erroneous beliefs 181
similarity-based approach 84 students’ memories 61
simulations 122 students’ perception 126
single-principle problems 71 student strategies 57
slowing down 22 student–student discourse 114, 131
social network analyses 128 Studio classroom 122
solving problems 38, 47, 65, 69, 70, 83, studio format 115
119, 132, 143 study habits 8, 141, 142, 157, 158, 162
spaced/distributed practice 157 studying 141, 152, 156, 170, 172, 173,
spaced practice 144 178, 182
sparse-feedback condition 52 study sessions 154
Specific Application of Physics study strategies 142, 156
principles 99 study time 157
speed 15, 21 subjective judgments 157
speeding up 22 summarizing 142, 156

11998_Master4.indd 198 2/11/2020 [Link] PM

 Index 199

summative tests 169, 170 top-down approach 70

supportive administration 186 topics covered 118
surface attributes 41, 42, 50, 64, 66, 68, traditional class 61
78, 148 traditional laboratories 120, 130
surface features 40, 67, 68, 84 traditional lectures 116, 118
surface similarities 40 transfer 100, 113, 141
survey data 128 transfer of learning 129, 130
synthesis problems 88, 89, 103, 151 transferring knowledge 145
treatment group 51
T
tutorials 122
tacit knowledge 5, 41, 64, 65, 67, 69, 70, Tutorials in Introductory Physics 32
77, 78, 179, 180 tutorial workbook 122
taking tests 177 tutoring systems 123
targeted studying 159 two-column solution 73, 94
teaching and learning 2 two-principle problems 57, 71
teaching and learning of physics 185 type of institution 126
teaching and learning practices 7
teaching by telling 23, 112, 129 U
teaching correct scientific concepts 14
teaching evaluations 4, 7 unbalanced force 29
teaching examples 2 under-confidence 155
teaching generalizations 13 underlining text 156
teaching interventions 25, 65, 102, underlying conceptual structure 66
131, 158, 179 underlying physics 44
teaching practices 65 underperforming students 156, 158,
teach the way you do research 6, 7, 128 162
technology-enhanced resources 116 under-representation of women 185
templates 92, 102 Understanding the Problem 93
tension 29, 74 undesirable learning behaviors 123
tenure portfolio 187 University of Illinois 127
test-enhanced learning 177 upper division physics 186
testing 8, 169, 171–176, 179 Useful Description 99
testing effect 169–171, 173, 176, 178, using models 132
179
test-potentiated learning 176, 182 V
tests 141, 181
textbook presentations 124 variable names 55
textbooks 76 vector 151
thematically-relevant 42 velocity 14, 32
theoretical description 94 video-based materials 124
think aloud 21 visual cognition 42
think-pair-share 31, 116, 119 visualizing problems 85

11998_Master4.indd 199 2/11/2020 [Link] PM

 200 The Science of Learning Physics

W work done by conservative and

non-conservative forces 75
web-based homework 69 worked examples 146, 147
web-based pre-lectures 126 work-energy theorem 39, 61, 75
well-organized memory 64 working collaboratively 127
what experts and novices notice 45 working memory 84
what experts notice 41 worksheets 32
what novices notice 42 Workshop for New Physics and
what, why and how of solving a Astronomy Faculty 186
problem 56 Workshop Physics 115, 121, 122
WISE strategy 93 writing strategies 57
word recall studies 174
work 70 Y
work and energy 75
work collaboratively 73 young faculty 186

11998_Master4.indd 200 2/11/2020 [Link] PM

 About the Authors

José P. Mestre is an emeritus Professor of Physics and Educational

Psychology at the University of Illinois at Urbana-Champaign.
Since earning his Ph.D. in theoretical nuclear physics, his research
has focused on the learning of physics, making many pioneering
contributions in areas such as the acquisition and use of knowl-
edge by experts and novices, transfer of learning, and problem
solving. He was among the first to publish scholarly articles on
the use of classroom polling technologies (clickers) to promote
active learning in large classes, and is a co-developer of Minds-On
Physics, an activity-based high school physics curriculum that is
heavily informed by learning research. Most recently, his research
has focused on applications of methodologies common in cog-
nitive science (e.g., eye-tracking) to study learning and informa-
tion processing by physics novices and experts. He has served
on many national committees and boards for organizations such
as the National Research Council, The College Board and Educa-
tional Testing Service and the American Association of Physics
Teachers, and has offered Congressional testimony on The Science
of Learning. He has published numerous research and review arti-
cles on science learning and teaching, and has co-authored or
co-edited 19 books. Mestre served as Associate Dean at the Univer-
sity of Massachusetts-Amherst in the College of Natural Sciences
and Mathematics, and both as Chair of the Department of Educa-
tional Psychology and as Associate Dean for Research at the Col-
lege of Education at the University of Illinois-Urbana/Champaign.
He is a Fellow of the American Physical Society with citation: “For
ground-breaking applications of principles and methodologies

201

11998_Master4.indd 201 2/11/2020 [Link] PM

 202 The Science of Learning Physics

from cognitive science to physics education research and for

elucidating expert-novice performance differences in physics
learning and problem solving.”

Jennifer L. Docktor is an Associate Professor of Physics at the

University of Wisconsin – La Crosse. After completing a physics
teacher preparation program at North Dakota State University she
earned her M.S. in High Energy Physics and her Ph.D. in Physics
Education Research at the University of Minnesota. Her doctoral
research focused on the Development and Validation of a Physics
Problem-Solving Assessment Rubric. She spent two years as a
postdoctoral fellow in Cognitive Science at the Beckman Institute
for Advanced Science and Technology at the University of Illi-
nois at Urbana-Champaign as part of a unique interdisciplinary
research group on physics learning and cognition. She has col-
laborated with José on a variety of projects including conceptual
problem solving in high school physics, categorization, and using
eye-tracking technology to study physics representations. In 2010,
they co-authored a commissioned paper for the National Acad-
emies report on Discipline-Based Education Research which was
later published in Physical Review as the article Synthesis of Dis-
cipline-Based Education Research in Physics. In addition to these
endeavors she is involved in several national efforts surrounding
physics teacher preparation including the Physics Teacher Educa-
tion Coalition (PhysTEC) and the project Get the Facts Out about
STEM Teacher Recruitment. She has served on national commit-
tees for the American Association of Physics Teachers and cur-
rently serves as editor-in-chief for the American Physical Society’s
Forum on Education newsletter.

11998_Master4.indd 202 2/11/2020 [Link] PM