MTED 5320

Tinkerplots and the Development of Quantitative Reasoning

Jose L. Bautista


The University of Texas at El Paso


Table of Contents

MTED 5320 .................................................................................................................................... 1
Introduction ..................................................................................................................................... 1
Advantages of Tinkerplots .............................................................................................................. 3
The Need ......................................................................................................................................... 5
Teachers Roles ............................................................................................................................... 7
Data based Inquiry .......................................................................................................................... 9
Problem Context ............................................................................................................................. 9
Data Representation ...................................................................................................................... 10
Implications of Using Tinkerplots in the Classroom .................................................................... 10
Bibliography ................................................................................................................................. 12


1

Introduction
Inferential statistical reasoning or quantitative reasoning is a statement about a population or
process that is inferred from a sample, along with an explicit level of confidence. Randomization
is usually introduced to young students (middle school) through chance devices like coins and
dice; however, it is very difficult for students to form a good conceptual understanding of the
activities in which they get engaged during data collection. Traditional pedagogies in which
students learn with "ready-made" mathematical knowledge often leads to students not being
receptive of the mathematical ideas.
Therefore, research in statistics education has deemed necessary to start by generating informal
statistical inferences in students before submerging them in formal statistical methods. This
would be the base in which students would support the development of more formal conceptual
and procedural understandings of statistical inferences at a later time of their lives. According to
a research study (Makar, Bakker, & Ben-Zvi, 2011, p. 152), informal statistical inference is
nurtured by statistical knowledge, knowledge about the problem context, and useful norms and
habits developed over time, and is supported by an inquiry-based environment (see figure 1). The
primary goal is to learn to make sense to form students that are better prepared to use
mathematics in everyday life and to answer any questions that might appear on a standardized
test.




2












Some studies suggest that computer visualization tools can enhance teachers content and
pedagogical knowledge of statistics which can in turn increase students learning (Meletiou-
Mavrotheris, Paparistodemou, & Stylianou, 2009). One of the most useful and known computer
programs used to explore issues associated with developing ideas of informal inference is
Tinkerplots. This computer software has the potential to present students with meaningful
investigations that lead to an appreciation of the types of questions that informal inference can
help to answer. Many of these investigations in which students get engaged could have not been
possible without the use of Tinkerplots and while some of them could, the software package can
save time and add creativity and student ownership to the production of evidence and the
creation of a final report answering the initial questions.
Statistical
Knowledge

Problem
Context

Useful
norms
Problem
Context
Habits
Sense
Making
Informal Statistical Inference
Elements
Inquiry-Based Environment
Figure 1: Informal Statistical Inference Model

3

Advantages of Tinkerplots
The ever increasing technological advances in our society bring also as a consequence an
increasing availability of mathematics education software. It is impossible and would be
negligent to neglect the importance that technology plays among the student population and how
students have learned to embrace technology in their daily lives. Tinkerplots is a dynamic
graphic software created with a constructivist view in mind especially for middle school
students. The software is relatively simple to use, data entry takes place directly using data cards
(see figure 3) or a table, or by importing files from spreadsheets or web sites. Data cards mimic
very closely the data cards that students are familiar with when working collaboratively in their
classroom. Data are first presented in a random fashion in a two dimensional form on the screen,
as can be observed in figure 4.


Figure 2: Example of Data Card


4


Figure 3: Tinkerplots screenshots

Students can create bins from which plots are generated by moving variables, called attributes,
using a drag-and-drop movement to drop the variable into the plot area. Dragging the icons to the
plot left or right creates more or fewer bins. Tools available for interpreting data presented in
plots include reference lines, dividers, hat plots and box plots, group size and percent, median
and mean, and labels.
The value of Tinkerplots as an educational software was evaluated (Fitzallen, 2007) in Australia
in 2007 and found to satisfy the evaluation criteria, which included: being accessible and easy to
use, assisting recall of knowledge and representation in multiple forms, facilitating transfer
between mathematical and natural language, providing extended memory when organizing and
reorganizing data, allowing multiple entry points for abstraction and concepts, and providing
visual representation for interpretation and expression (Watson & Donne, 2009, p. 2).
Also, the use of the hat plot for representational purposes has been found to be and engaging and
easy to use as a tool for data analysis (Watson J. M., 2008). An example of hat plots covering

5

about 50% of the data in the crowns is shown in figure 5. The low 25% and high 25% in the
brims for height data for samples of adult males and females is shown.

Figure 4: Example of hat plots

Tinkerplots allows students to understand beginning inference by exploring what is possible and
what could be possible.
In order for students to develop a data-oriented mindset and good data literacy skills, significant
change must be made to the typical instructional methods employed in the classroom to teach
statistical concepts. The implementation of Tinkerplots as a dynamic software has the high
potential, according to research, to allow students to explore data in ways that had not been
possible for them before.
The Need
The notion that one can use collected and available data to make claims or inferences about the
unknown is the most powerful tool in statistics. Statistics promote the formulation and

6

explanations of real life events or phenomena based on available or collected data. This leads
students to provide explanations that draw on knowledge about statistics on the context being
studied. Difficulties in studying formal statistics courses at a high school or university level have
induced teachers and researchers to explore data analysis with less formal methods with the
intentions to bridge and create a smooth transition between the latter and more formal statistical
inference methods.
The literature is filled with evidence of students difficulties in the area of statistical inference
methods. Traditional mathematical pedagogies usually guide students to obtain right answers
without allowing them to fully understand the mathematical concepts involved. Some studies
attribute that much difficulty for learners arises from the procedural focus given in many
statistics courses and the lack of direction towards reasoning and sense making (Madden, 2011,
p. 111). This is very important because employers very often need people that are capable of
drawing conclusions beyond available data to make judgments in the workplace, often without
formal statistical techniques. However, too often in schools, teachers focus their pedagogical
instruction of statistical learning on the graph, the conclusion, or the computation, but most of
the time the pedagogy employed neglects the part that is involved in sense making. In order to
generate informal inferential reasoning processes, students must be exposed to repeated
experiences with data over time to develop norms and habits of working collaboratively with
other students in an inquiry-based environment. The latter norms serve to induce students to
utilize statistical concepts as tools to seek evidence for the interpretations of data, but most
importantly, the inquiry process should be a sense-making process driven by doubt; this process
will then effectively lead to students giving inferences and explanations.

7

The easy and simple ways in which students can create different arrangements and
representations of the data collected in Tinkerplots, leading to the formulation of inferences,
make this software package a suitable choice for introducing students to informal statistical
inference methods. Research literature in the area of statistics indicates very poor statistical
intuition among most college-level students and adults; however, this is not the result of peoples
lack of intelligence to understand statistics but the deficient learning environments that
characterized our educational system.
Teachers Roles
There is a different issue associated with effective instruction of informal inferential statistics:
teachers who are not really prepared to teach a statistics course and that are not confident in the
level of understanding they possess about the subject. The writer of this report has himself heard
from in-service mathematics teachers that statistics is the most difficult subject to teach and to
understand for them. In one occasion during a graduate class discourse, an in-service teacher
asked a question about the simple relationship of how wide a bell shape is and the standard
deviation. For example, having two similar bell shapes with same mean, one having wider spread
than the other means that the data deviates more on the former than on the latter, just to view it in
very simple terms (see figure 2). This shows that if teachers have trouble having a good
conceptual understanding of formal simple statistical concepts, then it is expected that students
would have more trouble understanding the concepts and even much more trouble making sense
of the statistical methods. Thus, teachers who have limited experiences or have difficulty
understanding statistical processes, appear to fail to understand and appreciate what their
students experience as learners (Watson J. M., 2008, p. 80).

8


Figure 5: Dist. Equal means and Diff. Std. Dev.

Even though, many teachers have experience with professional development courses in
mathematics, many of them do not feel confident when teaching statistics, this creates a negative
effect which does not relate well to the goals of the investigations in which they engage along
with their students. Mathematics teachers are much more effective when they understand their
role as classroom managers to facilitate students professional growth, but they must be fully
prepared and must have a strong conceptual and procedural knowledge of the subject. Preparing
teachers for their roles as managers of the learning environment presents new challenges both to
novices and experienced mathematics teachers. The role of the teachers engaged in informal
statistical inference environments involves a lot more than just presenting the statistics
knowledge and pouring it into the students minds. A key aspect for teachers to be successful in
this field is to be highly knowledgeable about statistics so that they can present the statistical
concepts in a non-traditional manner, with the help of software packages like Tinkerplots. If the
aforementioned conditions are not met, then the learning environment is definitely doomed for
failure because the teacher lacks the problem-solving strategies that students need to be exposed

9

to. Thus, teachers must engage students in problem solving and applying their knowledge to
different situations. Therefore, teachers manage the classroom environment with the ultimate
goal of facilitating students' professional growth.
Being prepared to use technology and knowing and knowing how that technology can support
students conceptual understandings must become integral skills in every teachers professional
repertoire. However, research shows that the adoption of technology in the classroom is strongly
influenced by teachers attitudes and beliefs about technology (Angers & Machtmes, 2005, p.
771).
Data based Inquiry
Inquiry processes play an important role in informal inferential reasoning. In order to understand
how people learn something new, and particularly how students form inferences, it is of extreme
importance that students make sense of the data at hand. Inquiry is a fundamental process in all
human cognition (Makar, Bakker, & Ben-Zvi, 2011, p. 152). The inquiry processes that occur in
students minds are needed to develop reasoning and judgments when drawing conclusions
beyond the available data. This is an important aspect that helps students and their teachers to
make reasonable inferences without learning formal techniques.
Problem Context
Inquiry-based mental processes must derive from the uncertainty or doubt that the problem
context provides. This context allows students to find meaning in observed patterns about the
real situation by weighting up explanations and seeking competing explanations, in other words,
the context forms a connection between the abstract nature of the problem and how it can be

10

interpreted into a story. The use of visual stimulations, particularly randomization tests, in a
dynamical technological environment like Tinkerplots, is very helpful when comparing
distributions. Comparing distributions can serve as a vehicle to support increasingly
sophisticated statistical arguments while investigating characteristics of distributions; measures,
types, and ramifications of variability; and the construction and interpretation of sampling
distributions.
Data Representation
Tinkerplots randomization tests are used to find if certain differences in measures (like the
mean, medians, and other statistics of interest) from the experimental data could be considered
statistically significant. As a pedagogical tool, Tinkerplots provides a representation from which
students can reason about population differences based on samples. Thus, randomization tests
help students decide or infer whether perceived distribution differences are significant.
Tinkerplots provides the conditions to illuminate relationships among distribution, variability,
and sampling and to subject informal inferential reasoning to a more extended scrutiny by
learners sense making.
Implications of Using Tinkerplots in the Classroom
According to several research studies, mentioned in the body of the report, there are strong
indications that support the belief that using Tinkerplots in the statistics classroom can promote
active knowledge construction by encouraging students to build, refine, and reorganize their
prior understandings about statistics. The Tinkerplots software has been designed with young
students in mind, particularly middle school students; it can provide an inquiry-based learning

11

environment. Rather than having students repeatedly practice how to calculate measures such as
the mean and median by hand, instruction should focus on helping them understand how to use
those measures in making comparisons, predictions, and generalizations. Tinkerplots is a useful
pedagogical tool in achieving this purpose. It provides a medium for the design of activities that
integrate experiential and formal pieces of knowledge, allowing students to make direct
connections between the collected data and its formal representations. With Tinkerplots, students
can experiment with statistical ideas, learn to progressively articulate their informal theories, use
them to make conjectures, and then use the experimental results to test and modify those
conjectures.
Therefore, as suggested by numerous research articles, Tinkerplots is an effective software
program for the teaching and learning of early statistical reasoning that forms a strong support
for the more formal conceptual and procedural understanding of statistics in later years.


12

Bibliography
Angers, J., & Machtmes, K. (2005). An Ethnographic Case Study of Beliefs, Context Factors,
and Practices of Teachers Integrating Technology. The Qualitative Report, 10(4), 771-
794.
Madden, S. R. (2011). Statistically, Technologically, and Contextually Provocative Tasks:
Supporting Teachers Informal Inferential Reasoning . Mathematical Thinking and
Learning, 109-131.
Makar, K., Bakker, A., & Ben-Zvi, D. (2011). The Reasoning Behind Informal Statistical
Inference. Mathematical Thinking & Learning, 13(1/2), 152-173.
Meletiou-Mavrotheris, M., Paparistodemou, E., & Stylianou, D. (2009). Enhancing Statistics
Instruction in Elementary Schools: Integrating Technology in Professional Development.
The Monatana Mathematics Enthusiast, 57-78.
Watson, J. M. (2008). Exploring Beginning Inference with Novice Grade 7 Students. Statistics
Research Journal, 59-82.
Watson, J., & Donne, J. (2009). TinkerPlots as a Research Tool to Explore Student
Understanding. Technology Innovations in Statistics Educattion, 1-35.