Chapter 2

2
Building Equations and Functions
Focus in High School Mathematics: Reasoning and Sense Making (NCTM 2009) stresses the
importance of reasoning with algebraic symbols, equations, and functions. These skills are precisely
the ones that cause students so much difficulty in the transition from arithmetic to algebra. Indeed,
teachers’ assessments of the areas that cause beginning students to struggle are overwhelmingly uni-
form. The following areas often present challenges:

1. Expressing generality with algebraic notation, including function notation

2. Reasoning about slope, graphing lines, and finding equations of lines
3. Building and using algebraic functions
4. Setting up the appropriate equations to solve word problems

At first glance this list looks like a collection of

disparate topics. Yet, looking underneath the topics The habit of seeking and expressing regular-
and considering the kind of reasoning that would help ity in repeated calculations runs throughout
students master them reveals a remarkable similarity. the specific components that Focus in High
A key component of all of these topics is the reason- School Mathematics: Reasoning and Sense
Making identifies in the reasoning habits that
ing habit of seeking and expressing regularity in re- it describes. For example, those components
peated calculations. include—
This habit manifests itself when one is perform- • identifying relevant mathematical
ing the same calculation over and over and begins to concepts, procedures, or representations;
notice the “rhythm” in the operations. Articulating • seeking patterns and relationships;
this regularity leads to a generic algorithm, which is • looking for hidden structure;
typically expressed with algebraic symbols and can • making purposeful use of procedures;
be applied to any instance and transformed to reveal
• organizing the solution, including
additional meaning, often leading to a solution of the calculations.
problem at hand.
This chapter explores how this habit can be used
to bring coherence to three topics in the high school
curriculum:
1. Building equations to model situations
2. Finding lines of best fit
3. Calculating monthly loan payments

25
 26 Focus in High School Mathematics: Algebra

From Calculations to Equations

Focus in High School Mathematics: Reasoning and Sense Making calls for reasoned solving of
equations—seeing steps in the solution of an equation as logical deductions. However, before equations
can be solved, they have to be constructed, by using what that publication calls the meaningful use of
symbols. Teachers report that many students, even students who are quite skillful in solving linear and
quadratic equations, have a very hard time building equations that model particular situations.
For example, consider how hard it is for students to set up
an equation that they can use to solve an algebra word problem.
Teachers of algebra typically
comment, “My students can Reasons for their difficulties typically include the reading levels
solve the equations; setting them and the unfamiliar contexts of such problems. Still, there has
up is the hard part.” to be more to students’ difficulties than these surface features.
Consider, for example, the following two problems:

Problem 1: The driving distance from Boston to Chicago is 990 miles. Rico drives from
Boston to Chicago at an average speed of 50 mph and returns at an average speed of
60 mph. For how many hours is Rico on the road?

Problem 2: Rico drives from Boston to Chicago at an average speed of 50 mph and
returns at an average speed of 60 mph. Rico is on the road for 36 hours. What is the driving
distance from Boston to Chicago?

The problems have identical reading levels and context. But teachers report that many students
who can solve problem 1 are baffled by problem 2. A significant body of research can help to ex-
plain this phenomenon (Bransford, Brown, and Cocking 1999;
Breidenbach et al. 1992; Cuoco 1995; Piaget 1972; Sfard 1991;
Problems 1 and 2 and others like Sfard and Linchevski 1994; Slavit 1997).
them make no pretense of being Problem 1 can be solved with isolated calculations, as
rooted in reality. Indeed, their shown in figure 2.1. However, problem 2 requires that the stu-
puzzle-like quality makes them dent encapsulate these isolated individual calculations into a
ideal vehicles for developing the
coherent process—an algorithm that calculates the time on the
reasoning habits under consider-
ation. road from the distance traveled—so that they can invert the
algorithm (reasoned solving again) to come up with a distance
that will produce 36 hours.

Fig. 2.1. Isolated calculations for solving problem 1

Building Equations and Functions 27

In this situation, the reasoning habit of “expressing the

Students who solve this problem with the
rhythm” in a calculation can be of great use to them. The aid of a calculator typically hit the “=”
basic idea is for them to guess at an answer to problem 2 key very often.
and check their guess as if they were working on problem
1, keeping track of their steps. The purpose of the students’
guess is not to stumble on (or to approximate) the correct answer; rather, it is to help them construct
a “checking algorithm” that will work for any guess. So, students can make several guesses until they
are able to express their checking algorithm in algebraic symbols. The following example shows how
a student might approach this problem; figure 2.1 shows the student’s calculations.

Student: I began by guessing that the distance is 1000 miles. I then divided 1000 by 50 and
1000 by 60. Then I added the answers together to see if I got 36. I didn’t, so I made
another guess—950 miles. Let’s see: 950 divided by 50 plus 950 divided by 60. Is that
36? No, but a general method is evolving that might allow me to check any guess. My
guess-checker is

guess guess ?
+ = 36 .
50 60

So my equation is guess guess

+ = 36,
50 60

or, letting x stand for the unknown correct guess,

x x
+ = 36.
50 60

In the classroom
In the following vignette, two teachers sort out the dif- “Guess-and-check” has long been a
ference between the solution method described for problem 2 popular method for finding or approxi-
and traditional guess-and-check. Mr. Thomas Gradgrind and mating solutions to all kinds of prob-
lems. What we present here isn’t quite
Ms. Maria Agnesi are talking about their algebra classes. Tom
the same—the guesses are just scaffolds
is sharing his concerns about a lesson involving the relation- to help students build equations. The
ships among distance, rate, and time. real goal is to build a generic “guess
checker”—the equation that can be
solved to produce the exact solution.
Tom: Maria, I just don’t know what to do. Right
now in my class we are working on distance-
rate problems. We had already talked about
the relationship among distance, rate, and time. I then gave students a problem like
this: “The driving distance from Boston to Chicago is 990 miles. Rico drives from
Boston to Chicago at an average speed of 50 mph and returns at an average speed of
60 mph. How many hours was Rico on the road?” Almost every student was able to
come up with the correct solution.
Maria: How do you know students understood what they were doing?
28 Focus in High School Mathematics: Algebra

Tom: When I walked around to see what students were doing, I saw that they were dividing
the one-way distance by each respective speed and then adding both times to get the
total hours. I asked students to explain why they were dividing, and they were able to
talk about d = rt.
Maria: So what exactly is your concern?
Tom: After asking students to determine the total time for the problem, I switched the prob-
lem a bit. I gave students the same speeds as before but told them this time that Rico
was making a round trip from Fort Lauderdale, Florida, to Reston, Virginia. I asked
them to figure out the one-way distance between the two cities if the total driving time
was 38.5 hours. They didn’t even know how to begin the problem. So I ended up just
telling them how to set up the equation to solve the problem.
Maria: What understanding do you think your students have about the problem?
Tom: None. I gave them a formula of sorts that can help them solve these types of prob-
lems. What else was I supposed to do?
Maria: This is a great opportunity to help students develop as problem solvers while at the
same time giving them a chance to make meaning out of algebraic symbols. Let me
show you what I mean. Given two different rates, one each for the trip out and back,
your students were able to determine the total trip time. Well, have them use this
method to help solve the second problem.
Tom: I’m not sure that I follow. Students didn’t set up an equation initially, and they clearly
couldn’t set up an equation for the second problem.
Maria: Suggest to students that they “guess” a distance and use it to check if they are correct
by calculating if they get the same total driving time.
Tom: But how does guessing help them? I don’t want them to keep guessing and checking.
It’s not efficient, and they may never get the right answer.
Maria: The “guessing” is just the means for them to develop an algorithm. Have students
keep track of the steps they are using to check their guess. Here, let’s try one. Begin
with a guess of 500 miles and conjecture what students will do.
Tom: They will divide 500 by 50 and then divide 500 by 60 and add them together to get
the total time—just as they did for the initial problem.
Maria: Suggest they try another number for the distance between the cities, like maybe
800 miles. What will they do?
Tom: The same thing as before. They will divide 800 by 50 and then divide 800 by 60 and
add them together. Oh, I see what you’re getting at. After a couple of times, students
can begin to see a pattern. I can coach them to come up with a type of verbal descrip-
tion, like

Miles between cities Miles between cities

+ = 38.5 hours .
50 mph 60 mph
Building Equations and Functions 29

Maria: Now students can simply replace “miles between cities” with x and they have an equa-
tion where the variables and equation make sense to them. They have also developed a
method that will come in very handy in the future for setting up equations.

This habit of trying numerical examples until the structure of an algorithm becomes clear cap-
tures a very common process that is a useful tool throughout algebra: we carry out several concrete
examples of a process that we don’t quite “have in our heads” to find regularity and build a generic
algorithm that describes every instance of the calculation. As another example, let’s look at how this
same reasoning can be used to find equations of lines and other curves.

Equations of lines and other curves

Roger Howe (forthcoming) makes a careful
Suppose that a student who is new to alge- analysis of word problems, showing how arith-
bra and comes to it with no formulas is asked metic and algebraic approaches can be devel-
to find the equation of the vertical line l that oped and used in tandem.
passes through the point with coordinates (5, 4).
Students can draw the line, and, just as in the
word problem example, they can guess points and check to see if they are on l. For example, trying
some points, like (5, 1), (3, 4), (2, 2), and (5, 17), leads to a generic guess-checker: To see if a point
is on l, you check that its x-coordinate is 5. This leads to the guess-checker x 5 and the equation
x = 5.
This method works well for vertical and
horizontal lines, and even for special lines like
To be completely rigorous, students should the one that bisects the first and third quadrant.
check that a point is on l if and only if its x- But what about lines for which there is no simple
coordinate is 5. The equation x = 5 is often re- guess-checker? The idea is to find a geometric
ferred to as a point tester for l.
characterization of such a line and then to de-
velop a guess-checker based on that characteriza-
tion. One such characterization uses slope.
In first-year algebra, students study slope, and one fact about slope that often comes up is that
three points on the coordinate plane but not all on the same vertical line are collinear if and only if
the slope between any two of them is the same. Figure 2.2 shows three points that satisfy this condi-
tion and three points that do not.

Fig. 2.2. On the left, points A, B, and C are collinear. On the right, they are not.
 30 Focus in High School Mathematics: Algebra

In the figure, if we let m(A, B) denote the slope between A and B (calculated as change in y di-
vided by change in x), then the collinearity condition can be stated like this:

Three points A, B, and C that do not all lie on the same vertical line are collinear if and only if
m(A, B) = m(B, C).

To prove this characterization of collinearity, This criterion for collinearity can be used to find the
one needs some facts about similar triangles. In equation of a line between two points. Suppose, for
figure 2.2, the two triangles on the left are simi-
example, that students are asked to find an equation
lar; the two triangles on the right are not. 
for AB if A is the point (5, 1) and B is the point (–3,
6). Imagine again that they have no knowledge of y =
mx + b or related formalisms. They can reason as fol-
5
lows: The slope between A and B is − 8 , and we can guess at some points C and check to see whether
or not C is collinear with A and B by checking slopes:

C m(B, C) C on AB ?
3
−
(1, 3) 4 no
3
−
(7, 0) 5 no
5
−
(13, –4) 8 yes
1
−
(–6, 7) 3 no

How might the students check a generic guess, say C(x, y)? They could calculate the slope be-
5 5
tween C(x, y) and B(–3, 6), and see if the slope is − 8 . The guess-checker is m(B, C) − 8 , or

yy−−66 ? 55 .
==−−
xx++33 88

So an equation of AB is

y−6 5
=− .
x+3 8

From here, the students can simplify the equation to get it into a more standard form.

It is certainly true that algebra students need

This outline glosses over some important de-
tails that would need classroom discussion. For
to become fluent in understanding the correspond-
example, the special case when x = –3 needs ence between linear equations and their graphs. In
attention, and students should check this result many applications, they will need to be able to read
against the result obtained when one checks the the slope and y-intercept of a line from its equation,
slope from C to A instead of from C to B. and given these features, they will need to be able
to draw a line.
Building Equations and Functions 31

So, why not jump directly to the development of these

Students frequently think that
skills without the guess-checking activities? A number of y = 3x + 4 is a “code” that means
reasons support starting with an approach like the one out- “put a point at (0, 4), then go over
lined here: 1 and up 3, put a point there, and
then draw a line between these
1. Several research studies (Greenes et. al. 2007; two points.”
Goldenberg 1988, 1991) show that students who can
fluently graph equations like y = 3x + 4 often can’t use
the equation to see if a given point is on the graph. Building equations from the repeated
testing of numerical examples reinforces the “Cartesian connection” that a point is on the
graph of an equation if and only if its coordinates satisfy the equation.

2. This same reasoning habit can be applied to other equations and their graphs. For example,
to find an equation for the circle with center C(3, 7) and radius 5, students who are used to
thinking this way might ask, “How can I check to see if a given point P is on the circle?”
They might then follow up this question by asking, “Is the distance from P to C equal to 5?”
Students equipped with the Pythagorean theorem would be able to write down the equation
from this characterization long before learning about the formula (x – h)2 + (y – k)2 = r2.

3. The very act of articulating a guess-checking algorithm in a way that can be formulated with al-
gebraic symbols is a skill that will serve students well throughout mathematics and related fields.

Automaticity in graphing is very important. However,

jumping directly to the automatic applications of methods like Articulating a guess-checking
using “y = mx + b” can disconnect students’ skill in graphing algorithm as described will serve
equations from the underlying meaning that connects equa- students well in dealing with
algebra word problems and area
tions and their graphs. formulas. Eventually, students
should be able to go from a prob-
Fitting Lines to Data lem directly to an equation or
function that models the problem’s
Imagine a class in which students have developed auto- situation, but jumping directly to
“problems by type” or rules like
maticity with the connection between lines and their equa- 1
“let x= ...” or A = 2 (b1 + b2)h can
tions. One application of this set of skills is to provide some
disconnect the symbols from their
insight into the sometimes-mysterious calculator button that meaning for students.
calculates the line of best fit for a set of data. After students
have had appropriate informal experiences with data trends,
many high school curricula give a definition of a best-fit line
in a manner such as the following:

For a set of points {( xi , yi )}i =1 , the line of best fit minimizes the sum of the squares of the devia-
n

tions in y-values. In other words, it is a line with equation y = ax + b so that the sum

∑ ( y − ( ax + b ))
2
i i
Notice that a and b are the
i =1 variables here. Think of the set
of all possible lines dancing
through the data, each one with
is as small as possible.
its own “badness,” or lack of fit
(its particular sum of squares of
The actual derivation of the a and b that minimize this sum is deviations in y-heights from the
usually left for linear algebra or calculus. However, a little knowl- data points). The use of dynamic
edge of quadratic functions (and how to minimize them), along geometry software can make this
image precise (Cuoco and Gold-
with the habit of abstracting from calculations, can take students
enberg 1996).
quite a bit further.