Mathematics Education Research Journal 2006, Vol. 18, No.

1, 47–77

Effects of Classroom Instruction on Students’

Understanding of Quadratic Equations
Pongchawee Vaiyavutjamai M. A. (Ken) Clements
Chiang Mai University Illinois State University

Two hundred and thirty-one students in six Grade 9 classes in two government
secondary schools located near Chiang Mai, Thailand, attempted to solve the
same 18 quadratic equations before and after participating in 11 lessons on
quadratic equations. Data from the students’ written responses to the
equations, together with data in the form of transcripts of 36 interviews with 18
interviewees (a high performer, a medium performer, and a low performer
from each of the six classes), were analysed. Using a rubric for assessing
students’ understanding, the analysis revealed that at the post-teaching stage
students improved their performance on quadratic equations and had a better
understanding of associated concepts than they had at the pre-teaching stage.
However, many were still confused about the concepts of a variable and of a
“solution” to a quadratic equation. After the lessons, most students had
acquired neither an instrumental nor a relational understanding of the
mathematics associated with solving elementary quadratic equations.

This report explores the impact of traditional teaching approaches on Grade

9 students’ learning of mathematics associated with solving quadratic
equations. Data were gathered in two government secondary schools in
Thailand during the period from December 2002 to February 2003. The main
focus in the report is not on what transpired in the lessons themselves, but
rather on how, and to what extent, the lessons affected student performance
on, and understanding of, quadratic equations.
At the outset it will be useful to comment on what we mean by
“traditional teaching approaches”. Analyses of lessons within the large
collection of videotapes made for the 1995 TIMSS Video Study (Stigler &
Hiebert, 1999) and the 1999 TIMSS Video Study (Hiebert et al., 2003) revealed
that in all the participating countries except one – Japan was the exception –
most secondary school mathematics lessons proceeded according to a
review/introduction/model example/seatwork/ summary structure
(Hiebert et al., 2003; Hollingsworth, Lokan & McCrae, 2003). When, in this
article, the term “traditional” is used in relation to lessons it can be assumed
that the lessons were structured in that way.
In traditional lessons the teacher talks more than the students and
classroom discourse patterns feature many sequences of chorused answers,
with teachers channeling students’ thinking towards answers that they want
the students to give (Vaiyavutjamai, 2004a, 2004b; Voigt, 1994).
Vaiyavutjamai’s (2004a) detailed analysis of algebra classrooms in Grade 9
classes in Thailand, and Lim’s (2000) analyses of Grade 10 O-level algebra
classrooms in Brunei Darussalam, revealed that most teaching, by local as
well as by expatriate teachers, was of this elicitation variety, with exposition
48 Vaiyavutjamai & Clements

(“teacher telling”) also being common. Vaiyavutjamai (2204a) and Lim (2000)
reported that there was always a strong emphasis on symbol manipulation,
with less attention being given to the meanings of the symbols.
It would be wrong to think that such methods are confined to Asian
nations, for TIMSS video studies suggest that forms of traditional instruction
are still widely used in mathematics classes, including algebra classes, in
many nations. Yet researchers have shied away from investigating the
immediate effects of such teaching on student learning. This article responds
to the need to become more aware of those effects.

Review of Related Literature

Research into the Teaching and Learning of Quadratic Equations
Considering the importance of quadratic equations in the history of
mathematics and in secondary school mathematics curricula around the
world, it is surprising that research into the teaching and learning of
quadratic equations has been so sparse. Publications of the National Council
of Teachers of Mathematics (NCTM) in the United States, for example, have
rarely reported on research into the teaching and learning of quadratic
equations. Although articles on algebra in various NCTM research
publications (e.g., Kieran, 1992; Kieran & Chalouh, 1993; Wagner & Parker,
1993) have paid some attention to research into the teaching and learning of
linear equations, they have been silent on quadratic equations.
The 13 chapters in Stacey, Chick and Kendal’s (2004) edited volume on
The Future of the Teaching and Learning of Algebra provided commentary
from scholars from numerous countries on the past, present, and future
status of algebra in school curricula, and on possible renewal of pedagogy
through creative use of modern technology. None of the contributors to the
volume, however, provided a careful statement on the cognitive challenges
faced by students trying to solve quadratic equations. As far as we are aware,
such a statement is not to be found, anywhere.
It appears to be the case that difficulties that students experience in
learning to solve quadratic equations are not part of the pedagogical
content knowledge of secondary mathematics teachers or, for that matter,
of authors of textbooks or articles on the teaching and learning of algebra.
Vaiyavutjamai (2004a) reviewed sections on quadratic equations in
mathematics textbooks and teachers’ guides widely used in Thailand, and
found no reference to such difficulties. In his book on Teaching and Learning
Algebra, French (2002) made no reference to difficulties students experience
with quadratic equations. The same was true of Filloy and Sutherland
(1996) in their chapter on algebra in the International Handbook of
Mathematics Education.
Database searches on the Internet led to the identification of many
studies into the teaching and learning of linear equations, but hardly any on
quadratic equations. A search of Dissertation Abstract International for the
past 10 years identified just one dissertation concerned with quadratic
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 49

equations. That dissertation was by Sproule (2000), who investigated the

performance on different types of quadratic equations of high- and low-
ability upper-secondary students (in Grades 11 through 14) attending a
Northern Ireland grammar school. She concluded that only very high ability
students operated at adequate conceptual levels, a finding which raises the
question whether Grade 9 middle-secondary students in Thailand should be
expected to cope with the conceptual challenges presented by even the
simplest quadratic equation.
An ERIC search identified a study by Zaslavsky (1997) who investigated
misconceptions with respect to quadratic functions of more than 800 Grade
10 and Grade 11 students in 25 different schools in Israel. Zaslavsky’s
research emphasis was quadratic functions, however, and her report touched
only incidentally on students’ responses to quadratic equations. In this
article the emphasis is on how students responded to quadratic equations in
one variable that were presented in standard format. The investigation
sought to uncover how the students solved them, what meaning they gave
to the variable, and how they interpreted solutions.
In the chapters on algebra in the last two four-yearly research summary
publications of the Mathematics Education Research Group of Australasia
(Warren, 2000; Warren & Pierce, 2004), the word “quadratic” was used just
twice – when Warren and Pierce (2004) referred to a small study by Gray and
Thomas (2001) into the use of a graphics calculator and multiple
representations to explore quadratic equations. The Gray and Thomas study
involved a sample of 25 students aged 14–15 years, and results indicated that
the students did not improve their ability to solve quadratic equations.
The relative lack of research into the learning of quadratic equations has
meant that peculiarities associated with variables in quadratic equations
and, in particular, with the effects of these on student learning, have
remained hidden. Thomas and Tall (2001) distinguished, among other
things, between “algebra as generalised arithmetic” and “manipulation
algebra”, and commented that research indicated that students who
completed secondary education were usually able to substitute values
correctly for variables in expressions and equations, and were able to
interpret variables in symbolic and graphical contexts. However, student
thinking in such contexts appeared to be dominated by a need to achieve
procedural mastery, and usually there was no guarantee that relational
understanding was achieved.

Misconceptions Regarding Variables

There is considerable discussion in the international mathematics education
literature on beginning algebra students’ abilities to grasp the concept of a
variable in the context of linear equations (see, e.g., Booth, 1984; Filloy &
Rojano, 1984; Fujii, 2003; MacGregor, 1991; Stacey & MacGregor, 1997, 1999a).
The findings of Fujii (2003), who investigated the understanding of the
concept of variable of students in the United States and Japan, would appear
50 Vaiyavutjamai & Clements

to have direct relevance to likely students’ misconceptions with respect to

quadratic equations. Among the tasks that Fujii (2003) asked students to
attempt were the two shown in Figure 1:

First Task, from Fujii (2003)

Mary has the following problem to solve:
“Find the value(s) of x for the expression x + x + x = 12.”
She answered in the following manner (a) 2, 5, 5; (b) 10, 1, 1; (c) 4, 4, 4.
Which of her answers is/are correct? Circle the letters that are correct.
State the reason for your selection.

Second Task, from Fujii (2003)

Jon has the following problem to solve:
“Find the values of x and y in the expression x + y = 16.”
He answered in the following manner (a) 6, 10; (b) 9, 7; (c) 8, 8. Which of
his answers is/are correct? Circle the letters that are correct. State the
reason for your selection.

Figure 1. Two tasks from Fujii (2003).

Fujii reported that within a sample of 6th, 8th, and 9th graders in Georgia, in
the United States, only 11.5%, 11.5%, and 5.7% respectively gave correct
answers to both of these tasks. With 6th, 7th, 8th, 9th, 10th and 11th graders
in Japan, the results were similar, with 0.0%, 1.7%, 9.5%, 10.8%, 18.1%, and
24.8% respectively giving correct answers. According to Fujii (2003), for both
countries the percentages of correct answers were “disturbingly low”,
especially insofar as the percentages did not dramatically increase according
to the grades “as we may expect” (p. 53).
The main error made on the first task by students in Fujii’s (2003)
study, in both the United States and Japan, was that students felt that the
x’s on the left side of the equation need not necessarily represent the same
value. For the second task, many students thought that because x and y
were different letters they could never take the same value. If this kind of
thinking were to be translated to a task like “Solve the equation (x – 3)(x –
5) = 0” then one might expect some students to think that the value of x in
(x – 3) would be 3, and, simultaneously, the value of x in (x – 5) would be
5. In this article data relating to that conjecture are presented. It should also
be observed that students giving correct solutions (i.e., x = 3, 5) to (x – 3)(x
– 5) = 0 might think that the variable x had to take different values in the
two sets of parentheses. Thus, it was important to build in an interview
component into the research design.
Another aspect of equations and inequalities that often causes difficulty
with middle-secondary school students is that of checking solutions of
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 51

equations. The report of the Mathematical Association (1962), in the United

Kingdom, on the teaching of algebra in schools emphasised that part of the
problem is that students who generate solutions to an equation often do not
understand that the only number(s) which “make the equation true” are the
solutions. Students who do not realise this may find it difficult to “check”
“solution(s).” The problem is one of semantics, and gets disguised by
students’ participation in the ritual manipulations surrounding the processes
of finding solutions to equations.
Thus, for example, the student who correctly gets the solution x = 4
x–1 2
to the equation +4=9– (3x – 2) but then makes the following
3 5
statement as a check, needs to be brought to realise that from a formal logical
point of view the given check is illogical:
4–1 2(12 – 2)
+4=9–
3 5
... 1 + 4 = 9 – 4

... 5 = 5
In the first and second lines the “=” sign is being used as an expression
of hope, not as a statement of fact. In any case, “by merely following the same
procedure as was used to generate the solution, students are likely to repeat
any mistake they made (such as errors in sign, or in learning fractions)”
(Mathematical Association, 1962, pp. 23–24).
After being taught to solve linear equations presented in the form ax + b
= c, Form 2 (i.e., Grade 8) students in Thailand are expected to learn to solve
other equations that, although still linear, are not initially presented in
standard form. For example, in order to solve equations in the form ax + b =
cx + d, students are shown how to “get the x terms on the left-hand side, and
the other terms on the other side.” They are then expected to combine the x
terms into one term (strictly speaking by using the distributive law, but often
teachers prefer to use expressions like “you can add or subtract ‘like’ terms,
but you cannot add or subtract ‘unlike’ terms”). Students are then told to
“divide both sides by the coefficient of x” (which, mathematically speaking,
would be justified by a combination of the multiplication property of
equations and the inverse and associative laws for multiplication).
The TIMSS investigation of the late 1990s revealed that Grade 8 students
do not solve equations in the form ax + b = cx + d very well, with only 29%
of Thai Grade 8 students in Thailand getting the correct solution for 12x – 10
= 6x + 32, a result that was well below the international average (Mullis et al.,
2000). Stacey and MacGregor (1997, 1999a, 1999b), after recognising that
equations in the form ax + b = c are more easily solved than equations in the
form ax + b = cx + d, maintained that student performance on “ax + b = cx +
d” equations and associated problems provides a litmus test for the extent of
a student’s algebraic development.
52 Vaiyavutjamai & Clements

Part of the problem could be that with 12x – 10 = 6x + 32, for example,
some students believe that the x’s on the two sides of the equation represent
different values (e.g., 5 on the left side and 3 on the right side). Filloy and
Rojano (1984) reported that many 12- to 13-year-olds thought that whereas
the first x on the left side of the equation x + x/4 = 6 + x/4 had to be 6, the
x’s in the two x/4 terms could take any values. Similarly, students thought
that with x + 5 = x + x, the first x on the left and right sides could take any
value, but the second x on the right side had to be 5.
Hoch and Dreyfus (2004) argued that whereas, 30x 2 – 28x + 6, for
example, is equal to (5x – 3)(6x – 2), students without “structure sense” may
not realise that the quadratic trinomial and its factorised equivalent are
“different interpretations of the same structure” (p. 51). Data related to that
phenomenon have been reported by Vaiyavutjamai, Ellerton, and Clements
(2005) and Lim (2000), who found that when faced with (x – 3)(x – 5) = 0, for
example, many secondary-level mathematics students, some university
mathematics students, and even some teachers choose to write the left side
as x 2 – 8x + 15, then re-factorise before applying the null factor law.
There has been much research on the extent to which teachers, over the
world, use potentially harmful expressions such as “take the term over to the
other side (of the equation) and change its sign,” “multiplication becomes
division when it goes to the other side,” and “cross-multiplication gives …”
(Bodin & Capponi, 1996; Mathematical Association, 1962; Vaiyavutjamai,
2004a, 2004b). Although details are not given here of the first author’s
analyses of classroom observations of lessons on equations in the study (for
details of the analyses of data on the language discourses in Grade 9 algebra
classes in Thailand see Vaiyavutjamai, 2004a), interviews with selected
students are reported. These revealed that many students could not explain
what such expressions meant. They simply did what they thought their
teachers told them to do (which, often, was not the same as what their
teachers did tell them to do).

Effects of Traditional Teaching on Student Understanding

There is evidence that traditional, elicitation/exposition-type teaching in
mathematics classrooms isolates skills and fails to draw attention to
connections (see, e.g., Farrell & Farmer, 1988; Good, Grouws, & Ebemeier,
1983; Hiebert, 2003; Hiebert & Carpenter, 1992; Phomjwi et al., 1999; Skemp,
1976). As Hatano (1988) stated, over-emphasising skills is likely to result “in a
sacrifice of understanding and of the construction of conceptual knowledge”,
to the extent that it is difficult for understanding to be achieved at some later
time because “it is hard to unpack a merged specific rule to find the meaning
of any given step” (p. 64). Resnick and Ford (1981) concluded that research
indicated that presenting mathematical concepts through stand-alone
examples and repetitious practice does not foster understanding.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 53

Quadratic Equations in the Grade 9 Mathematics Curriculum

for Thailand
Chaysuwan (1996) reported that immediately after 661 Grade 9 students in
secondary schools in Bangkok had participated in lessons on quadratic
equations, 70% of their responses to standard quadratic equations tasks were
incorrect. Many students did not seem to have the prerequisite algebraic
skills needed to cope with quadratic equations. That raised the question
whether most Grade 9 students should be expected to study quadratic
equations. Until recently, it was a compulsory topic within M 012, a
mathematics unit taken by most secondary students in Thailand (Institute
for the Promotion of Teaching Science and Technology, 1998a, 1998b).

Relational and Instrumental Understanding

The authors decided to adopt, and adapt for the purposes of the study,
Skemp’s (1976) distinction between “instrumental” (or procedural)
understanding, and “relational” understanding. According to Skemp (1976),
instrumental understanding involves “rules without reason” (p. 20), and
relational understanding “knowing both what to do and why” (p. 20).
Skemp maintained that with relational understanding:
1. The means become independent of particular ends to be reached
thereby;
2. Building up a schema within a given area of knowledge becomes an
intrinsically satisfying goal in itself.
3. The more complete a pupil’s schema, the greater his feeling of
confidence in his own ability to find new ways of “getting there”
without outside help.
4. A schema is never complete. As our schemas change, so our
awareness of possibilities is thereby enlarged. Thus the process often
becomes self-continuing, and (by virtue of 3) self-rewarding. (p. 26)
The authors were aware of more recent approaches to defining and
investigating mathematical understanding in school children (e.g., Borgen &
Manu, 2002; Pirie & Kieren, 1994; Schoenfeld, 1992), but decided that
Skemp’s (1976) seminal instrumental/relational distinction would be
sufficient for the purposes of the study.

The Newman Interview Technique, and Additional Questions

Used by Booth
Given that the research study would aim at assessing student understanding
of quadratic equations it was essential that a valid and reliable procedure be
devised for assessing understanding in that context. The authors decided
that the extent of student understanding could best be evaluated by
comparing what students wrote, when attempting to solve quadratic
equations on pencil-and-paper tests, with what they said about the same
54 Vaiyavutjamai & Clements

equations in one-to-one interviews. A valid and reliable pencil-and-paper

instrument for assessing students’ knowledge and skills with respect to
quadratic equations was therefore needed, as was an associated pre- and
post-teaching interview strategy. For the latter, the authors decided to
develop an interview protocol that combined interview approaches
developed by Newman (1983) and Booth (1984).
Newman’s (1983) diagnostic interview technique. With the Newman
approach the interviewer makes five key requests of interviewees:
1. Please read the question to me.
2. Tell me, what does the question mean?
3. What will you need to do to answer this answer?
4. Now answer it, and tell me what you are thinking as you do it.
5. Now, write down your actual answer.
These requests are associated with error classifications that Newman called
Reading, Comprehension, Transformation, Process Skills, and Encoding.
Booth’s extended set of categories. In her study into the teaching and
learning of algebra in lower secondary classes in the United Kingdom in the
early 1980s, Booth (1984) used a semi-structured interview schedule that
extended the Newman approach to interviewing. Booth’s schedule included
requests equivalent to all five Newman requests, and in addition she made
requests aimed at finding out whether interviewees:
• knew what their answers meant in relation to the original question.
• could check their answers.
• would stick to their answers if challenged with other possibilities.
• could identify other questions similar to a question they had just
answered.
• could generalise solutions to solve more complex, but nonetheless
similar, tasks.
The authors thought that Newman requests could be used to investigate
interviewees’ instrumental understanding of quadratic equations, and
Booth’s additional requests could be used to explore the extent of their
relational understanding.

Design of the Study

Aim
The aim of the study was to investigate how traditional lessons on quadratic
equations, in which the emphasis was on teaching Grade 9 students to solve
quadratic equations by factorisation (and application of the null factor law),
by “completing the square”, and by the quadratic formula, influenced
student development of knowledge, skills, concepts, and understanding
with respect to quadratic equations.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 55

The Samples of Students and Teachers

Altogether, 231 students in six Grade 9 classes in two government secondary
schools in the Chiang Mai Province of Thailand, and four teachers,
participated in the study. During the 2002/2003 school year most, but not all,
Grade 9 students at one of the participating schools (“School X”) took M 012,
and were therefore expected to study quadratic equations. At the other
participating school (“School Y”) all Grade 9 students took M 012.
The six participating classes in the study were taking M 012. There were
109 students in three participating classes at School X and 122 in three classes
at School Y. Analyses of national data reported by the Office of the Basic
Education Commission (2003) suggested that if all Grade 9 students in
Thailand in 2002 were regarded as a population, the sample in this study
would be of “average mathematical ability”.
At the beginning of their Grade 8 year, students at School X and School
Y who participated in the study were “streamed” into high-, middle-, or low-
stream classes on the basis of overall academic performance in Grade 7.
These students remained in the same classes when they progressed to
Grade 9. At School X, “Teacher A”, taught three classes (the high-stream
Grade 9/1, the medium-stream Grade 9/2, and the low-stream Grade 9/3),
and at School Y, “Teacher B”, “Teacher C” and “Teacher D” taught the high-
stream Grade 9/4, the medium-stream Grade 9/5, and the low-stream Grade
9/6, respectively. Each teacher was experienced and well-qualified.

The Lessons
At the time this study was conducted (during the 2002/2003 school year) the
Grade 9 national mathematics curriculum for Thailand stipulated that M 012
students should receive 13 lessons on linear equations and inequalities and
11 lessons on quadratic equations (Institute for the Promotion of Teaching
Science and Technology, 1998b). Teachers who participated in the study
permitted the first author to observe and audiotape a total of 18 50-minute
lessons – one lesson on linear equations, one on linear inequalities and one
on quadratic equations, for each of the six classes.
The teachers agreed to teach “as they normally would” during lessons
observed, and to follow, as they usually did, approaches recommended in
the “official” Grade 9 student textbook and the associated teachers’ guide
(Institute for the Promotion of Teaching Science and Technology, 1998a,
1998b). The teachers gave permission for transcripts to be produced and
reported. During the 11 lessons on quadratic equations the teachers taught a
standard secondary school mathematics sequence of lessons on quadratic
equations encompassing factorisation and the null factor law, completing the
square, and the use of the quadratic formula. Analysis by the first author of
transcripts of the six lessons observed on quadratic equations revealed that
these lessons were all taught in a traditional way (Vaiyavutjamai, 2004a).
56 Vaiyavutjamai & Clements

Sources of Data
Student interview data. The main data source for this report was a set of 36
transcripts of interviews that the first author conducted with 18 student
interviewees (two interviews with each interviewee). The selection of
interviewees was based on relative within-class student performance on two
pencil-and-paper tests that had been administered to all six classes in August
2002 – a Language of Equations and Inequalities Test and a Linear Equations
and Inequalities Test (see Vaiyavutjamai, 2004a, for further details of these
tests). Each test was marked out of a possible 27, and so any overall mark
within the range 0 to 54 was possible. From each of the six classes the highest
performer, a medium performer, and the lowest performer (on the basis of
total score on the two tests) were selected to be part of the interview sample.
Eighteen of the interviews took place at the pre-teaching stage in
December 2002, immediately before the lessons on quadratic equations. The
other 18 interviews took place in January 2003, soon after the six classes had
participated in 11 lessons on quadratic equations. About 360 pages of
interview transcripts, in Thai, were generated from the audiotapes, and these
were translated into English by the first author.
Student performance data. In addition to interview data, pre- and post-
teaching performance data generated by student responses to the Quadratic
Equations Test were also analysed. This test, which comprised 18 standard
questions on quadratic equations, had a Cronbach alpha reliability of 0.90,
and the four teachers agreed that after the lessons on quadratic equations
their students should have been able to answer each question correctly
(Vaiyavutjamai, 2004a). The 18 questions are shown in Figure 2 (which
appears later in this article). For the purposes of performance analysis,
student responses to a question were scored 1 for a response deemed to be
correct, or 0 for a response deemed to be incorrect, and 0, also, if no response
was given. At both the pre- and post-teaching stages, all 231 participating
students were given ample time to complete all 18 questions.

The Four Interview Questions

In the interviews, the interviewer (the first author) explored the knowledge
and understanding of interviewees with respect to four so-called “interview
questions”.
• (x – 3)(x – 5) = 0
• x2 – x = 12
• x2 = 9
• 2x2 = 10x
These “interview questions” differed in subtle but mathematically important
ways. During interviews, interviewees were invited to solve the equations,
showing all their working, and to respond to questions asked by the
interviewer. Transcripts of excerpts from interviews in relation to the
equations (x – 3)(x – 5) = 0 and x2 – x = 12 are reproduced later in the article.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 57

Most of the participating Grade 9 students had their first-ever class

lessons on quadratic equations after they had attempted the (pre-teaching)
Quadratic Equations Test, and after the pre-teaching interviews for quadratic
equations had been conducted. It was expected that at the pre-teaching stage
only a small proportion of responses given by the 231 students to the
equations would be correct.
It was a matter of interest whether students confronted with a quadratic
equation at the post-teaching stage would know, before they attempted to
solve it, that the equation might have two solutions, or one solution, or
perhaps no solution. Would students tend to have a better appreciation of the
mathematical principles involved, for example, of how the null factor law, “if
ab = 0, then a = 0, or b = 0, or both a and b are zero,” might be applied? Would
they realise that in equations like (x – 3)(x – 5) = 0, x2 – x = 12, and 2x2 = 10x,
the two x’s represent the same variable? Would they know that a solution to
a quadratic equation was a number which, when substituted in the equation,
generated a true statement? These were among the questions probed in the
interviews associated with the Quadratic Equations Test.
During interviews the first author concentrated on aspects of the
problems that were not necessarily revealed by an examination of students’
pencil-and-paper responses. For example, with (x – 3)(x – 5) = 0, did students
think that the x in (x – 3) stood for a different variable from the x in (x – 5)?
Lim (2000) reported that many students attempted a similar question by
expanding the two brackets, refactorising, and then equating each factor to
zero. It was expected that the interviews would reveal whether a similar
tendency existed within the six classes involved in the present study. With
each of the equations “x2 – x = 12” and “2 x 2 = 10x” it was expected that the
interviews would reveal that, even at the post-teaching stage, some students
would think that the x in the x2 term and in the other x term stood for
different variables. It was also expected that some students would divide
both sides of 2x2 = 10x by x. And, with the “x2 = 9” question it was expected
that the most common error would be the single solution (x = 3) answer.

Developing a Rubric for Measuring Understanding

The authors believed that analysis of interview data generated by student
responses to the interview questions had the potential to make clear the
extent to which student understanding, as distinct from unconnected
knowledge, was present in a student’s schema. The authors set out,
therefore, to develop a rubric that would enable reliable distinctions to be
made between aspects of students’ understanding.

A Rubric for Assessing Understanding

The rubric that was developed and used in this study for allocating
“understanding scores” to students is now elaborated. An interviewee’s set
of responses to questions asked in relation to an “interview” question was
assessed on a 0, 1, 2, 3, or 4 basis, where the score was allocated according to
the following criteria:
58 Vaiyavutjamai & Clements

Score of 0: The interviewee does not comprehend the meaning of the

question. This could be due to an inability to decode one or
more symbols used in the question, or to grasp the overall
meaning of the question as it is presented.
Score of 1: The interviewee has some idea of the meaning of the
question, and is able to transform the task by choosing
appropriate mathematical procedures for answering it.
However, he/she is not able to apply those procedures
correctly.
Score of 2: The interviewee not only identifies an appropriate
sequence of procedures for the question, but also applies
the procedures accurately, or reasonably accurately, when
carrying out associated process skills. He/she is able to
encode his/her answer in an appropriate way, but cannot
explain the meaning of the answer obtained in relation to
the original question. Also, he/she is not aware of
mathematical principles that underlie the algorithms used
when attempting to solve the equation.
Score of 3: The interviewee can apply an appropriate procedure, or set
of procedures, accurately, has some awareness of the
mathematical principles that underlie the algorithms, and
has some idea of how the answer relates to the original
question. However, he/she is either not able to check
whether his/her answer is correct or, if he/she can check
the answer, cannot interpret the check.
Score of 4: The interviewee can apply an appropriate procedure
accurately, has some awareness of the mathematical
principles that underlie the algorithms, and knows how the
answer obtained relates to the original question. He/she
can check the answer and can link the various
representations of the answer(s) – written, verbal, and
symbols – to each other and to the original question.
Specific criteria for the four interview questions from the Quadratic
Equations Test are shown in Table 1.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 59

Table 1
Criteria for Assessing “Understanding Scores” for the Four Interview Questions

Question: Solve the Equation …

Score (x – 3)(x – 5) = 0 x2 – x = 12 x2 = 9 2x2 = 10x

0 Does not know Does not know Does not know Does not know
how many how many how many how many
solutions to expect, solutions to expect, solutions to expect, solutions to expect,
and does not and does not and does not and does not
comprehend the comprehend the comprehend the comprehend the
instruction “solve instruction “solve instruction “solve instruction “solve
the equation”. the equation”. the equation”. the equation”.

1 Knows there could Knows there could Knows there could Knows there could
be up to two be up to two be up to two be up to two
solutions. Applies solutions. Applies solutions, Applies solutions. Applies
an appropriate an appropriate an appropriate an appropriate
procedure (e.g., procedure (e.g., procedure (e.g., procedure (e.g.,
equate each bracket writes x 2 – x – 12 writes x 2 – 9 = 0, writes 2x 2 – 10x
to 0). But does not = 0, factorises, and factorises and = 0, factorises and
know why the then equates each equates each factor equates each factor
procedure works, bracket to 0). Does to 0, or writes x = to 0). Cannot
and cannot carry it not know why the ±√ 9 ). Cannot explain why the
out accurately. procedure works, explain why the procedure works
and does not apply procedure works and does not apply
it accurately. and does not apply it accurately.
it accurately.

2 Accurately carries Accurately carries Selects, and Selects and

out an appropriate out an appropriate accurately carries accurately carries
procedure (minor procedure (minor out, an appropriate out, an appropriate
careless error careless error procedure (minor procedure (minor
permitted). Cannot permitted). Cannot careless error careless error
state the null factor relate answers permitted). permitted).
law or relate to the original Cannot state Cannot state
answers to the equation. a mathematical a mathematical
original equation. reason for using the reason for using the
procedure, or relate procedure, or relate
answers to the answers to the
original equation. original equation.

3 Accurately carries Accurately carries Accurately carries Accurately carries

out an appropriate out an appropriate out an appropriate out an appropriate
procedure (minor procedure (minor procedure (minor procedure (minor
careless error careless error careless error careless error
permitted). Can permitted). Can permitted). Can permitted). Can
relate answers relate answers to relate answers to relate solutions
to the original original equation, original equation, to the original
equation, or, if or if asked to do so, or, if asked to do equation, or, if
asked to do so, can can check solutions so, can check asked to do so, can
check solutions meaningfully, but solutions check solutions
meaningfully, but cannot do both. meaningfully, but meaningfully, but
cannot do both. cannot do both. cannot do both.
60 Vaiyavutjamai & Clements

4 Accurately carries Accurately carries Accurately carries Accurately carries

out an appropriate out an appropriate out an appropriate out an appropriate
procedure. Knows procedure. Knows procedure. Knows procedure. Knows
what solutions what solutions what solutions what solutions
mean in relation mean in relation to mean in relation to mean in relation to
to the original original equation. original equation. original equation.
equation. If asked If asked to do so, If asked to do so, If asked to do so,
to do so, can check can check solutions can check solutions can check solutions
solutions meaningfully. meaningfully. meaningfully.
meaningfully.

Inter-rater reliability in assessing understanding. In order to check inter-

rater reliability, the extent to which two qualified persons independently
allocated the same understanding score to the same interview transcript was
ascertained. Since each of the 18 students answered the four interview
questions in both the pre- and post-teaching interviews, altogether 144
interview units were assessed.
The first author assessed the 144 Thai-language units of transcript, and
the second author independently assessed the corresponding 144 English-
language translations of the transcripts. The Pearson-product moment
correlation coefficient for the 144 pairs of scores for the interview units was
0.97. The two assessors gave the same score for 130 of the units, and differed
by 1 on the other 14 units. Whenever different scores were given, a consensus
score was reached, through discussion. Most differences in initial scores
arose as a result of linguistic factors. Post-grading discussion led to
refinement of criteria, and the final criteria were those shown in Table 1.

Results
Performance Trends
Figure 2 shows the percentages of the 231 students who gave correct
answers, at the pre- and post-teaching stages, to each of the 18 questions on
the Quadratic Equations Test. In every case, more students gave a correct
answer after the lessons than before. It can be seen that at the pre-teaching
stage about half of the students correctly solved the equations for Questions
7 and 14, probably because each of those equations had one solution only
and that solution could be obtained fairly easily by substitution. At the pre-
teaching stage, relatively few students correctly solved any of the other
equations.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 61

Figure 2. Pre- and post-teaching percentages correct for the 18 equations

comprising the Quadratic Equations Test (231 students).

Table 2 summarises pre- and post-teaching test means and standard

deviations for the six classes, and shows Cohen’s d effect sizes (Cohen, 1992)
for pre-/post-teaching comparisons of performance.
62 Vaiyavutjamai & Clements

Table 2
Performance Mean Scores, Standard Deviations, Effect Sizes (Quadratic
Equations Test)

Class Mean Standard Standard Pre-Post

Pre-Test Mean Deviation Deviation Effect Size
Score/18 Post-Test/18 Pre-Test Post-Test (Cohen’s d)

School X
Grade 9/1 1.75 9.49 2.17 4.82 2.07
Grade 9/2 1.08 4.90 1.00 4.02 1.30
Grade 9/3 0.73 2.67 0.91 3.08 0.85
School Y
Grade 9/4 3.70 7.91 3.29 4.58 1.05
Grade 9/5 2.18 4.45 2.89 4.11 0.64
Grade 9/6 0.55 6.66 0.83 4.89 1.74
Total 1.75 6.17 2.41 4.83 1.16

The high effect sizes might suggest that the lessons were “effective”. We
would argue, however, that such a conclusion is not warranted because it
ignores the fact that for five of the six classes the post-teaching means on the
Quadratic Equations Test were less than 9.0 (i.e., 50% of the maximum
possible score on the test). The overall post-teaching mean for the 231
students was 6.17 (out of a possible 18).
After the 11 lessons on quadratic equations most participating students
found it difficult to solve most of the questions on the Quadratic Equations
Test. The effect was not uniform across the classes, however. For example,
although most Grade 9/1 students benefited from Teacher A’s lessons on
quadratic equations, his lessons were not nearly so successful with Grade
9/2 and Grade 9/3 students. By comparison with the two medium-stream
classes, and with the other low-stream class, students in the low-stream
Grade 9/6 did well at the post-teaching stage.

Summary of Performance Data for the 18 Interviewees

Table 3 shows pre- and post-teaching scores of the 18 interviewees on the
Quadratic Equations Test. The selection of the interviewees was based on the
total of the pre-teaching scores on the Language of Equations and Inequalities
Test and the Linear Equations and Inequalities Test. With respect to Table 3 it
should be noted that Students 1, 4, 7, 10, 13, and 16 were high-performing
students relative to their classmates, and Students 2, 5, 8, 11, 14, and 17 were
medium-performing students. The other interviewees were low-performing
students. The percentile ranks for the 18 interviewees (within the sample of
231 students), based on the sum of their pre-teaching scores on Test 1 and
Test 2, are shown in the second column.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 63

Although the 18 interviewees were not selected randomly from the total
sample of 231 participating students, it can be seen from the percentile ranks
in the third column of Table 3 that the sample covered the full spectrum of
performance. The mean percentile rank of the 18 interviewees was 48.2.

Table 3
Performances of 18 Interviewees on 3 Tests (Language of Equations and
Inequalities, Linear Equations and Inequalities, and Quadratic Equations)
School, Percentile (Pre-T) (Pre-T) Pre-Teaching Post-
and Rank of Language of Linear Quadratic Teaching
Grade Interviewee Interviewee Equations Equations Equations Quadratic
(Within Total and and Test Equations
Sample of 231 Inequalities Inequalities Test
Students) Test Test
Score/27 Score/27 Score/18 Score/18

School X Student 1 99 23 20 1 12
Grade 9/1 Student 2 75 18 4 1 6
(high-stream) Student 3 24 7 2 0 3
Grade 9/2 Student 4 79 18 6 1 11
(medium- Student 5 48 11 3 1 3
stream) Student 6 00 0 0 0 6
Grade 9/3 Student 7 57 15 2 0 7
(low-stream) Student 8 19 6 2 1 3
Student 9 04 3 2 0 1
School Y Student 10 100 24 25 15 17
Grade 9/4 Student 11 81 16 10 5 11
(high-stream) Student 12 19 8 0 3 3
Grade 9/5 Student 13 95 21 14 11 12
(medium) Student 14 41 10 2 2 3
-stream Student 15 01 0 1 0 0
Grade 9/6 Student 16 88 17 12 1 14
(low-stream) Student 17 28 6 4 2 7
Student 18 01 3 0 1 2

Entries in Table 4 reveal that 11 of the 18 interviewees obtained a post-

teaching mean understanding score per question score less than 2.0. That is to
say, after the lessons they did not understand, relationally, much of what
they had been taught.
From Table 4 it can be seen that the mean understanding gains per person
per question (based on the sums of the pre- and post-teaching assessments for
the four interview questions) can be compared for the high-, medium- and
low-performing interviewees. Mean “per person per question”
understanding gains for the three groups were 1.54, 0.58, and 0.42,
respectively. Thus, high performers started off understanding more than
64 Vaiyavutjamai & Clements

other students and after the lessons their understanding had increased
relative to the other students.
Student 10 was clearly an outlier. Even at the pre-teaching stage she
demonstrated a strong understanding of the mathematics associated with
each of the four interview tasks. If her data were to be removed from the
analysis of pre-teaching scores then the mean would reduce from 3.39 to 2.65
and the standard deviation form 3.78 to 2.15, both very substantial effects. It
was decided, however, that since Student 10 was included in the interview
sample on the basis of a selection procedure that specifically included the
highest performing student in each of the six classes, it would be
inappropriate to exclude her interview data from the analyses.

Table 4
Summary of Understanding Scores for the 18 Interviewees
Interviewee Pre-Teaching Score on Sum of Post-Teaching Score on Sum of
Student Question Concerned Pre-T Question Concerned Post-T
Number with … Scores with … Scores
(x – 3) x2 – x x2 = 9 2x2 = 10x (x – 3) x2 – x x2 = 9 2x2 = 10x
(x – 5) = 12 (x – 5) = 12
=0 =0

(School X, Grade 9/1)

1 2 2 1 1 6 3 3 3 3 12
2 1 1 1 1 4 3 3 3 1 10
3 0 1 1 1 3 2 1 2 2 7
(School X, Grade 9/2)
4 0 0 1 0 1 4 3 4 4 15
5 2 1 1 1 5 1 2 4 0 7
6 0 0 0 0 0 0 1 0 0 1
(School X, Grade 9/3)
7 0 0 1 0 1 2 2 4 2 10
8 0 1 1 0 2 2 2 2 1 7
9 0 0 0 0 0 0 0 1 0 1
(School Y, Grade 9/4)
10 4 4 4 4 16 4 4 4 4 16
11 1 1 1 1 4 2 1 1 1 5
12 0 0 1 0 1 2 1 1 1 5
(School Y, Grade 9/5)
13 2 2 1 0 5 2 2 2 3 9
14 2 0 1 0 3 1 1 1 1 4
15 0 0 0 0 0 0 0 0 0 0
(School Y, Grade 9/6)
16 2 2 1 0 5 2 2 4 1 9
17 2 2 1 0 5 2 1 1 0 4
18 0 0 0 0 0 0 0 0 0 0
Total 18 17 17 9 61 32 29 37 24 122
Mean 1.00 0.94 0.94 0.50 3.39 1.78 1.61 2.06 1.33 6.78
SD 1.19 1.11 0.87 0.99 3.78 1.26 1.14 1.51 1.37 4.80
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 65

English-Language Translations of Excerpts of Thai

Interview Transcripts
Excerpts from English-language composite transcripts of two interviewees,
in relation to two of the four interview questions, are reproduced in Table 5.
Excerpts appear in pairs, with a pre-teaching interview excerpt being
visually accompanied by the corresponding post-teaching interview excerpt.
Both interview excerpts chosen could be associated with “understanding
scores” that approximated the average scores for the 18 interviewees before
the lessons and after the lessons were given. By reading the transcripts the
reader should get a “feel” for how much “typical” students knew and
understood about the question before and after the lessons. The symbol [...]
is used to indicate where sections of an excerpt have been omitted.

Excerpts of Interviews with Student 11 in Relation

to “(x – 3)(x – 5) = 0”
Student 11, who was involved in the interview that generated the transcript
in Table 5, was a medium-performing student in the high-stream Grade 9/4.
He gave the same correct answer, “x = 3, 5”, to (x – 3)(x – 5) = 0 on both the
pre- and post-teaching administrations of the Quadratic Equations Test. At the
pre-and post-teaching stages, he scored 5 and 11, respectively, out of a
possible 18 on the Quadratic Equations Test, thereby achieving a gain of 6 on
the Test.
Although Student 11 managed to obtain correct solutions to (x – 3)(x – 5)
= 0 at both stages he could not describe a correct method for checking his
answers, and he thought that the two x’s in the equation represented
different numbers. At the post-teaching stage he still used a substitution
method and still thought that the two x’s represented different numbers.
However, at that stage he knew how to check his answers and how his
solutions related to the original equation.

Table 5
English-Language Translations of Excerpts of Transcripts of Pre- and Post-
Teaching Interviews with Student 11 for the Question: “Solve the equation
(x – 3)(x – 5) = 0”

Pre-Teaching Interview Post-Teaching Interview

Interviewer: Are those x’s the same Interviewer: Are those x’s the same
variable? variable?
Student: No, they aren’t. Student: No. They are different.
Interviewer: They aren’t the same variable? Interviewer: So, what do you need to do to
So, what method will you need solve this equation?
to use to solve the problem? Student: Use the substitution method.
Student: The substitution method.
It is an easy method.
66 Vaiyavutjamai & Clements

Interviewer: What number will you Interviewer: You will use a substitution
substitute? method. What are the
Student: The first x is three and the numbers?
second x is five. Student: Three and five.
Interviewer: Why do you use that method? Interviewer: Please show me your working.
Student: I substituted numbers and got Student: [Wrote, on a piece of paper:
three minus three equals zero, 3 – 3 = 0, 5 – 5 = 0, 0 x 0 = 0]
five minus five equals zero. Three minus three equals zero.
Zero multiplied by zero equals Five minus five equals zero.
zero. It is a true sentence. Zero multiplied by zero equals
Interviewer: What are the answers to this zero. It is a true sentence.
question? Interviewer: What is your answer?
Student: Three and five. Student: Three and five.
Interviewer: All right! Please show me your
working.
Student: [Wrote, on a piece of paper:
(3 – 3)(5 – 5) = 0]
Interviewer: What are your answers?
Student: Three and five.

Interviewer: Can you check your answers? Interviewer: Can you check your answers?
Student: It equals zero. It is a true Student: Yes. I substitute the first
sentence. answer, three, for x.
Interviewer: How did you do it?
Student: I substituted the answers for x
here. Interviewer: All right! Please show me your
Interviewer: Where? working.
Student: I substituted them for x in the Student: [Wrote, on a piece of paper:
question. [Pointed to the line he (3 – 3)(3 – 5) = 0]
had written.] Three minus three equals zero.
Three minus five equals
negative two. Zero multiplied
by negative two equals zero.
Interviewer: Have you finished your check?
Student: The other one, five. [Wrote the
next line: (5 – 3)(5 – 5) = 0]
Five minus three equals two.
Five minus five equals zero.
Two multiplied by zero equals
zero.
Interviewer: Why did you use this method
to solve the equation?
Student: It is an easy method.

Interviewer: Please tell me your answers Interviewer: Please tell me your answers
again. again.
Student: Three and five. Student: Three and five
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 67

The authors agreed that Student 11 should receive a pre-teaching

understanding score of 1, even though he got both correct solutions. Student
11 substituted 3 for one of the x’s and 5 for the other, and thought that the
variable had different values within the two sets of parentheses. Although he
obtained correct solutions the procedure he used was not “appropriate”, and
he could not meaningfully relate solutions to the original equation. Both
authors therefore gave his response a score of 1, not 2. For the post-teaching
interview, both authors wondered whether Student 11 should receive a score
of 1 or 2. Although he still said the two x’s stood for different variables, when
checking his solutions he related solutions to the original equation. Both
authors tentatively allocated a score of 2, and after subsequent discussion
agreed that 2 should remain as the score.
For the 18 interviewees, the means of the pre- and post-teaching scores
for the “(x –3)(x – 5) = 0” task were 1.00 and 1.78, respectively.

Excerpts of Interviews with Student 8 in Relation to x 2 – x = 12

Student 8, who was involved in the interviews that generated the excerpts
shown in Table 6, was a medium-performing student in the low-stream
Grade 9/3. On the pre-and post-teaching administrations of the Quadratic
Equations Test he scored 1 and 3 respectively, out of a possible 18. On the pre-
teaching administration he gave an incorrect answer (“7 2– 2”) to the “x2 – x =
12” question, but his answer on the post-teaching administration, “x = 4, -3”,
was assessed as correct.
During the pre-teaching interview Student 8 tried to use a substitution
method to solve x 2 – x = 12. He got one correct solution, and could check that
answer. During the post-teaching interview, he correctly solved the equation
by factorisation. However, although he used the null factor law he thought
the x’s stood for different variables and did not know how to check his
answer.
The authors agreed that the interview data indicated that Student 8
should receive a pre-teaching understanding score of 1, and a post-teaching
score of 2 for this question. For the 18 interviewees, the mean of the pre-
teaching understanding scores awarded for the “x 2 – x = 12” question was
0.94, and the mean of the post-teaching scores was 1.61.
68 Vaiyavutjamai & Clements

Table 6
English-Language Translations of Excerpts from Pre- and Post-Teaching
Interviews with Student 8 in Relation to the Question: “Solve the equation
x 2 – x = 12”

Pre-Teaching Interview Post-Teaching Interview

Interviewer: Are both x’s the same variable? Interviewer: Are the x’s in the first and
Student: Yes. second terms the same
variable?
Student: No, they aren’t.
Interviewer: Which x do you need to find?
Student: That x [Pointed to the second
term of the question.]
Interviewer: You will find the second term.
Will you find the first term,
too?
Student: The first term is a square
number. I can factorise it.

Interviewer: What do you need to do Interviewer: What method will you need
to solve this question? to use to solve this question?
Student: [Quiet.] Student: Change … change twelve to
Interviewer: What method do you need the other side so that one side
to use? of the equation equals zero.
Student: Find a number that equals Interviewer: Why do you do that?
twelve. Student: I want to find the value of x.
Interviewer: How do you do that? Interviewer: Can you find the value of x
Student: Find two numbers which, when without making one side of the
you subtract, equals twelve. equation equal to zero?
Interviewer: Do it, and show me your Student: No.
answer. Interviewer: What do you do next?
Student: [He thought for about three Student: I factorise into two brackets.
minutes] Interviewer: That’s your method. Please
Interviewer: What did you get? show me your working.
Student: I got four squared minus four. Student: [Wrote, on a piece of paper:
[Wrote: = 42 – 4] x2 – x – 12 = 0
Interviewer: What is your answer? (x – 4)(x + 3) = 0
Student: Four. The solutions of the equation
are 4, – 3]
[…]
Interviewer: How did you get the answers?
Student: Like … may I write it down?
[Wrote the following:
(x – 4)(x + 3) = 0
x – 4 = 0 or x + 3 = 0
x = 4, x = – 3.]
Interviewer: Why does x minus four equal
zero or x plus three equal zero?
Student: The two brackets are
multiplied.
Interviewer: Why could you separate the
question into two equations?
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 69

Student: I found the product of the two

numbers, which equals twelve.
Interviewer: Can you explain to me why
you could get two equations?
Student: No, I can’t.
Interviewer: You just remember how to
do it?
Student: Yes.

Interviewer: Can you check your answer? Interviewer: Can you check your answer?
Student: Yes, I can. Student: [He thought for about half
Interviewer: How do you do it? a minute.] No, I can’t.
Student: Four squared equals four
multiplied by four, which
equals sixteen. Sixteen minus
four equals twelve.

Relationships Between Performance and Understanding

Entries in Table 7 summarise the 18 interviewees’ pre- and post-teaching
Quadratic Equations Test performances (out of 18, in each case) and their
understanding scores (out of 16) with respect to the four interview questions.
In particular, Table 7 shows:
• The interviewees’ performance scores on the pre-teaching
administration of the Quadratic Equations Test (out of a possible 18);
• The interviewees’ performance scores on the post-teaching
administration of the Quadratic Equations Test (out of a possible 18);
• The sum of the understanding scores of each interviewee for the four
interview questions at the pre-teaching stage (out of a possible 16);
• The sum of the understanding scores of each interviewee for the four
interview questions at the post-teaching stage (out of a possible 16).
Table 8 shows Pearson product-moment correlation coefficients between
four variables (pre-teaching performance score, pre-teaching understanding
score, post-teaching performance score, and post-teaching understanding
score) defined in the above dot points. Only one correlation coefficient
(between X 1 and Y2) is less than 0.5.
70 Vaiyavutjamai & Clements

Table 7
Performances of 18 Interviewees on the Quadratic Equations Test, and
Understanding Scores for the Interview Questions (Pre- and Post-Teaching)

Interviewee Pre- and Post-Teaching Total of Pre- and Post-Teaching

Student Performance Scores on the Understanding Scores on
Number Quadratic Equations Test the Four Interview Questions
(Max. Possible = 18) (Max. Possible = 16)
Pre-T Post-T Pre-T Post-T

(School X, Grade 9/1)

1 1 12 6 12
2 1 6 4 10
3 0 3 3 7
(School X, Grade 9/2)
4 1 11 1 15
5 1 3 5 7
6 0 6 0 1
(School X, Grade 9/3)
7 0 7 1 10
8 1 3 2 7
9 0 1 0 1
(School Y, Grade 9/4)
10 15 17 16 16
11 5 11 4 5
12 3 3 1 5
(School Y, Grade 9/5)
13 11 12 5 9
14 2 3 3 4
15 0 0 0 0
(School Y, Grade 9/6)
16 1 14 5 9
17 2 7 5 4
18 1 2 0 0
Total 45 121 61 122
Mean 2.50 6.72 3.39 6.78
S. D. 4.08 4.98 3.78 4.80
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 71

Table 8
Pearson Product-Moment Correlation Coefficients Between Two Performance
and Two Understanding Variables for 18 Interviewees (at the Pre- and Post-
Teaching Stages)

Variable X1 X2 Y1 Y2

X1 : Pre-T Scores on Quadratic Equations Test/18 1.00

X2 : Sum of Pre-T Scores on 4 Interview Questions/16 0.78 1.00
Y1 : Post-T Scores on Quadratic Equations Test/18 0.64 0.71 1.00
Y2 : Sum of Post-T Scores on 4 Interview Questions/16 0.45 0.64 0.75 1.00

It is not being argued, here, that entries in Table 7 and Table 8 support the
proposition that “understanding” caused high performance on the Quadratic
Equations Test, or “high performance” caused “understanding”. Rather, the
evidence supports the view that understanding and performance went
“hand in glove”.

A Fundamental Misconception
At the post-teaching stage, two of the 18 interviewees (Students 4 and 10)
convinced the interviewer, by the quality of their reasoning in interviews,
that they had achieved a strong relational understanding of quadratic
equations. They gave correct solutions to the four “interview questions” and
generally understood how the null factor law could be applied in the context
of quadratic equations. They did not think that the two x’s in (x – 3)(x – 5) =
0 represented different variables. When checking solutions, they did not
substitute x = 3 into (x – 3) and x = 5 into (x – 5) and did not conclude that
since 0 x 0 is equal to 0 their solutions were correct.
It is reasonable to assume that four of the 18 interviewees (Students 6, 9,
15 and 18) learned virtually nothing about quadratic equations from the
lessons. These four did not give correct solutions to any interview question,
did not know how to factorise quadratic trinomials, and did not know how
to check whether any of the “solutions” that they obtained were correct.
The other 12 interviewees gave correct answers to at least two of the
“interview questions”, but that did not mean that they achieved a relational
understanding of quadratic equations. Most obtained 3 and 5 as solutions to
(x – 3)(x – 5) = 0. But when asked to check their solutions, substituted x = 3
into (x – 3) and x = 5 into (x – 5) and concluded that since 0 x 0 = 0 their
solutions were correct. They believed that the two x’s in (x – 3)(x – 5) = 0
represented different variables and should take different values. This
misconception found expression not only with (x – 3)(x – 5) = 0, but also with
x2 – x = 12. Ten of the 12 students rearranged x2 – x = 12 to x2 – x – 12 = 0 and
(x – 4)(x + 3) = 0. They then equated x – 4 and x + 3 to zero and got the correct
solutions. But, they thought that the x in the x2 term in the original equation
72 Vaiyavutjamai & Clements

represented a different variable from the x in the other x term in the same
equation.
When asked to check their solutions, seven of the 12 students said they
did not know how to check. Three students “checked” in the (x – 4)(x + 3)
= 0 form of the equation, putting x equal to 4 in (x – 4) and x equal to (– 3 in
x + 3). After noting that 0 x 0 = 0 was true, they concluded that their
solutions were correct. The other two students substituted into x2 – x – 12
= 0, but let x equal one “solution” with x2, and x equal the other “solution”
for “– x.” They got 16 + 3 – 12 and wondered why this was not zero.
The authors believe that if all 231 students could have been interviewed,
the misconception that equations like (x – 3)(x – 5) = 0, x2 – x = 12, and 2x2 =
10x have two variables, not one, would have been shown to be guiding the
thinking of many, perhaps a majority, of the students. The misinterpretation
could have arisen from teacher statements, often made in lower-secondary
algebra classes, that expressions like 2x2 and 10x are “unlike terms”. It could
also have arisen from students misinterpreting their teachers’ statements
that quadratic equations can “have two different solutions”. In students’
minds this could mean that if two x’s appear in an equation then they should
take different values. That could explain why, even at the post-teaching
stage, relatively few students got both solutions to x2 = 9. In the words of one
of the interviewees, “in that equation x appears only once, and therefore
there is only one solution”.
The authors believe that at the post-teaching stage a minority of students
in the six classes grasped the concept of a variable in the context of quadratic
equations. Because of that fundamental misconception, many did not really
understand the null factor law, or how “solutions” to quadratic equations
could be checked.

Conclusions and Final Comments

The pre-teaching analyses of understanding scores indicated that before the
lessons on quadratic equations most interviewees had little understanding of
the mathematics associated with quadratic equations. That was to be
expected. After the lessons, the mean understanding scores for interview
questions increased, but for each question the mean of the 18 post-teaching
understanding scores was still less than 2.0. A score of 2.0 was assumed to
correspond to an accurate instrumental understanding of related concepts
and skills but an absence of relational understanding.
Although the lessons on quadratic equations helped most students to
perform better on the Quadratic Equations Test at the post-teaching stage, the
post-teaching test mean for the 231 students was only slightly more than 6
out of a possible 18. Analysis suggested that improvement was related to a
moderate increase in understanding, but the extent of improvement in
performance and understanding was disappointing.
With the obvious exception of Student 10 (the outlier) – curiously, there
is no evidence that Student 10 learned anything about quadratic equations
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 73

from the lessons that she did not already know before the lessons – the high-
performing interviewees tended to improve their understanding the most.
The failure of most medium-performing and low-performing interviewees to
develop relational understanding was reflected in relatively low mean
performance gains of students in Grade 9/2, Grade 9/3, and Grade 9/5.
Arguably, the failure of most students in those three classes to develop
relational understanding for most questions meant that any improvement in
test performance was based on rote-learned knowledge and skills.
Analysis of interview data revealed that many interviewees who
obtained correct solutions actually had serious misconceptions about what
quadratic equations actually are. Their answers were correct but, from a
mathematical point of view, they did not know what they were talking
about. Getting correct answers to quadratic equations on traditional pencil-
and-paper tests merely served to reinforce their misconceptions about the
nature of a variable within a quadratic equation.
The issues raised in this article present a real challenge to mathematics
teachers and researchers. Are there realistically feasible forms of teaching
that will result in students, and not just high-achieving students, learning
quadratic equations, and other mathematics topics, in a relational way?
Although that issue was not a primary focus of this report, a few concluding
comments in relation to it might be in order.
In school education, traditionally accepted sequences of content, and
teaching approaches, die hard. In many countries – perhaps most countries
– traditional elicitation/exposition approaches to teaching quadratic
equations are still widely used and preferred by teachers (Lim, 2000). Some
teachers, and education researchers, believe that a teaching approach which
places the study of equations, including quadratic equations, within the
study of functions – the so-called “functions” approach – is far more likely
than traditional elicitation/exposition approaches to solving equations to
induce relational understanding within students. The same teachers and
education researchers believe that the functions approach is especially likely
to be successful if it is enriched by the use of modern technology,
like graphics calculators (e.g., Drijvers & Doorman, 1996; Kirshner & Awtry,
2004; Mourão, 2002; Schwarz & Hershkowitz, 1999; Zazkis, Liljedahl, &
Gadowsky, 2003).
Perhaps at the beginning of this new millennium, the power and
accessibility of graphics calculators and computer algebra systems –
technology which can readily generate graphs of even complicated
functions, and permit the solutions of equations to be quickly linked to
graphs of associated functions – make it more likely that a functions
approach will win the hearts and minds of most secondary mathematics
teachers (and their students). Some studies have provided tentative support
for that view (see, e.g., Schwarz & Hershkowitz, 1999).
The now well-documented failure of many – almost certainly, most –
middle-secondary students, across the world, to cope with the mathematical
demands of quadratic equations suggests that curriculum designers should
74 Vaiyavutjamai & Clements

delay the inclusion of quadratic equations into curricula until Grade 10,
Grade 11, or Grade 12. At this time of flux, when technology is raising
questions about what mathematics should be studied, and when, we leave
that issue to future researchers. We would comment, in closing, though, that
the analyses for this present study, and for Lim’s (2000) study, suggested that
most middle-secondary students who participate in lessons on quadratic
equations taught in traditional ways fall well short of acquiring a relational
understanding of what they are taught.

References
Bodin, A., & Capponi, B. (1996). Junior secondary school practices. In A. J. Bishop, M.
A. Clements, C. Keitel, J. Kilpatrick & C. Laborde (Eds.), International handbook
of mathematics education (pp. 565–614). Dordrecht, The Netherlands: Kluwer
Academic Publishers.
Booth, L. (1984). Algebra: Children’s strategies and errors. Windsor, UK: NFER-Nelson.
Borgen, K. L., & Manu, S. S. (2002). What do students really understand? Journal of
Mathematical Behavior, 21(2), 151–165.
Chaysuwan, S. (1996). The construction of a diagnostic test on basic knowledge of algebra
for Mathayom Suksa three students in the Bangkok Metropolis. Unpublished M.Ed
thesis, Kasetsart University (Thailand).
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
Drïjvers, P., & Doorman, M. (1996). The graphics calculator in mathematics education.
Journal of Mathematical Behavior, 14, 425–440.
Farrell, M. A., & Farmer, W. A. (1988). Secondary mathematics instruction: An integrated
approach. Providence, Canada: Janson Publications.
Filloy, E., & Rojano, T. (1984). From an arithmetical to an algebraic thought. A clinical
study with 12–13 years old. In North American Chapter of the International
Group for the Psychology of Mathematics Education (Ed.), Proceedings of the
Sixth Annual Meeting (pp. 51–56). Madison, WI: North American Chapter of the
International Group for the Psychology of Mathematics Education
Filloy, E., & Sutherland, R. (1996). Designing curricula for teaching and learning
algebra. In A. J. Bishop, K. Clements, C. Keitel, J. Kilpatrick & C. Laborde (Eds.),
International handbook of mathematics education (pp. 139–160). Dordrecht, The
Netherlands: Kluwer Academic Press.
French, D. (2002). Teaching and learning algebra. London: Continuum.
Fujii, T. (2003). Probing students’ understanding of variables through cognitive
conflict: Is the concept of a variable so difficult for students to understand? In N.
A. Pateman, B. J. Dougherty & J. Zilliox (Eds.), Proceedings of the 2003 joint
meeting of the PME and MNENA (Vol. 1, pp. 49–65). Honolulu: International
Group for the Psychology of Mathematics Education.
Good, T. L., Grouws, D. A., & Ebemeier, H. (1983). Active mathematics teaching. New
York: Longman.
Gray, R., & Thomas, M. O. J. (2001). Quadratic equation representations and graphic
calculators: Procedural and conceptual interactions. In J. Bobis, B. Perry & M.
Mitchelmore (Eds.), Numeracy and beyond (Proceedings of the 24th Conference
for the Mathematics Education Research Group of Australasia, (pp. 257–264).
Sydney: MERGA.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 75

Hatano, G. (1988). Social and motivational bases for mathematical understanding. In

G. B. Saxe & M. Gearhart (Eds.), Children’s mathematics (pp. 55–70). San
Francisco: Jossey-Bass.
Hiebert, J. (2003). What research says about the NCTM Standards. In J. Kilpatrick, W.
G. Martin, & D. Schifter (Eds.), A research companion to Principles and Standards
for School Mathematics (pp. 5–24). Reston, VA: National Council of Teachers of
Mathematics.
Hiebert, J., & Carpenter, T. P. (1992). Learning and teaching with understanding. In D.
A. Grouws (Ed.), Handbook of research on mathematics teaching and learning (pp.
65–97). New York: Macmillan.
Hiebert, J., Gallimore, R., Garnier, H., Givvin, K. B., Hollingsworth, H., Jacobs, J.,
Chui, A. M-Y, Wearne, D., Smith, M., Kersting, N., Manaster, A., Tseng, E.,
Etterbeek, W., Manaster, C., Gonzales, P., & Stigler, J. (2003). Teaching mathematics
in seven countries: Results from the TIMSS 1999 Video Study. Washington, D.C.:
National Center for Educational Statistics (U.S. Department of Education).
Hoch, M., & Dreyfus, T. (2004). Structure sense in high school algebra: The effects of
brackets. In M. J. Høines & A. B. Fuglestad (Eds.), Proceedings of the 28th
Conference of the International Group for the Psychology of Mathematics Education
(Vol. 3, pp. 49–56). Bergen, Norway: PME.
Hollingsworth, H., Lokan, J., & McCrae, B. (2003). Teaching mathematics in Australia:
Results from the TIMSS Video Study. Camberwell, Australia: Australian Council
for Educational Research.
Institute for the Promotion of Teaching Science and Technology (1998a). Mathematics
M 012: Lower-secondary level (10th ed.). Bangkok: Author.
Institute for the Promotion of Teaching Science and Technology (1998b). Mathematics
teachers’ guide M 012: Lower-secondary level (3rd ed.). Bangkok: Author.
Kieran, C. (1992). The learning and teaching of school algebra. In D. A. Grouws (Ed.),
Handbook of research on mathematics teaching and learning (pp. 390–419). New
York: Macmillan.
Kieran, C., & Chalouh, L. (1993). Prealgebra: The transition from arithmetic to
algebra. In P. S. Wilson (Ed.), Research ideas for the classroom: Middle grades
mathematics (pp. 119–139). New York: Macmillan.
Kirshner, D., & Awtry, T. (2004). Visual salience of algebraic transformations. Journal
for Research in Mathematics Education, 35, 224–257.
Lim, T. H. (2000). The teaching and learning of algebraic equations and factorisation in O-
level Mathematics: A case study. Unpublished M.Ed dissertation, Universiti
Brunei Darussalam.
MacGregor, M. (1991). Making sense of algebra: Cognitive processes influencing
comprehension. Geelong, Australia: Deakin University.
Mathematical Association. (1962). The teaching of algebra in schools. London: G. Bell &
Sons.
Mourão, A. P. (2002). Quadratic function and imagery: Alice’s case. In A. D. Cockburn
& E. Nardi (Eds.), Proceedings of the 26th Conference of the International Group for
the Psychology of Mathematics Education (Vol. 3, pp. 377–384). Norwich, UK:
PME.
Mullis, I. V. S., Martin, M. O., Gonzalez, E. J., Gregory, K. D., Garden, R. A., O’Connor,
K. M., Chrostowski, S. J., & Smith, T. A. (2000). TIMSS 1999 international
mathematics report. Boston: International Association for the Evaluation of
Educational Achievement.
76 Vaiyavutjamai & Clements

Newman, A. (1983). The Newman language of mathematics kit. Sydney: Harcourt, Brace
and Jovanovich.
Office of the Basic Education Commission (2003). Report of results on the General
Achievement Test (for 2003). Retrieved February 21, 2004 from:
http://bet.bed.go.th/
Phomjwi, S., Wihokto, P., Jindanuluk, T., Phibun, S., Wihokto, P., & Kruapanit, S.
(1999). Research into the strengths and weaknesses of the learning and teaching of
secondary school mathematics. Bangkok: Institute for the Promotion of Teaching
Science and Technology.
Pirie, S., & Kieren, T. (1994). Growth in mathematical understanding: How can we
characterise it and how can we represent it? Educational Studies in Mathematics,
26, 165–190.
Resnick, L. B., & Ford, W. W. (1981). The psychology of mathematics for instruction.
Hillsdale, NJ: Lawrence Erlbaum.
Schoenfeld, A. (1992). On paradigms and methods: What do you do when ones you
know don’t do what you want them to do? Journal of the Learning Sciences, 2(2),
179–214.
Schwarz, B. B., & Hershkowitz, R. (1999). Prototypes: Brakes or levers in learning the
function concept? The role of computer tools. Journal for Research in Mathematics
Education, 30, 362–389.
Skemp, R. R. (1976). Relational understanding and instrumental understanding.
Mathematics Teaching, 77, 20–26.
Sproule, O. E. (2000). The development of concepts of linear and quadratic equations.
Unpublished PhD thesis, Queen’s University of Belfast.
Stacey, K., Chick, H., & Kendal, M. (Eds.). (2004). The future of the teaching and learning
of algebra. Boston: Kluwer Academic Publishers.
Stacey, K., & MacGregor, M. (1997). Ideas about symbolism that students bring to
algebra. Mathematics Teacher, 90, 110–113.
Stacey, K., & MacGregor, M. (1999a). Taking the algebraic thinking out of algebra.
Mathematics Education Research Journal, 11(1), 25–38.
Stacey, K., & MacGregor, M. (1999b). Implications for mathematics education policy
of research on algebra learning. Australian Journal of Education, 43(1), 58–71.
Stigler, J. W., & Hiebert, J. (1999). The teaching gap: Best ideas from the world’s teachers
for improving education in the classroom. New York: Free Press.
Thomas, M. O. J., & Tall, D. (2001). The long-term cognitive development of symbolic
algebra. In H. Chick, K. Stacey, J. Vincent & J. T. Zilliox (Eds.), The future of the
teaching and learning of algebra (Proceedings of the 12th ICMI Study Conference,
pp. 590–597). Melbourne: University of Melbourne.
Vaiyavutjamai, P. (2004a). Factors influencing the teaching and learning of equations and
inequations in two government secondary schools in Thailand. Unpublished PhD
thesis, Universiti Brunei Darussalam.
Vaiyavutjamai, P. (2004b). Influences on student learning of teaching methods used
by secondary mathematics teachers. Journal of Applied Research in Education, 8,
37–52.
Vaiyavutjamai, P., Ellerton, N. F., & Clements, M. A. (2005). Influences on student
learning of quadratic equations. In P. Clarkson, A. Downton, D. Gronn, M.
Horne, A. McDonough, R. Pierce, & A. Roche (Eds.), Building connections:
Research, theory and practice (Vol. 2, pp. 735–742). Sydney: MERGA.
Effects of Classroom Instruction on Students’ Understanding of Quadratic Equations 77

Voigt, J. (1994). Thematic patterns of interactions and sociomathematical norms. In

P. Cobb & H. Bauersfeld (Eds.), The emergence of mathematical meaning:
Interaction in classroom cultures (pp. 163–201). Hillsdale, NJ: Lawrence Erlbaum.
Wagner, S., & Parker, S. (1993). Advancing algebra. In P. S. Wilson (Ed.), Research ideas
for the classroom: High school mathematics (pp. 119–139). New York: Macmillan.
Warren, E. (2000). Research in teaching and learning algebra. In K. Owens &
J. Mousley (Eds.), Research in mathematics education in Australasia 1996–1999
(pp. 161–180). Sydney: MERGA.
Warren, E., & Pierce, R. (2004). Learning and teaching algebra. In B. Perry, G.
Anthony, & C. Diezmann (Eds.), Research in mathematics education in Australasia
2000–2003 (pp. 291–312). Flaxton, Australia: Mathematics Education Research
Group of Australasia.
Zaslavsky, O. (1997). Conceptual obstacles in the learning of quadratic functions.
Focus on Learning Problems in Mathematics, 19(1), 20–44.
Zazkis, T., Liljedahl, P., & Gadowsky, K. (2003). Conceptions of function translation:
Obstacles, intuitions and rerouting. Journal of Mathematical Behavior, 22, 437–450.

Authors
Pongchawee Vaiyavutjamai, Department of Secondary Education, Faculty of
Education, Chiang Mai University, Chiang Mai, Thailand, 50200. Email:
<[email protected]>
M. A. (Ken) Clements, Illinois State University, Department of Mathematics,
Campus Box 4520, Normal, Illinois 61790-4520. USA. E-mail: <[email protected]>