SHOULD

MORE

THAN

APPROACH BE

USED

FOR

STUDENTS' ERRORS?

AREAS,

VOLUMES

ONE

AND

THE

THEORETICAL
ANALYZING
CASE

OF

INTEGRATION

PESSIA TSAMIR

This article addressesthe question: Should more than one

theorybe used for analyzingsourcesof students'errors?For
this purpose,this articleexaminesthe contributionsof Stavy
and Tirosh's intuitive rules theory and Fischbein's theory
regardingthe algorithmic,the intuitiveand theformal components of mathematical knowledge, for looking into
students'solutions to problems dealing with definite integrals, areas and volumes. I highlight the different sources
offered by these two theories for students'erroneoussolutions (e.g., ifb-a = d-c then

jf(x)dx=ff(x)dx),

andconsequentlydiscuss the benefitsof differenttheoretical

frameworksto the understandingof the question "why do
studentserrin certainways?".
There is a wide consensus that data regardingstudents'
ways of thinking should be used in both the designing of
mathematicsinstructionand in its teaching (e.g., NCTM,
2000; Noddings, 1992; Tirosh, 2000). For this purpose,
teachershave to be familiar with students'common errors
and with their possible sources (knowingthat and knowing
why, cf. Even and Tirosh, 2003). However, why students
frequentlyerr in specific ways is a complex question, and,
thus, mathematics educators often use theoretical frameworks as connecting organizersfor investigating students'
tendencies to err (e.g., Tall [1]; Zazkis, 1999). Typically,
the analysis of the data, in studies dedicated to this issue,
uses a single theoreticalframework.
Should more than one theoreticalframeworkbe used for
analyzing sources of students' errorsfor a certain topic? I
shall examinethe benefits of using two theories,e.g., Stavy
and Tirosh's (e.g., 2000) and Fischbein's (e.g., 1993) theoreticalapproachesto the analysis of students'solutions for,
specifically, integrationproblems.Stavy and Tirosh's intuitive rules theory is a task-oriented model, claiming that
studentshave typical (correctand incorrect)ways for solving tasks that are structuredin specific ways, regardlessof
theircontent.Fischbein'sthree-knowledge-components
theory [2], on the other hand, offers a content-oriented
interpretationof students'mathematicalperformances,indicating ways in which students may make content-based
connections between mathematical, scientific and daily
issues thatare not necessarilyrelated.
28

The following sections include:

a briefdescriptionof the two theoreticalapproaches
dataabouterroneoussolutionsto integrals
a discussionof the datain lightof the two theoretical
approaches,subsequentlyaddressingthe question:
Shouldmore thanone theoreticalapproachbe used
for analyzingsourcesof students'errors?

The two theoretical approaches

Whatis the intuitiverules theory?: Stavy and Tirosh (e.g.,

2000) formulated the intuitive rules theory, showing that
students often react to scientifically unrelated but structurallysimilar tasks in line with three intuitive rules: more
A - moreB, same A - same B and everythingcan be divided.
The rules were considered to be intuitive since they have
the characteristicsof intuitive knowledge, i.e., immediacy,
obviousness, self-evidence, confidence and perseverance
(cf. Fischbein, 1987).
Here,I focus on the intuitiverulesameA - same B. It was
identified in students'reactionsto comparisontasks, when
they were presented with two systems that are equal in
respectto one observedcharacteristicA (A, = A2),andasked
to comparethese systems with regardto anothercharacteristic B (whereBj * B2).A commonincorrectresponseto such
tasks is: "A, = A2thereforeBj = B2"or same A - same B.
Students were found to use the intuitive rule same A same B in various topics. For example, they claimed that
polygons with the same perimetersmusthave the same area,
and vice versa: same area - same perimeter,same perimeter - same area (e.g., Dembo, Levin and Siegler, 1997;
Hoffer and Hoffer, 1992; Menon, 1998; Reinke, 1997); and
in their solutions to areas-and-volumes tasks, students
tended to claim same surface area - same volume (e.g.,
Stavy,Tiroshand Ronen [3]).
Whatis Fischbein 's theory?: In his seminal analysis of
students'mathematicalperformances,Fischbein(e.g., 1987;
1993) relatedto three componentsof knowledge: algorithmic, formal, and intuitive. According to Fischbein,
algorithmicknowledge is the ability to activateprocedures
in solving problems and understandwhy these procedures
"work".Formal knowledgerefers to the wider perspective
of the mathematicalrealm - what is accepted as valid and

For the Learningof Mathematics27, 2 (July,2007)

FLM PublishingAssociation, Edmonton,Alberta,Canada

This content downloaded from 202.94.83.67 on Tue, 26 Aug 2014 10:27:45 UTC
All use subject to JSTOR Terms and Conditions

how to validatestatementsin a mathematicalcontext - and

intuitiveknowledgeis described as an immediate self-evident cognition - students are sure, feeling no need of
validation.Intuitiveknowledgemay lead the learnerto generalizationsthatgo beyond the given.
Fischbeinexplainedthatthe threecomponentsareusually
inseparable, and, often, the intuitive backgroundhinders
the formal interpretationor the use of algorithms.He presented and analyzed the sources of a number of rigid
algorithmicprocedures,which he labeled algorithmicmodels. Forexample,students'tendenciesto claim that(a + b)5=
a5+ b5or sin(a + P) = sina + sin|3,were interpretedas evolving from the applicationof the distributivelaw (Fischbein,
1993; Fischbeinand Barash,1993).
Errors in integrals

found

in calculus-lesson

segments

This section describesfive lesson segments,taughtby experienced mathematics teachers, in classes of 25-28 12th
graderswho discussedproblemsdealing with integrals.
Segment 1: Parameters, areas and volumes
Studentswere workingin pairson a problem(Figure 1); the
teacherand a visiting prospectiveteacher,Betty, responded
to requestsfor help.
1. Dan [4]: That'sit. . . we finished all we had to do.
2. Daffy: No. . . no. . . we still have the last item.
A straightline y = ajc(a > 0) intersectsthe parabolay =
-je2+ 4jc in two points: (0, 0) and another point, P.
Anotherstraightline, perpendicularto the jc-axis,passes
throughP.
1. Calculate- Forwhich value of 'a' will the area
enclosedbetweenthe perpendicularline, y = ajc
and the jc-axisbe maximall
2. Calculate- For which value of 'a' will the volumecreatedby the rotationof the previousarea
aroundthe Jt-axisbe maximall

Segment 2: Adjacent areas

A problemdiscussed duringthis lesson was:
The area limited (enclosed) between the graphs of the functions
y

= V5-JC and = V.v-1

y

and the x-axis, rotates around the x-axis.

-^

(1) Find the marked area. (2) Find the volume of the rotating figure.

Figure2: Problem2.
The solutionof this problemis somewhatsurprising,since
in the calculationof the area:
S - 5, . 5, - nfj(x~^~\~)dx + ;r//(5"x)dx -^'J + ^-j8 -^Js
3
3
3
1

Sj andS2areequal,andin thecalculationof therelatedvolume:

3
5
V = V, V, - .t/(x - 1)dx + JiJ(5 - x)dx -2* + In - 4* ,
1
3

V! and V2 are equal as well.

_
The studentseasily arrivedat the areaof 2 times (2/3)^8,
andone student(55. Gal) wrotean expressionfor the volume:
y = jij(x-l)dx +;rjf(5-x)dx= ;

several studentsvoiced theirsolutionsfor this expression:

58. Danny:I got V^ It's 2jt . . . so, I actuallygot BOTH
. . . V2 is also 2ji. Sure,it's equal to V,.
60. Ron: They're [V! and V2] the same ... the areas
were also equal ... So, V equalstwo times Vj that's
two times 2jr . . .
61. Danny:I said it ... That'swhat I was saying . . .
Clearly, the students' "equal volumes" conclusions were
based on their "equalareas"solutions.Danny and Ron confidently deducedthatV, - V2because S, - S2.
Segment 3: Areas and volumes for/(x) = 2x
A problemdiscussed in this lesson was:

Figure1: Problem1.
3. Dan: [sharply]It's exactly the same solution as in
Item 1... The same value of 'a' will lead to the maximum area, and to the maximum of the [related]
volume.
4. Daffy: [hesitant]ehh...
5. Betty: [to Dan] You can check your solution by
comparingit with the solution in the book [i.e., (1)
a = 3/4, (2) a = 8/5].
6. Dan: [scornfully]There might be a mistake in the
book. . .
Dan was confidentin the correctnessof his erroneousstatement same value of la 'for maximal area - same value of
'a yformaximal volume solution, to the point of rejecting
Betty's suggestion to re-examine his solutions in light of
the differentsolution in the textbook.

Figure3: Problem3.
One of the students,Ron, presenteda conventionalcorrect
solution on the blackboard:

S = J(2x)dx- [2- 1 = [x2 = 12.25-0.25= 12

2
--<
L

J0.5

29

This content downloaded from 202.94.83.67 on Tue, 26 Aug 2014 10:27:45 UTC
All use subject to JSTOR Terms and Conditions

Aftera shortwhile, anotherstudent(25. Edna)was surprised

to see that she reached the same, correct solution (12),
althoughshe mistakenlycalculated
/ insteadof / .
She expectedthe differentlimits to yield differentareas.The
teacherasked a volunteerto check whetherthe areas were
the same. Ann said:
27. Ann: Yeah ... What's to be done? Like ... it's the
same function and the same distance ehh . . . from
X!to x2 ... so it's the same area . . .
Ann believed that, since it is the same function and since
Xj- x2 is the same in both cases (which was a mistake),the
areashould also be the same. The teacherinsisted thatAnn
show a traditionalcalculation.She reacted:
29. Ann: Show . . . like calculate ... do the whole thing
.... Yeah ... I thinklcan ... OK ...OK ... I'll doit
the long way . . . [walksup to the boardand writes]:

4
Sf(2*)&-

r 214
2-

-16-4

= 12

You see ... It's the same . . . It's also 12.

Then, the class was asked to solve part 2 of the problem
(Figure3). Ron dictatedhis solution and the teacherwrote
on the blackboard:

V=nf(2x)2dx
V=

jiJ4x2dx

= 4ji -

^-20-^(64-8)-^
Suddenlythe teacherrealizedthatthe originallimits were
between x = 0.5 and x = 3.5. She asked: "Whatabout the
second [original]volume?",and studentsresponded:
41. Ann:It's the same areas... it SHOULDalso be the
same volume . . . about 75jt ... we can do without
the calculations. . .
42. Gal: It's the same formula... on the same function
. . . and in both ... the areais 12 ... It's equal ... the
volume . . .
Ann and Gal clearly and confidently expressed ideas, connecting the equal sizes of areas with the equality of related
volumes, stating that "we can do without the calculations
. . .".Althougha simple, familiarcalculationcould show that
the volumes differed in spite of the equal areas, neither of
the studentsbotheredactually to do this calculation. They
were confidentin the correctnessof theirintuitivesolutions.
Segment 4: Composite trigonometric function
Students were working on the first part of Problem 4 (see
Figure4). One studentarrivedat a negative areaand said "I
can't find what's wrong",so the teacherinvited him to present his solution.

1. Find the area enclosed between f(x) = cos2(x) and

the x-axis, between x = 0 and x = jt/2.
2. Findthe volume createdby rotatingthis areaaround
the x-axis.
Figure4: Problem4.
1. Erez: [writes] /(cosx)*"[~T~| =~3
2. Gil: [cuts in] No ... no ... you forgot the ... to
divide by the derivative of the inner function . . .
can I show? [approachesthe boardand writes]:
,
(cosx)3
r.
J

-3sinx

3. T: [to Gil] Why is this so?

4. Gil: We did things like that . . . here we divide by
(-sin jc) . . . the innerderivative.
5. T: [to Gil] Whatis your solution?How much is the
area?
6. Gil: I did not substitute [the limits] ... You ... I
have to do just that . . .
Although the class had solved similar tasks by using the
expressioncos(2jc)= 2cos2jc- 1, studentsstill tendedto grasp
the integral of any composite function of the type \f(x)]n
where/(jc) is not necessarilylinear

(I^r,Erezor-f-,Gil).
/?+ !
( +

0/W

Erez sensed that something was wrong due to his negative

result. Gil, however, did not bother to substitutethe numbers, so he did not notice the zero-denominator that he
would have reached.
Segment 5: Areas, volumes and intersection points
Studentswere asked to solve the following problem:
1. Find the area enclosed between y = (1 - x)(x - 5)
and the x-axis, between x = 0 and x = 3.
2. Findthe volume createdby rotatingthis areaaround
the x-axis.
Figure5: Problem5.
Theproblemwas presentedwithno accompanyingdrawing.
After several minutes of individual work in class, Eran
(contribution7) said that he got 15.75, while Shirley (contribution8) interrupted,saying that she got 3. The teacher
invited them to the blackboard:
Eran:[writes]
3
S f(l - v)(.v - 5)dx -

W-x)dxj{x-5)(k-

hum-

(-1.5) (-10.5) - 15.75

Shirley:[writes]

30

This content downloaded from 202.94.83.67 on Tue, 26 Aug 2014 10:27:45 UTC
All use subject to JSTOR Terms and Conditions

3
S ? (1 - x)(x - 5)dx
f (_5 + 6X _ x" )dx -

-5r + 6

-3

2 3Jo

11. Eran: So ... the area is the integral between the

given limits . . . zero and three ... of [points to the
function]. . . andI did it withoutopeningthe parentheses . . . integralof each [factor]betweenzero and
three ...
12. Shirley: [interrupts]But you have to simplify the
expression ... on this expression I calculated the
integralfor the area . . .
13. Gil: [fromhis place] I calculatedtwo integrals ...
from zero to one andfrom one to three . . . and I did
absolutevalues ...
14. Eran:So it's like from zero to three . . .
Both Eranand Shirley erroneouslyignoredthe intersection
point at x = 1. Insteadof
| ffixytc|+| j/(x)A| they calculated /(*)*
and Eranalso believed that
//(*) g(x)dx- ff(x)dx Jg(x)dx .
All in all, the lesson segments illustrate students' diffitasks.
culties with definiteintegrals-area-volume

Analysis of the data in light of the two theoretical approaches

The main aim of this articleis to examinethe pros and cons

of implementing more than one theory for analyzing students' mathematical reasoning. To explore how the two
theoretical approaches, i.e., the intuitive rules theory and
Fischbeinystheory,may contributeto our understandingof
students' errors in definite integrals, areas and volumes
tasks, I addressthe questions:Whatwere students'errorsin
definite integralstasks?;What are possible sources for students'errors?andShouldmorethanone theoreticalapproach
be used for analyzingstudents'mathematicalsolutions?
What were students9errors in definite integrals tasks?
A numberof integration-relatedand area-volume related
errorswere evident in students'solutions. In segments 1, 2
and 3, studentsrepeatedlyexpressederroneoussame area same volumeideas (e.g., Stavy,Tiroshand Ronen [3]). The
novelty of the datahere is thatthe area-volume connections
are associatedwith integrals.
In Segment 1, there is a new variant of the belief that
equal areas, when revolving around the jc-axis, result in
equalvolumes. Studentsclaimedthat the value of '"a*which
yields the maximal area is equal to the value of a' which
yields the related maximal volume. In Segment 2, equal
areaswere revolving aroundthe jc-axisand indeed created
equal volumes. However, ratherthan calculating and perhaps pointing to the uniqueness of the equal-volumes
solution, the students spontaneously deduced the equality

of the volumes fromthe equalityof the areas.Then,Ann and

Gal erroneouslystatedthatwhen equalareasrevolve around
the x-axis, they createequal volumes
(i.e., if //w*-/(ikfcthen njfwdx-nfgwdx).
In Segment 3, Ednaabsentmindedlycalculated
]f{2x)c/xinstead of ]}2x)dx

and was surprisedwhen she realizedthat in spite of the different limits she reached the same solution. She expected
an integrationof the samefunction withdifferentboundaries
to yield differentsolutions
(Z.?.,ifa*b,c*dthen ff(x)dx*Jf(x)dx).
Later on, in the same lesson, Ann expressed anothererroneous idea, that an integration of a function between the
limits a, b and between the limits c, d where b - a = d - c
should yield the same solution
(i.e., ifb-a = d-c then //(*)<&- ff{x)dx ).
In Segment 4, two more erroneous formulas are presented:
^ (Gil).
Jl/W*-^
/[/(x)r(fc_[^(Erez)and
Finally,in Segment 5, Eranused a pseudo formula(Vinner,
1997):
ff(x)'g(x)dx=ff(x)dx'Jg(x)dx>

and Eranand Shirleyclaimed thatthe areaenclosed byfix),

the x-axis, x = a, and x = b, is

S=ff(x)dx ,
ignoringa significantintersectionpoint with the jc-axis(see
also, Orton, 1983).
to observe"howstuWhile it seems quitestraightforward
dents err" the analysis of why they err, and why in this
particular way, is much more demanding.The next section
offers suggestions of sources for the identifiederrorsreferring to two theoreticalmodels simultaneously.
What are possible sources for students9errors?
This section shows how the intuitive rules theory and Fischbein's theoretical approach provide us with possible
sources for students' errors.A closer look at the data from
the intuitiverulestheoryperspective,yields thatseveralerroneous solutions had a recurrentsame A - same B pattern:
same value of 'a' for max area - same value of 'a' for max
volume (Dan, Segment 1);same area- same volume (Danny
andRon, Segment2; Ann, Segment3); same function- same
interval- same area (Ann, Segment 3); andsame formulasame function- same area- same volume (Gal, Segment3).
Does this structurein the solutionsimply thatthe students
were thinkingin termsof the intuitiverulesameA - same Bl
As mentionedbefore, Stavy and Tirosh (e.g., 2000) identified students'tendenciesto give same A - same B solutions
to comparisontasksthatpresenttwo entitiesthatareequalin
a certainrespectA (Aj = A2),while askingto comparethese
entities with regard to another aspect B (where B! * B2).
Commonincorrectsolutions aresame A - same B, and they
typically carrycharacteristicsof immediacy, obviousness,
self-evidence, confidence and perseverance.Bearingthis in
31

This content downloaded from 202.94.83.67 on Tue, 26 Aug 2014 10:27:45 UTC
All use subject to JSTOR Terms and Conditions

mind, and re-examiningthe datain orderto see whetherthe

intuitiverules theory is appropriatefor analyzing students'
solutionsin the presentedlesson-sections,the studentswere
not presentedwith a comparisontask, but with "calculate"
problems.Still, in all the cases, while explicitly dealingwith
calculations, the students preferredto make shortcuts and
jumped to conclusions as solutions for the factors in question from an irrelevantgiven or from an irrelevantsolution
they had reachedbefore.Thatis, the students,implicitlyand
on their own initiative, conducted comparisons, although
they were neitherasked to nor led to do so. This may point
to the stronglypreservedpower of the erroneousideas they
held. The students also voiced their same-same solutions
with no hesitationsand as immediatereactionsto the problems posed. The confidence they had in the correctnessof
theirsolutionsandtheirtendencyto regardtheirsuggestions
as being self-evidentwere expressedin theirtone andin their
phrasings(e.g.,Ann said in lesson 3, "itSHOULDalso be the
same volume" and "we can do without the calculations").
Most striking was Dan's firm rejection of Betty's (the
prospectiveteacher's) suggestion to re-examine his samesame solutionin light of the differentsolutionprintedin his
book.He was so surethathe hadreachedthe rightconclusion
that not only did he not agree to check it, but he assumed
furtherthatdie solutionin the book was incorrect.
In sum, the students might have implicitly solved selfdesigned comparison tasks, whose structure might have
influencedthe students'reasoning,eliciting answersin line
with the intuitiverule same A - same B.
On the otherhand,an examinationof the datawith reference to Fischbein's three-knowledge-componentstheory
(e.g., Fischbein,1987; 1993), indicatesthatsome of the previously mentioned,erroneoussolutionscan be interpretedby
addressingstudents'intuitive,content-embeddedtendencies
to view areas and volumes of given figures as completely
Thatis, it couldbe thatDannyandRon (Segment
interrelated.
2) andAnn andGal (Segment3) intuitivelyover-generalized
connectionsbetween areas and relatedvolumes, and, thus,
believedthatwhenthe areasareequal,the relatedvolumesare
equaltoo (see also Stavy,TiroshandRonen [3]).
Similarly,it could be thatDan's (Segment 1) belief that a
parameterthat formulates a maximal area should yield a
maximal, related volume evolved from his grasp of areas
and volumes being totally dependent.It is furtherpossible
that students'tendency to claim same function-sameintervdl-same area v .#., 27. Ann, Segment 3), was rooted in
anotherintuitive, content-embeddedbelief regardingconnectionsbetweenperimetersandareas.Possibly,the way the
graphof the functionandthe axis surroundthe enclosed area
reminded students of a perimeter, which they intuitively
connected to the related area (see also Dembo, Levin and
Siegler, 1997; Hoffer and Hoffer, 1992; Menon, 1998;
Reinke, 1997;Tsamirand Mandel,2000).
Consequently,several errorsin Segments 1, 2 and 3, previously interpretedas possibly evolving from the implicit
structureof comparisontasks and from students'use of the
intuitiverule same A - same B, could be rootedin students'
intuitivegraspof contentrelatedissues regardingperimeterarea-volumeconnections.

However, sources for all the errorsidentified in the data

have not yet been offered. For example, what are possible
sources for Erez's claim that
and for Gil's claim that
(Segments)?
/[/wrA.J^r;)
These cannot be interpretedby the intuitive rule same A same B9butFischbein'notionof algorithmicmodelsmayoffer
possiblesourcesfor the lattererroneoussolutions.Thesesolutions indicatestudents'tendenciesto use previouselystudied,
notnecessarilyrelevantalgorithms(see also,Orton,1983;Ferarri-Mundi, 1994). The specific mathematicalconditions
underwhich the algorithmwas valid were ignored,thus the
new implementationof the algorithmis erroneous,andgives
rise to differentalgorithmicmodels: the polynomial model
wherestudentsperformthe integralof a compositefunction,
while drawingon theirworkwith polynomials
the composite-on-linearmodel where studentstreat
JAg(x)]dx as f.f[ax+b}dx( j[M*-j^

);

and the distributivelaw model

(in Segment5: If(x) g(x)dx~ff(x)dx'fg(x)dx,

see also Fischbeinand Barash, 1993).
Clearly,bothStavyandTirosh'sandFischbein'smodelscontributedto the analysis of sources of students'errorsin this
study.A questionthatarisesis, Shouldmore thanone model
solutions?
commonlybe usedto analyzestudents'mathematical
Should more than one model be used for analysing students9mathematical solutions?
I will first state that, in my view, using more thanone theoretical framework for analyzing students' solutions is
important.As shown in the analysesof students'solutionsto
integration-area-volumetasks, applying two models may
yield threetypesof situationwherethe datacanbe interpreted
(a) by bothmodels (b) by only one model, or (c) by none.
In the lessons described,a numberof erroneoussolutions
could be explainedby meansof each of the two models.This
phenomenon may point to cases where we cannot be certain regardingthe sources for students' errors.It could be
that students erred merely because they used the intuitive
rule same A - same B but it could also be that, ratherthan
being influencedby the structureof the task, the reasonsfor
the errorswere content-oriented,i.e., mal-understandingof
the notionsareaandvolume. Some studentscould have been
influenced by both the intuitive rule and their poor familiarity with the notions at hand, and, thus, they might have
become extremely certainabout their erroneoussolutions.
Here, the use of two models providesus with multipleperspectives for possible sources of students'difficulties.
However,not all of the errorscould be analyzedby both
models. Several errorscould be interpretedonly in light of
Fischbein 'salgorithmic model. That is to say, the intuitive
rules theorycould not suggest sources for these errors,and,
withoutFischbein'stheory,we might have missed the theo-

32

This content downloaded from 202.94.83.67 on Tue, 26 Aug 2014 10:27:45 UTC
All use subject to JSTOR Terms and Conditions

retical interpretation of the errors, merely viewing them

instead as instances of confusion. Fischbein's theoretical
"lens" brings to the surface details, regarding students' difficulties, which might be ignored without them. Here, the
benefits of having more than one model to interpret the data
seem quite obvious.
Although the two models seem beneficial in the analysis
of the data, their use did not provide answers regarding the
sources of all apparent errors. Some of the errors, like
Edna's different boundaries - different definite integral solution, her mistaken substitution of 2 and 4 instead of 0.5 and
3.5, in lesson 1, and her assumption that the differences
between 2 and 4 equals the difference between 0.5 and 3.5,
cannot be explained by any of the two models suggested in
this article. While the different boundaries - different definite integral solution can be regarded as being in line with
a new intuitive rule, different A - different B, offering an
extension to the intuitive rules theory. The 2 to 4 instead of
0.5 to 3.5 errors are either a result of confusion or of a general, yet unfamiliar phenomenon that needs to be interpreted
by another theoretical model. Clearly, the richness of our
understanding of students' ways of thinking was significantly, yet insufficiently, increased by the double-model
implementation. Thus, my strong claim is that it is important
to use a variety of theoretical approaches when analyzing
data regarding students' mathematical reasoning.
This article offers the examination of the given data in light
of two cognitive frameworks. In another article, Even and
Schwarz (2003) showed how analyses based on a cognitive
theoretical orientation and a socio-cultural orientation, led to
different understandings of the same lesson. The two theories
explained, in different yet complementary ways, why students
exhibited unexpected mathematical behaviors. While the cognitive analysis pointed to students' cognitive difficulties in
addressing different representations, the socio-cultural analysis pointed to differences in the teacher's and the students'
motives, beliefs and norms regarding school mathematics.
In conclusion, along with my strong belief that it might be
beneficial to extend the use of multiple theories in the interpretations of data regarding students' mathematical solutions,
I would like to indicate a point that needs some careful consideration. It is well known that theories impart interpretive
and predictive powers regarding students' difficulties (e.g.,
Guba and Lincoln, 1994; Stavy and Tirosh, 2000). It seems
that the use of various theories for the analyses of research
data may contribute both to the strength of the theories and
the interpretations of the data. However, it should be noted
that theory and the interpretations of research data seem to
be trapped in a vicious circle:
research findings are interpretedin ways that confirm the
theories that serve as research lenses, and correspondingly support the theories. (Even and Schwartz, 2003)
In both articles, the different theories played a compatible
and complementary role. What new insights can be gained
by using other or additional theoretical frameworks for analyzing the data? Might we get incompatible interpretations,
and, if yes, how should we go about it? Clearly, these questions call for research to accompany the wider use of
theories in the analysis of data.

Notes
[1] Tall. D. (2005) 'A theoryof mathematicalgrowththroughembodiment,
symbolism and proof, plenary lecture for the International Colloquium
on MathematicalLearningfrom Early Childhoodto Adulthood,5-7 July,
Belgium.
[2] Fischbein addressed three components of mathematicalknowledge:
algorithmic,formaland intuitive.However,the label the three-knowledgecomponentstheoryis mine.
[3] Stavy, R., Tirosh, D. and Ronen, I. (1996) 'Overgeneralizations of
schemes: the case of conservation', paper presented at the international
seminar,Thegrowing mind,Geneva, Switzerland.
[4] In transcripts,"1. Dan"indicatesthatDan is the pseudonymof the student and 1 places the contributionin the sequence of the lesson; similarly,
in "28.T\ T' standsfor the teachers'contribution,and '28' placesthatcontributionin the sequence of the lesson.

References
Dembo, Y., Levin, I. and Siegler, R. (1997) 'A comparisonof the geometric reasoningof studentsattendingIsraeliultra-orthodoxandmainstream
schools', DevelopmentalPsychology33, 92-103.
Even, R. and Schwarz,B. (2003) 'Implicationsof competinginterpretations
of practice to researchand theory in mathematicseducation', Educational Studiesin Mathematics54, 283-313.
Ferrini-MundyJ. and GrahamK. (1994) 'Research in calculus learning:
understandingof limits, derivatives and integrals', in Kaput, J. and
Dubinksy, E. (eds), Research issues in undergraduate mathematics
learning - preliminaryanalyses and results,MathematicalAssociation
of AmericaNotes, 33, pp. 31-45.
Fischbein,E. (1987) Intuitionin science and mathematics:an educational
approach.Dordrecht,The Netherlands,Reidel.
Fischbein,E. (1993) 'Theinteractionbetweenthe formal,the algorithmicand
the intuitivecomponentsin a mathematicalactivity'in Biehler,R., Scholz,
R., Straser,R. and Winkehnann,B. (eds), Didactics of mathematicsas a
scientificdiscipline,Dordrecht,The Netherlands,Kluwer,pp. 231-245.
Fischbein,E. and Barash,A. (1993) 'Algorithmicmodels and theirmisuse
in solving algebraicproblems', Proceedings of the seventeenthannual
conferenceof the InternationalGroupfor the Psychologyof Mathematics Education,1, Tsukuba,Japan,pp. 162-172.
Guba, E. and Lincoln, Y. (1994) 'Competing paradigms in qualitative
research',in Denzin, N. and Lincoln, Y. (eds), Handbookof qualitative
research,CA, Sage, pp. 105-117.
HofferA. and Hoffer, S. (1992) 'Geometryand visual thinking'in Post, T.
(d., second edition), Teachingmathematicsin grades K-8: researchbased methods,Boston, MA, Allyn and Bacon.
Menon, R. (1998) 'Preservice teachers' understandingof perimeterand
area', School Science and Mathematics98(7), 361-368.
Mundy,J. (1984) 'Analysisof errorsof firstyearcalculusstudents',in Bell,
A., Love, B. and Kilpatrick,J. (eds), Theory,researchand practice in
mathematicseducation,Proceedingsof thefifth InternationalCongress
on MathematicalEducation,Nottingham,UK, pp. 170-172.
NCTM (2000) Principles and standardsfor school mathematics,Reston,
VA, NationalCouncilof Teachersof Mathematics.
Noddings, N. (1992) 'Professionalizationand mathematicsteaching', in
Grouws,D. (d.), Handbookof Researchon MathematicsTeachingand
Learning,New York,NY, Macmillan,pp. 197-208.
Orton, A. (1983) 'Students' understandingof integration', Educational
Studiesin Mathematics14, 19-38.
Reinke, K. (1997) 'Area and perimeter:prospectiveteachers'confusion',
School Science and Mathematics97(2), 75-77.
Stavy,R., andTirosh,D. (2000) How students'(misunderstandscience and
mathematics:intuitiverules, New York,NY, TeachersCollege Press.
Tirosh,D. (2000) 'Enhancingprospectiveteachers'knowledgeof children's
conceptions:the case of division of fractions',Journalfor Researchin
MathematicsEducation31, 5-25.
Tsamir,P. and Mandel,N. (2000) 'The intuitiverule same A - same B: the
case of area and perimeter', Proceedings of the twenty-fourthannual
conferenceof the InternationalGroupfor the Psychologyof Mathematics Education,4, Hiroshima,Japan,pp. 225-232.
Vinner, S. (1997) 'The pseudo-conceptual and the pseudo-analytical
thought processes in mathematics learning', Educational Studies in
Mathematics34, 97-129.
Zazkis, R. (1999) 'Intuitiverules in numbertheory:example of "themore
of A, the moreof B" rule implementation',EducationalStudiesin Mathematics40, 197-209.

33

This content downloaded from 202.94.83.67 on Tue, 26 Aug 2014 10:27:45 UTC
All use subject to JSTOR Terms and Conditions