The Relationship Between Posing and Solving Division-With-Remainder Problems Among Flemish Upper Elementary School Children
The Relationship Between Posing and Solving Division-With-Remainder Problems Among Flemish Upper Elementary School Children
The Relationship Between Posing and Solving Division-With-Remainder Problems Among Flemish Upper Elementary School Children
INTRODUCTION
(see Van Dooren, Verschaffel, Greer, & De Bock, 2006; Verschaffel, Greer, &
De Corte, 2000, for overviews of this research). More specifically, instead Of
going through the different phases Of a genuine modelling process
(understanding. modelling, analysis. interpretation. e,'aluation. and
communication) they go through a superficial process whereby a particular
mathematical model is triggered more or less automatically by certain
conspicuous elements in the problem situation (e.g., certain key words in the
problem statement), there is no serious attempt to understand and analyse the
problem situation, and the result is immediately computed and communicated as
the answer, without referring back to the original problem situation to verity that
it is a meaningful response to the original question. or to check its
reasonableness.
One of the most frequently studied types Of non-routine problems in this line Of
research are the division-with-remainder (DWR) problems, i.e. problems wherein the
dividend and the divisor are both integers and wherein the division results in a non-
integer outcome that still needs to be interpreted and e,'aluated in relation to the real-
world constraints of the problem setting. This kind of problem was used with a stratified
sample of 13-year-olds in the Third National Assessment of Educational Pmgrvss
(A14EP) in the U.S. (Carpenter, Lindquist, Matthews, & Silver, 1983): "An army bus
holds 36 soldiers. If 1128 soldiers are being bussed to their training site, how many buses
are needed?" Of about 700/0 of the students who correctly carried out the division 1128
by 36 to get a quotient of31 and remainder of 12, only 23% (ofthe total number of
students) gave the answer as 32 buses, 19% gave the answer as 31 buses, and 29% gave
the answer as "31 , remainder 12". A similar division problem appeared on the 1983
version of the California Assessment Program (CAP) Mathematics for Grade 6, and was
answered correctly by about 35% of the sixth-graders in Califomia (Silver et al., 1993).
As in the NAEP assessment, most students erred by giving nonwhole-number answers or
rounding the outcome of the division to its nearest number predecessor. Since
then, numerous studies involving this type Of problem have been done in different
countries, basically leading to similar results and conclusions (for an overview see
Verschaffel et al., 2000). In an early reflection upon these results, Hilton (1984, p. 8)
asserted that:
the separation of division from its context is an appalling feature of traditional drill
arithmetic. the solution to the division problem I(HH) 12 should depend on the
context ofthe problem and not the grade of the student.
- 532
ET AL.
- 533
— 532 remainder 10
— 532.83 remainder 4
- 532.83333
- About 530_
However. Streenand himself never reported systematic findings or even
anecdotal data on children's reactions to this nice problem-posing task presented
in the context of an ascertaining study or a design experiment.
However, Streenand himself never reported systematic findings or even
anecdotal data on children's reactions to this nice problem-posing task presented
in the context of an ascertaining study or a design experiment.
From the studies on problem posing, we can conclude that this activity has a
potential value in assessing and improving children's problem-solving capacities.
However, an important feature of the problem-posing research up to now is that it
has investigated the relationship between problem posing and solving especially,
and even almost exclusively, for standard word problems that can be
unproblematically modelled by one ore more operations with the numbers given
in the problem Statement. Stated differently, no research investigated this
relationship for problematic problem situations, wherein students have to use their
commonsense knowledge and real-world experience when trying to model and
solve the problem (Verschaffel et al., 2000). At the basis or the present study was
the expectation that, especially for this kind of non-standard problems, looking at
children's problem-posing processes and skills may have a great diagnostic value,
because of the unfamiliar or non-routine nature of word problem posing (as
opposed to the ritual scholastic activity of solving school word problems, which
elicits (quasi-) automatised behaviour) and because of its potential to reveal
modelling difficulties during the phases of the modelling cycle that are most
interesting from a modelling perspective, namely the establishment of an
appropriate link between the understanding of the problem situation and the
development or a mathematical model modelling) and the interpretation of the
result of the computational work in terms of the original problem situation (—
interpretation).
Based on the above literature review and the results Of a pilot study (Chen, Van
Dooren, Chen, & Verschaffel, 2005), we made the following predictions. First, as
in the previous studies on realistic mathematical modelling, we expected to find
that upper elementary school children in Flanders would show great difficulties
with solving such problems realistically. Second, these findings were expected to
show up also in students' problem-posing performance, although we had no clear
expectations about children's relative performance on the two tasks. On the one
hand, it could be argued that, because of the unfamiliarity of the problem-posing
task, performance on this task will be worse than on the parallel problem-solving
task. On the Other hand, because the problem-posing task confronted children
with the fact that the same division can result in different responses (while this
idea was left implicit and had to be discovered by the children themselves in the
problemsolving task), it could be argued that performance on problem solving
THE RELATIONSHIP BETWEEN POSING AND SOLVING DW'R PROBLEMS
would be worse. Third, as in the studies on the relationship between solving and
posing of standard word problems, we expected a significant positive relationship
between students' ability to pose and solve non-standard problems. Finally, we
wanted to assess the impact of a number Of subject variables, such as pupils' age,
mathematieal ability, and gender, as well as of some task variables, namely order
of test administration and interv•al between the two tests. With respect to the
latter two task variables, we anticipated that children's performance on the
problem-solving test would be better when they had to do this test after the
problem-posing test, and that their performance on the problem-posing test would
be better when they had to do this test after the problem-solving test, and that the
transfer effect would be greater when the interval between the two tests was
smaller.
METHOD
Participants
Participants were 245 children from two elementary schools in Flanders: 84
fourth graders (9-year-olds), 92 fifth graders (10-year-olds), and 69 sixth graders
(11year-olds). There was about an equal number of boys and girls in the sample
and also in the three different grade levels. All children were administered a
mathematics achievement test (Dudal, 2002). This test distinguishes five
achievement levels ranging from level A to level F. Pupils with an A or B scorc
were considered as mathematically strong, those with a C score as medium, and
those with D or E as weak
Materials
In the problem-posing task, students were instructed to generate three questions
or stories. In the problem-solving task, they were asked to solve three problems
that were closely related to the questions or stories to be generated. Besides
producing a numerical answer, they were also explicitly asked to write down the
solution proc or some explanation for each problem-solving task
Problem-posing (PP) task. Invent three story problems that belong to the
numerical problem 100 + ? such that the answer to the story problem is,
respectively (a) 13 (PPI)
(b) 12 (PP2)
(c) 12.5 (PP3).
Problem-solving (PS) task. Solve the following three problems and write down
how you arrived at the answer:
(a) 100 children are being transported by minibuses to a summer camp at the
seaside. Each minibus can hold a maximum of 8 children. How many
minibuses are needed? (answer: 13 buses) (PSI)
VERSCHAEEEL
(c) A tailor sold a large piece of cloth with a length of 50 meters. He wants to
cut it into 4 pieces of the same length. How long is each piece? (answer: 12.5
meters) (PS3).
TO minimise the possible role Of specific number effects, we worked with
two number clusters: 100+ 8 = ? (see above examples) and 70 + 4 = ? (answer,
respectively 18, 17 or 17.5) in the pp task. For the same reason, we worked with
two number clusters for each PS task: 100* 8 ? (answer: 13) and 70 + 4 ?
(answer.• 18) for PSI. 80+ 6 = ? (ansuer: 13) and 50+ 4 = (answer: 12) for PS2.
and 4 ? (answer: 12.5) and 1004 8 = ? (answer: 12.5) for PS3.
148
Data Coding
Pupils' answers to the three PS items were coded using a schema that was an
elaboration of the classification schema developed by Verschaffel et al. (1994).
The classification schema comprised 14 categories, which were, in a second
stage, reduced to three general categories: realistic reaction, no reaction, other
reaction.
THE RELATIONSHIP BETWEEN POSING AND SOLVING DW'R PROBLEMS
Realistic reactions (RR) comprise all cases wherein a pupil either gave the
(most) correct numerical response that also took into account the real-world
aspects of the problem context, as well as cases wherein there was a clear
indication that (s)he tried to take into account these real-world aspects, without,
however, giving the mathematically and situationally (most) accurate numerical
answer.
No reactions (NR) were all those cases wherein a pupil did not give a
numerical response and did not give any further written comment that indicated
that (s)he was aware Of the realistic modelhng difficulty that prevented him/her
from answering the problem.
Other reactions (OR) were all other cases without any indication that the pupil
was aware of the realistic modelling difficulty, for example mathematieally
correct but situationally inaccurate responses, computational errors, etc.
Here are examples of answers to PSI, illustrating the coding of answers as RR
and OR.
which were, just as the categories Of the PS coding schema, reduced to three
main categories: realistic reaction, no reaction, and other reaction.
Realistic reactions (RR) comprise all cases wherein a pupil generated a
mathematically and situationally accurate problem, but also those cases wherein
the child posed another problem, an incomplete problem or even no problem at
ET AL.
all, but expressed in whatever way clear awareness Of the realistic modelling
problem involved in the task.
No reactions (NR) were all those cases wherein a pupil did not generate any
problem and also did not give any further explication that suggested awareness
of the realistic modelling difficulty that prevented him/her from succeeding in
the problem-posing task.
Other reactions (OR) were all remaining cases, including mathematically
correct but situationally inaccurate problems, mathematically and situationally
inaccurate problems, incomplete problems without any indication of awareness
of the realistic modelling difficulty, etc.
Typical examples Of student-generated problems for PPI, together with their
code, are given below.
Realistic reactions
"100 students take boats One boat takes at most 8 students. How many boats
are needed?" (mathematically and situationally accurate problem)
100 children need transportation to go to the theatre. They go by car. A car
can hold 8 children. So, 13 cars are needed." (problem with an answer in stead
of a question at the end, but providing clear evidence of a situationally correct
interpretation)
— "Teacher Laura has 100 sweets and wants to divide them among her 8 pupils.
How many sweets does each child get? Afterwards they all additionally get half
Of a sweet." (even though the problem cannot be simply be modelled by 100 +
8 ?, it was scored as RR because Of the pupil's clear and meaningful struggling
with the realistic modelling complexity)Other reactions
•'1n the class of Teacher Tine there are 100 8 children. How many children are
there in the class of Teacher Tine?" (besides the fact that the problem
comprises an unacceptable computational command (namely 100 + 8), the
situation does not call tar rounding up to the next natural number (13)
"A man buys 100 flowers. He wants to distribute them equally among his 8
grandchildren. How many flowers does each grandchild get?" (problem with
an appropriate underlying mathematical model. namely 100 * 8 ?, but leading
to another response than the requested one, namely 12 instead of 13
"100 children are participating in a running competition. 8 children fall during
the competition. How many children reach the finish?" (problem leading to an
inappropriate mathematical model. namely I (H) — 8 = 'P)
"The teacher asks Jan to make the following sum in the mathematics class:
100 + 8 = ? Do you know the result?" (no real word problem, since the required
operation (100 * 8 2) is simply given as a symbolic problem, without a "real
context").
RESULTS
for the buses item, although there are also a few studies that report opposite
results (see Verschaffel et al., 2000). Moreover, we were somewhat surprised that
PS3 did not yield more RRs, taking into account that, for this item, the outcome
of the computation is at the same time the (correct) answer to the word problem
(and, can, thus, according to Our terminology, be considered as a "standard
problem").
A more detailed 100k at the responses revealed that almost all RRs belonged
to the subcategory of both mathematically and situationally accurate responses
(for PSI, PS2, and PS3, respectively, 97%, 98%, and 95% of all RRs, and 97%
overall). Of the "other responses" on PSI, PS2, and PS3, 57%, 410/0, and 52%
respectively, and 52% overall, were computationally correct but situationally
inaccurate. so, about half of the ORS were the direct and clear result of a
situationally incorrect interpretation of a correctly computed division with a
remainder, whereas for the Other half the role or the realistic modclling difficulty
in the actual response to the word problem remained unclear.
Table l. Percentages (and absolute numbers) ofrealistic reactions (RRs). no reactions
(NRs), and other reactions (ORS) for rhe three problem-solving (PS) items
71% 56%
(107) (173) (138) (418)
NR 2% 3% 4% 3%
(7) (21)
27%
(133) (65) (98) (296)
The results of Table 2, which show the percentage of pupils who succeeded in
solving three, two, and one problem, indicate that the Flemish upper elementary
school pupils performed rather weakly on these non-standard problems requiring
realistic Only a quarter of the students produced three situationally
accurate answers or reacted three times in a way that reveals attention to the
realistic modelling complexity involved in the problem.
VERSCHAEVEI.
Table 2. Percentages (and absolute numbers) ofpupils who produced Ihree. two. one.
and zero realistic reactions (RRs) in the problem-solving (PS) task
ORR Total
100%
(70) (69) (70) (36) (245)
Resultsfor the PP Task
Table 3 shows the percentages Of appropriate math stories for the three pp items.
Overall, there were 27% RRs, which is considerably lower than 57% RRs for the
corresponding PS task. As for the PS task. the number of children who gave no
reaction at all (NRs) was small, meaning that the majority of the reactions
belonged to the Category of ORS.
ET AL.
The PPI, pp2 and pp3 item elicited, respectively, 13%, 39%, and 31% realistic
reactions (RR so, as for the PP task, the number Of RRs for PP2 was
considerably higher than for PPI , and was even somewhat higher than for PP3.
A closer look revealed that the percentages of RRs to PPI pp2, and pp3 that
were both mathematically and situationally accurate was, respectively, 22%,
65%, and 84% (overall percentage: 66%). Of the Other subcategories Of
reactions that were coded as RRs, problems that were not (completely) accurate
from a mathematical point of view but that showed that the pupil had struggled
with the realistic modelling difficulty, were the most frequent category. For the
PPI item, such reaclions were responsible ror 77% Of all RRs.
Of the ORS, problems that were mathematically correct but situationally
inaccurate — such as generating the word problem "A man buys 100 flowers. He
wants to distribute them equally among his 8 grandchildren. How many flowers
does each grandchild get?" for PPI — was the most frequent subcategory (28%
overall; 33% for PPI, 21% for PP2, and 28% for PP3, respectively). Another
subcategory that occurred almost as frequently was the construction of word
problem that led to another mathematical outcome than the correct one or that
involved Other numbers as givens than the ones provided in the task, such as
'There are I (K) words written on the blackboard. The teachers wipes Off 8
words. How many words are there still on the blackboard?" for pp2 (26% overall;
23%, 33%, and 25% for PPI, pp2, and pp3, respectively).
The results Of Table 4, which show the percentage Of pupils who succeeded
in producing three, two, one, and none RRs to the problem-posing task, indicate
that the Flemish pupils have great difficulties in posing word problems that do
not simply require the generation of a problem situation that correctly matches a
given number sentence but also demands a situationally accurate interpretation
Of the outcome. Only 60/0 of the students produced three realistic answers,
which means that they either gave three times the most realistic response or
demonstrated at least awareness of the realistic modelling complexity.
Apparently, thinking of meaningful contexts that ask for rounding up, rounding
down, and keeping as a decimal the outcome of a division with remainder, and
distinguishing between these different contexts, is a very difficult task for
children.
Table 4. percentage (and absolute number) ofpupils who produced three, one, and zero
realistic reactions (RRs) in the problem-posing (PP) task
Total
6% 190/0 26% 1000/0
THE RELATIONSHIP BETWEEN POSING AND SOLVING DW'R PROBLEMS
Effect of age. Percentages (and absolute numbers) of RRs of the fourth, fifth and
sixth graders on the PS and pp tests are given in Table 5. The logistic regression
analysis showed that, as expected, the percentage of RRs for PS increased with
age, 22(2) 29.02, p < .01. The contrast analysis shows that sixth graders
performed better than fifth graders, = 19.88, p < .01, who in turn performed
better than the fourth graders, 1(1) = 51.34,p < .01.
On PP we also found a main effect of age, 1(2) 13.02, p _01_ Again, we
found significantly better performances for sixth graders than for fifth graders,
= 5.45, p = .02, and for fifth graders than for fourth graders, = 4.53, p = .03.
There also was an interaction effect between type of task and age, l) 12.39, p
< .01. The difference between PS and pp was significant in each grade, z2(l)
23.37, p < .01 for sixth grade, i(l) — 20.49, p < .01 for fifth grade, and —
4.64, p = .03 for fourth grade, but differences in percentage ofRRs between PS
and PP increased with grade (a difference in favour of PS of 12%, 35%, and 43%
for fourth, fifth, and sixth graders, respectively), mainly due to the fact that, with
age, PS performance grew considerably more than PP performance
49% 37%
(284) (72) (62)
36%
(151) (31) (20)
As far as PP is concerned, the logistic regression analysis revealed also a
significant effect of mathematical achievement, X2(2) 20.30, p .01, again with
mathematically strong pupils producing more RRs than mathematically medium
ones, 10.26, p < .01; the difference between this group and the weaker ones was
statistically not significant, i(l) = 3.39, p = _07_ Finally, there was no significant
interaction effect between type of task and mathematical achievement.
The finding that realistic problem-solving and problem-posing performance is
related to mathematical achievement in combination with the finding that there
was an overrepresentation of mathematically strong pupils in our sample, implies
that our results concerning Flemish upper elementary school pupils performance
154
THE RELATIONSHIP BETWEEN POSING AND SOLVING DW'R PROBLEMS
Effect of gender. Given that some previous studies have suggested that gender
might affect pupils' problem-solving and/or problem-posing competence (Leder,
1992) anWor have an impact on their disposition towards realistic mathematical
modelling (Verschaffel, 2002), we also analysed the effect of that factor (see
Table 7). Even though the performance Of the boys was somewhat better on both
tasks, we found no statistically significant effect of gender on PS, l) 0.02, p
— .89, or on PP, z2(l) — 0.61, p — .43. Also the interaction of gender and
mathematical achiev'ement level was not significant either for PS, 2.42, p .30,
or for PP, = 1.71, p = .43, indicating that gender effects were absent both for the
stronger, and for the medium and weaker pupils.
Table 7. Percentages (and absolute numbers) ofrealistic reactions (RRs) on the
problemsolving (PS) andproblem-posing ( PP) 'asks according 10 gender
Effect oforder. As explained in the Methods section, about half Of the pupils got
the pp test first and the PS test second (PP-PS group), whereas the other half got
these two tests in the reverse order (PS-pp group), Percentages (and absolute
numbers) ofRRs on PS and PP for both groups of pupils (PR-PS and PS-PP) are
given in Table 8.
Table S. Percentages (and absolute numbers) ofrealistic reactions (RRs) on the
problemsolving (PS) and the problem-posing (PP) tasks for the pupilsfrom the PS-pp
and the
(168) (250)
The expectation was that the pupils from the PP-PS group would perform
better on the PS task than those from the PS-PP group. Similarly, we expected
the pupils from the PP-PS group would perform weaker on the PP task than those
from the PS-pp group. As Table 8 shows, however, the differences were small
VERSCHAFFEL ET AL.
and they were not statistically significant, z2(l) = 0.00, p = .95 for PS
performance, and ) = 0.64, p = .42 for pp performance, so, unexpectedly, for
neither Of the tasks was there a (positive) transfer effect from one task to the
Other.
Finally, we also analysed the relationship between PS and PP at the item level.
This analysis revealed that the general significant correlation was also found for
each of the three items, although correlations were quite small (0.18, 0.15, and
0.23 for task l, 2, and 3 respectively).
GENERAL DISCUSSION
Recent research has documented a close relationship between problem posing and
problem solving in (elementary) arithmetic. However, most studies have
investigated the relationship between problem posing and problem solving only
by means of standard problem situations without realistic modelling complexities.
The present study aimed at filling that gap by investigating this relationship for
nonstandard problem situations, and, more particularly, division-with-remainder
problem (DWR) situations, which have elicited already quite a lot Of problem-
solving research in the past, but, until recently, were not subjected to a systematic
analysis from a problem-posing perspective.
156
THE RELATIONSHIP BETWEEN POSING AND SOLVING DW'R PROBLEMS
Groups of fourth, fifth, and sixth grade pupils from elementary schools in
Flanders were administered a set Of 3 DWR-items both in a problem-solving
(PS) and a problem-posing (PP) task setting. The analysis focused on (a) the
realistic nature Of pupils' reactions to both kinds of tasks, (b) the impact of
various subject and task variables on pupils' performance, and (c) the relationship
between performances on the two types of tasks.
First, pupils' rather weak overall performance on the PS task, and the finding
that this relatively weak performance was to a large extent due to the fact that
they provided a lot of mathematically correct but situationally inappropriate
responses, largely confirmed the conclusion from previous research with DWR-
problems (for an overview see Verschafiél et al., 2000), namely that students
have great difficulty with problem situations about division that require the
activation of realistic considerations and sense-making to give a proper
interpretation of a non-integer quotient. Complementary to that previous research,
our results indicate that these problems with realistic modellung not only apply to
DWR problems for which the situationally most reasonable answer is the whole
number that is next or previous to the computational outcome ofthe division, as
was found in previous studies, but also to DWR problems with the decimal
number that directly results from the calculation as their situationally most
approprlate response.
Second, our study yielded further evidence for upper elementary school
pupils' dimculties with realistic mathematical modelling Of school word
problems by showing that they perform weakly not only when having to solve
word problems that demand realistic interpretations, but also when they have to
pose such problems starting from given divisions requiring certain numerical
responses. Actually, pupils' performance on this problem-posing task was even
considerably worse than for the problem-solving task — not only in general
terms but also for each of the three items separately. The finding that children in
general also performed weakly and behaved unrealistically on the unfamiliar and
challenging problem-posing task suggests that pupils' tendency to neglect their
real-world knowledge when doing mathematical modelling is not simply due to
their lack Of alertness or mindfulness when doing (seemingly) routine word
problems, but is (also) caused by genuine cognitive difflculties with flexibly
switching between the mathematical operation Of division and real-world
situations that can be modelled by it. However, the results for the PP test also
revealed that pupils also made a lot Of Other errors that were not directly due to
realistic modelling difficulties but rather to their unfamiliarity with problem
posing as such.
Third, the analysis Of the effects Of the different subject and task variables
revealed that the ability to properly take into account real-world considerations
when solving or generating DWR-problems increased, as expected, with age and
mathematical achievement leeel. On the other hand, no gender effect was found.
Neither did we find a (positive) transfer effect ofthe solution of either task
towards the other task, not even among those pupils who got the second task
only a few days after the first one.
VERSCHAFFEL ET AL.
ACKNOWLEDGEMENT
158
THE RELATIONSHIP BETWEEN POSING AND SOLVING DW'R PROBLEMS
REFERENCES
Brown, S. & Walter, M, ( 1993). problem posing: Reflections and Hillsdale,
NJ Erlbaum
Cai, J.. & Hwang. S. (2002). Generalized and generative thinking in U.S. and Chinese students' mathe
solving and problem posing, Journal of Behavior, 401—421.
Cai. J.. & Silver. E. A. (1995) Solution processes and interpretations Of solutions in solving divisionwith-
remainder story problems: Do Chinese and US students have similar difficulties? Journalfor Research in
Carpenter. T. P.. Lindquist. M. W. Matthews. & Silver. E. A_ (1983)_ Results of the third NAEP
mathernatics assessment: Secondary school. Mathematics reacher, 76,
Chen. Van W,. Chen. L. relationship and solving division With remainder
problems among Chinese elementary school children. Mediterranean Journalfor Research in 'futhemarics
Education.
De Cone. E. & Verschaffel. L. (1996). An empirical test of the impact ofoperations intuitive models of
operations on solving word problems With a multiplicative structure. Learning and Instruction, 6, 219-
242,
Dudal. P. (2002). Leerlingvolgsysteem Wiskunde: Toelsen 4 Basisboek [Mathematics Achievement Tests:
Tests for Grade 4 - Basic Book.l. Antwerpen, Belgium'Ape1doorn, Netherlands: VCI.B & Garant,
Ellenon, N F, (1986). Children's made-up mathematics problems: A new Frspective on talented
mathematicians. Educational Studies tn Mathematics, 17, 261—271
English, L D. ( lie development of fifth-grade cluldren•s problem-posing abilities. Educational Studies
34.
English, L. D. (1997b). promoting a problem Btsing classroom. Teaching Children Mathematics, 4, 172-
178.
English, L, D. ( 's problem Within and Re— searo% in Mathemalics EUucanon, 29, 83
106.
Problem in presented the annual meeting Of the American Research
Association. San Diego, CA
Greer, B. ( 1993). The modelling perspective on wor(lk] problems. Journal ofMathema1icaI Behavior,
Hilton. P. (1984). Current trends in mathematics and future trends in mathematics education_ For the
Ia•arning ofMathematics, I J, 2—8.
Kilpatrick. J (1987) problem formulating: Where do problems come from? In A. H. Schoenfeld
(FAI), Cogni'ive science and mathemalics educurion (pp. 1 23—147/ Hillsdale, NY Erlbaurn
(i (1992b Mathematics and gender: Changing perspectives, In D, X Grouws Handbook of research on
mathematics reaching and learning (pp. 597-622). New York: Macmillan.
Ministry of Fulucation of Peoples' Republic of China (NCSM) (2001b Chinese Varional Curriculum
Slanåards on Mathematics, Beijing, China; Beijing Normal University Publishing House.
National Council of Teachers of Mathematics. principles and standards for school mathemat
Reston, VA: NCTM
Onnvjkkelingsdoelen en eind,'ermen, Informariemap de onderw\jspraktijk. Gewoon basisonderw\js
IStandards. Documentation for practitioners, Elementary educationl, (1998) Brussel: Ministerie van de
Vlaamse Gcmecnschap, Departcment Ondcrwijs. Afdcling Informatic cn Documentatie
Rudnitsky, A., Etheredge, S, Freeman, S, & Gilbert, T. 1995). Learmng to solve addition and subtraction
word problems through a structure-plus-writing approach. Journalfor Research in Mathematics
Education, 26, 467486.
Silver, E. A. ( 1994b On mathematical problem posing. Fur the Learning ofMathematics, 141), 19—28,
Silver. E. A.. & Cai, J. (1996). An analysis Of arithmetic problem posing by middle school
students. Journalfor Research Mathematics Educalion, 2 7, 521—539.
Silver, E. A., Shapiro, L. J., & DeuLsch, A. ( 1993) Sense making and the solution ofdivision problems
involving remainders: An examination Of middle school students' solution processes and their inter•
pretations of solutions. Journalfor Research in Murhemutics 24, 1 17—135.
Streefland. L. ( 1988b Rec•onstructivc learning In A. Borbas (Ed.). Proceedings of rhe twelfth interna'ional
conference for the Psycho/ogv of Mathematics Education Vol. I, pp. 7S—91 Veszprem, Hungary: OOK
Printing House.
159
Van Dooren. W.. Verschaffel. L, Greer. B.. & De Bock, D. (2CÅ)6), Mcxlelling for life: Developing
adaptive expertise in mathematical modelling from an early age. In L. Vcrschaffel. F_ Dochy. M.
Bexkaerts & S. Vcysniadou ( Eds», Instructionalpsychologv: Pas'. presenl andfuture trends (pp. 91— 1
12). oxford, U.K.' Elsevver.
Versehaffel, L_ (2002). Taking the modeling seriously at the elementary school level: Promises and pitfalls
(Plenary leeture[ In X Cix•kbum & F. Nardi Proceedings 26th cmnual conference ofthe International
Grow'for the psychology of Marhemarics Education (Vol.
School Of Education and Professional Development. University Of East Anglia, UK.
Verschaffel. & fk• Conc. E. (1997)_ Teaching realistic mathematical modeling in the elementary
school, A teaching experiment With fifth graders. Journalfor Research in Mathematics
Education, 28. 577-601.
Verschaffel, l. , De Corte, E.. & l_asure, S. ( 1994) Realistic consideranons in mathematical nuxkling of
sch•K'I arithmetic word problems. Learning and 4, 273—294
Vcrschaffel. I. , De Corte. E. 161sure. S. Van Vaerenbergh. G., Bogaerts. H . Ratinckx. (1999) Sign
evaluation Of a learning environment for and problem solving children; /.
Versehaffcl. L, Dc Corte, E, I,mvyck. Y, I)herl. S., & Vandeput. L_ (2000), Supporting mathematical
problem solving and in upper elementary by Knowledge Deliverable of project NW 201 7 Cl.-Net:
Supported Collaborative Learning Networks in Prunary and Secondary Education.
I.. Greer. .
Nether— lands: Swcts & Zettlmger.
Lieven Verschaffel
Centrefor Instructional Psychologv and Technology
Kalholieke Universitei/ Leuven
Belgium
Wim Van Dooren
Centrefor Instructional Psychologv and Technologv
Kalholieke Universileit Leuven
Belgium
Limin Chen
College of Teachers • Professional Development
Shenyang Normal Universiry
China
Katrien Stessen.s
Centrefor Instructional Psychologv and Technology
Katholieke Universiteir Leuven
Belgium