Learning and Individual Differences 49 (2016) 297–304

Contents lists available at ScienceDirect

Learning and Individual Differences

journal homepage: www.elsevier.com/locate/lindif

Children's multiplication and division shortcuts: Increasing shortcut use

depends on how the shortcuts are evaluated☆
Katherine M. Robinson a,⁎, Adam K. Dubé b, Jacqueline-Ann Beatch c
a
Department of Psychology, University of Regina, Canada
b
Department of Educational and Counselling Psychology, McGill University, Canada
c
Department of Psychology, University of Calgary, Canada

a r t i c l e i n f o a b s t r a c t

Article history: The goals of this study were to investigate whether children's use of two arithmetic shortcuts based on the un-
Received 26 June 2015 derstanding of the operations of multiplication and division could be increased and to examine how children's
Received in revised form 4 March 2016 evaluations of the shortcuts interacted with the brief task aimed at promoting shortcut use. Grade 6, 7, and 8 stu-
Accepted 24 June 2016
dents solved two sets of inversion (e.g., d × e ÷ e) and associativity (e.g., d × e ÷ f) problems. Children with a good
understanding of the relation between multiplication and division can use conceptually-based shortcuts to solve
Keywords:
Arithmetic
both types of problems. Students were also given a brief task demonstrating the inversion shortcut (the answer is
Inversion the first number and no calculations are required) and the associativity shortcut (dividing first and then multi-
Associativity plying) and were asked to compare each shortcut to a left-to-right procedure (i.e., multiplying and then divid-
Conceptual knowledge ing). Half the participants were given the demonstration task between the problem sets and the other half
Multiplication after the sets. Inversion and associativity shortcut use increased by Grade 8 and improved across problem sets
Division for both the Demonstration Middle and Demonstration Last groups. The demonstration task successfully promot-
ed subsequent shortcut use but the participants who positively evaluated the shortcuts compared to the left-to-
right procedure in the demonstration task had greater subsequent shortcut use. Conceptually-based shortcut use
on multiplication and division problems was increased using a brief demonstration task but children's subse-
quent use of shortcuts depended on their evaluations of the shortcuts.
Crown Copyright © 2016 Published by Elsevier Inc. All rights reserved.

1. Introduction the need for research to improve conceptual understanding which is

critical for the development of mathematical knowledge (Kilpatrick,
Numerous studies from mathematics education and cognitive devel- Swafford, and Findell, 2001), and for later algebraic reasoning (Nunes
opment have demonstrated the difficulties that children have with con- et al., 2008).
ceptual knowledge of arithmetic (Prather and Alibali, 2009). We define The challenge of conducting research on children's conceptual
conceptual knowledge as the understanding of the underlying structure knowledge is in assessing that knowledge (Bisanz and LeFevre, 1990)
of arithmetic and that the understanding of this structure can be used as but two concepts provide important information about children's
the basis for problem solving procedures (also see Crooks and Alibali, knowledge. The inversion concept, extensively researched on addition
2014). The focus of the current study is on children's understanding of and subtraction problems, is the understanding that addition and sub-
the relationships between arithmetic operations which are particularly traction and multiplication and division are pairs of operations that
challenging for children (Gilmore and Papadatou-Pastou, 2009). These are inversely related to one another (Starkey and Gelman, 1982).
difficulties, especially with concepts relating to the operations of multi- When children understand that inverse relationship, they can apply
plication and division (Robinson, Ninowski, and Gray, 2006b), have their knowledge by using a procedure called the inversion shortcut on
rarely been investigated (Robinson and LeFevre, 2012) but highlight problems such as 3 + 24 − 24 or 3 × 24 ÷ 24 (Bisanz and LeFevre,
1990). In these problems, adding and subtracting or multiplying and di-
viding by the same number results in the first number remaining un-
☆ This research was supported, in part, by an NSERC discovery grant. Thank you to Jen changed. The inversion shortcut requires no calculations and is fast
Gibson, John Brand, and Anna Maslany for assistance with the collection and coding of and accurate (Bisanz and LeFevre, 1990). Shortcut use is commonly
the data. Many thanks to the children, teachers, and principals from the participating used as an index of conceptual understanding (Crooks and Alibali,
schools.
⁎ Corresponding author at: Campion College at the University of Regina, 3737 Wascana
2014; Gilmore and Papadatou-Pastou, 2009; Prather and Alibali, 2009)
Parkway, Regina, Saskatchewan S4S 0A2, Canada. but may underestimate conceptual knowledge if children do not apply
E-mail address: [email protected] (K.M. Robinson). their knowledge during problem solving (Canobi, 2009).

http://dx.doi.org/10.1016/j.lindif.2016.06.014
1041-6080/Crown Copyright © 2016 Published by Elsevier Inc. All rights reserved.
298 K.M. Robinson et al. / Learning and Individual Differences 49 (2016) 297–304

The associativity concept is the understanding that addition and demonstration task, recognized the superiority of a shortcut versus a
subtraction and multiplication and division are pairs of operations left-to-right procedure had much greater subsequent shortcut use
which are associatively related to one another, and therefore each oper- than children who did not. The demonstration task shows promise for
ation within a pair can be solved in any order (Klein and Bisanz, 2000). being an effective tool for quickly promoting shortcut use and under-
When children understand the associative relationship, they can apply standing but its effectiveness depends on children's evaluations of the
their knowledge by using a procedure called the associativity shortcut shortcuts.
to simplify the problem solving process on problems such as No study has explicitly attempted to promote the inversion or asso-
3 + 24 − 20 or 3 × 24 ÷ 6 by solving 24 − 20 or 24 ÷ 6 first ciativity concepts for multiplication and division three-term problems.
(Robinson and Dubé, 2009a). Solving the subtraction or division compo- Robinson and Dubé (2009b) investigated repeated exposure to multi-
nent first simplifies problem solving as children are now dealing with plication and division inversion problems using the same design as
smaller numbers (e.g., 3 × 24 ÷ 6 becomes 3 × 4) and results in reduc- Siegler and Stern (1998). They found that over a third of Grade 6 stu-
tions in solution times and fewer retrieval or calculation errors dents who were repeatedly exposed to multiplication and division in-
(Robinson and LeFevre, 2012). version problems failed to discover the multiplication and division
Students' weak understanding of the relationships between mul- inversion shortcut. This suggests that the multiplication and division
tiplication and division is “unacceptable and indicates a substantive version of the inversion concept is more complex and that repeated
gap in the mathematics curricula that must be addressed (National exposure alone is insufficient to promote children's conceptual
Mathematics Advisory Panel, 2008, pp. 4–39).” Little research has understanding.
been conducted on the multiplication and division inversion and as- The demonstration task used by Robinson and Dubé (2012, 2013) on
sociativity concepts (Robinson and LeFevre, 2012) let alone on how addition and subtraction problems is promising as it can be easily
to strengthen children's understanding of these concepts. What is adapted to multiplication and division problems. Robinson and Dubé
known is that both children and adults consistently use fewer (2009c) used the demonstration task to assess conceptual understand-
conceptually-based shortcuts on multiplicative than additive ing of inversion for multiplication and division but only after problem
versions of inversion and associative problems (Robinson and solving. Unlike addition and subtraction, participants in Grades 6
Ninowski, 2003; Robinson et al., 2006b), use the associativity short- through 8 who compared the inversion shortcut to a left-to-right proce-
cut less than the inversion shortcut (Robinson et al., 2006b), that dure only marginally preferred the inversion shortcut and inversion
large individual differences exist in both additive and multiplicative shortcut use was lower than found on addition and subtraction prob-
shortcut use, and that additive shortcut use reaches adult levels in lems (see also Robinson et al., 2006b). Taken together, the findings
middle childhood but multiplicative shortcut use does not reach from these studies are disconcerting because children near the end of
adult levels until adolescence (Dubé, 2014). Given these difficulties, middle school need to be prepared to learn algebra and they should
it is particularly important to increase understanding of multiplicative be fluent with multiplication and division by Grade 5 (National
concepts. Mathematics Advisory Panel, 2008). To be successful in algebra, one im-
No research has been conducted on how to increase children's un- portant factor is the understanding of the relationships amongst opera-
derstanding of the multiplication and division concepts of inversion tions (Kilpatrick et al., 2001). The results of work on the multiplication
and associativity unlike the addition and subtraction inversion and as- and division concepts suggest that children's understanding of these re-
sociativity concepts. For addition and subtraction inversion problems, lationships is weak and therefore it is critical to find a way to increase
repeated exposure (Siegler and Stern, 1998), training with concrete ob- this understanding.
jects (Lai, Baroody, and Johnson, 2008), and visual and oral shortcut In previous research, children's attitudes towards the inversion and
demonstrations (Nunes, Bryant, Hallett, Bell, and Evans, 2009) have all associativity shortcuts have differed across individuals (Robinson and
increased shortcut use. Dubé, 2009c, 2012, 2013). While some children considered the short-
An alternative, and much briefer, approach was used by Robinson cuts to be clever approaches to problem solving, others considered the
and Dubé (2012, 2013) and was based on the premise that the ability shortcuts to be a form of cheating (i.e., skipping procedural steps).
to recognize the validity of a conceptually-based procedure or shortcut That children have strong feelings about how mathematics should be
implies knowledge of the concept and that the ability to also recognize performed is not surprising even though it has rarely been investigated
its superiority to a familiar and well-learned left-to-right procedure in the domains of either procedural or conceptual knowledge of arith-
indicates even stronger understanding (Bisanz, Watchorn, Piatt, metic. Other research has shown that, as schooling proceeds, children
and Sherman, 2009; Canobi, Reeve, and Pattison, 1998; Prather and develop diverging beliefs, attitudes, motivations, and anxieties towards
Alibali, 2009). Robinson and Dubé (2009a) assessed children's un- mathematics (Beilock, 2008; Lepper, Corpus, and Iyengan, 2005; Martin,
derstanding of addition and subtraction inversion and associativity Anderson, Bobis, Vellar, and Way, 2012; McLeod, 1993; Wolters, 2004).
concepts using two tasks. First, Grade 2 through 4 children's concep- Thus, we expected that the demonstration task would help increase
tual understanding of inversion and associativity was assessed via conceptually-based shortcut use on multiplication and division prob-
shortcut use during problem solving. Second, children were then lems as it did on addition and subtraction problems but that it would
shown inversion and associativity problems and given a verbal dem- be most effective for students who positively evaluated the shortcuts
onstration for each problem type of how one fictitious child used a (Robinson and Dubé, 2012, 2013). However, given that previous work
shortcut, and a verbal demonstration of how another fictitious child has shown that children use the inversion and associativity shortcuts
used a left-to-right procedure. Participants were asked to compare the much less frequently on multiplication and division problems than on
pair of procedures on each problem type and decide which approach addition and subtraction problems and are also more skeptical about
(the shortcut or the left-to-right procedure) was better. Most children the shortcuts themselves (Robinson and Dubé, 2009c; Robinson et al.,
preferred the inversion shortcut but only marginally preferred the asso- 2006b), it is possible that the demonstration task will not be sufficient
ciativity shortcut. to promote shortcut use.
Robinson and Dubé (2012, 2013), also with addition and subtrac- The goals of the current study were to establish for the first time,
tion, used the same problem solving and demonstration tasks but had (a) whether understanding of the inverse and associative relationships
half the Grade 2 through 5 participants complete the demonstration between multiplication and division can be successfully promoted using
task in the middle of the problem solving task. The demonstration a task previously successful for promoting the same relationships be-
task increased later shortcut use, particularly for the associativity tween addition and subtraction and (b) whether the success of this
shortcut. Importantly, however, children's evaluation of the demon- task would be mitigated by children's evaluations of the conceptually-
strated procedures mitigated shortcut use. Children who, during the based shortcuts.
 K.M. Robinson et al. / Learning and Individual Differences 49 (2016) 297–304 299

2. Method test was used for post-hoc effects. Accuracy, median solution latencies
for correct responses, and cut-offs (when participants were not able to
2.1. Participants solve the problem within the 5-second limit) for each procedure used
on the inversion and associativity problems are reported in Appendix
Participants included twenty-six Grade 6 (11 boys, 15 girls) (mean B but not discussed further as they matched the verbal reports (see
age = 11 years, 4 months), thirty-one Grade 7 (14 boys, 17 girls) also Dubé, 2014; Dubé and Robinson, 2010a, 2010b; Robinson and
(mean age = 12 years, 4 months), and thirty-two Grade 8 students Dubé, 2009a, 2009b, 2009c, 2012, 2013; Robinson and Ninowski,
(15 boys, 17 girls) (mean age = 13 years, 5 months). Participants 2003; Robinson et al., 2006b). Accuracy rates were higher and solution
were from a large Canadian city, were predominantly Caucasian, and latencies faster for shortcuts than for the left-to-right procedure.
from middle SES families. The study took place in the first half of the The subsequent analyses inform our hypotheses as follows. First,
school year. analyses of the problem-solving task detailed the frequency of inversion
and associativity shortcut use to determine if the demonstration task it-
2.2. Materials and procedure self promoted subsequent shortcut use. Second, analyses of the demon-
stration task detailed whether children's evaluations of the shortcuts
In one individually-administered session, all participants completed affected subsequent shortcut use.
two tasks: a problem solving task comprised of two sets of problems
and the demonstration task. Participants were randomly assigned to 3.1. Problem-solving task
one of two groups. The Demonstration Middle group solved the first
set of the problem solving task, completed the demonstration task, Two coders independently coded 10% of the participants' proce-
and then ended with the second set of problems. The Demonstration dures, Cohen's Κ for both inversion and associativity problems was
Last group solved the first and second sets of the problem solving task 0.95. The remaining verbal reports were coded by one of the two coders.
with a pause between the sets and then ended with the demonstration Disagreements and ambiguous responses were resolved by discussion.
task.
The two sets of the problem solving task each comprised of 8 inver-
3.1.1. Inversion
sion problems of the form d × e ÷ e and 8 associativity problems of the
Verbal reports of solution procedures on the inversion problems
form d × e ÷ f (see Appendix A). The associativity problems were de-
were classified into the three types (see Table 1) used in previous stud-
signed so dividing the second number by the third number would result
ies (e.g., Dubé, 2014; Robinson and Ninowski, 2003; Robinson et al.,
in a whole number. Half the problems of each type were small
2006b; Robinson and Dubé, 2009c). In the left-to-right procedure, par-
(d × e b 25) (e.g., 3 × 2 ÷ 2 and 3 × 4 ÷ 2) and half were large
ticipants solved the problem by first finding the solution to the first
(d × e N 25) (e.g., 6 × 8 ÷ 8 and 6 × 8 ÷ 4). No more than two problems
part of the problem and then the second part (e.g., on 9 × 6 ÷ 6, “I mul-
of each type or size were presented consecutively. Problems were pre-
tiplied 9 × 6 which is 54 and then 54 divided by 6 is 9”). The inversion
sented on a laptop using e-prime software. Participants had a maximum
shortcut involved stating the first number of the problem without any
of five seconds to solve each problem based on research showing that
calculations (e.g., “The answer is 9 because the 6s cancel each other
solution latencies for shortcut use are usually b 5 s (Robinson and
out”). The negation procedure was a combination of the left-to-right
Dubé, 2009c; Robinson et al., 2006b). For each problem, solution laten-
procedure and the inversion shortcut as participants would solve the
cies, accuracy, and immediately retrospective reports of problem solv-
first part of the problem and then realize that as the third number was
ing procedure (i.e., “How did you get that answer?”) were collected. If
the same as the second, dividing by the third number negated the mul-
participants were unable to solve the problem in less than five seconds,
tiplication just performed (Bisanz and LeFevre, 1990) (e.g., “I did the
they were asked what procedure they were trying to use to solve the
9 × 6 which equals 54 but then I realized that it was divided by the
problem (i.e., “How were you trying to get the answer?”). No feedback
was provided.
In the demonstration task, participants were shown two inversion Table 1
problems and two associativity problems. For inversion, on one problem Percentage use of inversion, negation, and left-to-right procedures on inversion problems
participants were told how a fictitious child had solved the problem by across problem sets for the Demonstration Middle and Demonstration Last groups in
using the inversion shortcut (“When X solved this problem, s/he said Grades 6, 7, and 8.

that the answer would be the first number because when you multiply Procedure
and divide by the same number, the answer is always the first num-
Inversion Negation Left-to-right
ber”). On the other problem, participants were told how a fictitious
Set 1
child had solved the problem using a left-to-right procedure (“When
Grade 6
X solved this problem, s/he multiplied the first two numbers together Demonstration Middle 2.7 (3.4) 19.6 (3.8) 76.8 (4.5)
and then divided that number by the third number”). For associativity, Demonstration Last 8.3 (3.6) 16.7 (4.1) 72.9 (4.4)
on one problem, participants were told how a fictitious child had solved Grade 7
the problem by using the associativity shortcut (“When X solved this Demonstration Middle 13.3 (3.2) 26.6 (3.6) 60.2 (4.2)
Demonstration Last 10.8 (3.3) 32.5 (3.7) 55.0 (4.4)
problem, s/he divided the second number by the third number and Grade 8
then took the answer and multiplied it by the first number”). On the Demonstration Middle 23.3 (3.3) 9.2 (3.7) 60.0 (4.4)
other problem, the left-to-right procedure was used again. The order Demonstration Last 37.5 (3.1) 22.8 (3.5) 39.0 (4.1)
of the problem types (inversion or associativity) and demonstrated pro-
Set 2
cedures (shortcut or left-to-right) was counterbalanced as closely as Grade 6
possible within each group. Participants were asked which procedure Demonstration Middle 43.8 (4.4) 13.4 (3.5) 40.2 (4.5)
they preferred on the inversion problems and on the associativity prob- Demonstration Last 12.5 (4.8) 21.9 (3.8) 65.6 (4.8)
Grade 7
lems. No feedback was provided.
Demonstration Middle 53.1 (4.1) 15.6 (3.3) 31.3 (4.2)
Demonstration Last 26.7 (4.3) 29.2 (3.4) 43.3 (4.3)
3. Results Grade 8
Demonstration Middle 63.9 (4.3) 5.9 (3.4) 29.4 (4.3)
Boys and girls were collapsed together as no significant results in- Demonstration Last 53.7 (4.0) 19.9 (3.2) 26.5 (4.0)

volving gender were found. Tukey's Honestly Significant Difference Note. Standard errors are in parentheses.
300 K.M. Robinson et al. / Learning and Individual Differences 49 (2016) 297–304

same number again so it goes back to 9 again”). No differences involving

problem set, group, or grade were found for negation so this combina-
tion procedure is not discussed further. All remaining procedures
(e.g., guessing or only multiplying the first two numbers and not divid-
ing by the third number) were infrequent (b1%).
A 3 (Grade: 6, 7, and 8) × 2 (Problem Set: 1 and 2) × 2 (Condition:
Demonstration Middle and Demonstration Last) analysis of variance
was performed on inversion shortcut use. A main effect of grade was
found, F(2, 83) = 6.06, MSE = 1874.05, p = 0.023, η2p = 0.13 (17.0%,
26.6%, and 44.5% for Grades 6, 7, and 8, respectively), with Grade 8 hav-
ing the highest shortcut use followed by Grade 6 and 7 which did not
differ from each other, HSD = 16.0. Inversion use increased across set
(17.1% vs. 43.8%), F(1, 83) = 60.84, MSE = 495.72, p b 0.001, η2p =
0.42, but was qualified by an interaction with condition, F(1, 83) =
18.42, MSE = 495.72, p b 0.001, η2p = 0.18 (see top panel of Fig. 1).
Both conditions increased shortcut use. In Set 1 there was no difference
between conditions but in Set 2, the Demonstration Middle group had
higher use than the Demonstration Last group, HSD = 9.1. To get a bet-
ter sense of how the Demonstration Middle benefited from completing
the demonstration task between the problem sets, individual differ-
ences in shortcut use between the two problem sets were examined.
The top panel of Fig. 2 shows the increase in shortcut use from the
first to the second problem sets. More than 40% of the Demonstration
Middle group doubled their shortcut use compared to b 10% of the Dem-
onstration Last group.

3.1.2. Associativity
Verbal reports of solution procedures on the associativity problems
were classified into the two types (see Table 2) used in previous studies
(e.g., Robinson and Dubé, 2009c). The left-to-right procedure was the
same as for the inversion problems. For the associativity shortcut,

Fig. 2. Percentage increase in inversion shortcut use (top panel) and associativity shortcut
use (bottom panel) between problem sets for the Demonstration Middle and
Demonstration Last groups.

participants divided the second number by the third number and then
multiplied the answer by the first number (e.g., on 9 × 6 ÷ 3, “I took
the 6 and divided it by the 3 and got 2 and then I multiplied that number

Table 2
Percentage use of associativity and left-to-right procedures on associativity problems
across problem sets for the Demonstration Middle and Demonstration Last groups in
Grades 6, 7, and 8.

Procedure

Associativity Left-to-right

Set 1
Grade 6
Demonstration Middle 0 (0) 100.0 (2.6)
Demonstration Last 0 (0) 100.0 (2.8)
Grade 7
Demonstration Middle 10.2 (2.3) 88.3 (2.4)
Demonstration Last 5.0 (2.4) 95.0 (2.5)
Grade 8
Demonstration Middle 10.0 (2.4) 88.3 (2.5)
Demonstration Last 19.1 (2.3) 80.1 (2.3)

Set 2
Grade 6
Demonstration Middle 15.2 (3.6) 83.9 (3.7)
Demonstration Last 4.2 (3.9) 95.8 (4.0)
Grade 7
Demonstration Middle 32.0 (3.4) 66.4 (3.4)
Demonstration Last 7.5 (3.5) 91.7 (3.6)
Grade 8
Demonstration Middle 40.8 (3.5) 56.7 (3.6)
Fig. 1. Inversion shortcut use (top panel) and associativity shortcut use (bottom panel) on Demonstration Last 19.1 (3.3) 80.9 (3.3)
each problem set for the Demonstration Middle and Demonstration Last groups. Note. Standard errors are in parentheses.
 K.M. Robinson et al. / Learning and Individual Differences 49 (2016) 297–304 301

by the first number and got 18”). All remaining procedures

(e.g., guessing or only multiplying) were infrequent (b 1%).
A 3 (Grade) × 2 (Problem Set) × 2 (Condition) ANOVA was per-
formed on associativity shortcut use. A main effect of grade was
found, F(2, 83) = 3.32, MSE = 1343.61, p = 0.041, η2p = 0.07. Grade 8
students used the shortcut more than Grade 6 and 7 students who did
not differ (4.8%, 13.7%, and 28.4%, for Grades 6, 7, and 8, respectively),
HSD = 13.6. Associativity use increased across set, F(1, 83) = 17.99,
MSE = 369.00, p b 0.001, η2p = 0.18, but was qualified by an interaction
with Condition, F(1, 83) = 12.07, MSE = 369.00, p = 0.001, η2p = 0.13
(see bottom panel of Fig. 1). Unlike the inversion shortcut, associativity
shortcut use only increased across set for the Demonstration Middle
group. In Set 1 there was no difference between conditions but in Set
2, the Demonstration Last group had higher use than the no demonstra-
tion group in which shortcut use remained the same, HSD = 7.1. To get
a better sense of how the Demonstration Middle group benefited from
completing the demonstration task between the problem sets, individ-
ual differences in shortcut use between the two problem sets were ex-
amined. The bottom panel of Fig. 2 shows that almost 25% of the
Demonstration Middle group doubled their shortcut use compared to
hardly any participants in the Demonstration Last group.

3.2. Demonstration task

In the demonstration task, all participants were asked to compare

the demonstrated procedures and state which procedure they
preferred.
On the inversion problems, two participants in Grade 6, one in Grade
7, and one in Grade 8 could not decide which procedure they preferred
and one in Grade 8 did not provide a response. Exactly 75.0% preferred
the inversion shortcut. A chi-square analysis indicated that there was no
grade effect, χ2(2, N = 84) = 2.04, p = 0.360 (66.7%, 73.3%, and 83.3%
for Grades 6, 7, and 8, respectively).
On the associativity problems, one participant in Grade 6, one in Fig. 3. Inversion shortcut use (top panel) and associativity shortcut use (bottom panel) on
Grade 7, and two participants in Grade 8 could not decide which proce- each problem set for participants in the Demonstration Middle group who preferred the
dure they preferred. Overall, 52.4% of participants preferred the associa- shortcut or the left-to-right procedure in the demonstration task.

tivity shortcut. A chi-square analysis indicated that there was a grade

effect, χ2(2, N = 82) = 6.33, p = 0.042, with shortcut preference
3.2.2. Associativity
highest for Grade 8 students (45.8%, 40.0%, and 71.4% for Grades 6, 7,
A 2 (Preference: Associativity and Left-to-Right) × 2 (Problem Set)
and 8, respectively).
ANOVA was performed on associativity shortcut use during the problem
For a more detailed examination of the impact of the demonstration
solving task. There were main effects of preference, F(1, 39) = 5.18,
task on later problem solving, the Demonstration Middle group's perfor-
MSE = 1278.12, p = 0.028, η2p = 0.11, and set, F(1, 39) = 27.41,
mance in the second problem set was examined with participants from
MSE = 336.56, p b 0.001, η2p = 0.41 (see bottom panel of Fig. 3). Overall,
all grades collapsed together. No differences were found in shortcut
there was more shortcut use by participants who preferred the shortcut
preference between the Demonstration Middle and Demonstration
in the demonstration task and shortcut use increased from the first to
Last groups suggesting there was no benefit of more problem solving
the second set. Preference and problem set interacted, F(1, 39) =
on performance on the demonstration task.
24.42, MSE = 366.56, p b 0.001, η2p = 0.39, with the same pattern as in-
version. In the first set, there was no difference in associativity shortcut
3.2.1. Inversion use between the preference groups but in the second set there was
A 2 (Preference: Inversion and Left-to-Right) × 2 (Problem Set) more shortcut use by the associativity preference group as shortcut
ANOVA was performed on inversion shortcut use during the problem use increased across problem sets, HSD = 21.8.
solving task. There were main effects of preference, F(1, 38) = 6.26,
MSE = 1591.78, p = 0.017, η2p = 0.14, and set, F(1, 38) = 21.50,
MSE = 585.61, p b 0.001, η2p = 0.36 (see top panel of Fig. 3). There 4. Discussion
was more inversion shortcut use by participants who preferred the
shortcut in the demonstration task and inversion shortcut use increased Across both problem sets, both conditions, and all grades, shortcut
from the first to the second set of problems. Preference and problem set use was higher for inversion than and in the demonstration task, partic-
interacted, F(1, 38) = 7.33, MSE = 585.61, p = 0.010, η2p = 0.16. In the ipants were more likely to prefer the inversion than the associativity
first set, there was no difference in shortcut use between the preference shortcut. Although older children in Grade 8 were more likely to use
groups but in the second set there was more shortcut use by the inver- conceptually-based shortcuts than their younger counterparts, their
sion preference group as shortcut use increased across problem sets for overall shortcut use remained well under 65%. These results are consis-
the inversion preference group but not for the left-to-right preference tent with previous findings that multiplicative conceptually-based
group, HSD = 24.6. These results must be interpreted with some cau- shortcut use, even by the late middle school years, remains weak.
tion as, even collapsing across grade (n = 40), there were only nine par- Despite this, the demonstration task was successful in increasing subse-
ticipants who did not prefer the inversion shortcut. quent shortcut use during problem solving, supporting our first
302 K.M. Robinson et al. / Learning and Individual Differences 49 (2016) 297–304

hypothesis. Interestingly, and in line with previous findings about the procedures are available or problem characteristics change (Lemaire
greater difficulty that both children and adults have with the associativ- and Lecacheur, 2011; McNeil, Rittle-Johnson, Hattikudur, and Petersen,
ity than the inversion shortcut (e.g., Robinson and Ninowski, 2003; 2010). High levels of practice with a procedure may inhibit flexibility
Robinson et al., 2006b), inversion shortcut use increased across both and Newton, Star, and Lynch (2010) proposed that flexible problem
conditions from the first to the second problem set but associativity solvers use efficient procedures that take into account problem solving
shortcut use only increased if the participants were given the demon- structure.
stration task between the problem sets. This suggests that for the inver- Recently, McNeil (2014) proposed the change-resistance account of
sion shortcut, simply more exposure to inversion problems promotes children's mathematical difficulties stating that not only can children be
shortcut use but that for associativity, participants need to have their at- entrenched in their old problem solving methods but that this practice
tention drawn to the associativity shortcut. Finally, consistent with also narrows their interpretation of new problems, making practice
Robinson and Dubé's (2012, 2013) findings with addition and subtrac- alone insufficient for learning new ways to solve a problem. Change-
tion, children's evaluations of the multiplication and division resistance accounts for why simple practice with inversion problems
conceptually-based shortcuts also mattered, particularly for the associa- does not increase shortcut use and the notion of mindlessness or the
tivity shortcut which was more likely to be used subsequently if a par- “Einstellung effect” may explain why some participants in the Demon-
ticipant evaluated the shortcut positively. stration Middle group did not change their problem solving procedures
The current study is the first to investigate whether a brief demon- in light of new information.
stration can promote the use of inversion and associativity shortcuts Conversely, children may not be mindless in their problem solving,
on multiplication and division three-term problems. The demonstration but rather mindful and skeptical evaluators of new mathematical infor-
task was successful but most particularly for participants who preferred mation. Children were more likely to use the procedure (shortcut or
shortcuts. Few studies have examined the effects of participants' evalu- left-to-right) that they judged as superior, suggesting that the children
ation of different solution methods on learning using either academic were evaluating the procedures and choosing those procedures that
tasks or with child participants. Even fewer studies have investigated they judged to be better. The issue may not only be why some children
whether children's evaluations of mathematical information influences mindlessly use old problem solving methods but also why some chil-
their problem solving procedures although there is extensive research dren judge inefficient procedures to be superior.
demonstrating the learning benefits of comparing problem solving
procedures (e.g., Matthews and Rittle-Johnson, 2009; Rittle-Johnson & 4.2. Implications and limitations
Star, 2007) as well as research indicating that children can critique
information before adopting it (Gelman, 2009). The current study The first implication of the results is to reinforce previous findings
showed that children's evaluations of conceptually-based shortcuts mit- that students have a relatively weak understanding of the relationships
igated their subsequent problem solving procedures, suggesting that between multiplication and division. Although older students in the
simply presenting information is not sufficient to promote change and current study used the shortcuts more frequently, shortcut use never
that children can be critical consumers of mathematics instruction. predominated over more effortful procedures even after the demon-
The results support the notion that conceptual understanding of the stration. This weak understanding of the relationship between multipli-
relationships between operations, as assessed by the use of the inver- cation and division seems counterintuitive given that multiplication and
sion and associativity shortcuts during problem solving, can be easily division are so often linked together in instruction. Children are typical-
and quickly increased. Adding the inversion and associativity shortcuts ly taught simple division as the opposite or reverse of multiplication
to children's problem solving repertoires through a brief and simple (NCTM, 2000. Saskatchewan Ministry of Education, 2009) and make
demonstration will promote further learning in the domain of mathe- use of this inverse understanding when solving simple division prob-
matics meeting a critical goal of instruction (Schneider and Stern, 2009). lems (Robinson et al., 2006a). This knowledge does not appear to trans-
fer to three-term multiplication and division problems. Teachers may
4.1. Individual differences in shortcut acquisition overestimate students' understanding of the relationships between
multiplication and division as they do children's understanding of
The finding that a brief demonstration of alternate solution methods equivalence (Sherman, 2007). Two possible reasons for why children's
is sufficient to promote shortcut acquisition was affected by whether inverse knowledge does not transfer to three-term problems might be
participants positively evaluated the shortcut compared to the left-to- because children are less familiar with solving three-term problems
right procedure. Participants in the Demonstration Middle group who (Canobi, 2004) and are poorer at solving them (Smedslund, 1962,
preferred the shortcut did not differ from those who preferred the 1966) but these reasons do not appear to impact inversion shortcut
left-to-right procedure in their shortcut use before the demonstration use by younger children on addition and subtraction inversion prob-
task. After the demonstration, participants who preferred the shortcut lems. Studies such as the current one have strong educational value as
markedly increased their shortcut use while those who preferred the they serve to identify effective ways of teaching mathematical concepts
left-to-right procedure did not. Even for the shortcut preference (Nunes et al., 2009) and help tailor instruction to the specific operations
group, average shortcut use after the demonstration did not over- being taught.
whelmingly predominate as inversion shortcut use remained under The second implication is that the inversion and associativity short-
70% and associativity shortcut use under 50% despite many participants cuts can be quickly and easily increased by demonstrating alternative
doubling their shortcut use after the demonstration. solution methods including the inversion and associativity shortcuts
Why students used the left-to-right procedure during problem solv- and asking students to evaluate the methods. Thus, the demonstration
ing and/or preferred the left-to-right procedure during the demonstra- task which has been used primarily as a measure of children's conceptu-
tion task may be that by Grade 6, participants have many years of al knowledge (Crooks and Alibali, 2014; Prather and Alibali, 2009) is
solving horizontally presented problems using a left-to-right approach. also an effective and brief instruction tool for promoting conceptually-
Anecdotally, students who preferred the left-to-right procedure often based shortcuts or problem solving procedures.
spontaneously commented on its familiarity, consistent with previous The information provided in our task, however, was limited and fu-
studies (e.g., Robinson and Dubé, 2009c, 2012, 2013). Older students ture research using a task which provides more information about the
may become “entrenched” in using “tried and true” problem solving shortcuts and the concepts behind the shortcuts may lead to even great-
procedures (McNeil, 2007). The “Einstellung effect” or “mindlessness” er benefits. Further, the current study examined only shortcut use im-
has been identified as when well-learned and successful problem solv- mediately after the demonstration task so whether children would
ing procedures continue to be used even when superior solution still continue to use the shortcuts after a delay needs to be investigated.
 K.M. Robinson et al. / Learning and Individual Differences 49 (2016) 297–304 303

In addition, the demonstration task may only have been effective as stu- References
dents in the Demonstration Middle group had the immediate opportu-
Beilock, S. L. (2008). Math performance in stressful situations. Current Directions in
nity to use the shortcuts. However, that even so little information Psychological Science, 17, 339–343. http://dx.doi.org/10.1011/j.1467-8721.2008.
presented in such a short period of time promoted shortcut use is en- 000602.
couraging and suggests that although children's conceptual knowledge Bisanz, J., & LeFevre, J. -A. (1990). Strategic and nonstrategic processing in the develop-
ment of mathematical cognition. In D. F. Bjorklund (Ed.), Children's strategies: Contem-
of multiplication and division might be weak, it may also be fairly porary views of cognitive development (pp. 213–244). Hillsdale, NJ: Erlbaum.
straightforward to increase, particularly for students who evaluate Bisanz, J., Watchorn, R. P. D., Piatt, C., & Sherman, J. (2009). On “understanding” children's
conceptually-based shortcuts more positively than traditional and developing use of inversion. Mathematical Thinking and Learning, 11, 10–24. http://dx.
doi.org/10.1080/10986060802583907.
well-learned problem solving procedures. Canobi, K. H. (2004). Individual differences in children's addition and subtraction knowl-
The third implication is that because of individual differences in re- edge. Cognitive Development, 19, 81–93. http://dx.doi.org/10.1016/j.cogdev.2003.10.
ceptivity to the merits of a conceptually-based shortcut, more direct in- 001.
Canobi, K. H. (2009). Concept-procedure interactions in children's addition and subtrac-
struction may be required for some students and instructional studies
tion. Journal of Experimental Child Psychology, 102, 131–149. http://dx.doi.org/10.
need to consider how students bring their own attitudes to the learning 1016/j.jecp.2008.07.008.
context about how problem solving is supposed to be done. This finding Canobi, K. H., Reeve, R. A., & Pattison, P. E. (1998). The role of conceptual understanding in
highlights that not only are there individual differences in children's children's addition problem solving. Developmental Psychology, 34, 882–891. http://
dx.doi.org/10.1037/0012-1649.34.5.882.
conceptual knowledge of multiplication and division but that there are Crooks, N. M., & Alibali, M. W. (2014). Defining and measuring conceptual knowledge in
also marked individual differences in children's attitudes towards mathematics. Developmental Review, 34, 344–377. http://dx.doi.org/10.1016/j.dr.
conceptually-based shortcuts. More than one type of instruction may 2014.10.001.
Dubé, A. K. (2014). Adolescents' understanding of inversion and associativity. Learning
be needed to promote the development of children's understanding of and Individual Differences, 36, 49–59. http://dx.doi.org/10.1016/j.lindif.2014.09.002.
arithmetic concepts. It is not only the information presented but also Dubé, A. K., & Robinson, K. M. (2010a). The relationship between adults's conceptual un-
how it is interpreted by individuals that determine the efficacy of math- derstanding of inversion and associativity. Canadian Journal of Experimental
Psychology, 64, 60–66. http://dx.doi.org/10.1037/a0017756.
ematical instruction aimed at improving children's understanding of Dubé, A. K., & Robinson, K. M. (2010b). Accounting for individual variability in inversion
critical mathematical concepts (McNeil, Fyfe, Petersen, Dunwiddie, shortcut use. Learning and Individual Differences, 20, 687–693. http://dx.doi.org/10.
and Brletic-Shipley, 2011). Overall, although children's understanding 1016/j.lindif.2010.09.009.
Gelman, S. A. (2009). Learning from others: Children's construction of concepts. Annual
of multiplication and division concepts is quite weak, particularly rela-
Review of Psychology, 60, 115–140. http://dx.doi.org/10.1146/annurev.psych.59.
tive to addition and subtraction concepts, promoting conceptual under- 103006.093659.
standing may be quite straightforward – at least for some individuals. Gilmore, C. K., & Papadatou-Pastou, M. (2009). Patterns of individual differences in con-
ceptual understanding and arithmetical skill: A meta-analysis. Mathematical
Thinking and Learning, 11, 25–40. http://dx.doi.org/10.1080/10986060802583923.
Kilpatrick, J., Swafford, J., & Findell, B. (2001). Adding it up: Helping children learn mathe-
Appendix A matics. Washington, DC: National Academy Press.
Klein, J. S., & Bisanz, J. (2000). Preschoolers doing arithmetic: The concepts are willing but
the working memory is weak. Canadian Journal of Experimental Psychology, 54,
Set 1 Set 2 105–116. http://dx.doi.org/10.1037/h0087333.
Lai, M. -L., Baroody, A. J., & Johnson, A. R. (2008). Fostering Taiwanese preschoolers' un-
Inversion problems
derstanding of the addition-subtraction inverse principle. Cognitive Development, 23,
4×6÷6 8×2÷2
216–235. http://dx.doi.org/10.1016/j.cogdev.2007.06.002.
8×5÷5 7×6÷6
Lemaire, P., & Lecacheur, M. (2011). Age-related changes in children's executive functions
7×4÷4 6×9÷9 and strategy selection: A study in computational estimation. Cognitive Development,
8×6÷6 5×7÷7 26, 282–294. http://dx.doi.org/10.1016/j.cogdev.2011.01.002.
6×2÷2 6×4÷4 Lepper, M. R., Corpus, H. J., & Iyengan, S. S. (2005). Intrinsic and extrinsic motivation-
4×5÷5 4×3÷3 al orientations in the classroom: Age differences and academic correlates. Journal
7×3÷3 3×5÷5 of Educational Psychology, 97, 184–196. http://dx.doi.org/10.1037/0022-0663.97.2.
9×4÷4 8×4÷4 184.
Martin, A. J., Anderson, J., Bobis, J., Vellar, R., & Way, J. (2012). Switching on and switching
Associativity problems off in mathematics: An ecological study of future intent and disengagement among
6×8÷4 5×9÷3 middle school students. Journal of Educational Psychology, 104, 1–18. http://dx.doi.
2×9÷3 2×8÷4 org/10.1037/a0025988.
3×8÷4 3×8÷2 Matthews, P., & Rittle-Johnson, B. (2009). In pursuit of knowledge: Comparing self-
7×8÷2 7×9÷3 explanations, concepts, and procedures as pedagogical tools. Journal of Experimental
5×4÷2 3×4÷2 Child Psychology, 104, 1–21. http://dx.doi.org/10.1016/j.jecp.2008.08.004.
4×9÷3 7×8÷4 McLeod, D. B. (1993). Connecting research to teaching. Mathematics Teacher, 86, 761–763.
5×6÷2 9×4÷2 McNeil, N. M. (2007). U-shaped development in math: 7-year-olds outperform 9-year-
olds on equivalence problems. Developmental Psychology, 43, 687–695. http://dx.
4×6÷3 2×6÷3
doi.org/10.1037/0012-1649.43.3.687.
McNeil, N. M. (2014). A change–resistance account of children's difficulties understand-
ing mathematical equivalence. Child Development Perspectives, 8, 42–47. http://dx.
doi.org/10.1111/cdep.12062.
McNeil, N. M., Rittle-Johnson, B., Hattikudur, S., & Petersen, L. A. (2010). Continuity in rep-
Appendix B
resentation between children and adults: Arithmetic knowledge hinders undergrad-
uates' algebraic problem solving. Journal of Cognition and Development, 11, 437–457.
Percentage use, accuracy, and cut-offs, and mean correct solution http://dx.doi.org/10.1080/15248372.2010.516421.
latencies (in ms) of each procedure on inversion and associativity McNeil, N. M., Fyfe, E. R., Petersen, L. A., Dunwiddie, A. E., & Brletic-Shipley, H. (2011).
Benefits of practicing 4 = 2 + 2: Nontraditional problem formats facilitate children's
problems. understanding of mathematical equivalence. Child Development, 82, 1620–1633.
http://dx.doi.org/10.1111/j.1467-8624.2011.01622.x.
National Council of Teachers of Mathematics (2000). Principles and standards for school
Use Accuracy Cut-offs Solution latencies
mathematics. Reston, VA: NCTM.
Inversion problems National Mathematics Advisory Panel (2008). Foundations for success: The final report of
Shortcut 29.1 (1.2) 90.2 (2.3) 7.7 (2.2) 1818 (61) the National Mathematics Advisory Panel. Washington, D.C.: U.S. Department of
Negation 19.4 (1.0) 93.3 (2.6) 5.4 (2.6) 2988 (69) Education.
Newton, K. J., Star, J. R., & Lynch, K. (2010). Understanding the flexibility of struggling al-
Left-to-right 50.0 (1.3) 44.7 (1.5) 49.3 (1.5) 2848 (55)
gebra students. Mathematical Thinking and Learning, 12, 282–305. http://dx.doi.org/
Associativity problems
10.1080/10986065.2010.482150.
Shortcut 13.7 (0.9) 51.4 (5.1) 44.2 (5.2) 2469 (123)
Nunes, T., Bryant, P., Burman, D., Bell, D., Evans, D., Hallett, D., & Montgomery, L. (2008).
Left-to-right 85.6 (0.9) 36.4 (1.4) 57.1 (1.4) 2993 (46) Deaf children's understanding of inverse relationships. In M. Marschark, & P. C.
Hauser (Eds.), Deaf cognition: Foundations and outcomes (pp. 201–225). New York,
Note. Standard errors are in parentheses. NY: Oxford.
304 K.M. Robinson et al. / Learning and Individual Differences 49 (2016) 297–304

Nunes, T., Bryant, P., Hallett, D., Bell, D., & Evans, D. (2009). Teaching children about the Robinson, K. M., Arbuthnott, K. D., Rose, D., McCarron, M. C., Globa, C. A., & Phonexay, S. D.
inverse relation between addition and subtraction. Mathematical Thinking and (2006a). Children's strategies for solving simple division problems. Journal of
Learning, 11, 61–78. http://dx.doi.org/10.1080/10986060802583980. Experimental Child Psychology, 93, 224–238. http://dx.doi.org/10.1016/j.jecp.2005.09.
Prather, R. W., & Alibali, M. W. (2009). The development of arithmetic principle knowl- 002.
edge: How do we know what learners know? Developmental Review, 29, 221–248 Robinson, K. M., Ninowski, J. E., & Gray, M. L. (2006b). Children's understanding of the ar-
10/1016/j.dr.2009.09.001. ithmetic concepts of inversion and associativity. Journal of Experimental Child
Rittle-Johnson, B., & Star, J. (2007). Does comparing solution methods facilitate conceptu- Psychology, 94, 349–362. http://dx.doi.org/10.1016/j.jecp.2006.03.004.
al and procedural knowledge: An experimental study on learning to solve equations. Saskatchewan Ministry of Education (2009). Saskatchewan curriculum: Mathematics 3.
Journal of Educational Psychology, 99, 561–574. http://dx.doi.org/10.1037/0022-0663. Saskatchewan: Curriculum and E-Learning Branch.
99.3.561. Schneider, M., & Stern, E. (2009). The inverse relation of addition and subtraction: A
Robinson, K. M., & Dubé, A. K. (2009a). Children's understanding of addition and subtrac- knowledge integration perspective. Mathematical Thinking and Learning, 11(1),
tion concepts. Journal of Experimental Child Psychology, 103, 532–545. http://dx.doi. 92–101. http://dx.doi.org/10.1080/10986060802584012.
org/10.1016/j.jecp.2008.12.002. Sherman, J. (2007, April). From failure to success on equivalence problems and how
Robinson, K. M., & Dubé, A. K. (2009b). A microgenetic study of the multiplication and di- teachers perceive the process. Talk given at the biennial conference for the Society for
vision inversion concept. Canadian Journal of Experimental Psychology, 63, 193–200. Research in Child Development. Boston: MA.
http://dx.doi.org/10.1037/a0013908. Siegler, R. S., & Stern, E. (1998). Conscious and unconscious strategy discoveries: A
Robinson, K. M., & Dubé, A. K. (2009c). Children's understanding of the inverse relation microgenetic analysis. Journal of Experimental Psychology: General, 127, 377–397.
between multiplication and division. Cognitive Development, 24, 310–321. http://dx. Smedslund, J. (1962). The acquisition of conservation of substance and weight in children.
doi.org/10.1016/j.cogdev.2008.11.001. Scandinavian Journal of Psychology, 3, 69–77.
Robinson, K. M., & Dubé, A. K. (2012). Children's use of arithmetic shortcuts: The role of Smedslund, J. (1966). Microanalysis of concrete reasoning: The difficulty of some combi-
attitudes in strategy choice. Child Development Research, 2012. http://dx.doi.org/10. nations of addition and subtraction of one unit. Scandinavian Journal of Psychology, 7,
1155/2012/459385 10 pages. 145–156.
Robinson, K. M., & Dubé, A. K. (2013). Children's additive concepts: Promoting under- Starkey, P., & Gelman, R. (1982). The development of addition and subtraction abilities
standing and the role of inhibition. Learning and Individual Differences, 23, 101–107. prior to formal schooling in arithmetic. In T. P. Carpenter, J. M. Moser, & T. A.
http://dx.doi.org/10.1016/j.lindif.2012.07.016. Romberg (Eds.), Addition and subtraction: A cognitive perspective (pp. 99–116).
Robinson, K. M., & LeFevre, J. (2012). The inverse relation between multiplication and di- Hillsdale, NJ: Erlbaum.
vision: Concepts, procedures, and a cognitive framework. Educational Studies in Wolters, C. A. (2004). Advancing achievement goal theory: Using goal structures and goal
Mathematics, 79, 409–428. http://dx.doi.org/10.1007/s10649-011-9330-5. orientation to predict students' motivation, cognition, and achievement. Journal of
Robinson, K. M., & Ninowski, J. E. (2003). Adults' understanding of inversion concepts: Educational Psychology, 96, 236–250. http://dx.doi.org/10.1037/0022-0663.96.2.236.
How does performance on addition and subtraction inversion problems compare to
performance on multiplication and division inversion problems? Canadian Journal
of Experimental Psychology, 57, 321–330. http://dx.doi.org/10.1037/h0087435.