British J of Psychology - 2011 - Torbeyns - Use of Indirect Addition in Adults Mental Subtraction in The Number Domain Up

20448295, 2011, 3, Downloaded from https://bpspsychub.onlinelibrary.wiley.com/doi/10.1111/j.2044-8295.2011.02019.x by The Open University, Wiley Online Library on [12/03/2023].
See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
585
The
British
Psychological
British Journal of Psychology (2011), 102, 585–597

C 2011 The British Psychological Society
Society
www.wileyonlinelibrary.com
Use of indirect addition in adults’ mental

subtraction in the number domain up to 1,000
Joke Torbeyns1,2 ∗ , Bert De Smedt1 , Greet Peters1 ,

Pol Ghesquière1 and Lieven Verschaffel1
1
Katholieke Universiteit Leuven, Belgium
2
GROEP T - Leuven Education College, Belgium
This study examined adults’ use of indirect addition and direct subtraction strategies
on multi-digit subtractions in the number domain up to 1,000. Seventy students who
differed in their level of arithmetic ability solved multi-digit subtractions in one choice
and two no-choice conditions. Against the background of recent findings in elementary
subtraction, we manipulated the size of the subtrahend compared to the difference and
only selected items with large distances between these two integers. Results revealed
that adults frequently and efficiently apply indirect addition on multi-digit subtractions,
yet adults with higher arithmetic ability performed more efficiently than those with
lower arithmetic ability. In both groups, indirect addition was more efficient than direct
subtraction both on subtractions with a subtrahend much larger than the difference (e.g.,
713 − 695) and on subtractions with a subtrahend much smaller than the difference (e.g.,
613 − 67). Unexpectedly, only adults with lower arithmetic ability fitted their strategy
choices to their individual strategy performance skills. Results are interpreted in terms
of mathematical and cognitive perspectives on strategy efficiency and adaptiveness.
For a long time, researchers in cognitive and educational psychology have investigated
the variety of strategies that people use to mentally solve symbolic subtraction problems
(e.g., Campbell, 2005; Siegler, 2003; Verschaffel, Greer, & De Corte, 2007). Most studies
focused on adults’ strategy use in the domain of single-digit subtraction (Campbell, 2008;
Campbell & Xue, 2001; LeFevre, DeStefano, Penner-Wilger, & Daley, 2006; Peters, De
Smedt, Torbeyns, Ghesquière, & Verschaffel, 2010a; Seyler, Kirk, & Ashcraft, 2003) and
revealed that adults often use indirect addition strategies to solve these subtractions. Few
researchers, however, have investigated adults’ strategy use on multi-digit subtractions
(Geary, Frensch, & Wiley, 1993; Peters, De Smedt, Torbeyns, Ghesquière, & Verschaffel,
2010b; Torbeyns, Ghesquière, & Verschaffel, 2009) and little is known about the use of
indirect addition in this number domain. The present study therefore aimed at investi-
gating adults’ use of indirect addition in multi-digit subtraction more systematically. In
∗ Correspondence should be addressed to Joke Torbeyns, Centre for Instructional Psychology and Technology, Vesaliusstraat 2,
Box 3770, 3000 Leuven, Belgium (e-mail: joke.torbeyns@ped.kuleuven.be).
DOI:10.1111/j.2044-8295.2011.02019.x
20448295, 2011, 3, Downloaded from https://bpspsychub.onlinelibrary.wiley.com/doi/10.1111/j.2044-8295.2011.02019.x by The Open University, Wiley Online Library on [12/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
586 Joke Torbeyns et al.
the remainder of the introduction, we summarize the research on adults’ use of indirect
addition and present the design and hypotheses of the current study.
Indirect addition strategy use

When confronted with multi-digit subtractions, such as 341 − 163, people can basically
choose between two types of strategies, namely direct subtraction and indirect addition
strategies. Direct subtraction involves the straightforward use of the subtraction
operation to solve subtractions, by directly subtracting the subtrahend from the minuend
(e.g., 341− 163 = ?; 341 − 100 = 241, 241 − 60 = 181, 181 − 3 = 178). Indirect
addition is characterized by the use of the complementary addition operation to solve a
subtraction problem, by determining how much should be added to the subtrahend to
arrive at the minuend (e.g., 341 − 163 = ?; 163 + 37 = 200, 200 + 100 = 300, 300 +
41 = 341, so the answer is 37 + 100 + 41 or 178).1,2 Although adults’ use of indirect
addition has been extensively studied in single-digit subtraction, its characteristics and
use on multi-digit subtractions hardly received any research attention. Moreover, the
existing body of data mainly provided anecdotal reports of the application of indirect
addition on complex subtractions (Geary et al., 1993). To the best of our knowledge,
only Torbeyns et al. (2009) provided a systematic attempt to analyse the characteristics
of adults’ indirect addition use on complex subtractions.
Torbeyns et al. (2009) examined adults’ strategy use on symbolic subtractions in
the number domain up to 1,000. They examined three types of subtractions, namely
subtractions with a small difference between the two given integers (e.g., 812 −783 = ?),
subtractions with a medium difference (e.g., 821 − 475 = ?), and subtractions with a
large difference (e.g., 813 − 176 = ?). This study revealed that adults use indirect addition
very frequently and efficiently on all three types of subtractions, and that it was used
more frequently and faster (without any loss in accuracy) than direct subtraction.
These findings are in line with those of Campbell (2008) in the domain of single-
digit subtraction. He showed that adults use the complementary addition operation to
answer subtraction problems when confronted with large single-digit subtractions (i.e.,
subtractions involving the crossing of the 10, as 13 − 8 = ?). However, recent data by
Peters et al. (2010a, 2010b) suggest that adults switch between indirect addition and
direct subtraction depending on the relative size of the subtrahend. Peters et al. (2010a)
showed that adults solve large single-digit subtractions where the subtrahend (S) is
larger than the difference (D) (S > D; e.g., 12 − 9 = ?) by means of indirect addition, but
answer large single-digit subtractions where the subtrahend is smaller than the difference
(S < D; e.g., 12 − 3 = ?) by means of direct subtraction. Importantly, this strategy switch
1 Both direct subtraction and indirect addition can be executed in various ways, each referring to a specific retrieval or
procedural strategy. For example, 341 − 163 can be solved with a large variety of direct subtraction strategies, including
(without being exhaustive) sequentially subtracting the hundreds, tens, and units of the subtrahend from the minuend (i.e.,
341 − 100 = 241; 241 − 60 = 181; 181 − 3 = 178), decomposing the subtrahend into parts that can be subtracted easily
from the un-split minuend (i.e., 341 − 141 = 200; 200 − 22 = 178), or splitting both the minuend and the subtrahend
into hundreds, tens, and units (i.e., 300 − 100 = 200; 40 − 60 = −20; 1 − 3 = −2; 200 − 20 − 2 = 178). Likewise,
the use of indirect addition might involve the sequential addition of hundreds, tens, and units to the subtrahend to reach the
minuend (i.e., 163 + 100 = 263; 263 + 70 = 333; 333 + 8 = 341), or jumping towards the next hundred until the
difference is bridged (i.e., 163 + 37 = 200; 200 + 100 = 300; 300 + 41 = 341). Taking into account the major aims of
the present study, we classified strategies as direct subtraction and indirect addition on the basis of the operation performed
during the solution process, but did not further distinguish the rich variety of direct subtraction and indirect addition strategies
that can be applied on complex subtractions.
2 It should be noted here that there is also a third type of strategy, namely indirect subtraction, whereby one finds the
solution by determining how much has to be decreased or subtracted from the larger given number to get the smaller one
(e.g., 341 − 163 = ?; 341 − ? = 163) (De Corte & Verschaffel, 1987).
Indirect addition in adults 587
was only observed when the numerical distance between S and D was relatively large
(e.g., 3 and 9 in 12 − 3 = ?) and not when S and D were relatively close to each other
(e.g., 6 and 7 in 13 − 6 = ?). Peters et al. (2010b) recently replicated these results on
two-digit subtractions in the number domain up to 100: when the numerical distance
between S and D is large, adults solve two-digit subtractions where S > D (e.g., 77 −
68 = ?) with indirect addition, but two-digit subtractions where S < D (e.g., 83 − 4 = ?)
with direct subtraction.
The findings by Peters et al. (2010a, 2010b) might challenge the results of Torbeyns
et al. (2009) because they might explain the remarkably high frequency of indirect
addition use on (large difference) subtractions. The large-difference subtractions in the
latter study were characterized by S < D, but the relative distance between S and D was
not systematically manipulated and might have been too small to (strongly) elicit the use
of direct subtraction and reveal its efficiency on these subtractions. Moreover, Torbeyns
et al. (2009) selected only a limited number of items per subtraction type (maximum
four items per subtraction type) and presented these subtractions to a relatively small
sample of participants, which might have jeopardized the reliability of their results.
The present study

The present study tried to replicate and extend the findings by Torbeyns et al. (2009)
by administering a more extreme set of subtractions with systematic manipulation of
the relative distance between the subtrahend and the difference. We also administered a
larger number of items per subtraction type and selected a larger sample of participants,
who differed in their level of arithmetic ability. We aimed at investigating whether adults’
dominant and efficient use of indirect addition still holds when confronted with multi-
digit subtractions with large distances between the operands. In line with Torbeyns et al.
(2009), we focused on complex subtractions in the number domain up to 1,000 and
defined small-difference subtractions as subtractions where the subtrahend is larger than
the difference (S > D); large-difference subtractions were (again) defined as subtractions
where the subtrahend is smaller than the difference (S < D). Different from Torbeyns
et al. (2009), we only included subtractions with a very large distance between S and
D, that is, subtractions with the minuend being a three-digit number and either S or D a
two-digit number. On the basis of Peters et al. (2010a, 2010b), we expected that these
subtractions would be most favourable towards a switch between direct subtraction and
indirect addition.
As in Torbeyns et al.’s (2009) study, Lemaire and Siegler’s (1995) model of strategy
change served as the theoretical framework. This model distinguishes four parameters
of strategic competence. The first parameter, strategy repertoire, refers to the types of
strategies an individual applies to solve a series of items. The second parameter, strategy
distribution, refers to the frequency of each strategy. The third parameter, strategy
efficiency, involves the accuracy and speed of strategy execution. The fourth parameter,
strategy selection, refers to the adaptiveness of individual strategy choices, that is, do
people select those strategies that are most efficient for them.
We also used Siegler and Lemaire’s (1997, see also Luwel, Onghena, Torbeyns,
Schillemans, & Verschaffel, 2010) choice/no-choice method to investigate strategy
efficiency data in an unbiased way and to assess the adaptive nature of strategy choices
on the basis of individual strategy performance characteristics. This method requires
assessing each participant in a choice condition, where participants can choose which
strategy they use to solve each item, and one or more no-choice condition(s), where
participants are forced to solve all items with one particular strategy. This makes
selective assignments of strategies to items and/or participants impossible, and thus

yields unbiased data about strategy efficiency. A comparison of the efficiency data of
the different strategies in the no-choice conditions with the strategy choices made
in the choice condition allows inference about whether participants are adaptive in
their strategy choice, that is, whether they take into account their individual strategy
performance characteristics (as evidenced by the information obtained in the no-choice
conditions) during the strategy choice process (in the choice condition).
Against the background of Lemaire and Siegler’s (1995) model of strategy change,
we expected that indirect addition would be part of participants’ strategy repertoire
(Parameter 1). If adults always use indirect addition to solve large single-digit (Campbell,
2008) and multi-digit subtractions (Torbeyns et al., 2009), then indirect addition would
be the most frequent strategy in the choice condition (Parameter 2) and the most
efficient strategy in the no-choice conditions (Parameter 3) for all types of subtraction
problems. If, however, adults switch between direct subtraction and indirect addition
depending on the size of the subtrahend compared to the difference (Peters et al.,
2010a, 2010b), then differences between S > D and S < D subtractions are expected: on
S > D subtractions, indirect addition should be the most frequent strategy in the choice
condition (Parameter 2) and the most efficient strategy in the no-choice conditions
(Parameter 3), whereas on S < D subtractions, direct subtraction should be the most
frequent strategy in the choice condition (Parameter 2) and the most efficient strategy
in the no-choice conditions (Parameter 3). Next, we anticipated that adults would select
their strategies adaptively, that is, that indirect addition would be used in the choice
condition most frequently by those adults who master indirect addition better than
direct subtraction (as evidenced by the information obtained in the no-choice conditions)
(Parameter 4). Finally, we hypothesized that adults’ arithmetic ability would influence
their strategic behaviour (Shrager & Siegler, 1998; Siegler, 2000), that is, that adults with
higher arithmetic ability would execute both direct subtraction and indirect addition
more efficiently than adults with lower arithmetic fluency (Parameter 3) and that the
former would adapt their strategy choices more to their individual strategy performance
characteristics than the latter (Parameter 4).
Method
Participants
Participants were 70 healthy higher education students (35 men, 35 women). Their
age ranged from 18 to 26 years (M = 21; SD = 2). We administered participants’
arithmetic ability using the arithmetic sub-test of the Wechsler Adult Intelligence Scale-
Revised (WAIS-R; Wechsler, 1981). This sub-test involves 20 word problems that need
to be answered within a (variable) time limit, with a maximum sub-test score of 22.
We distinguished among two arithmetic ability groups on the basis of their scores on
the arithmetic sub-test of WAIS-R. The group of adults with higher arithmetic ability
included those participants scoring 18 or more (= a score higher than the median score);
adults with lower arithmetic ability received a score of 17 or less (= the median score
or lower); F(1, 69) = 118.92, p < .01. Table 1 describes the sex, arithmetic scores, and
ages for the two groups.
Materials and design

All participants mentally solved subtractions in the number domain up to 1,000. Two
types of experimental items, each involving 10 trials, were presented. All items consisted
Table 1. Sex, score (maximum = 22) and age (in years) per arithmetic ability group
Arithmetic sub-test score Age
Group Male Female M SD Range M SD
Lower 10 25 14.34 2.62 8–17 22.08 2

Higher 25 10 19.54 1.04 18–22 22.54 2
Note. Lower = lower arithmetic ability group; higher = higher arithmetic ability group.
of a three-digit minuend with values ranging from 501 to 951 (two items per hundreds
for each item type). To prevent the use of quasi-automatic strategies – and thus strategies
that are difficult to report reliably (Kirk & Ashcraft, 2001) – all items required two trades,
namely a trade from the hundreds to the tens and a trade from the tens to the ones. The
first item type, S > D subtractions, were subtractions with a three-digit subtrahend and
a two-digit difference (e.g., 713 − 695 = ?). The second item type, S < D subtractions,
involved subtractions with a two-digit subtrahend and a three-digit difference (e.g.,
613 − 67 = ?).
Participants solved the series of subtractions individually in one choice and two
no-choice conditions. In the choice condition, participants were instructed to solve all
subtractions with their preferred strategy, which could be either direct subtraction,
that is, calculating the outcome by subtracting the subtrahend from the minuend, or
indirect addition, that is, determining how much should be added to the subtrahend to
arrive at the minuend. Seven buffer items were added to the set of experimental items to
increase the variety of the problems on which participants had to make a strategy choice,
which would make the strategy choice process more natural. These buffer items were
subtractions with a relatively small distance between the subtrahend and the difference,
that is, subtractions with a three-digit subtrahend and a three-digit difference, differing
maximum 100 from half of the minuend (e.g., 873 − 487 = ?). Two no-choice conditions
were further administered: one in which participants had to solve all subtractions by
directly subtracting the subtrahend from the minuend and one in which they had to
answer all subtractions by adding up from the subtrahend to the minuend.
All problems were presented horizontally in the middle of a computer screen.
All items were presented by means of the Affect 3.0 computer program (Hermans,
Clarysse, Baeyens, & Spruyt, 2002). Participants solved the series of items individually
in a quiet room in the presence of a trained experimenter. The experimenter instructed
participants to solve the subtractions both accurately and fast. Participants were also
asked to think aloud during each trial, allowing the experimenter to make notes on
the solution process. When the participant had completely stated his/her answer, the
experimenter, who was seated next to the participant, immediately pressed the space
bar of an external keyboard connected to the PC to register the reaction time. After
registration of the reaction time, the answer was entered on the keyboard by the
experimenter and the next trial was initiated.
All participants started with the choice condition, in which they were required to
choose either direct subtraction or indirect addition. Half of the participants continued
with the direct subtraction no-choice condition, and ended with the indirect addition
no-choice condition. For the other participants, the order of the no-choice conditions
was reversed. Each condition started with two practice items to familiarize participants
with the testing procedure and the two types of strategies. The first practice item was
solved by the experimenter, the second one by the participant, using the requested
strategy (no-choice conditions) or strategies (choice condition). After completion of the
computer task, participants answered the arithmetic sub-test of the WAIS-R to assess
their arithmetic ability.
Analysis
The analysis was restricted to the 20 experimental items; so, the seven buffer items
were excluded from the analysis. Transcripts of the verbal reports were used to classify
strategies as direct subtraction or indirect addition. This classification was done by two
independent researchers, who agreed on all trials. On the basis of these reports, we
determined the frequency of strategy use in the choice condition. Efficiency of strategy
execution was calculated on the basis of the accuracy data and the speed data from the
correctly solved trials gathered in the no-choice conditions.
We analysed strategy frequency and efficiency using repeated measurements ANOVA
(Analysis of Variance) by means of the proc mixed procedure of the SAS computer
program (Littell, Milliken, Stroup, Wolfinger, & Schabenberger, 2006). Post hoc analyses
were corrected for multiple comparisons with Tukey–Kramer adjustments. The adaptiv-
ity of adults’ strategy choices was determined on the basis of their individual strategy
performance characteristics. We compared the strategy frequency data in the choice
condition with the efficiency data gathered in the no-choice conditions. In line with
Siegler and Lemaire (1997), we calculated the correlation between the percentage of
use of indirect addition in the choice condition and the differences in accuracy and
speed between the indirect addition and direct subtraction condition (with participant
as the unit of analysis).
Results
Initial analyses indicated that the order of the no-choice conditions did not influence
participants’ task performances. Therefore, we grouped the data from the no-choice
conditions in all further analyses. All reported post hoc comparisons are statistically
significant at an alpha level of .05, unless otherwise noted.
Strategy distribution in the choice condition

Most participants used indirect addition at least once in the choice condition: 49/70
participants (70%) used both indirect addition and direct subtraction; 18/70 participants
(26%) answered all subtractions through indirect addition; 3/70 participants (4%) never
applied indirect addition in the choice condition. We observed no group differences in
participants’ strategy use, ␹ 2 (n = 70; df = 2) = 1.73, p = .42.
Indirect addition was used more frequently than direct subtraction in the choice
condition, t(69) = −6.63, p < .01 (see Table 2). We further analysed the relative
frequency of indirect addition in the choice condition using repeated measurements
ANOVA, with subtraction type (S > D vs. S < D) as within-subjects factor and arithmetic
ability (lower vs. higher) as between-subjects factor. As expected, participants applied
indirect addition more frequently on S > D than on S < D subtractions, F(1, 68) = 30.70,
p < .01. There were no group differences in the relative frequency of indirect addition
use in general, F(1, 68) = 0.71, p = .40. The group × subtraction type interaction was
also not significant, F(1, 68) = 2.87, p = .09.
Table 2. Frequency of indirect addition (in percentages) per group and item type in the choice condition
S⬎D S⬍D Total
M SD M SD M SD
Lower 78.00 32.79 61.14 40.13 69.57 37.36

Higher 90.86 20.05 59.14 34.16 75.00 32.07
All 84.43 27.75 60.14 37.01 72.29 34.79
Note. Lower = lower arithmetic ability group; higher = higher arithmetic ability group. S , subtrahend,
D , difference.
Strategy efficiency in the no-choice conditions

To evaluate differences in the efficiency between the no-choice conditions, we calculated
repeated measures ANOVAs, with condition (indirect addition vs. direct subtraction) and
subtraction type (S > D vs. S < D) as within-subject factors, and arithmetic ability (lower
vs. higher) as between-subjects factor. Tables 3 and 4 provide descriptive statistics for
accuracy and speed on correctly solved trials in the no-choice conditions, respectively,
for each subtraction type and arithmetic ability group.
Accuracy analyses revealed that participants answered subtractions with different
accuracies in the two no-choice conditions, F(1, 204) = 30.05, p < .01. The obligatory
use of indirect addition (M = .96, SD = 0.08) resulted in a higher accuracy level than the
obligatory use of direct subtraction (M = .90, SD = 0.12). The accuracy of responding in
Table 3. Accuracy (in proportion correct) per group and item type in the no-choice conditions
Indirect addition Direct subtraction
S⬎D S⬍D S⬎D S⬍D
M SD M SD M SD M SD
Lower 0.99 0.03 0.91 0.11 0.90 0.15 0.86 0.13

Higher 0.99 0.03 0.95 0.10 0.96 0.07 0.88 0.10
All 0.99 0.03 0.93 0.11 0.93 0.12 0.87 0.12
Note. Lower = lower arithmetic ability group; higher = higher arithmetic ability group. S, subtrahend;
D, difference.
Table 4. Speed (in seconds) per group and item type in the no-choice conditions
Indirect addition Direct subtraction
S⬎D S⬍D S⬎D S⬍D
M SD M SD M SD M SD
Lower 6.27 2.91 12.77 5.16 15.13 7.15 16.62 9.29

Higher 4.38 1.62 8.99 2.74 9.77 3.02 11.17 2.78
All 5.32 2.52 10.88 4.52 12.45 6.08 13.89 7.34
Note. Lower = lower arithmetic ability group; higher = higher arithmetic ability group. S, subtrahend;
D, difference.
the no-choice conditions also differed between the subtraction types, F(1, 204) = 27.22,
p < .01. Participants solved S > D subtractions (M = .96, SD = 0.09) in the no-choice
conditions more accurately than S < D subtractions (M = .90, SD = 0.11). Furthermore,
adults with higher arithmetic ability (M = .94, SD = 0.09) tended to answer the items
in the no-choice conditions more accurately than adults with lower arithmetic ability
(M = .92, SD = 0.12), F(1, 68) = 3.86, p = .05. Finally, the condition × subtraction
type interaction, F(1, 204) = 0.04, p = .84, and the condition × group interaction,
F(1, 204) = 1.26, p = .26, did not reach significance. In other words, indirect addition was
executed more accurately than direct subtraction on both S > D and S < D subtractions
in both arithmetic ability groups. The condition × subtraction type × group interaction
was also not significant, F(1, 204) = 2.73, p = .10, indicating that indirect addition was
executed more accurately than direct subtraction on both S > D and S < D subtractions
in the two arithmetic ability groups (see Table 3).
Turning to the speed data (see Table 4), our analyses revealed that the speed of
responding differed between the indirect addition and direct subtraction condition,
F(1, 204) = 177.14, p < .01. Participants answered the items about 5 s faster when
they had to use indirect addition (M = 8.10s, SD = 4.59) than when they were required
to apply direct subtraction (M = 13.17s, SD = 6.75). The speed of responding in the
no-choice conditions also differed between the subtraction types, F(1, 204) = 84.47,
p < .01, and the arithmetic ability groups, F(1, 68) = 17.13, p < .01. S > D subtractions
(M = 8.89s, SD = 5.86) were answered faster in the no-choice conditions than S <
D subtractions (M = 12.39s, SD = 6.26); and adults with higher arithmetic ability
(M = 8.58s, SD = 3.63) executed the strategies faster in the no-choice conditions
than adults with lower arithmetic fluency (M = 12.70s, SD = 7.61). There was also a
significant condition × subtraction type interaction, F(1, 204) = 29.28, p < .01. In the
direct subtraction condition, S > D subtractions were answered with the same speed
as S < D subtractions, whereas in the indirect addition condition, S > D subtractions
were solved faster than the S < D subtractions. Moreover, indirect addition was executed
faster than direct subtraction on both subtraction types. We further observed a significant
condition × group interaction, F(1, 204) = 11.36, p < .01. Although both higher and
lower arithmetic ability adults solved the subtractions faster in the indirect addition
condition (M High = 6.69s, SDHigh = 3.23; M Low = 9.52s, SDLow = 5.29) than in the direct
subtraction condition (M High = 10.47s, SDHigh = 2.97; M Low = 15.88s, SDLow = 8.26),
the speed differences between the no-choice conditions were larger for the lower than
for the higher arithmetic ability group. Finally, the condition × subtraction type ×
group interaction was not significant, F(1, 204) = 1.40, p = .24, indicating that indirect
addition led to faster responses than direct subtraction on both subtraction types in the
two arithmetic ability groups.
Strategy adaptivity in the choice condition

We assessed the adaptive nature of adults’ strategy choices in the choice condition
by correlating, for each arithmetic ability group separately, the frequency of indi-
rect addition in the choice condition with the efficiency data gathered in the no-
choice conditions. Unexpectedly, the correlations between the frequency of indirect
addition in the choice condition and the accuracy and speed (for correctly solved
subtractions) differences between the indirect addition and direct subtraction condition
were not significant in the higher arithmetic ability group, r(34) = .16, p = .35, and
r(34) = −0.06, p = .72, respectively. Participants with lower arithmetic ability did not
take into account their individual strategy accuracy characteristics during the strategy
selection process either, r(34) = .01, p = .99, but they flexibly fitted their strategy choices
to their individual strategy speed characteristics, r(34) = .40, p = .02. In other words,
the greater the difference in speed between indirect addition and direct subtraction for
a participant (as evidenced by the data in the no-choice condition), the more frequently
(s)he selected and applied indirect addition in the choice condition.
Discussion
Recent studies on adults’ strategy use on elementary subtraction indicate that they often
apply indirect addition to solve single-digit subtractions (Campbell, 2008; Campbell &
Xue, 2001; LeFevre et al., 2006; Peters et al., 2010a; Seyler et al., 2003). Surprisingly
few studies have examined the use of indirect addition on multi-digit subtractions (Geary
et al., 1993; Peters et al., 2010b; Torbeyns et al., 2009). The present study aimed at
systematically analysing the frequency, efficiency, and flexibility of indirect addition on
multi-digit subtractions in adults who differed in their level of arithmetic ability. Against
the background of recent data in elementary subtraction by Peters et al. (2010a, 2010b),
who showed that adults use indirect addition only when the numerical distance between
the difference and the subtrahend is sufficiently large, we extended earlier findings by
Torbeyns et al. (2009) in complex subtraction, by administering a set of subtractions
with a much larger distance between the subtrahend and the difference.
In line with Torbeyns et al. (2009), the present findings indicate that adults frequently
and efficiently apply indirect addition on subtractions with numbers up to 1,000.
Extending Torbeyns et al. (2009), the present study demonstrates that indirect addition
is frequently and efficiently used on complex subtractions by adults of both higher and
lower arithmetic ability. In both the ability groups, indirect addition was applied most
frequently on S > D subtractions in the choice condition. These S > D subtractions were
solved more accurately and faster with indirect addition than with direct subtraction.
Interestingly, indirect addition was also applied more frequently than direct subtraction
on the S < D subtractions in the choice condition and these subtractions were also
answered more efficiently with indirect addition than with direct subtraction. These
findings extend those of Torbeyns et al. (2009) and indicate that indirect addition
is the dominant and most efficient strategy for multi-digit subtraction, even when
confronted with subtractions with a very large difference between the subtrahend
and the difference, and also for adults with lower arithmetic ability.
Although these findings are generally in line with those of Campbell (2008) for single-
digit subtractions and of Torbeyns et al. (2009), who both showed that indirect addition
was the dominant strategy to solve subtractions, they are not in line with the recent
findings of Peters et al. (2010a, 2010b), who demonstrated that adults switch between
indirect addition and direct subtraction depending of the numerical distance between
the subtrahend and the difference. How can we explain the dominant and efficient use
of the indirect addition strategy on multi-digit subtractions in the present study?
A first explanation for the higher frequency and efficiency of indirect addition as
compared to direct subtraction relates to differences in the intrinsic difficulty of the
underlying addition and subtraction processes. The process of adding a number might
be easier than the process of taking away a similar number. Therefore, people prefer
solutions that refer to addition facts or rely on additive operations when confronted
with subtraction, and they are more efficient in doing so. Although it is difficult to
provide direct evidence for this claim, there are several elements from various sources
in support of it. First, from a mathematical perspective, the additive operation is

fundamental for the development of the natural numbers, the most basic number system
(Kilpatrick, Swafford, & Findell, 2001). Natural numbers are constructed on the basis
of the principle of recurrence, which means that every natural number is constructed
on the basis of the previous number as n + 1 (Burrill, 1967; Dedekind, 1963; Ifrah,
1985; Mainzer, 1995). In other words, the system of natural numbers is created and
defined in an additive order, almost automatically leading to the priority of addition to
subtraction. Second, many psychological studies in numerical cognition have shown
that addition facts and operations are acquired quicker and solved more efficiently
than the corresponding subtractions. This difference holds not only for single-digit
(Campbell & Xue, 2001), but also for multi-digit arithmetic (Fehr, Code, & Herrmann,
2007; Fuson, 1992; Verschaffel et al., 2007). Third, the priority of addition to subtraction
is also reflected in children’s curricula, textbooks, and teacher practices for elementary
arithmetic. Worldwide, mathematics teachers first teach their pupils how to add numbers
and only afterwards, that is, after sufficient practice and mastery of the addition operation,
teach how to subtract (Kilpatrick et al., 2001). Typically, the operation of subtraction
is introduced by reference to the complementary addition operation (Campbell, 2008).
The earlier and more intensive practice of the addition operation in the classroom may
further contribute to addition being more familiar and easier than subtraction. Future
studies should therefore try to examine to which extent adults’ preference for and
proficiency in solving subtraction problems by means of indirect addition is linked to
measures of their personal beliefs about the relative easiness of addition as compared to
subtraction, their feelings of familiarity with and competence in both operations, and
objective measures of their performance on both operations.
A second explanation relates to the number of carry and borrow operations that
had to be performed when using indirect addition or direct subtraction. Cognitive
psychological studies revealed that the number of trades influences the accuracy and
speed of responding in complex addition and subtraction (Ashcraft & Kirk, 2001; Geary
et al., 1993; Imbo, Vandierendonck, & De Rammelaere, 2007; Imbo, Vandierendonck,
& Vergauwe, 2007): performances become slower and less accurate as more carry or
borrow operations need to be performed. In the present study, all subtraction items
were defined as subtractions requiring two trades to prevent the use of quasi-automatic
strategies. When using the direct subtraction strategy on these items, participants had
to execute two borrow operations (e.g., 756 − 78 = ?; 756 − 70 = 686, 686 − 8 =
678). By contrast, when applying the indirect addition strategy, no carry operations were
required (i.e., 78 + 22 = 100, 100 + 600 = 700, 700 + 56 = 756, so the answer is
22 + 600 + 56 = 678). To evaluate the viability of this second explanation, future
studies on the frequency and efficiency of indirect addition and direct subtraction need
to include subtractions that do not require borrowing (e.g., 756 − 32 = ?), using the
dual-task paradigm.
Further extending the study of Torbeyns et al. (2009), we investigated the effect
of arithmetic ability on adults’ strategy competencies. As expected and in line with
Peters et al. (2010a), the present findings reveal differences in the efficiency of strategy
execution, favouring adults with higher arithmetic ability. We also observed differences
in the adaptive nature of the strategy choices: although neither the higher nor the lower
arithmetic ability group took into account individual strategy accuracy characteristics,
it – unexpectedly – were not the participants with higher arithmetic ability but those
with lower arithmetic ability who flexibly fitted their individual strategy choices to their
strategy speed characteristics.
This unexpected result might first of all be a statistical artefact, due to the absence
of variance both in arithmetic ability and in strategy speed in the higher arithmetic
ability group. To test the viability of this first explanation, future studies should select
samples with a larger variability in arithmetic ability level, by including not only
young adults from higher education (as was the case in this study), but also young
adults who did decide not to follow higher education courses and even people with
mathematical difficulties. An alternative explanation refers to the (objectively observed
and/or subjectively experienced) differences in strategy efficiency. Although indirect
addition was the most efficient strategy for both lower and higher arithmetic ability
groups, the efficiency differences between indirect addition and direct subtraction were
largest for participants with lower arithmetic ability. The efficiency differences between
the two strategies might have been too small for the adults with higher arithmetic ability
to trigger a flexible strategy switch. Indeed, whatever strategy the higher arithmetic
ability adults used, it enabled them to answer the subtractions accurately and fast. Taking
into account the high accuracy and speed of both strategies and the efficiency cost that
might accompany a switch between the different strategies, behaving ‘inflexibly’ might
have been most adaptive for them (cf. Imbo & LeFevre, 2009; Lemaire & Lecacheur,
2010; Luwel, Schillemans, Onghena, & Verschaffel, 2009). Future studies on the strategy
switch cost effect in the domain of complex subtraction and on the task, subject, and
context variables that may influence the role of this effect in people’s strategy choices
are needed to test this second explanation.
Finally, the current findings only involve adults. Previous studies on elementary
school children’s strategy use in symbolic subtraction indicate that they often have
considerable difficulties in learning and applying this strategy (De Smedt, Torbeyns,
Stassens, Ghesquière, & Verschaffel, 2010; Dowker, 2009; Heirdsfield & Cooper, 2004;
Selter, 2001). Against the background of the dominant use of indirect addition on
elementary and complex subtractions in adults, future studies should try to unravel
why so many elementary school children stick so strongly to direct subtraction and
move so slowly in the direction of indirect addition.
Acknowledgements
Bert De Smedt is a Postdoctoral Fellow of Research Foundation Flanders (Belgium); Greet
Peters is a Research Assistant for the Research Foundation Flanders (Belgium). This research
was partially supported by Grant GOA 2006/01 ‘Developing adaptive expertise in mathe-
matics education’ from the Research Fund K.U. Leuven, Belgium. The authors would like to
thank Saskia Nicolaas for her assistance during data collection.
References
Ashcraft, M. H., & Kirk, E. P. (2001). The relationships among working memory, math anxiety,
and performance. Journal of Experimental Psychology: General, 130, 224–237.
Burrill, C. W. (1967). Foundations of real numbers. New York: McGraw-Hill.
Campbell, J. I. D. (Ed.) (2005). Handbook of mathematical cognition. New York: Psychology
Press.
Campbell, J. I. D. (2008). Subtraction by addition. Memory & Cognition, 36, 1094–1102.
Campbell, J. I. D., & Xue, Q. (2001). Cognitive arithmetic across cultures. Journal of Experimental
Psychology: General, 130, 299–315.
De Corte, E., & Verschaffel, L. (1987). The effect of semantic structure on first graders’ solution
strategies of elementary addition and subtraction word problems. Journal for Research in
Mathematics Education, 18, 363–381.
Dedekind, R. (1963). Essays on the theory of numbers. Dover, U.K.: Kessinger.
De Smedt, B., Torbeyns, J., Stassens, N., Ghesquière, P., & Verschaffel, L. (2010). Frequency,
efficiency and flexibility of indirect addition in two learning environments. Learning and
Instruction, 20, 205–215. doi:10.1016/j.learninstruc.2009.02.020
Dowker, A. (2009). Use of derived fact strategies by children with mathematical difficulties.
Cognitive Development, 24, 401–410.
Fehr, T., Code, C., & Herrmann, M. (2007). Common brain regions underlying different arithmetic
operations as revealed by conjunct fMRI-BOLD activation. Brain Research, 1172, 93–102.
Fuson, K. C. (1992). Research on whole number addition and subtraction. In D. A. Grouws (Ed.),
Handbook of research on mathematics teaching and learning: A project of the National
Council of Teachers of Mathematics (pp. 243–275). New York: MacMillan.
Geary, D. C., Frensch, P. A., & Wiley, J. G. (1993). Simple and complex mental subtraction: Strategy
choice and speed-of-processing differences in younger and older adults. Psychology & Aging,
8, 242–256.
Heirdsfield, A. M., & Cooper, T. J. (2004). Factors affecting the process of proficient mental addition
and subtraction: Case studies of flexible and inflexible computers. Journal of Mathematical
Behaviour, 23, 443–463.
Hermans, D., Clarysse, J., Baeyens, F., & Spruyt, A. (2002). Affect (Versie 3.0) [computer software].
Leuven, Belgium: K.U. Leuven, Faculty of Psychology and Educational Sciences.
Ifrah, G. (1985). From one to zero: A universal history of numbers. New York: Viking Press.
Imbo, I., & LeFevre, J.-A. (2009). Cultural differences in complex addition: Efficient Chinese versus
adaptive Belgians and Canadians. Journal of Experimental Psychology: Learning, Memory
and Cognition, 35, 1465–1476.
Imbo, I., Vandierendonck, A., & De Rammelaere, S. (2007). The role of working memory in the
carry operation of mental arithmetic: Number and value of the carry. Quarterly Journal of
Experimental Psychology, 60, 708–731.
Imbo, I., Vandierendonck, A., & Vergauwe, E. (2007). The role of working memory in carrying
and borrowing. Psychological Research, 71, 467–483.
Kilpatrick, J., Swafford, J., & Findell, B. (2001). Adding it up. Helping children learn mathematics.
Washington, DC: National Academy Press.
Kirk, E. P., & Ashcraft, M. H. (2001). Telling stories: The perils and promise of using verbal
reports to study math strategies. Journal of Experimental Psychology: Learning, Memory,
LeFevre, J., DeStefano, D., Penner-Wilger, M., & Daley, K. E. (2006). Selection of procedures in
mental subtraction. Canadian Journal of Experimental Psychology, 60, 209–220.
Lemaire, P., & Lecacheur, M. (2010). Strategy switch costs in arithmetic problem solving. Memory
Lemaire, P., & Siegler, R. S. (1995). Four aspects of strategic change: Contributions to children’s
learning of multiplication. Journal of Experimental Psychology: General, 124, 83–97.
Littell, R. C., Milliken, G. A., Stroup, W. W., Wolfinger, R. D., & Schabenberger, O. (2006). SAS R
for mixed models, second edition. Cary, NC: SAS Institute Inc.
Luwel, K., Onghena, P., Torbeyns, J., Schillemans, V., & Verschaffel, L. (2010). Strengths
and weaknesses of the choice/no-choice method in research on strategy use. European
Psychologist, 14, 351–362.
Luwel, K., Schillemans, V., Onghena, P., & Verschaffel, L. (2009). Does switching between
strategies within the same task involve a cost? British Journal of Psychology, 100, 753–771.
Mainzer, K. (1995). Real numbers. In J. H. Ewing, H. Hermes, K. Lamotke, & H. L. Orde (Eds.),
Numbers (pp. 27–54). New York: Springer.
Peters, G., De Smedt, B., Torbeyns, J., Ghesquière, P., & Verschaffel, L. (2010a). Using addition to
solve large subtractions in the number domain up to 20. Acta Psychologica, 133, 163–169.
Peters, G., De Smedt, B., Torbeyns, J., Ghesquière, P., & Verschaffel, L. (2010b). Adults’ use of
subtraction by addition. Acta Psychologica, 135, 323–329.
Selter, C. (2001). Addition and subtraction of three-digit numbers: German elementary children’s
success, methods, and strategies. Educational Studies in Mathematics, 47, 145–173.
Seyler, D. J., Kirk, E. P., & Ashcraft, M. H. (2003). Elementary subtraction. Journal of Experimental
Psychology: Learning, Memory, & Cognition, 29, 1339–1352.
Shrager, J., & Siegler, R. S. (1998). SCADS: A model of children’s strategy choices and strategy
discoveries. Psychological Science, 9, 405–410.
Siegler, R. S. (2000). The rebirth of children’s learning. Child Development, 71, 26–35.
Siegler, R. S. (2003). Implications of cognitive science research for mathematics education. In J.
Kilpatrick, W. B. Martin, & D. E. Schifter (Eds.), A research companion to principles and
standards for school mathematics (pp. 219–233). Reston, VA: National Council of Teachers
of Mathematics.
Siegler, R. S., & Lemaire, P. (1997). Older and younger adults’ strategy choices in multiplication:
Testing predictions of ASCM using the choice/no-choice method. Journal of Experimental
Psychology: General, 126, 71–92.
Torbeyns, J., Ghesquière, P., & Verschaffel, L. (2009). Efficiency and flexibility of indirect addition
in the domain of multi-digit subtraction. Learning and Instruction, 19, 1–12.
Verschaffel, L., Greer, B., & De Corte, E. (2007). Whole number concepts and operations. In
F. Lester (Ed.), Handbook of research in mathematics teaching and learning (2nd ed.)
(pp. 557–628). New York: MacMillan.
Wechsler, D. (1981). WAIS-R. Wechsler Adult Intelligence Scale-Revised. San Antonio, TX: The
Psychological Corporation.
Received 2 July 2009; revised version received 8 December 2010

British J of Psychology - 2011 - Torbeyns - Use of Indirect Addition in Adults Mental Subtraction in The Number Domain Up

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

British J of Psychology - 2011 - Torbeyns - Use of Indirect Addition in Adults Mental Subtraction in The Number Domain Up

Uploaded by

Copyright:

Available Formats

20448295, 2011, 3, Downloaded from https://bpspsychub.onlinelibrary.wiley.com/doi/10.1111/j.2044-8295.2011.02019.x by The Open University, Wiley Online Library on [12/03/2023].

Use of indirect addition in adults’ mental

Joke Torbeyns1,2 ∗ , Bert De Smedt1 , Greet Peters1 ,

Indirect addition strategy use

The present study

selective assignments of strategies to items and/or participants impossible, and thus

Materials and design

Arithmetic sub-test score Age

Group Male Female M SD Range M SD

Lower 10 25 14.34 2.62 8–17 22.08 2

Strategy distribution in the choice condition

S⬎D S⬍D Total

Lower 78.00 32.79 61.14 40.13 69.57 37.36

Strategy efficiency in the no-choice conditions

Indirect addition Direct subtraction

S⬎D S⬍D S⬎D S⬍D

Lower 0.99 0.03 0.91 0.11 0.90 0.15 0.86 0.13

Indirect addition Direct subtraction

S⬎D S⬍D S⬎D S⬍D

Lower 6.27 2.91 12.77 5.16 15.13 7.15 16.62 9.29

Strategy adaptivity in the choice condition

in support of it. First, from a mathematical perspective, the additive operation is

Received 2 July 2009; revised version received 8 December 2010

You might also like