Professional Documents
Culture Documents
Cognitive Styles and Perfomance On Complex Tasks
Cognitive Styles and Perfomance On Complex Tasks
Cognitive Styles and Perfomance On Complex Tasks
a r t i c l e
i n f o
Article history:
Received 19 May 2014
Received in revised form 3 April 2015
Accepted 24 July 2015
Available online xxxx
Keywords:
Cognitive style
Assimilator
Explorer
Structured tasks
Motivation
a b s t r a c t
This study looks at the conditional relationship between the AssimilatorExplorer (AE) cognitive styles and performance on complex, structured tasks. We predicted that the achievement motive should moderate the style
performance relationship. Eighty-three participants with mean age 18.5 years completed a cognitive style test,
a measure of the achievement motive, and a uid intelligence test (control variable). The dependent variable
was scored on a remote association test (RAT) with items based on complex word pairs. Results showed that Explorers performed better when they scored higher on the achievement motive. Assimilators performed better
when they scored lower on the achievement motive. The idea of optimal motivation was supported by the pattern of interaction.
2015 Elsevier Inc. All rights reserved.
1. Introduction
The AssimilatorExplorer styles (AE styles) describe individual differences in cognitive strategy preferences. The AE styles can be seen as
an important moderator in research on creativity and problem solving
and the purpose of the present paper is to further investigate how
these styles may moderate problem solving performance. The theory
posits that the strategy preferences associated with the AE styles describe competence implications for different types of tasks along a structuredunstructured (high novelty) dimension (Kaufmann, 1979). More
recent positions argue that the task competence described by the AE
styles will be moderated by the achievement motive (Martinsen,
1994). However, while previous studies have emphasized performance
on complex, unstructured tasks, we presently focus on complex, structured tasks. This represents a further test of the interactive nature of
the AE styles.
The theory of the AssimilatorExplorer styles was proposed by
Kaufmann (1979, 1995) and later elaborated by Martinsen (1993,
1994, 1995a,b). It has been placed in the WholistAnalyst category of
style constructs (Riding & Rayner, 1998). Assimilators are more rulebound in problem solving behavior and inclined to interpret new events
in terms of existing knowledge. Explorers seek novelty, which manifests
as a search for new types of solutions and new ways of solving problems
even without external pressure to do so.
The AE style construct has been operationalized as a continuum
where higher scores describe Explorers and lower scores describe
Corresponding author at: Department for Leadership and Organizational Behavior,
Norwegian Business School, Nydalsveien 37, 0484 Oslo, Norway.
E-mail address: oyvind.martinsen@bi.no (.L. Martinsen).
http://dx.doi.org/10.1016/j.lindif.2015.07.013
1041-6080/ 2015 Elsevier Inc. All rights reserved.
Please cite this article as: Martinsen, .L., & Furnham, A., Cognitive styles and performance on complex, structured tasks, Learning and Individual
Differences (2015), http://dx.doi.org/10.1016/j.lindif.2015.07.013
.L. Martinsen, A. Furnham / Learning and Individual Differences xxx (2015) xxxxxx
control for intelligence to provide evidence for the validity of style constructs. We included a measure of uid intelligence, which has been
found to have a high loading on the G-factor (Undheim, 1981).
2. Method
2.1. Sample
Eighty-three students (54 males and 27 females) from a Norwegian
upper high school participated. Their mean age was 18.5 years. Upper
high school in Norway lasts three years and includes students between
the ages of 1617 and 1819 years.
2.2. Measures
2.2.1. AE inventory
The revised AE inventory (Kaufmann & Martinsen, 1992) was used.
The scale is continuous and Explorers have higher scores and Assimilators lower scores on the inventory. Each item has a ve-point response
scale and the present version of the inventory has 30 items. The inventory has been validated in previous studies (Martinsen & Diseth, 2011;
Martinsen & Kaufmann, 2000). Alpha for the AE inventory was .91 in
the present study.
Please cite this article as: Martinsen, .L., & Furnham, A., Cognitive styles and performance on complex, structured tasks, Learning and Individual
Differences (2015), http://dx.doi.org/10.1016/j.lindif.2015.07.013
.L. Martinsen, A. Furnham / Learning and Individual Differences xxx (2015) xxxxxx
Table 1
Correlations, means, and standard deviations between predictors, control variables, and product terms. For gender: males = 1 and females = 2.
1. Gender
2. Verbal analogies
3. Mf
4. Ms (C)
5. AE (C)
6. Ms (C) Ms (C)
7. Ms (C) AE (C)
8. RAT
Mean
Std
Dev
10.57
23.82
.00
.01
39.73
14.04
4.59
4.47
7.96
6.34
12.48
40.84
80.24
2.66
.189
.081
.043
.040
.010
.020
.281
.106
.103
.246
.024
.074
.154
.598
.192
.177
.019
.056
.183
.087
.105
.191
.084
.232
.155
.372
.292
.122
3. Results
As can be seen in Table 1, there are negative correlations between A
E, verbal analogies, and RAT, however, only the rst of these is signicant. The correlations between the AE styles and the two motives are
in the lower range and insignicant, but similar to the results of previous studies (Martinsen, 1994).
In order to test the posited (H1) interactive effect between the AE
styles and Ms on RAT, scores on AE and Ms were centered and multiplied according to the procedure outlined by Aiken and West (1991).
We used a hierarchical regression analysis, which emphasizes the signicance of the R square increment for the product term as evidence
for interaction. Two persons had missing data on one of the included
variables. In these analyses, the initial test of the Ms by AE interactive
effect was not signicant. The data were therefore investigated further
and revealed a signicant inverted U relationship between Ms and
RAT. Such a curvilinear effect can in itself support the hypothesis of optimal motivation for the task (Yerkes & Dodson, 1908), where complex
tasks are better solved with moderate motive strength.
When controlling for this curvilinear relationship in our further
analyses, the postulated joint effect between the AE styles and Ms
was clearly signicant and in support of our hypothesis. The results
can be seen in Table 2.
As shown in Table 2, the interactive effect of the AE styles and Ms
explains 4.2% of the variance and the full model explains nearly 20% of
the variance in the dependent variable.
When inserting our control variables for gender, verbal analogies,
and Mf rst in the equation, both the Ms Ms curvilinear effect and
the Ms AE interactive effects were still signicant. The increment
in R2, because of Ms AE, was now 5.5% (p = .016). The full model
(gender, verbal analogies, Mf, AE, Ms, Ms Ms, Ms AE) explained
33% of the variance (27% adjusted) in the RAT scores.
Table 2
Summary of hierarchical regression analysis to test the interaction between the AE styles
and the Motive to Approach Success (MS) on the Remote Associates Test (RAT). AE and
Ms were centered. N = 81.
Ms
AE
Ms * Ms
Ms * AE
Rsq
F
Rsq
F
Note:
p b .05.
p b .01.
Step 1
Step 2
Step 3
.167
.132
.191
.154
.322
.182
.110
.402
.227
.042
4.06
.198
4.74
.053
2.22
.102
9.44
.155
4.79
The pattern of the posited joint effect is shown in Fig. 1. Here it is evident that the shape of Ms is well tted to an inverted U at various levels
of the AE styles. It is also clear that Assimilators perform better at lower
levels of Ms and Explorers perform better at higher levels of Ms.
To further test the interaction, we did simple slope analyses. A significant interaction term in a two-way interaction design signies a difference between the slopes for a predictor at different levels of a
moderator. In the present case, this means that the slopes for Ms were
signicantly different at higher and lower levels of the AE strategy dispositions. When testing the simple slopes for the Ms at one standard deviation below the mean on the AE scale, following the recommended
procedure by Aiken and West (1991), the simple slope for Ms was negative and signicant (simple slope = .17, SE = 4.77, t (77) = 2.88, p =
.002 (one tailed)). The simple slope for Ms at one standard deviation
above the mean on the AE scale was, however, not signicant. When
including the control variables: verbal analogies, gender, and Mf, the
simple slopes for Ms were signicant at two standard deviations
above the mean (explorer strategy dispositions) on the AE inventory
(simple slope = .16, SE = .091, t (73) = 1.79, p = .037 (one tailed)).
The simple slopes for Ms were negative and signicant at one standard
deviation below the mean (assimilator strategy dispositions) on the AE
inventory. These results imply that arousal of Ms generally impaired
performance for Assimilators, while only high-scoring Explorers improved their performance because of strong Ms scores.
4. Discussion
The present theory posits that the performance on complex tasks
should be inuenced by the existence of search constraints (task structure), general heuristics associated with cognitive style, and motive
arousal. Within this framework our study was constructed to expand
upon previous ndings by Martinsen (1994), where insight tasks without salient search constraints were included. In light of the theory, those
tasks clearly favored participants with the Explorer style. In contrast to
these studies, the present study included adapted RAT tasks designed
to favor the heuristic competence of Assimilators through providing
cues in the problem situation and in the problem instructions.
As posited, the results showed a reversed pattern of interaction compared to what we have found in our previous studies and were in line
with the theory's prediction. Both Explorers and Assimilators tended
to perform well on the dependent tasks, but their success depended
on their scores on the motive to approach success. Consequently, our
idea of optimal motivation as dened by task competence and motive
strength in interplay was supported by the pattern of interaction. In
order to detect the predicted interaction, it was necessary to control
for a curvilinear relationship between Ms and task performance. This
inverted U relationship can clearly be attributed to mechanisms related
to optimal arousal dating back to Yerkes and Dodson (1908).
Please cite this article as: Martinsen, .L., & Furnham, A., Cognitive styles and performance on complex, structured tasks, Learning and Individual
Differences (2015), http://dx.doi.org/10.1016/j.lindif.2015.07.013
.L. Martinsen, A. Furnham / Learning and Individual Differences xxx (2015) xxxxxx
Fig. 1. Three-dimensional plot of the interaction between the AE styles and the Motive to
Approach Success (Ms). Scores on the vertical (z-) axis represent scores on the dependent
variable while scores on Ms are located on the x-axis and the scores on AE are located on
the y-axis.
4.1. Conclusion
The present study provided further evidence that cognitive style
may be a valid predictor of task performance but only in a complex interplay with other inuences.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions.
London: Sage.
Atkinson, J. W. (1974). Strength of motivation and efciency of performance. In J. W.
Atkinson, & J. O. Raynor (Eds.), Motivation and achievement (pp. 193218). New
York: Wiley.
Bowden, E. M., Jung-Beeman, M., Fleck, J., & Kounios, J. (2005). New approaches to
demystifying insight. Trends in Cognitive Science, 9, 322328.
Please cite this article as: Martinsen, .L., & Furnham, A., Cognitive styles and performance on complex, structured tasks, Learning and Individual
Differences (2015), http://dx.doi.org/10.1016/j.lindif.2015.07.013