Sex differences on the Progressive Matrices are influenced by sex differences on spatial ability
Abstract
The Progressive Matrices Test (PM) is considered an excellent measure of general intelligence or g. There are several studies showing that, on average, males score higher than females on the PM. However, some investigations have demonstrated that there are no sex differences in g. Why are there these contradictory findings? The visuo-spatial format of the PM could be favourable to the male performance, because (on average) males score higher than females on overall spatial ability. The present study administered the Advanced Progressive Matrices Test (APM) and the Spatial Rotation Test from the Primary Mental Abilities Battery (PMA) to a sample of 239 university undergraduates. Males outperformed females on both tests. However, the male advantage on the APM turned out to be non-significant when sex differences on spatial rotation were statistically controlled. It is suggested that sex differences on the PM could be a by-product of its visuo-spatial format.
Figures
Full-text (PDF)
Available from: Roberto Colom, Mar 25, 2014
Sex differences on the Progressive Matrices are influenced
by sex differences on spatial ability
Roberto Colom
*
, Sergio Escorial, Irene Rebollo
Facultad de Psicolog
ıa, Universidad Aut
onoma de Madrid, 28049 Madrid, Spain
Received 15 May 2003; received in revised form 29 October 2003; accepted 15 December 2003
Available online 22 January 2004
Abstract
The Progressive Matrices Test (PM) is considered an excellent measure of general intelligence or g. There
are several studies showing that, on average, males score higher than females on the PM. However, some
investigations have demonstrated that there are no sex differences in g. Why are there these contradictory
findings? The visuo-spatial format of the PM could be favourable to the male performance, because (on
average) males score higher than females on overall spatial ability. The present study administered the
Advanced Progressive Matrices Test (APM) and the Spatial Rotation Test from the Primary Mental
Abilities Battery (PMA) to a sample of 239 university undergraduates. Males outperformed females on
both tests. However, the male advantage on the APM turned out to be non-significant when sex differences
on spatial rotation were statistically controlled. It is suggested that sex differences on the PM could be a
by-product of its visuo-spatial format.
Ó2004 Elsevier Ltd. All rights reserved.
Keywords: Progressive matrices test; Spatial rotation; Sex differences; General intelligence
1. Introduction
The Progressive Matrices Test (PM) is an excellent measure of general intelligence or g
(Mackintosh, 1998; McLaurin, Jenkins, Farrar, & Rumore, 1973; Jensen, 1998; Paul, 1985). Lynn
and Irwing (submitted) have reported a meta-analysis of 87 studies showing that males outper-
form females on the PM. Their results disconfirm the frequent assertion that there are no sex
*
Corresponding author. Tel.: +34-91-397-4141; fax: +34-91-397-5215.
E-mail address: roberto.colom@uam.es (R. Colom).
0191-8869/$ - see front matter Ó2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.paid.2003.12.014
Personality and Individual Differences 37 (2004) 1289–1293
www.elsevier.com/locate/paid
differences on the PM. Given that this test is considered one of the best measures of general
intelligence or g, then it can be stated that, on average, males have a higher gthan females.
However, there are two findings that reject the conclusion that males have a higher gthan
females. First, Jensen (1998), Colom, Juan-Espinosa, Abad, and Garc
ıa (2000), as well as Colom,
Garc
ıa, Juan-Espinosa, and Abad (2002) have shown that there are no sex differences on g.
Second, Colom and Garc
ıa-L
opez (2002) demonstrated that sex differences on gestimated by a
single test are ‘‘contaminated’’ by the test format. These researchers found a difference favourable
to males in the Advanced Progressive Matrices Test (APM), but a difference favourable to females
in the Inductive Reasoning Test from the Primary Mental Abilities Battery (PMA-R). Both tests
are excellent measures of g, but differ in that the APM is constructed relying on a visuo-spatial
format, while the PMA-R is constructed relying on a verbal format. Furthermore, Colom and
Garc
ıa-L
opez (2002) failed to find a sex difference in what Jensen (1980) and Cattell (1980)
consider the best available measure of g, namely, the Culture Fair Intelligence Test.
The possibility that sex differences on the PM could be the result of its visuo-spatial format
should not be neglected. Several studies emphasized the visuo-spatial bias of the PM (Burke, 1958;
Hunt, 1974; DeShon, Chan, & Weissbein, 1995; Lim, 1994; Abad, Colom, Rebollo, & Escorial, in
press). This bias could favour male performance, because the PM could rely both on reasoning
and spatial abilities. Given that males have an average higher overall spatial ability than females
(Neisser et al., 1996), then the former could obtain higher scores on the PM relying on their higher
ability to process visuo-spatial information.
The present study administered the Advanced Progressive Matrices Test (APM) and the spatial
rotation test from the Primary Mental Abilities Battery (PMA-S) to a sample of university
undergraduates. Two related issues are investigated: (1) is there a sex difference on the APM? (2)
does the difference remain when sex differences on spatial ability are statistically controlled?
2. Method
2.1. Participants, measures, and procedures
239 psychology undergraduates were tested. 119 were females and 120 were males. The mean
age of females was 19.58 (SD ¼1.26, range ¼18–24). The mean age of males was 20.23
(SD ¼1.82, range ¼18–28).
The APM was administered with a time limit of 40 min. The PMA-S was administered
according to instructions given in the test manual. Both tests were collectively administered in
groups of no more than 30 people. The students received course credit for their participation.
Reliabilities were 0.75 for the APM and 0.92 for the PMA-S.
3. Results
Table 1 shows the descriptive statistics for males and females on the APM and the PMA-S. The
sex difference in dunits (the difference between the two means divided by the standard deviation)
and IQ points are also shown.
1290 R. Colom et al. / Personality and Individual Differences 37 (2004) 1289–1293
The Pearson rbetween the APM and the PMA-S for the combined sample was 0.34
ðp<0:001Þ, for males 0.35 ðp<0:001Þ, and for females 0.27 ðp<0:01Þ.
The statistical significance of the difference was tested by analysis of variance. Male scores were
statistically higher than those of females on the APM [Fð1;238Þ¼4:956, MC ¼92.961, p¼0:027]
and on the PMA-S [Fð1;238Þ¼19:295, MC ¼2085.098, p<0:001]. The sex difference on the
APM was equivalent to 4.3 IQ points, and on the PMA-S was equivalent to 8.5 IQ points.
The analysis was repeated for the APM, but controlling for sex differences on the PMA-S by
analysis of covariance. The sex variable was not significant [Fð1;238Þ¼0:733, MC ¼12.451,
p¼0:393]. Thus, sex differences on the APM turned out to be non-significant when sex differences
on spatial ability were statistically removed.
4. Discussion
The present study shows a male advantage on the APM equivalent to 4.3 IQ points. Colom and
Garc
ıa-L
opez (2002) tested 301 females and 303 males (both university undergraduates) on the
APM, finding a male advantage equivalent to 4.2 IQ points. Abad et al. (in press) tested 1069
males and 901 females (applicants to a Spanish University) on the APM, finding a male advantage
of 4.05 IQ points. Lynn and Irwing (submitted) found (among adults) a male advantage equiv-
alent to 5 IQ points on the PM. Those data are favourable to the conclusion that males out-
perform females on the PM, and, therefore, that the former have a higher gthan the latter.
However, Abad et al. (in press) have demonstrated (through DIF analysis) that several items
from the APM are biased against females. The bias derives from the visuo-spatial format of the
test. Presumably, malesÕhigher spatial ability contributes to increase their scores on the APM.
LimÕs (1994) study supports this statement. This researcher found that females (but not males)
showed a significant APM loading on a spatial factor defined by several spatial measures.
These pieces of evidence suggest that the evaluation of sex differences on gmust not be based on
a single test, as Colom and Garc
ıa-L
opez (2002) have noticed, because the results will be
inconclusive. They wrote: ‘‘researchers must be careful in selecting the markers-measures––of
central abilities like fluid intelligence, which is supposed to be the core of intelligent behaviour. A
‘‘gross’’ selection can lead to confusing results and misleading conclusions’’ (p. 450).
In order to avoid the use of a single test to measure g, Jensen (1998) proposed the method
of correlated vectors. Colom et al. (2000, 2002) applied this method to test for the presence
of sex differences on g. Their results strongly support the view that there are no sex differences
on g.
Table 1
Descriptive statistics for males and females on the APM and the PMA-Sa
Tests Males Females dIQ
Mean SD N Mean SD N
APM 24.57 4.52 120 23.32 4.13 119 0.29 4.3
PMA-S 29.02 11.51 120 23.11 9.14 119 0.57 8.5
a
The sex difference in dunits and IQ points are also shown.
R. Colom et al. / Personality and Individual Differences 37 (2004) 1289–1293 1291
There is increasing evidence showing that males and females do not differ in general intelligence
or g. Several procedures have demonstrated this beyond any reasonable doubt (see Fern
andez-
Ballesteros & Colom, in press). Nevertheless, there are some researchers still supporting the view
that there are sex differences in general intelligence (Lynn, 1999). The present article demonstrates
that there are reasons to think that a single measure is not an appropriate way to reach a strong
conclusion.
Colom, Contreras, Arend, Garc
ıa-Leal, and Santacreu (in press) have demonstrated that there
is a great difference between the superficial appearance of a test and its cognitive requirements (see
also Jensen, 1998). Those researchers assessed 1593 university graduates (794 females and 799
males) through two computerized tests, one thought to reflect verbal reasoning and one thought to
reflect dynamic spatial performance. Their results showed that males outperformed females on
both tests. However, sex differences on verbal reasoning turned out to be non-significant when sex
differences on dynamic spatial performance were statistically removed. The finding was inter-
preted from the demonstrated fact that the verbal reasoning test requires spatial processing
(Colom, Contreras, Arend, Botella, & Santacreu, 2002; Arend et al., 2003).
Summing up, the common practice of measuring sex differences on cognitive abilities from the
tests’superficial characteristics can lead to confusing and inconclusive results. The analysis of the
tests’cognitive requirements seems inescapable. If the resolution of a given test requires other
abilities in addition to the central one, then the comparison of males and females could be con-
fused by the secondary ability. The results of the present study tell this frequently forgotten story
once again.
References
Abad, F.J., Colom, R., Rebollo, I., & Escorial, S. (in press). Sex differential item functioning in the RavenÕs Advanced
Progressive Matrices Test: Evidence for bias. Personality and Individual Differences.
Arend, I., Colom, R., Botella, J., Contreras, Ma. J., Rubio, V., & Santacreu, J. (2003). Quantifying cognitive
complexity: Evidence from a reasoning task. Personality and Individual Differences, 35(3), 659–669.
Burke, H. R. (1958). RavenÕs progressive matrices: A review and critical evaluation. Journal of Genetic Psychology, 93,
199–228.
Cattell, R. B. (1980). They talk of some strict testing of us––Pish. Behavioral and Brain Sciences, 3, 336–337.
Colom, R., Contreras, Ma. J., Arend, I., Botella, J., & Santacreu, J. (2002). Series de tres t
erminos y modelos de
representaci
on [Three-term series and models of representation]. Psicothema, 14(3), 651–658.
Colom, R., Contreras, Ma.J., Arend, I., Garc
ıa-Leal, O., & Santacreu, J. (in press). Sex differences in verbal reasoning
are mediated by sex differences in spatial ability. The Psychological Record.
Colom, R., & Garc
ıa-L
opez, O. (2002). Sex differences in fluid intelligence among high-school graduates. Personality
and Individual Differences, 32(3), 445–451.
Colom, R., Garc
ıa, L. F., Juan-Espinosa, M., & Abad, F. J. (2002). Null sex differences in general intelligence: Evidence
from the WAIS-III. Spanish Journal of Psychology, 5(1), 29–35.
Colom, R., Juan-Espinosa, M., Abad, F., & Garc
ıa, L. F. (2000). Negligible sex differences in general intelligence.
Intelligence, 28(1), 57–68.
DeShon, R. P., Chan, D., & Weissbein, D. A. (1995). Verbal overshadowing effects on RavenÕs Advanced Progressive
Matrices: Evidence from multidimensional performance determinants. Intelligence, 21, 135–155.
Fern
andez-Ballesteros, R., & Colom, R. (in press). The psychology of human intelligence in Spain. In: R. J. Sternberg
(Ed.), International handbook of the psychology of human intelligence. Cambridge: Cambridge University Press.
1292 R. Colom et al. / Personality and Individual Differences 37 (2004) 1289–1293
Hunt, E. B. (1974). Quote the Raven? Nevermore!. In L. W. Gregg (Ed.), Knowledge and cognition. Hillsdale, NJ:
Erlbaum.
Jensen, A. R. (1980). AuthorÕs response. Pr
ecis of Bias in mental testing. Behavioral and Brain Sciences, 3, 359–368.
Jensen, A. R. (1998). The gfactor. New York: Praeger.
Lim, T. K. (1994). Gender-related differences in intelligence: Application of confirmatory factor analysis. Intelligence,
19, 179–192.
Lynn, R. (1999). Sex differences in intelligence and brain size: A developmental theory. Intelligence, 27, 1–12.
Lynn, R., & Irwing, P. (submitted). Sex differences on the progressive matrices: A meta-analysis.
Mackintosh, N. J. (1998). IQ and human intelligence. Oxford: Oxford University Press.
McLaurin, W., Jenkins, J., Farrar, W., & Rumore, M. (1973). Correlations of IQ on verbal and non-verbal tests of
intelligence. Psychological Reports, 22, 821–822.
Neisser, U., Boodoo, G., Bouchard, T. J., Boyking, W., Brody, N., Ceci, S., Halpern, D., Loehlin, J., Stenberg, R., &
Urbina, S. (1996). Intelligence: Knowns and unknowns. American Psychologist, 51(2), 77–101.
Paul, S. M. (1985). The Advanced RavenÕs Progressive Matrices: Normative data for an American University
population and examination of the relationship with SpearmanÕsg.Journal of Experimental Education, 54, 95–100.
R. Colom et al. / Personality and Individual Differences 37 (2004) 1289–1293 1293
- CitationsCitations27
- ReferencesReferences25
- "The two tasks differ from each other, however, by concerning different reasoning abilities. Whereas modular arithmetic represents a test of mathematical abilities, it has been suggested that performance at Raven's matrices depends on visual-spatial abilities besides representing general reasoning ability (Colom & Garcia-Lopez, 2002; Colom, Escorial, & Rebollo, 2004; Flynn, 1998 ). We assumed the activation of failurerelated concepts to influence performance on both tasks. "
[Show abstract] [Hide abstract] ABSTRACT: http://www.tandfonline.com/doi/full/10.1080/00220973.2015.1094649#.Vso1RCm8ky4 We investigated processes underlying performance decrements of highly test-anxious persons. Three experiments contrasted conditions that differed in the degree of activation of concepts related to failure. Participants memorized a list of words either containing words related to failure or containing no words related to failure in Experiment 1. In Experiment 2, all participants initially memorized words related to failure, but one experimental condition subsequently required the intentional forgetting of these words, thereby, decreasing the words’ activation in memory. The performance of highly test-anxious participants in unannounced cognitive ability tests following the memorizing phase profited from the activation of failure-related concepts. Experiment 3 replicated this finding by contrasting a condition priming failure–related concepts with a neutral control condition. The results point to a mechanism of suppressing worrying thoughts that is detrimental to test performances of highly test-anxious persons. Bringing the feared threat of failure to their awareness prevents that suppression.- "However, in PET studies highlighting the role of left striatal dopamine in working memory, it has been argued that the verbal nature of the administered tasks could explain left-lateralized results [Cools et al., 2008]. In our study, Gf is measured by tests with a high visuospatial load such as Raven's Advanced Progressive Matrices Test [Abad et al., 2004; Ackerman et al., 2002; Colom et al., 2004] , whereas Gv evidently pertains to the spatial domain . Therefore, the spatial load of the tests used here could account for our findings, given that visuospatial cognitive abilities are more likely to be associated with the right hemisphere [Jordan et al., 2002; Mottaghy, 2006; Suchan, 2008]. "
[Show abstract] [Hide abstract] ABSTRACT: Neuroimaging studies have revealed associations between intelligence and brain morphology. However, researchers have focused primarily on the anatomical features of the cerebral cortex, whereas subcortical structures, such as the basal ganglia (BG), have often been neglected despite extensive functional evidence on their relation with higher-order cognition. Here we performed shape analyses to understand how individual differences in BG local morphology account for variability in cognitive performance. Structural MRI was acquired in 104 young adults (45 men, 59 women, mean age = 19.83, SD = 1.64), and the outer surface of striatal structures (caudate, nucleus accumbens, and putamen), globus pallidus, and thalamus was estimated for each subject and hemisphere. Further, nine cognitive tests were used to measure fluid (Gf), crystallized (Gc), and spatial intelligence (Gv). Latent scores for these factors were computed by means of confirmatory factor analysis and regressed vertex-wise against subcortical shape (local displacements of vertex position), controlling for age, sex, and adjusted for brain size. Significant results (FDR < 5%) were found for Gf and Gv, but not Gc, for the right striatal structures and thalamus. The main results show a relative enlargement of the rostral putamen, which is functionally connected to the right dorsolateral prefrontal cortex and other intelligence-related prefrontal areas. Hum Brain Mapp, 2013. © 2013 Wiley Periodicals, Inc.- "c o m / l o c a t e / p a i d may primarily engage visuospatial processes. Several authors have indeed suggested that the small overall male advantage on Raven's Matrices is due to the (partially) visuospatial nature of the test (Colom, Escorial, & Rebollo, 2004; Flynn, 1998). Mackintosh and Bennett (2005) investigated this possibility by correlating participants' scores on a mental rotation test (MRT-A: Vandenberg & Kuse, 1978) with their scores on analytic (1) and analytic (2) items. "
[Show abstract] [Hide abstract] ABSTRACT: A high Autism-Spectrum Quotient (AQ) score (Baron-Cohen et al. in J Autism Dev Disord 31(1):5-17, 2001) is increasingly used as a proxy in empirical studies of perceptual mechanisms in autism. Several investigations have assessed perception in non-autistic people measured for AQ, claiming the same relationship exists between performance on perceptual tasks in high-AQ individuals as observed in autism. We question whether the similarity in performance by high-AQ individuals and autistics reflects the same underlying perceptual cause in the context of two visual search tasks administered to a large sample of typical individuals assessed for AQ. Our results indicate otherwise and that deploying the AQ as a proxy for autism introduces unsubstantiated assumptions about high-AQ individuals, the endophenotypes they express, and their relationship to Autism Spectrum Conditions (ASC) individuals.
People who read this publication also read
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.
This publication is from a journal that may support self archiving.
Learn more