Intellect Factor

Journal of Research in Personality 34, 198–207 (2000) doi:10.1006/jrpe.1999.2276, available online at http://www.idealibrary.com on

Fluid Intelligence, Crystallized Intelligence, and the Openness/ Intellect Factor Michael C. Ashton, Kibeom Lee, and Philip A. Vernon University of Western Ontario, London, Ontario, Canada

and Kerry L. Jang University of British Columbia, Vancouver, British Columbia, Canada We investigated the relations of the Big Five personality factor of Intellect or Openness to Experience with the crystallized and fluid aspects of measured intelligence. Approximately 500 participants completed the Personality Research Form (PRF) and the Multidimensional Aptitude Battery (MAB). An Openness/Intellect scale, defined as the sum of four PRF scales known to be markers of the Openness/ Intellect factor (Understanding, Sentience, Change, and Autonomy), was substantially correlated with MAB subtests that assess crystallized knowledge (the Verbal Scale subtests except Arithmetic), but only weakly correlated with MAB subtests that assess fluid ability (the Performance Scale subtests and Arithmetic). Among the fluid ability subtests, those containing pictures or meaningful visual stimuli were somewhat correlated with Openness/Intellect, whereas those containing numbers or abstract shapes were virtually uncorrelated with Openness/Intellect.  2000 Academic Press

Key Words: personality structure; Big Five; fluid intelligence; crystallized intelligence; intellect; openness to experience.

Personality structure is often summarized in terms of the ‘‘Big Five’’ factors (see Digman, 1990; Goldberg, 1993; McCrae & John, 1992). Four of these dimensions are known as Extraversion, Agreeableness, Conscientiousness, and Emotional Stability (versus Neuroticism), and the fifth is known as Openness to Experience or as Intellect. This study investigates the relations between the ‘‘fifth factor’’ of personality and different aspects of measured intelligence. Address correspondence and reprint requests to Michael C. Ashton, Department of Psychology, Brock University, St. Catharines, Ontario, L2S 3A1 Canada. E-mail: mashton@spartan. ac.brocku.ca. 198 0092-6566/00 $35.00 Copyright  2000 by Academic Press All rights of reproduction in any form reserved.

INTELLIGENCE AND THE FIFTH FACTOR

199

Conceptualizations of the Fifth Factor Costa and McCrae’s (1985, 1992a) Openness to Experience construct involves the tendency to fantasize (Fantasy), aesthetic sensitivity (Aesthetics), awareness of one’s emotions (Feelings), preference for novelty (Actions), intellectual curiosity (Ideas), and preference for nontraditional values (Values). The lexical Intellect factor involves artistic imagination, introspective reflection, and intellectual knowledge (Goldberg, 1994; Saucier, 1994) in addition to independence and nonconformity (Caprara & Perugini, 1994; De Raad, Hendriks, & Hofstee, 1992). Johnson (1994) and Saucier (1994) have noted that the aspects of Openness to Experience that define that factor most strongly—namely Aesthetics and Ideas—are also the aspects most strongly related to lexical Intellect. As a compromise between the two conceptualizations, we refer to the fifth personality factor as Openness/Intellect. Relations between Openness/Intellect and Measured Intelligence Previous research has indicated that Openness/Intellect—usually measured by the NEO Personality Inventory(-Revised) [NEO-PI(-R); Costa & McCrae, 1985, 1992a]—is correlated with intelligence test scores. For example, McCrae (1993–1994) reported a .33 correlation between NEO-PI-R Openness to Experience and the Full Scale IQ of the Wechsler Adult Intelligence Scale (WAIS), and Holland, Dollinger, Holland, and MacDonald (1995) reported a .42 correlation between NEO-PI Openness to Experience and the WAIS-Revised Full Scale IQ. Although the correlation between Openness/Intellect and measured intelligence is well established, an interesting question involves the relation of Openness/Intellect to different aspects of intelligence. For example, one might expect that tasks measuring vocabulary or general knowledge would be substantially correlated with personality traits such as intellectual curiosity, whereas tasks demanding reasoning or efficient information-processing would be nearly independent of ‘‘Typical Intellectual Engagement’’ (Goff & Ackerman, 1992). Goff and Ackerman (1992), using Cattell’s (1963) distinction between ‘‘crystallized’’ and ‘‘fluid’’ abilities, predicted that crystallized intelligence (measured by vocabulary-related tasks) would be more strongly related to NEO-PI-R Openness to Experience facets than would fluid intelligence (measured by reasoning tasks). The results supported those predictions: crystallized ability correlated .32 with Openness to Ideas, whereas fluid ability correlated only .13 with Openness to Ideas (Ackerman & Goff, 1994). If the Openness/Intellect factor is, in fact, more strongly correlated with crystallized than with fluid abilities, then one would expect that factor to correlate fairly strongly with the WAIS(-R) Verbal Scale subtests, which mainly demand verbal knowledge, but correlate very weakly with the Performance Scale subtests, which demand nonverbal reasoning, spatial visualiza-

200

ASHTON ET AL.

tion, and perceptual speed. However, in the studies by McCrae (1993–1994) and Holland et al. (1995) described above, NEO-PI(-R) Openness to Experience was correlated about equally with the WAIS(-R) Verbal Scale and Performance Scale subtests. Thus, previous research has yielded conflicting results regarding the relations between Openness/Intellect and the fluid and crystallized aspects of intelligence. The purpose of the present study was to investigate the relations between fluid intelligence, crystallized intelligence, and the Openness/Intellect personality factor. By virtue of the large sample size available in the present investigation, these data may provide a reliable indication of the size of those correlations. METHOD Participants Participants were sibling pairs from several Canadian cities who volunteered for the Western Ontario Twin Project (see, e.g., Harris, Vernon, & Jang, 1998; Vernon & Jang, 1993). Of the 516 participants, 508 completed all measures of interest to the present study. Seventy-four percent were women. Ages ranged from 13 to 45 years, with an average of 23. The sample included identical twins (36%), fraternal twins (25%), and nontwin siblings (38%). Because siblings, especially twins, tend to be somewhat similar in personality and mental abilities, results from this sample are somewhat less generalizable than would be the case if participants were unrelated. Nevertheless, the sample size is so large that its results are likely to be quite robust.

Measures of Personality and Intelligence Our study used the scales of the Personality Research Form (PRF; Jackson, 1984a) and the subtests of the Multidimensional Aptitude Battery (MAB; Jackson, 1984b) as markers of Openness/Intellect and of crystallized and fluid intelligence. The following four PRF scales, each 16 items in length, were used as markers of the Openness/Intellect personality factor in this study: Understanding, which measures intellectual curiosity; Sentience, which measures aesthetic sensitivity; Change, which measures variety seeking; and Autonomy, which measures independence. Several lines of evidence supported the selection of these scales as marker variables of the fifth factor. For example, Saucier (1999) found that the PRF scales having the highest correlations with a lexical Intellect factor were Autonomy (.49), Understanding (.48), Change (.39), and Sentience (.31). Moreover, NEO-PIR Openness to Experience has been found to correlate strongly with Sentience (.55), Understanding (.54), and Change (.40); those PRF scales correspond closely to the NEO-PI-R Openness facets of Aesthetics, Ideas, and Actions, respectively (Costa & McCrae, 1988, Table 2, p. 262). Costa and McCrae (1992b) have also noted the relevance of PRF Autonomy to such aspects of Openness as rebelliousness and nonconformity, which have represented the core of the fifth lexical factors obtained in Dutch (De Raad et al., 1992) and Italian (Caprara & Perugini, 1994). With regard to the domain of mental abilities, the MAB measures a set of constructs similar to those measured by the Wechsler tests. The MAB Verbal Scale contains several subtests— Vocabulary, Information, Similarities, and Comprehension—that mainly assess crystallized knowledge, although the latter two subtests also involve an element of judgment or reasoning. In contrast, the MAB Performance Scale subtests—Digit Symbol, Picture Completion, Spatial,

INTELLIGENCE AND THE FIFTH FACTOR

201

Picture Arrangement, and Object Assembly—mainly require fluid reasoning or information processing, although Picture Completion also demands some general knowledge. The MAB Arithmetic subtest, which contains written words and is part of the Verbal Scale, is a test of quantitative reasoning and computation that does not assess knowledge of mathematical techniques beyond basic arithmetic. We therefore used the MAB Verbal Scale subtests (minus Arithmetic) as measures of crystallized intelligence and the MAB Performance Scale subtests (plus Arithmetic) as measures of fluid intelligence.

Procedure Participants completed the MAB in a timed, supervised laboratory session. They completed the PRF and several other measures at home and then mailed those materials back to the researchers. Participants were paid $20 for their participation.

RESULTS AND DISCUSSION Descriptives The means (and standard deviations) for the participants’ Full Scale, Verbal Scale, and Performance Scale IQs were 110.0 (13.0), 106.7 (11.1), and 113.2 (14.5). Thus, compared with population values of 100 (15), the participants’ IQs tended to be above average, with somewhat restricted variability for the Verbal Scale. Participants’ scores on the PRF scales produced means and standard deviations close to normative values (Jackson, 1984a). Composite Scale Scores We calculated an overall, 64-item Openness/Intellect scale score by summing raw scores across the four PRF scales. We also calculated several composite scores for intelligence variables by adding standardized scores on the appropriate MAB subtests. Specifically, a crystallized intelligence composite was derived from the Verbal Scale subtests (excluding Arithmetic), and a fluid intelligence composite was derived from the Performance Scale subtests (plus Arithmetic). The crystallized and fluid composites intercorrelated .63. An overall intelligence composite was also created by adding standardized scores on all 10 MAB subtests. The composite scales were virtually identical to factor scores obtained from principal-axis factoring of the relevant variables. The Openness/Intellect composite correlated .99 with the first unrotated factor derived from the four PRF scales. The crystallized and fluid composites correlated .99 with two corresponding oblimin-rotated factors derived from the 10 MAB subtests, and the overall intelligence composite correlated .99 with the first unrotated MAB factor. After inspecting the PRF–MAB correlations (see below), we also calculated scores on two fluid intelligence subcomposites. We computed a ‘‘fluid– pictorial’’ composite as the sum of the Picture Completion, Picture Arrangement, and Object Assembly subtests, and a ‘‘fluid–nonpictorial’’ composite as the sum of the Arithmetic, Spatial, and Digit Symbol subtests. The two

202

ASHTON ET AL.

TABLE 1 Correlations between Multidimensional Aptitude Battery (MAB) Variables and Openness/ Intellect-Related Personality Research From (PRF) Variables PRF variables MAB variable

Total Understanding Sentience Change Autonomy Openness/Intellect

Individual subtests Information Vocabulary Similarities Comprehension Picture Completion Picture Arrangement Object Assembly Spatial Arithmetic Digit Symbol

.44* .46* .38* .32* .23* .19* .18* .11 .15* .03

.21* .17* .20* .15* .22* .13 .10 .02 ⫺.01 .02

Composite variables Crystallized Intelligence Fluid Intelligence Fluid–Pictorial Fluid–Nonpictorial Overall Intelligence

.46* .20* .25* .12 .35*

.21* .11 .18* .01 .17*

.09 .02 .15* .10 .10 .11 .05 .05 .00 .05 .10 .08 .10 .04 .10

.24* .22* .15* .17* .15* .09 .08 .06 .06 ⫺.04

.36* .32* .32* .27* .25* .19* .15* .09 .08 .02

.23* .09 .13 .03 .17*

.37* .18* .24* .08 .29*

Note. N ⫽ 508. *p ⬍ .001 (one-tailed). See text for variable descriptions.

fluid subcomposites intercorrelated .64. The fluid–pictorial and fluid–nonpictorial composites correlated .61 and .52, respectively, with the crystallized composite. Correlations of MAB Variables with PRF Openness/Intellect Variables Correlations of the MAB variables with the Openness/Intellect-related PRF scales are given in Table 1. Of the PRF scales, Understanding showed particularly strong relations with the crystallized MAB subtests, yielding a correlation of .46 with the crystallized intelligence composite. PRF Sentience and PRF Autonomy correlated .21 and .23, respectively, with crystallized intelligence, but PRF Change correlated only .10 with that variable. The Openness/Intellect composite scale correlated .37 with the crystallized intelligence composite. (The latter correlation is not entirely due to PRF Understanding; an Openness/Intellect composite derived as the sum of the other three scales still correlated .25 with crystallized intelligence.) The correlations of the PRF Openness/Intellect-related scales with the fluid MAB subtests were only about half as large as the correlations with the

203

INTELLIGENCE AND THE FIFTH FACTOR

TABLE 2 Correlations between Multidimensional Aptitude Battery (MAB) Composite Variables and Remaining Personality Research Form (PRF) Scales MAB composite variables PRF Scale Abasement Achievement Affiliation Aggression Cognitive Structure Defendence Desirability Dominance Endurance Exhibition Harmavoidance Impulsivity Nurturance Order Play Social Recognition Succorance

Crystallized Intelligence

Fluid Intelligence

Fluid (Pictorial)

Fluid (Nonpictorial)

Overall Intelligence

⫺.02 .16* ⫺.17* ⫺.01 ⫺.03 ⫺.12 .10 .19* .13 .05 ⫺.17* .00 ⫺.08 ⫺.14 ⫺.06 ⫺.04 ⫺.18*

⫺.09 .13 ⫺.04 .01 .07 ⫺.02 .11 .22* .15* .05 ⫺.19* ⫺.04 ⫺.06 ⫺.03 .00 .06 ⫺.10

⫺.08 .12 ⫺.05 .05 .06 ⫺.03 .13 .21* .15* .04 ⫺.21* ⫺.03 ⫺.04 ⫺.04 .00 .04 ⫺.11

⫺.09 .11 ⫺.03 ⫺.03 .07 .01 .06 .19* .12 .05 ⫺.13 ⫺.04 ⫺.08 ⫺.02 .00 .07 ⫺.08

⫺.06 .16* ⫺.11 .00 .03 ⫺.07 .11 .23* .15* .06 ⫺.20* ⫺.02 ⫺.08 ⫺.09 ⫺.03 .02 ⫺.15*

Note. N ⫽ 508. *p ⬍ .001 (one-tailed). See text for descriptions of MAB composite variables.

crystallized MAB subtests. This was also true of overall Openness/Intellect, which correlated only .18 with the fluid intelligence composite. One particularly interesting aspect of these results involved differences among the MAB fluid subtests in their correlations with the personality variables. The ‘‘pictorial’’ fluid subtests (Picture Completion, Picture Arrangement, and Object Assembly), which involve pictures of objects, living things, or social situations, were moderately correlated with the PRF Openness/Intellect-related scales. In contrast, the ‘‘nonpictorial’’ fluid subtests (Arithmetic, Spatial, and Digit Symbol), which involve the mental manipulation of numbers or abstract shapes, were virtually uncorrelated with the personality variables. The Openness/Intellect personality variables generally were also correlated with overall intelligence, but less strongly than with crystallized intelligence alone. Correlations of MAB Composite Variables with Remaining PRF Scales Table 2 shows the correlations of the MAB composite variables with the remaining PRF scales. A few PRF scales showed modest significant correlations with one or more of the intelligence composites, but there was no ten-

204

ASHTON ET AL.

dency for scales thought to define the same Big Five factor to show similar patterns of correlations. For example, various studies have suggested that Extraversion is defined by PRF scales such as Affiliation, Dominance, Exhibition, and Play (e.g., Ashton, Jackson, Helmes, & Paunonen, 1998; Costa & McCrae, 1988; Saucier, 1999). As seen in Table 2, Dominance correlated positively with the intelligence composites, but Exhibition and Play were uncorrelated with those variables, and Affiliation showed small negative correlations. Interestingly, two scales in Table 2 whose patterns of correlations were similar to those of the Openness/Intellect-related scales were Succorance and Affiliation, both of which showed significant negative correlations with the crystallized composite, but near-zero correlations with the fluid–nonpictorial composite. These same scales were found by Saucier (1999) to be negatively correlated with a lexical Intellect factor (⫺.31 and ⫺.29, respectively). This suggests that these scales’ correlations with the intelligence variables are attributable to negative secondary loadings on the Openness/Intellect factor. GENERAL DISCUSSION Relations with Previous Research The results of this study support the hypotheses of Goff and Ackerman (1992) regarding the relations among fluid intelligence, crystallized intelligence, and personality traits relevant to Openness/Intellect. In contrast, the moderately large correlations between Openness/Intellect-related traits and tasks assessing fluid ability, as reported by Holland et al. (1995) and McCrae (1993–1994), did not occur in the present study. One possible explanation for this difference is sampling error, given the small sample sizes (less than 90) in both the Holland et al. (1995) and McCrae (1993–1994) studies. Another explanation is that some WAIS Performance Scale subtests, such as Block Design and Object Assembly, involve an element of creativity (see Costa & McCrae, 1992b; Wechsler, 1958, p. 80) that may cause intellectually open individuals to become more highly engaged in those tasks. The corresponding MAB Performance Scale subtests, which are administered in multiple-choice format, may lack that aspect of creativity. The results of the present study also support McCrae’s (1987, 1993–1994) contention that Openness to Experience and intelligence are separate constructs. Even though measures of Openness/Intellect were correlated fairly strongly with measures of crystallized knowledge, those correlations—like the similar correlations of Openness/Intellect with creativity-related divergent thinking tasks (McCrae, 1987)—should be viewed as theoretically appropriate, given the plausible link between Openness/Intellect and the acquisition of vocabulary and general information.

INTELLIGENCE AND THE FIFTH FACTOR

205

Pictorial and Nonpictorial Fluid Ability Tasks In contrast to the substantial correlations between Openness/Intellect and crystallized knowledge, correlations between Openness/Intellect and some fluid ability tasks were very weak. Among the fluid subtests, correlations with Openness/Intellect were higher for the ‘‘pictorial’’ than for the ‘‘nonpictorial’’ subtests. The former correlations might reflect a tendency for intellectually open individuals to become especially absorbed in tasks that contain meaningful visual stimuli, and perhaps to have developed a specific talent for discovering meaning through their interests in reading, art, and science. The latter correlations suggest that Openness/Intellect is essentially orthogonal to the ability to process information of an abstract spatial or quantitative nature. Consequently, one might view overall measured intelligence as a compound of two independent and more fundamental traits (see also Gustafsson, 1984; Undheim, 1981): first, an abstract reasoning or information-processing ability (‘‘fluid g’’), which influences performance on all tasks, and second, the personality dimension of Openness/Intellect, which influences performance on crystallized and pictorial fluid tasks. PRF Understanding versus the Other Openness/Intellect-Related Scales The differences among the Openness/Intellect-related PRF scales, in terms of their correlations with the intelligence variables, also deserve mention. One particularly striking result was the especially strong correlation between PRF Understanding and crystallized ability; this was much higher than the correlations yielded by the other PRF scales, and higher also than the correlation yielded by the overall Openness/Intellect composite. One likely reason for the particularly strong correlations involving the PRF Understanding scale involves the specific item content of that scale: many Understanding items describe preferences for artistic, literary, and scientific activities, and such interests would naturally be expected to correlate with the general knowledge that is assessed by the crystallized intelligence subtests. In addition to trait-specific variance, however, another likely reason for the high correlations obtained by the PRF Understanding scale involves its particularly high loading on the Openness/Intellect factor. For example, Ashton et al. (1998) found the Understanding scale to load .74 on a factor that was highly correlated with lexical Intellect; no other PRF scale loaded as high as .60 on that factor. Similarly, Understanding was the only scale that was found by both Costa and McCrae (1988) and Saucier (1999) to correlate above .45 with markers of Openness/Intellect. Thus, the high correlations between PRF Understanding and crystallized intelligence are probably attributable partly to the specific item content of the Understanding scale, but partly also to the substantially greater saturation of that scale with Openness/ Intellect variance.

206

ASHTON ET AL.

Consistent with this interpretation, the other Openness/Intellect-related scales all showed similar patterns of correlations with the mental-ability variables. Moreover, the correlations of PRF Autonomy and PRF Sentience with crystallized intelligence were higher than those of any other PRF variable except Understanding, even though neither of those scales contains any items describing intellectual pursuits. This finding is difficult to explain except in terms of the moderate loadings of those scales on an Openness/Intellect factor. CONCLUSION The results of this study indicate that the personality factor of Intellect or Openness to Experience is substantially correlated with measures of crystallized intelligence, but rather weakly correlated with measures of fluid intelligence, particularly when the latter measures involve numbers or abstract figures rather than pictures or other meaningful visual stimuli. REFERENCES Ackerman, P. L., & Goff, M. (1994). Typical intellectual engagement and personality: Reply to Rocklin (1994). Journal of Educational Psychology, 86, 150–153. Ashton, M. C., Jackson, D. N., Helmes, E., & Paunonen, S. V. (1998). Joint factor analysis of the Personality Research Form and the Jackson Personality Inventory: Comparisons with the Big Five. Journal of Research in Personality, 32, 243–250. Caprara, G. V., & Perugini, M. (1994). Personality described by adjectives: The generalizability of the Big Five to the Italian lexical context. European Journal of Personality, 8, 357–369. Cattell, R. B. (1963). Theory of fluid and crystallized intelligence: A critical experiment. Journal of Educational Psychology, 54, 1–22. Costa, P. T., Jr., & McCrae, R. R. (1985). NEO Personality Inventory manual. Odessa, FL: Psychological Assessment Resources. Costa, P. T., Jr., & McCrae, R. R. (1988). From catalog to classification: Murray’s needs and the five-factor model. Journal of Personality and Social Psychology, 55, 258–265. Costa, P. T., Jr., & McCrae, R. R. (1992a). Revised NEO Personality Inventory and FiveFactor Inventory professional manual. Odessa, FL: Psychological Assessment Resources. Costa, P. T., Jr., & McCrae, R. R. (1992b). Four ways five factors are basic. Personality and Individual Differences, 13, 653–665. De Raad, B., Hendriks, A. A. J., & Hofstee, W. K. B. (1992). Towards a refined structure of personality traits. European Journal of Personality, 6, 301–319. Digman, J. M. (1990). Personality structure: Emergence of the five-factor model. Annual Review of Psychology, 41, 417–440. Goff, M., & Ackerman, P. L. (1992). Personality-intelligence relations: Assessing typical intellectual engagement. Journal of Educational Psychology, 84, 537–552. Goldberg, L. R. (1993). The structure of phenotypic personality traits. American Psychologist, 48, 26–34. Goldberg, L. R. (1994). Resolving a scientific embarrassment: A comment on the articles in this special issue. European Journal of Personality, 8, 351–356.

INTELLIGENCE AND THE FIFTH FACTOR

207

Gustafsson, J.-E. (1984). A unifying model for the structure of intellectual abilities. Intelligence, 8, 179–203. Harris, J. A., Vernon, P. A., & Jang, K. L. (1998). A multivariate genetic analysis of correlations between intelligence and personality. Developmental Neuropsychology, 14, 127– 142. Holland, D. C., Dollinger, S. J., Holland, C. J., & MacDonald, D. A. (1995). The relationship between psychometric intelligence and the five-factor model of personality in a rehabilitation sample. Journal of Clinical Psychology, 51, 79–88. Jackson, D. N. (1984a). Personality Research Form manual (3rd ed.). Port Huron, MI: Research Psychologists Press. Jackson, D. N. (1984b). Multidimensional Aptitude Battery manual. Port Huron, MI: Research Psychologists Press. Johnson, J. A. (1994). Clarification of Factor Five with the help of the AB5C model. European Journal of Personality, 8, 311–314. McCrae, R. R. (1987). Creativity, divergent thinking, and Openness to Experience. Journal of Personality and Social Psychology, 52, 1258–1265. McCrae, R. R. (1993–1994). Openness to Experience as a basic dimension of personality. Imagination, Cognition, and Personality, 13, 39–55. McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model and its applications. Journal of Personality, 60, 175–215. Saucier, G. (1994). Trapnell versus the lexical factor: More ado about nothing? European Journal of Personality, 8, 291–298. Saucier, G. (1999). A hierarchy of personality factors in the natural language: Integrating the Big Five, Big Three, and higher-order solutions. Unpublished manuscript, University of Oregon. Undheim, J. O. (1981). On intelligence II: A neo-Spearman model to replace Cattell’s theory of fluid and crystallized intelligence. Scandinavian Journal of Psychology, 22, 171–179. Vernon, P. A., & Jang, K. L. (1993). Self-rated vs. ‘‘actual’’ personality similarity in monozygotic and dizygotic twins and non-twin siblings. Personality and Individual Differences, 15, 219–220. Wechsler, D. (1958). The measurement and appraisal of adult intelligence (4th ed.). Baltimore: Williams & Wilkins.