Psychological Bulletin Copyright 1997 by the American Psychological Association, Inc.

1997, Vol. 121, No. 2, 219-245 0033-2909/97/$3.00

Intelligence, Personality, and Interests: Evidence for Overlapping Traits

Phillip L. Ackerman and Eric D. Heggestad

University of Minnesota, Twin Cities

The authors review the development of the modern paradigm for intelligence assessment and applica-
tion and consider the differentiation between intelligence-as-maximal performance and intelligence-
as-typical performance. They review theories of intelligence, personality, and interest as a means to
establish potential overlap. Consideration of intelligence-as-typical performance provides a basis for
evaluation of intelligence-personality and intelligence-interest relations. Evaluation of relations
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

among personality constructs, vocational interests, and intellectual abilities provides evidence for
This document is copyrighted by the American Psychological Association or one of its allied publishers.

communality across the domains of personality of J. L. Holland's (1959) model of vocational inter-
ests. The authors provide an extensive meta-analysis of personality-intellectual ability correlations,
and a review of interest-intellectual ability associations. They identify 4 trait complexes: social,
clerical/conventional, science/math, and intellectual/cultural.

In this article, we briefly review theoretical approaches to extensive review by Peterson, 1925). However, while there were
intellect, personality, and interests that make contact across these many earlier instances of psychologists making use of mental
seemingly disparate domains. We also review the empirical evi- tests (e.g., J. McK. Cattell, 1890; Gallon, 1883/1928), the be-
dence and theoretical arguments for an approach to adult intel- ginnings of the modern paradigm for intelligence testing can be
lect that goes beyond the traditional paradigm. The review first identified in two of Binet and Simon's (1905/1961, 1908/1961)
focuses on the description of the traditional paradigm for intelli- classic articles. They described a set of higher order mental tests
gence assessment of children. We next consider the extension that could be administered to children of various ages in an
of the paradigm to adult intellectual assessment. From this foun- effort to predict academic success or failure (e.g., see the
dation, we review a separation of the constructs of inteUigence- abridged translation reprinted by lenkins & Paterson in Binet &
as-maximal performance and intelligence-as-typical perfor- Simon, 1908/1961; also see the translation by Town in Binet &
mance. In an attempt to bridge the separation of maximal and Simon, 1911/1915).
typical performance, we review the literature on the commonal- The term paradigm is too eagerly applied and often overused
ity among personality constructs and intellectual abilities in in psychology, but the kind of mental testing proceduralized by
adults and provide a set of meta-analytic results. We also review Binet and Simon (1905/1961, 1908/1961) and their followers
the literature on the relations between interest constructs and readily meets many of the classificatory requirements for the
intellectual abilities in adults. existence of a scientific paradigm (Kuhn, 1970). Binet and
Simon described the type of procedures that allow assessment
Intelligence Testing as a Paradigm
of intelligence, specifically they distinguished between three
Assessment of Intelligence of Children different methods: medical, which focuses on physiology and
A comprehensive review of the early history of intelligence pathology; pedagogical, which determines intelligence on the
testing is beyond the scope of this article (although see an basis of the examinee's knowledge; and psychological, which
makes direct observations of intelligence (as translated by Kite
in Binet & Simon, 1908/1961).
Phillip L. Ackerman and Eric D. Heggestad, Department of Psychol-
Binet and Simon (1905/1961, 1908/1961) argued that they
ogy, University of Minnesota, Twin Cities.
wished to separate "natural intelligence from instruction" by
This research was partially supported by U.S. Air Force Office of
"disregarding, in so far as possible, the degree of instruction
Scientific Research Grant F49620-93-1-0206.
Phillip L. Ackerman would like to gratefully acknowledge early dis-
which the subject possesses" (as translated by Kite in Binet &
cussions and correspondence with J. B. Carroll, L. 3. Cronbach, C. Hert- Simon, 1908/1961, p. 93). That is, Binet and Simon attempted
zog, and F. L. Schmidt about some of the ideas contained in this article. to provide an estimate of individual differences in intellectual
We thank A. T. Church for providing raw correlations from his 1994 ability, which was, to a great degree, separated from influences
article, "Relating the Tellegen and Five-Factor Models of Personality of experience, social privilege, and other confounds of socioeco-
Structure.'' We also thank A. Tellegen for several productive discussions nomic status. Binet and Simon provided cogent arguments for
on the structure of personality and for his help in sorting personality the psychological method over the pedagogical method to assess
scales into trait classes. We also thank M. Goff for collecting and re-
school children's intelligence.
viewing many of the studies used in the personality-intelligence meta-
In an earlier article, Binet and Simon (1896, as cited by
analysis.
Correspondence concerning this article should be addressed to Phillip
Carroll, 1993) considered the use of a wide variety of tests,
L. Ackerman, Department of Psychology, University of Minnesota, N218 such as simple psychophysical procedures (like those of Gallon,
Elliott Hall, 75 East River Road, Minneapolis, Minnesota 55455. Elec- 1883/1928; and J. McK. Cattell, 1890), but ultimately rejected
tronic mail may be sent via Internet to [email protected]. those in favor of tests of higher mental processes, which had

219
 220 ACKERMAN AND HEGGESTAD

more substantial validity for predicting academic success (e.g. 1896) early work is that his studies focused exclusively on
following directions, memory, counting, etc.). For Binet and college students. Specifically, J. McK, Cattell and Farrand
Simon (1908/1961), intelligence was best assessed by a large
requested the Freshmen of the School of Arts and the School of
battery of tests that focused on attention, memory, thinking, and
Mines to come by appointment. About one-half of them came, and
other cognitive processes.
all were interested in the tests and agreed without hesitation to
Moreover, Bincl and Simon (1911/1915) described the spe-
repeal them at the end of the Sophomore and Senior years, (p. 624)
cific procedures for constructing the testing situation. Examina-
tions were to take place in "an isolated, quiet room" (as trans- They did not present any additional information about the condi-
lated by Town, p. 63). The examiner was instructed to tions of testing, but the clear indication from the description of
the study was that examinees were volunteers and that no special
meet the child pleasantly, do not stare at him when questioning
procedures were used to obtain maximal performance.
him; if he seems timid, constantly reassure him, not only by a
pleasant manner, but using one of the tests which seems most like
In contrast, in a classic investigation, Sharp (1899; in E. B.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

a game (pictures or making change). Encourage constantly in a Titchener's laboratory) only examined the abilities of "seven
This document is copyrighted by the American Psychological Association or one of its allied publishers.

kind way throughout the examination; show satisfaction with the advanced students in the Sage School of Philosophy . . . all of
answers whatever they are. Never criticise [src], and do not lose whom had had training in introspection" (p. 349), clearly a
time by making a lesson of i t . . .. Encouragement should be given group thai had motivation for maximal performance. In addition,
by the tone of voice, or by words wholly devoid of meaning which
serve only to stimulate: Go on! Quickly now! Hurry up! Good! certain of the tests which are especially adapted for collective study
Very good! Perfect! Marvelous! etc. (p. 63) were given not only to these groups, but also, by the aid of Prof.
Titchener, to the less advanced students taking the undergraduate
Elsewhere in the description of test procedures, Binet and (junior year) course in Experimental Psychology, (p. 350)
Simon (1911/1915) discussed means to obtain motivated re-
sponding from the examinees, for example: As in the J. McK. Cattell and Farrand (1896) study, the test
scores in Sharp's study had no direct impact on the students'
What should be done [if a child does not respond]? The help of academic status.
the teacher is often useful. If she is intelligent, she knows what to In another classic investigation of adult ability, Wissler
say to her children to reassure them and arouse their courage. A (1901) conducted ' 'yearly testing [of] sixty to seventy freshmen
caress to one, a reprimand to another, and all goes well. (p. 12) of Columbia College and repeated with those who remain to
the end of the senior year" (p. 4). In this case, though, the
In all these directions, Binet and Simon (1911/1915) made it participant sample was composed of students in a psychology
clear that the purpose of the intelligence test was to assess the course who were ' 'carefully instructed in the methods of proce-
child's maximal performance, that is, to estimate the child's dure and taken though the tests, both as participant and ob-
capabilities. server" (p. 43). The means toward the investigator eliciting
In summary, then, the elements of the modem intelligence motivation for the tests were only described as the following:
paradigm, put forth by Binet and Simon (1911/1915), were as ' 'At the beginning of the tests in the psychological laboratory
follows: (a) tests of higher order mental processes, (b) elimina- a few words are said to the student concerning the object of the
tion (as far as possible) from consideration of knowledge ac- tests and the value of the results" (p. 4).
quired through specific instruction, (c) elicitation of maximal Carothers (1922) reported that, as early as 1915, clear meth-
effort on the part of the examinee, and (d) school achievement ods were available to attempt to elicit maximal performance
as the fundamental criterion for external validation.
from college students who are given intelligence tests. For her
To get a sense of the paradigmatic nature of the Binet-Simon study of Barnard freshman, she sent
(1911/1915) approach toward eliciting maximal effort, one
need look no further than a review by Whipple (1914) on mental letters . . . to individual students in the class, reminding them of
and physical tests. In the chapter ' 'General Rules for the Con- the examination, and an account, written by Professor Hollingworth,
duct of Tests," Whipple stated, "in particular, attention may be of the widespread use of similar tests by reliable business firms
called here to ... the emotional attitude of the participants and their value in selecting candidates for positions along various
lines, appeared in the college weekly, (p. 18)
toward the investigation, their ability exactly to comprehend
what is wanted of them . . . and their willingness to do their
It is perhaps not at all surprising, then, that "out of a class of
best throughout the test" (p. 5, italics added).
about one hundred and forty freshmen during 1915-16, one
hundred were tested" (p. 18).
Assessment of Intelligence of Adults In 1915, Yerkes, Bridges, and Hardick described their devel-
opment of an adult mental ability test. Their description of the
1890s-1918
testing method was as follows:
Although Gallon (1883/1928) was probably the first investi-
He should then explain briefly what he is going to do, and what is
gator who performed large-scale testing of adults, the tests used
expected of the subject. For example, the examiner may say that
by Gallon (hearing, sensitivity, etc.) do not meet the higher he is going to ask some questions, and that the subject must try to
mental processes part of the modern conceptualization of intelli- answer them as well as he can; that some of the questions will be
gence. James McK. Cattell followed in Gallon's footsteps, but very easy and some more difficult; that the questions should be
one notable difference of Cattell's (J. McK. Cattell & Farrand, answered promptly, and that he should try to answer even those
 INTELLIGENCE, PERSONALITY, AND INTERESTS 221

that he is uncertain about, since a poor answer is better than none. from a "partial basis for admission" (p. 26) to vocational
(p. 139, italics added) counseling and creation of special sections of courses.
Thorndike (1920) reported on the development of the Thorn-
In the early history of intelligence testing, the creation and dike Intelligence Examination for High School Graduates, cre-
application of the Army Alpha Examination and, to some de- ated in 1919 for Columbia University, "where it is an optional
gree, the nonverbal Army Beta Examination stand out as the means of entrance for boys suitably recommended" (p. 330).
vehicles to fix the paradigm for adult ability assessment. In all, Of particular interest is Thorndike's somewhat flippant descrip-
1,700,000 men were tested with one of these two tests (about tion of the testing procedure:
15-20% of the male conscript-aged population of the United
States, according to a rough extrapolation of the 1910 and 1920 The administration of the examination consists simply of giving
Census figures). The following are the directions from the \oa- out the blanks, and instructing the candidates to go ahead at certain
kum and Yerkes (1920) manual: times even if they have not finished the work to date. In general, a
candidate does test after test without awaiting instructions, (p. 335)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

In giving the following directions E. [the examiner] should speak

As reported by Whipple (1922), "Professor Thorndike main-
rather slowly, distinctly, and with proper emphasis. He should expect
tains that his tests (which take about 3 hours to complete, in
and demand perfect order and prompt response to commands.
one sitting) show not only a man's intelligence, but also his
When everything is ready E. proceeds as follows: "Attention!" ability to stick to a long and, at the end, somewhat distasteful
The purpose of this examination is to see how well you can remem-
task" (p. 259). Also, "from another institution, it was reported
ber, think, and carry out what you are told to do . . .. %u are not
that two or three students fainted under the three-hour strain,
expected to make a perfect grade, but do the very best you can.
and the faculty became indignant at this alleged imposition of
Now, in the Army a man often has to listen to commands and then hardship" (p. 260). Whipple also reported that "many of the
carry them out exactly. I am going to give you some commands to students [at Michigan] are very keen to take mental tests; that
see how well you can carry them out. Listen closely. Ask no ques-
they are anxious to learn their standing" (p. 261). Undoubtedly,
tions, (pp. 53-54)
the paradigm for examiners attempting to elicit maximal perfor-
mance from examinees in a college environment had been fully
With such unambiguous instructions, there could be no doubt
established for adults by the early 1920s.
that the examiners intended for the examinees to provide their
maximal effort on the tests. However, the development and appli-
cation of the Army Alpha Examination did serve to broaden the Summary
standard intelligence test paradigm because the arm of the test As the modern intelligence testing paradigm formed around
was not to predict school achievement but rather to predict the Binet-Simon (1911/1915) approach and the adult exten-
occupational performance. sions during the first 2 decades of this century, aspects of the
testing content, procedures, and criteria were explored and for-
7979-7925 malized. Despite an explosive increase in the frequency of stud-
ies on intelligence assessment and application during this period,
After World War 1 (WW1), the Army Alpha Examination investigators quickly converged on specific intelligence assess-
and other intelligence tests (e.g., L. L. Thurstone's, 1919, Intel- ment content (i.e., higher order mental processes), procedures
ligence Test IV; and E. L. Thorndike's, 1920, 1921, College (i.e., encouragement for examinees to provide maximal effort),
Entrance Tests) became commonplace to test college students. and validational criteria (i.e., school and college grades and
Thurstone's exam had "been given to 6805 Engineering occupational performance).
freshmen, 5495 Liberal Arts Freshmen, and to 1575 Normal Finally, as Whipple (1922) asserted,
School Freshmen" (Pintner, 1923, p. 269). Similarly, Pintner
it ought to be made clear at the outset that no psychologist is foolish
reported that the Army Alpha Examination had been adminis-
enough to suppose that native intelligence is the sole factor in
tered to an aggregate total of 11,700 college students, at such
academic success; all that is contended is that it is one factor, and
locales as Brown University, Ohio State University, Syracuse
probably the most important single factor, and that it is measurable
University, University of Minnesota, University of North Dakota, by wholesale rapid methods with a reasonable degree of precision,
and so on. Toops (1926) conducted a survey of 110 colleges (p. 262)
and universities around the United States and showed how intel-
ligence testing of undergraduates had blossomed in just 6 years,
Maximal Versus Typical Performance
from 1918 to 1923-1924. Only 8 (7%) of 110 colleges and
universities were using intelligence tests in 1918, but 60% of Many early investigators identified the mental test procedure
the colleges were using intelligence tests and 12% were using as one designed to elicit maximal effort on the part of the
tests "experimentally" or "temporarily" as of the 1923-1924 examinee (e.g., Terman's, 1924, distinction between acute and
academic year. Even for the remaining 28% of the colleges and chronic behaviors). However, as Ackerman's (1994) review
universities in the survey, several were using tests for "students noted, Cronbach (1949) first explicitly categorized ability tests
in certain departments, but not to the entire student body" (p. as measures of maximal performance, in contrast to personality
26). Although no college or university surveyed reported using tests, which provide measures of typical performance. Subse-
the tests exclusively for selection purposes, the colleges and quently, Fiske and Butler (1963; Butler & Fiske, 1955) provided
universities reported that these tests were used for everything a formalized contrast and a rationale to match maximal and
 222 ACKERMAN AND HEGGESTAD

typical behavior with ability and personality measures, respec- ysis of a wide corpus of data in the literature.1 The structure of
tively. For ability testing, abilities is shown in Figure 1. It consists of one third-order
factor (General Intelligence), seven second-order factors (Fluid
first, we want a pure measure, one that is determined almost wholly Intelligence [Of], Crystallized Intelligence [Gc], Ideational
by one thing, the subject's capacity, rather than a measure which
Fluency, Knowledge and Achievement, Learning and Memory,
is affected by several influences. Second, we measure maximum
Perceptual Speed, and Visual Perception), and two first-order
performance because it is probably more stable than performance
factors (Math Reasoning and Closure). With the exception of
under more lifelike conditions. (Fiske & Butler, 1963, p. 253)
General Intelligence, each factor is presented with a list of com-
For personality testing, "we are ordinarily concerned with the monly found test types that load highly on that factor. The
typical (modal or mean?) strength of this tendency [to respond hierarchical arrangement indicates that the underlying factor
in a given way] because this provides the best estimate of what structure is oblique (often called "positive manifold"), with
a person is most likely to do" (Fiske & Butler, 1963, p. 258). correlations among narrower factors indicating the presence of
With this explicit treatment of two respective testing para- common broad factors. Note that, although Carroll listed many
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

more factors than we selected, our focus is the generation of a

This document is copyrighted by the American Psychological Association or one of its allied publishers.

digms, clearly the description of an ability testing paradigm

begs the question, "Why shouldn't investigators be interested structure and list of abilities that reflect those sampled in studies
in predicting, from ability testing, what a person is most likely of personality-intelligence and interest-intelligence relations.
to do?'' That is, an understanding of what an individual is likely
to do is based partly on personality characteristics but also on Personality Assessment and Theory
how much intellectual effort the individual is likely to put forth,
A review of the literature on personality assessment generally
whether in school or at a job—namely, the individual's typical
confirms Goldberg's (1971) observation that "the most potent
intellectual engagement (TIE). When considered in the context
source of variance in the determination of the [personality]
of an individual's typical behavior and in concert with investiga-
constructs for past scales and inventories has been sheer histori-
tors' informal observations about the circumstances when ability
cal accident'' (p. 335). That is, the literature is rife with isolated
test performance and achievement did not match, clearly person-
personality measures of varying levels of breadth, often with no
ality and intelligence are not necessarily orthogonal domains.
linkage to any personality theory. From this perspective, there
In addition, conative traits (e.g., interests) may also play a role
are competing orientations in the classification of personality
in the long-term development and expression of intellect, espe-
measures for later analysis: One goal is taxonomic complete-
cially as the individual reaches and continues to develop in
ness, the other goal is theory coherence. If one were to orient
adulthood. As such, we now consider the theoretical and empiri-
toward the first goal, then a complete list of narrow and broad
cal basis for personality—intelligence and interest—intelligence
personality scales could be collected for later analysis—how-
relations and the implications of these relations for development
ever, the comprehensiveness of this orientation would come at
and expression of adult intellect.
the cost of a failure to provide an organizing framework. In
contrast, one or more personality theories could be adopted, and
Intelligence Theory and Ability Structure personality scales then could be classified into theory—construct
A formal review of intelligence theory, as independent from categories. With this approach, though, a large number of mea-
theories that attempt to link intelligence with personality and sures would be excluded from later analysis because they do
interests, is beyond the scope of this article, and there are already not readily fit into the theoretical frameworks. We adopted a
excellent sources that provide this basic information (e.g., see "satisficing" solution, a hybrid approach: (1) Select a small
Carroll, 1993; and Steinberg, 1990). However, to conduct a number of prevalent factor-oriented theories of personality and
meta-analysis of personality and ability relations, it is necessary create a lowest common denominator list of personality con-
to classify constructs for each domain. Fortunately, there is a structs, (2) classify personality scales into the respective person-
consensus in the field about how abilities relate to one another ality constructs, and (3) organize as many of the remaining
and how the basic arrangement of cognitive abilities can be scales into logically coherent constructs, even if they exist out-
represented in a hierarchical structure. Although there are sev- side of the extant personality theories. A brief review of preva-
eral competing hierarchical theories of intelligence, most theo- lent personality theories follows.2
ries specify a general ability at the highest node (i.e., general At the outset, note that there is considerably less consensus
intelligence), followed by broad group factors at a second node about the structure of individual differences in personality than
and narrow group factors at lower nodes. Theories that fit this in the ability domain; controversy rages over many aspects of
kind of framework are ubiquitous, even if the precise description
of broad group factors differ between theories. For details of
' Carroll's review lists far more ability factors than we consider. We
these hierarchical theories of intelligence, see, for example,
have for the most part, though, limited our discussion to abilities that
R. B. Cattell (1971/1987); Gustafsson (1984); Horn (1965);
are both well researched and central to most competing hierarchical
Snow, Kyllonen, and Marshalek (1984); and Vernon (1950). It
theories of intelligence. A few additional factors were included in our
is also generally possible to roughly translate each of these
consideration to accommodate a significant number of cross-trait corre-
approaches to the others, depending on the type of spatial repre- lational studies.
sentation one desires (e.g., factor analytic or multidimensional 2
Note that our interests lie in (he nonpathological population, and, as
scaling). For our purposes, we adapt a structure and list of such, we have excluded from consideration theories of personality that
ability traits from Carroll's (1993) extensive review and reanal- deal primarily with psychopathology.
 INTELLIGENCE, PERSONALITY, AND INTERESTS 223

-- Closure
Math Reasoning
Closure Speed
/Quantitative Reasoning Visual ,,'"
Closure Flexibility
Math Problems Perception "
Visualization
Spatial Relations
Gf Closure Speed
(Fluid Closure Flexibility
Serial Perceptual Integration
Intelligence) Spatial Scanning
Sequential Reasoning Imagery
Inductive Reasoning
Quantitative Reasoning
Piagetian Reasoning Perceptual Speed
Number Computation
RT and other Elementary Cognitive Tasks
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Stroop
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Clerical Speed
Digit/Symbol

General
Intelligence Learning and Memory
Memory Span
Associative Memory
Free Recall Memory
Meaningful Memory
Visual Memory

Knowledge and Achievement

General School Achievement
Verbal Information and Knowledge
Information and Knowledge, Math and Science
Technical and Mechanical Knowledge
Ideational Knowledge of Behavioral Content
Gc Fluency
(Crystallized
Ideational Fluency
Intelligence) Naming Facility
Verbal Comprehension Expreisional Fluency
Lexical Knowledge Word Fluency
Reading Comprehension Creativity
Reading Speed Figural Fluency
"Cloze" Figural Flexibility
Spelling
Phonetic Coding
Grammatical Sensitivity
Foreign Language
Communication
Listening
Oral Production
Oral Style
Writing

Figure 1. A list and structure of ability constructs. Derived from information in Carroll (1993). Third-
order construct = General Intelligence; second-order constructs shown with solid lines; first-order construct
shown with dotted lines.

the research and theory development enterprises (see Block, McCrae's work), and the lowest order of Tellegen's (1982) 11-
1995a, 1995b; Church, 1994; Church & Burke, 1994; Costa & trait framework.
McCrae, 1992a, 1992b, 1995; H. J. Eysenck, 1992; Goldberg &
Saucier, 1995; and Tellegen & Waller, in press, for some exam-
ples). Nonetheless, for our task, we consider three broad theo- Eysenck's Theory
retical perspectives that have received substantial competitive
support in the literature and, at least across the theories, have H. J. Eysenck's (1947) original theory only included two
provided both breadth and a reasonable amount of specificity, factors of personality, namely Neuroticism and Extroversion.
so we can consider relations of both broad and relatively narrow Later Eysenck (1970) added the Psychoticism factor. His theory
personality traits with ability and interest traits. The three per- is set apart from other factor-based theories because it makes
sonality theories or models are H. J. Eysenck's three-factor the- several direct connections to physiological processes (e.g.,
ory, the Five-Factor approach (FFA; exemplified by Costa & arousal) and because he claimed that these personality factors
 224 ACKERMAN AND HEGGESTAD

are orthogonal to intellectual ability factors (e.g., see H. J. with three higher order factors of Negative Emotionality
Eysenck, 1994; and H. J. Eysenck & Eysenck, 1969). (NEM), Positive Emotionality (PEM), and Constraint. Telle-
gen's inventory is designed to measure the higher order traits;
Five-Factor Approach however, it also provides estimates of 11 lower order traits that
have been shown to have substantial commonality with traits
Although there were earlier attempts to simplify the factor identified by H.I. Eysenck (1947), with traits identified by
space of personality (e.g., Tupes & Christal, 1961), theoretical adherents of the FFA, and others. The lowest level traits identi-
and empirical research conducted during the 1970s and 1980s fied by Tellegen are Well-Being, Social Potency, Achievement,
eventually prompted several investigators to converge on a set of Social Closeness, Stress Reaction, Alienation, Aggression, Con-
five factors that appear across numerous personality inventories trol, Harm—Avoidance, Traditionalism, and Absorption. In gen-
(e.g., Goldberg, 1990; see also Digman, 1990, for a succinct eral, there is a moderate-to-close correspondence between the
review of the broad approaches to personality traits that con- Tellegen higher order traits and Eysenck's measures: NEM «
verge on five common factors). The five factors are broadly Neuroticism, Constraint « Psychoticism, and PEM « Extrover-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

classified (although the terms have slightly different meanings) sion (see Tellegen & Waller, in press).
This document is copyrighted by the American Psychological Association or one of its allied publishers.

as (1) Extroversion, (2) Agreeableness, (3) Conscientiousness,

(4) Neuroticism, and (5) Intellect. Relevant to our interests,
Digman reported that H. J. Eysenck's (1947) Psychoticism An Integrative Framework
straddles the Agreeableness and Conscientiousness factors and
that R. B. Cattell's (1946) Intelligence factor is aligned with Drawing on the literature that relates the Tellegen, H. J.
the Intellect factor and that the Extroversion factor subsumes Eysenck, and Five-Factor approaches, we are able to depict
H. J. Eysenck's Extroversion factor and R. B. Cattell's Exvia the prominent relations among each set of constructs (e.g., see
factor. McCrae (1994; McCrae & Costa, in press) has argued Church, 1994; and Church & Burke, 1994). In Figure 2, we
that there is substantial overlap between his conceptualization present a specification of these interrelations among personality
of Openness to Experience (Intellect in Digman's taxonomy), trait measures in a way tha: includes all three perspectives. The
Conscientiousness, and intellectual ability, amorphously defined connections between the Tellegen scales or factors and the FFA
(see, e.g., McCrae, 1994, Figure 1, p. 255). are derived from three sources: correlations between Tellegen's
(1982) Multidimensional Personality Questionnaire (MPQ) and
the Costa and McCrea's (1992c) NEO-Personality Inventory
Tellegen 's Framework
(NEO-PI, N = 575 college students; Church, 1994), between
Tellegen's theory of personality traits is essentially hierarchi- the MPQ and FFA trait descriptors (N = 1,015 community
cal (e.g., see Church, 1994; and Tellegen & Waller, in press). adults; Tellegen & Waller, in press), and between the MPQ and

Eysenck factors and

Tellegen factors

Five-factor approach

Neuroticism Openness

Stress Well- Alienation Aggression Social Closeness Control Achievement Traditionalism Absorption Harm-
Reaction Being and Avoidance
Social Potency

Tellegen lowest order traits

Figure 2. Personality constructs and their relations: lowest order constructs (from Tellegen. 1982); five-
factor approach (FFA) constructs (from Costa & McCrae, 1992c; Digman, 1990; and others); highest order
constructs (from H. J. Eysenck, 1970; and Tellegen & Waller, in press). Lines indicate both positive and
negative correlational (not necessarily causal) relations. Solid lines indicate relations between Tellegen and
FFA constructs. Dotted lines indicate relations between Tellegen and FFA constructs and H. J. Eysenck
constructs. NEM = Negative Emotionality; PEM = Positive Emotionality
 INTELLIGENCE, PERSONALITY, AND INTERESTS 225

the NEO-Five-Factor Inventory (a short form of the NEO-PI, tests and verbal tests to several groups of adolescents, for a total
N — 150 college students; Kanfer, Ackerman, & Heggestad, of 374 participants. In addition to the demonstration of a general
1996). Associations between the FFA and the Eysenck factors intelligence factor (g), a verbal factor (v), and a practical intel-
are derived from the data and discussions reported by Costa and ligence factor (F)—which had high loadings from the perfor-
McCrae (1992a, 1992b) and H. J. Eysenck (1991). Associa- mance tests and presumably represented spatial/mechanical
tions between Tellegen's higher order factors and Eysenck's abilities—Alexander found two "residual" factors underlying
factors derived from Tellegen and Waller's (in press) study of school grades. The first factor was identified as X (similar to
the MPQ and the Eysenck Personality Questionnaire (EPQ; N Webb's, 1915, W character factor) or rather an "'interest' in
= 155 college men and women) show the following loadings school work" (Alexander, 1935, p. 126) factor, also defined as
of the EPQ: Extroversion on PEM (.78), Neuroticism on NEM "persistence" and "will to succeed" (p. 128). The second
(.69), and Psychoticism on Constraint (.50). By providing this factor, Z, was related to "Shop Work, Mathematics, the number
framework, we are able to triangulate our later discussion of tests, and English" (p. 128) and was only described as "a
both specific and broad personality traits by locating the trait factor of some importance in school achievement" (p. 128).
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

in Figure 2 and examining the other traits with which it shows The first major review of the personality-intelligence domain
This document is copyrighted by the American Psychological Association or one of its allied publishers.

substantial communality. was reported by Lorge (1940). Unfortunately, the only thing
that one can take away from this short review (actual studies
Personality and Intellectual Abilities reviewed and specific coefficients were not presented) is the
confusion about personality traits and the lack of precision with
Intelligence and Character which they were measured during the first half of this century.
Lorge reviewed "some 200 coefficients between intelligence
1900-1940
and some scale of personality function" (p. 277). He reported
One of the first studies of the associations between personality that
variables and intelligence was reported by Karl Pearson (1906—
the correlations of intelligence with measures of personality range
1907). He reported the correlations between teacher judgments
from +.79 to -.49 with a median at +.04 . . .. Half the correla-
of intelligence and "mental characters" (Temper, Popularity,
tions, on the basis of absolute size, range between .00 and .15 and
Self-Consciousness, Shyness, Conscientiousness, and Quiet
only one fourth of them are greater than .30. (pp. 277-278)
Habits) for a sample of 5,000 school-aged boys and girls. The
results are of only passing interest; teacher judgments of both Although he did break out some correlations by specific per-
sets of traits undoubtedly were subject to halo effects. Nonethe- sonality scale (e.g., psychoneurotic tendency and Introversion-
less, note that Pearson reported correlations with rated intelli- Extroversion), again no regular correlational patterns could be
gence: Conscientiousness, .45; Popularity, .26; Temper, .21; Shy- detected. The only personality measures that showed regular
ness, .10, Self-Consciousness, .07; and Quiet Habits, .06. positive correlations with intelligence were tests of "moral"
Another early investigation of the relations between character judgment or knowledge; but as Lorge (1940) reported, the
and intelligence was by Webb (1915). For Webb, character knowledge aspect of these scales probably resulted in spuriously
included "emotional and volitional, social and moral qualities" high associations with the knowledge components of intelli-
(p. 2). As with Pearson's (1906-1907) earlier investigation, gence. Nonetheless, Lorge.concluded that "the correlations be-
character and intelligence assessments were provided by teacher tween intelligence and measures of personality may underesti-
ratings. In fact, the high correlations between judged intelligence mate the role of intellect in personality" (p. 281).
and judged character caused Webb to note what would later be
called halo effects, namely,
1940-I960
it throws light upon the nature of estimates of "General Intelli-
First, in an American Psychological Association (APA) pre-
gence," as supplied by teachers and others. These have been shown
sentation (1940), and later in an APA Division of Clinical and
to be not pure measures of intelligence proper (as the "g" is), hut
to be biased in various manners and in varying degrees for different Abnormal Psychology address, published in the American Psy-
judges, in favour of individuals who possess other desirable (or chologist (1950), Wechsler attempted to broaden the consider-
criterial) qualities besides the actual intelligence in question. ation of intelligence beyond the kinds of scales that were in-
(Webb, 1915, p. 75) cluded in standardized tests, that is, to encompass "cognitive,
conative, and non-intellective intelligence" (p. 82). Although
Nonetheless, Webb (1915) described several components of Wechsler was firm in his beliefs about the importance of person-
character (such as cheerfulness, depression, aesthetic feeling, ality and interest components for intelligence, the data available
esteem, etc.) and a general factor, W: to him were admittedly ' 'meager' and confusing. He cited the
work by Webb (1915) and Alexander (1935) and Lorge's
Its nature is best conceived, in light of our present evidence, to be
(1940) review. However, at the time of his article, there was
in some close relation to "persistence of motives"; i.e., to depend
upon the consistency of action resulting from deliberate volition, no general consensus agreement about the important factors of
i.e., from will. It appears to coincide more with Ach's conception personality or interests. As a result, it was not possible to docu-
of will than with either "perseveration" or the "secondary func- ment specific relations among these conative, affective, and cog-
tion." (p. 76) nitive/intellectual traits. He closed with the following:

Alexander (1935) administered a large series of performance One need not be afraid or ashamed to acknowledge impulse, instinct
 226 ACKpRMAN AND HEGGESTAD

and temperament as basic factors in general intelligence . . ..My am quite a fast reader," "I like science," and "I like to read
main point has been that general intelligence cannot be equated about history"); many of the other questions (mainly from the
with intellectual ability, but must be regarded as a manifestation of MMPI) address nonintellectual issues, such as general health
the personality as a whole. (Wechsler, 1950, p. 83)
or pathology (e.g., "I have had no difficulty in starting or hold-
ing my urine" and "My skin seems to be unusually sensitive
Up to the end of the 1930s, little coherence among theory to touch'').
and measures of personality existed, thus making a useful sum- Nonetheless, Cough's (1957, 1987) CPI includes the scale
mary of personality-intelligence relations impossible. However, of Intellectual-Efficiency (le), along with two related scales,
in the 1940s, several multiple trait theories and assessment bat- Achievement via Conformance (Ac) and Achievement via Inde-
teries were developed, and several of those batteries continue to pendence (Ai). A person who is high in le is described as
be used today. Prominent among personality measures is the "efficient in use of intellectual abilities; can keep on a task
Minnesota Multiphasic Personality Inventory (MMPI), al- where others might get bored or discouraged" (Gough, 1987,
though its focus on the clinical population accords little that p. 7). A person who is high in Ac is described as having a
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

can be considered directly relevant to general personality- "strong drive to do well, likes to work in settings where tasks
This document is copyrighted by the American Psychological Association or one of its allied publishers.

intelligence relations in nonpathological samples. and expectations are clearly defined" (p. 7). A person who is
There were, however, three broad approaches that provide high in Ai is described as having a "strong drive to do well;
grounds for investigation of personality-intelligence relations, likes to work in settings that are vague, poorly defined, and
namely those of R. B. Cattell, H. J. Eysenck, and the aggregated lacking in clear-cut methods and standards" (p. 7).
FFA perspective. In addition, these and other approaches
spawned measures developed for the nonpathological population Intellectence
(e.g., the California Psychological Inventory [CPI]) to provide
assessments of broad and narrow personality factors. A body of work by Welsh (e.g., 1975) focuses on a construct
R. B. Cattell's theory of personality traits has a distinction called "Intellectence," which he defined as "the personality
to provide a plethora of factors, along with original terms for dimension related to performance on intellectual measures such
most of the traits. In his first major effort to classify personality as the CMT [Terman's Concept Mastery Test]" (p. 69). Welsh
traits, R. B. Cattell (1946) introduced 12 factors. Later, Cattell converged on this construct (and a contrasting construct, "Ori-
(1949) refined the original factor descriptions and introduced gence") through analysis of personality measures (e.g., the Ad-
four additional factors, for a total of 16 "source traits." Rather jective Check List and MMPI) and interest measures (e.g.,
than an introduction of all of Cattell's terms, it is useful to Strong Vocational Interest Blank [SVIB]) in his investigations
mention two factors that, from their descriptions, may be related of creativity. He described individuals high in Intellectence as
to intellectual abilities (Institute for Personality and Ability Test-
somewhat introversive . . . more objective in outlook and responds
ing staff, 1986): Factor B, concrete thinking versus abstract
to people in the world around him and to their attitudes and ideas,
thinking, and Factor I, tough minded versus tender minded (Har-
although he tends to maintain some social and personal distance
ria vs. Premsia).
from them. Most of his responses are intellectualized or rationalized
Clearly, R. B. Cattell's Factor B is seen to be related to and he seldom acts impulsively, (p. 105)
intellectual ability. Unfortunately, the questions on the 16 Per-
sonality Factor Questionnaire (16PF) that pertain to Factor B However, although Welsh's construct has much in common with
are objective ability questions (e.g., analogies and vocabulary) similarly named personality constructs (e.g., R. B. Cattell's In-
rather than self-report, personality-type questions. As such, Cat- telligence personality factor, Gough's le measure, the FFA
tell provided no independent personality evaluation of intellect. Openness to Experience/Culture factor, Hogan and Hogan's
In contrast, R. B. Cattell's Factor I has the potential to exam- [1995] Intelligence factor, etc.), his research is impossible to
ine personality-intelligence relations. For example, Premsia is interpret in our context for several reasons; the most prominent
described by Horn (1965) as "sensitivity," with a negative load- among them is that assessment of Intellectence was often con-
ing of Mechanical Knowledge ability (i.e., "the 'sensitivity' is taminated by ability test scores (e.g., use of the CMT as a
that of a person who prefers English to mathematics, 'imagina- measure of Intellectence) or because Intellectence scores were
tive novels' to 'realistic accounts of military or political battles,' confounded by their being derived from difference scores be-
brandy to beer, ballet to burlesque, etc." [p. 294]). This trait, tween other standard measures of personality and interest traits
though, appears to share much content with the FFA construct (see, e.g., Welsh, 1971, 1975, and 1977).
of Openness to Experience/Culture.
Gough (1953) introduced the term "intellectual efficiency"
TIE
to describe a 52-item scale developed for the purpose of provid-
ing a non-ability-test assessment of intellect (with a criterion The construct of intelligence-as-typical performance was put
of IQ test scores). Gough reported that the questions for this forth by Ackerman (1994) to develop a parallel ability construc-
scale were selected "on a priori and theoretical grounds, to tion to the maximal versus typical distinction that Cronbach
covary with intellect" (p. 246). Although the correlations re- (1949) established for ability versus personality measures. That
ported by Gough were relatively high (median validity coeffi- is, Ackerman suggested that one reason why intelligence tests
cient of .47 in cross-validation samples), the content of the scale do not highly correlate with measures of advanced academic or
appears somewhat muddy from a construct-validity perspective. occupational performance is because intelligence is measured by
Some of the items could be clearly related to intellect (e.g., "I the maximal paradigm and long-term academic and occupational
 INTELLIGENCE, PERSONALITY, AND INTERESTS 227

performance takes place in a typical environment. He postulated 1972) anxiety (Mean r = .53), and transitory ("A state") anxi-
that a measure of intelligence-as-typical performance would be ety (Mean r = .45), as well as components of Worry (Mean r
more highly associated with crystallized abilities (i.e. resource- = .57) and Emotionality (Mean r = .54). For further details on
insensitive abilities such as knowledge), whereas intelligence- Test Anxiety traits, see, for example, Liebert and Morris (1967)
as-maximal performance would be more highly related to fluid and Spielberger. Given that general traits of Anxiety, Worry, and
abilities (i.e., more resource-dependent abilities such as memory Emotionality would generally be encompassed within Stress
or abstract reasoning). Reaction or Neuroticism domains (see, e.g., Church, 1994), one
To test the inferences about the construct of intelligence-as- might be tempted to simply identify Test Anxiety as a subfactor
typical performance, Goff and Ackerman (1992) created a self- of these broader traits. Indeed, Spielberger, Anton, and Bedell
report measure of TIE, essentially a personality scale (with items (1976) stated that Test Anxiety is a component of Trait Anxiety.
such as "I would enjoy hearing the details about discoveries in However, Hembree's meta-analysis also makes it clear that Test
any field" and "I read a great deal"), and correlated scores on Anxiety is malleable to a nontrivial degree (mean treatment
the scale with measures of ability, academic achievement, and effects as large as 1 SD improvement were found), thus sug-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

personality. The results of the first study, with 138 undergraduate gesting that Test Anxiety is less stable than the personality traits
This document is copyrighted by the American Psychological Association or one of its allied publishers.

students, validated the TIE measure—it correlated positively with which it correlates. In addition, it has been suggested (A.
and significantly with a Gc composite and correlated essentially Tellegen, personal communication, September 11, 1995) that
zero with a Gf composite. In comparison with personality mea- Test Anxiety may not be a personality trait per se but rather a
sures, TIE shows a substantial correlation with the Openness to set of attitudes toward test taking, which are highly influenced
Experience factor of the NEO-PI (r = .65) and a smaller, but by feedback from prior ability-testing situations. Because of the
significant, correlation with Conscientiousness (r = .27). Later uncertainty of the status of Test Anxiety as a personality trait,
investigations (e.g., see Ackerman & Goff, 1994) included anal- when we consider Test Anxiety-ability relations, we keep this
yses with additional participants (N — 455) and provided an construct separate from the other personality traits.
essentially identical correlation between TIE and Openness to
Experience (r — .65; see discussion by Rocklin, 1994).
Summary
The TIE measure has been administered to multiple samples,
along with several intelligence-aptitude batteries. For example, The theoretical and empirical literature on personality and
in a study of cognitive and noncognitive determinants and conse- intelligence reflect two general approaches: One approach infers
quences of complex skill acquisition, Ackerman, Kanfer, and broad (but unspecified) personality-intelligence relations, the
Goff (1995) administered the TIE scale and 15 ability tests to other specific personality-intelligence relations. The former ap-
93 university students and graduates. Correlations between TIE proach (such as that of Webb, 1915; and Wechsler, 1940, 1950)
scores and four composites were as follows: Verbal Ability (.49), provides a rationale for investigators searching broadly for per-
Perceptual Speed (.21), Math Ability (.12), and Spatial Ability sonality-intelligence relations. The latter approach involves a
(.06). In a field study of skill acquisition (Ackerman & Kanfer, small set of personality traits that historically have been linked
1994) among 213 U.S. Navy trainees, the TIE measure was explicitly to intellectual abilities (e.g., Openness to Experience,
administered and compared with Armed Services Vocational Intellectence, and Test Anxiety) and provides a rationale for
Aptitude Battery (ASVAB) scores. Correlations between TIE closer investigation of specific trait relations. In addition to
and ASXSU3 scales follow, with the scales identified by their the general approaches, the TIE approach is unique because it
factor membership (see Ackerman, 1988): Verbal Ability factor specifies that this personality trait shows larger correlations with
(Paragraph Comprehension [.33], General Science [.31], and particular intellectual abilities (e.g., Gc, fluency, and knowledge)
Word Knowledge [.29]); Vocational-Technical Information fac- than other abilities (e.g., Gf, Visual Perception, and Perceptual
tor (Mechanical Comprehension [.22], Electronics Information Speed). With this background, we performed a series of meta-
[.18], and Auto and Shop Information [-.07]); Math factor analyses to examine broad and narrow personality-intelligence
(Arithmetic Reasoning [.24] and Mathematics Knowledge relations.
[.22]); and Perceptual Speed (Coding Speed [.13] and Numerical
Operations [.04]). Along with the earlier results from Goff and
Meta-Analysis of Personality and Intelligence
Ackerman (1992), these results indicate that the TIE scale is
more closely associated with verbal/crystallized abilities than
Method
math/fluid abilities.
Literature Search
Test Anxiety
Studies for possible inclusion in the meta-analysis were initially iden-
tified by searching several computerized databases. Broadly denned
In a thorough review and meta-analysis of 562 studies,
searches were performed using the Educational Resources Information
Hembree (1988) examined the "correlates, causes, effects, and
Center's ERIC on CD-ROM (1968-1994), American Psychological As-
treatment of Test Anxiety'' (p. 47). From his review, Hembree
sociation's on-line FirstSearch (1962-1973), and the National Technical
determined that, first of all, significant negative correlations Information Service (NTIS) CD-ROM Database (1974-1994), on the
were found between Test Anxiety and measures of intelligence terms personality, intelligence, ability, and abilities (with the last three
(Mean r = —.23). Second, scales of Test Anxiety were found to terms conjoined by or statements). A more comprehensive search was
be substantially and significantly related to measures of General conducted using PsycLJT (1974-1994), in which all pairwise combina-
Anxiety (Mean r = .56), chronic ("A trait"; see Spielberger, tions of 47 personality terms and 15 ability terms were searched. The
 228 ACKERMAN AND HEGGESTAD

personality search terms consisted of trait constructs (e.g., impulsivity miscellaneous category). Thus, 191 ability scales were available for the
and introversion) and. personality battery names (e.g., 16PF), whereas meta-analysis. A complete listing of the ability scales and classifications
the ability search terms included ability constructs (e.g., spatial and is available from this article's authors.
numerical) and names or abbreviations of ability and aptitude tests (e.g.,
differential aptitude test [DAT] and scholastic aptitude test [SAT]). In
addition to the computerized searches, additional studies were identified
Classification of the Personality Scales
by (a) an examination of titles (aided by the topical index) included in The taxonomy of personality used for our sort starts with the 11
the 6,736-item Bibliography on Human Intelligence (Wright, 1969), (b) factors from A. Tellegen's MPQ, the 5 factors from the FFA, and the 3
an examination of promising studies found in the reference lists of those factors (Psychoticism, Neuroticism, and Extroversion) of H. J. Eysenck
studies identified in the computerized searches, (c) an examination of (although Extroversion identified in the FFA was deemed reasonably
personality and ability battery manuals, (d) an ERIC bibliography on equivalent to Eysenck's Extroversion—see Figure 2—and Stress Reac-
Test Anxiety compiled by Hembree (1988), and (e) a computer search tion and Neuroticism were equivalent across the construct space). Also,
of titles from Phillip L. Ackerman's files (1,500+ items). three additional factors were included, namely, Intellectence, Test Anxi-
Each article identified in these search procedures was subjected to an ety, and TIE. One miscellaneous personality trait category was added
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

initial screening in which the abstract (for those articles identified in the
This document is copyrighted by the American Psychological Association or one of its allied publishers.

to capture any scales that were not classifiable within this list of the 11
computerized searches), or article (for those articles which we did not + 5 + 3 + 3 (—3 for the overlapping Stress Reaction-Neuroticism and
have abstracts), was examined for indications dial both ability and per- Extroversion factors), for a total of 19 personality traits. A description
sonality constructs were assessed and reported. When the initial screen- of each of die unique 268 personality scales was obtained from manuals
ing suggested the possibility of useful data, the entire article was ob- (whenever available) or the study that used the scale. We and A. Tellegen
tained for a more detailed evaluation. A total of 1,073 articles was individually sorted cards that contained each scale description into the
obtained for further review (for additional details, see Ackerman & Goff, 19 personality trait groups and the miscellaneous category—again with
1995). the provision that the scales were to be placed in the lowest level trait
Studies were excluded from the meta-analysis if (a) participants were category appropriate to the scale. Given the limited amount of informa-
drawn from clinical or artificially or extremely restricted populations; tion available on many scales, interrater agreement was generally accept-
(b) participants were younger than 10 years of age; (c) an extreme groups able, ranging from 61% (Cohen's K = .57) to 79% (Cohen's K = .77).
design was used; (d) the only measures of personality reported were Scales were assigned to categories on the basis of the agreement of two
performance-type assessments (e.g., Tdata by R. B. Cattell, 1946, due raters. In the event that all three raters disagreed (24 scales, 10%), we
to method variance shared by such assessments and many intellectual placed scales into a category on the basis of a discussion. The number
ability tests), state measures (e.g., affect), cognitive styles (e.g., field of scales within each category varied from 32 (Stress Reaction) to 1
dependence), attitudes (e.g., political extremism), developmental assess- (TIE). The titles and descriptions of 90 scales were sufficiently uninfor-
ments (e.g., ego or moral development), or measures of psychopathology; mative or fell outside of the 19 taxonomic traits and were thus excluded.
or (e) the only measures of ability reported were those of creativity (due Each scale was carefully examined to ensure the direction of scoring.
to generally poor consensus on just how creativity can be assessed and Correlations from those scales scored in a direction opposite of our
the generally poor reliability of creativity measures; McNemar, 1964; categories (e.g., a scale with high scores representing Introversion rather
Weisberg, 1986) or psychomotor abilities. Finally, studies of masculin- than Extroversion) were reversed prior to inclusion in the analysis. A
ity-femininity were excluded because a prior meta-analysis has ad- total of 178 scales remained for the meta-analysis. A complete listing
dressed relations between masculinity-femininity and intellectual abili- of personality scales and classifications is available from this article's
ties (Signorella & Jamison, 1986) and because a consideration of gender authors. With a complete crossing of 10 ability traits and 19 personality
differences in this analysis would create an unworkable number of traits, there were a possible 190 meta-analyses to examine.
comparisons.3
One-hundred-thirty-five studies remained after evaluation with the
inclusion criteria. These studies contained a total of 188 independent
Procedure
samples, 2,033 correlations, and 64,592 participants. The studies that
Missing data. In some studies (e.g., Johnson, Nagoshi, Ahern, &
contributed correlations to the meta-analysis, along with the number of Wilson, 1983), investigators chose to report only significant correlations,
independent samples greater than one contributed (in parentheses) by
although an entire matrix of correlations was apparently computed.
each study, are indicated by an asterisk in the Reference section. Whereas several methods exist for dealing with such cases (see Bush-
man, 1994; and Bushman & Wang, 1995, for alternative procedures),

Classification of the Ability Tests

3
In a meta-analysis, Signorella and Jamison examined relations
The hierarchical taxonomy of abilities derived from Carroll's (1993) among the personality or self-concept constructs of masculinity-femi-
review, as shown in Figure 1, was used to sort ability scales into ability ninity, masculine and feminine aspects of androgyny measures, and an-
trait categories. Thus, scales were sorted into 10 categories and 1 miscel- drogyny as a balance between these scores and some mental rotation,
laneous ability category (for scales that did not fit within the taxon- spatial visualization, mathematical ability, and verbal ability assessments
omy)—with the provision that a scale would be sorted into the lowest similar to the abilities included in our meta-analysis. For spatial abilities,
level possible. Each trait was prepared on a card (along witii the list they found some suggestion that higher masculinity-femininity or mas-
of marker ability tests from Carroll). We placed the 245 ability scale culinity scores were associated with better performance, particularly for
descriptions on cards and then individually sorted them into the 11 women. However, for spatial abilities, effect sizes were substantively
categories. Interrater agreement (between Phillip L. Ackerman and Eric small, ranging from near 0% to 4% common variance. They found
D. Heggested) was quite high (96%, Cohen's K = .95). A few disagree- statistically significant although substantively modest effect sizes, ac-
ments were discussed and resolved, with occasional additional descrip- counting for approximately 1-2% common variance between masculin-
tive information obtained from the Mental Measurements Yearbook (e.g., ity-femininity scores and mathematical tasks. There were essentially
Conoley & Impara, 1995). Of the 245 scales sorted, 54 were either not negligible correlations that involved any of the personality variables and
classifiable or did not fit into the taxonomy (so were placed in the verbal abilities or androgyny and any of the abilities.
 INTELLIGENCE, PERSONALITY, AND INTERESTS 229

we decided to provide an estimate of the missing correlations; that is, where 1.96 is the two-tailed critical z value for a 95% CI, N is the total
when the correlation was reported as nonsignificant and the sign of the sample size, and k is the number of studies. The lower and upper bounds
correlation was unknown, the expected value, zero, was used as the of the estimated population correlation are obtained through a z-to-r
estimate.4 transformation of zL(owei) to pn^, and Zu^pen to pepper)-
Correction for attenuation due to unreliability of measures. The Test of heterogeneity. Assessments of heterogeneity of correlation
standard correction of unreliability of measures was used; coefficients were made within each personality-ability pairing, after
within-study aggregation, following the procedure of Hedges and Olkin
(1985, pp. 234-235),
(1)

G = I («. ~ (5)
where r* is the correlation between personality variable x and ability
variable y for study /, rw and ryy are the estimated reliabilities of personal-
ity variable x and ability variable y, respectively, and r^ is the observed where /i, is the sample size from study i, zf* is the Fisher's z-transformed
correlation between those variables. Reliability estimates came from a aggregated correlation from study i, and zis the weighted average corre-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

lation within the personality-ability cell. The Q statistic is distributed

This document is copyrighted by the American Psychological Association or one of its allied publishers.

variety of sources. Test-retest reliability estimates provided in test man-

uals or articles that describe the development of the scales were consid- as x2 with k — \ degree of freedom (where k represents the number of
ered most desirable and were used wherever possible. When test-retest zr-)- A significant Q suggests that more than one distribution may under-
reliability estimates were not available, however, tests of internal consis- lie the sample of correlations (i.e., heterogeneity) and that the investiga-
tency from the same sources were used. In a few cases, when these tor may not want to pool the data into a single estimate. Because of its
sources could not be located, or when no reliability estimates were relation to the chi-square distribution, however, the Q statistic is sensitive
reported, an attempt was made to locate reliability estimates reported in to sample size. Thus, with large enough samples, even small variations
other research studies that used these measures. Through these methods, among the sample correlations lead to significant heterogeneity statistics.
reliability estimates were available for 133 of the 178 personality vari- If samples are large, Hedges and Olkin suggested that it may be accept-
ables and 117 of the 191 ability measures. For those cases in which able to examine the pooled estimate, even if the Q statistic is significant.
reliability estimates remained unavailable, the mean value of the reliabil- Because many of the cells in our analyses contained samples with large
ity estimates that had been obtained was used (separately for personality sample sizes, the alpha level was set to .01 for the heterogeneity tests.
and ability).
Aggregation of within-sample effect sizes. Several studies used mul-
tiple measures of personality or ability that fell within a single personal- Results
ity-ability classification (e.g., Clarke, 1985). Rather than allowing a
The obtained results of the meta-analysis are shown in Table
single sample to contribute more than one correlation to a personality-
ability classification (thereby ignoring the nonindependence of the rela- 1. There are two aspects of the table that clearly stand out. First,
tions) or simply choosing one of the correlations at random (thereby for many cells, there are either no data or only a few studies
losing the information contained in the other relations), we chose to use that provide data. Second, many of the estimated population
an aggregated estimate by computing the mean correlation (using Fish- correlations are small, even if statistically significant.
er's z-transformed correlations). A mean estimate of reliability was also Further examination of the table shows that there are several
computed and used for correcting the estimated correlation for trait pairings that have a substantial accumulation of studies and
unreliability. samples. For example, Stress Reaction and Extroversion are
Correlational analysis. An analysis similar to the type described by
each represented by as many as 66 studies (Stress Reaction and
Hedges and Olkin (1985) was performed. Because individual differences
Gc) or 63 studies (Extroversion and Gc), with combined samples
constructs are under consideration, nearly all studies reported associa-
of 19,820 and 24,280 participants, respectively. In addition,
tions as correlations instead of other statistics. The effect size measure
for this analysis is therefore the weighted mean correlation after correc- many of the estimated population correlations are significantly
tion of attenuation due to unreliability of measures: different from zero. However, close inspection indicates that
many trait-pair cells also indicate significant heterogeneity of
correlations, suggesting that more than one population correla-
tion may underlie the data, which implicates the presence of
(2)
moderator variables. These issues are discussed in more detail
(«i - 3)
in Specific Findings.

where zr* is the Fisher's z-transformed correlation between personality

4
variable x and ability variable y for study (' after a correction for attenua- Out of 2,033, 185 (9%) of the correlations were "missing," and
tion due to unreliability of measures, n, is the sample size for that their expected value of zero was used in the analysis. The missing
correlation, and k is the number of aggregated correlations in the cell. correlations came from seven samples in three studies, were distributed
Fisher's z-transfonned correlations were used for aggregation to elimi- over 32 cells, and included eight different abilities categories and nine
nate bias that systematically underestimates effect size. The estimated different personality categories. The impact of these missing values on
population correlation (p) is obtained through a z-to-r transformation the conclusions of the analysis is trivial. That is, when these missing
off. values were removed, the estimates of rho were unchanged in 16 of the
Confidence intervals. Confidence intervals (CIs) for $ were calcu- 32 cells and changed by more than .02 or more in only 6 of the affected
lated using formulas by Hedges and Olkin (1985, p. 227): cells (the largest change was .05). The estimates that changed by .02
or more did not occur consistently in any one personality or ability
L = Z- (1.96AW - 3k) (3) grouping. Furthermore, the removal of the missing data did not change
any of the conclusions on the significance associated with the estimated
- 3k), (4) values of rho.
 230 ACKERMAN AND HEGOESTAD

"w
1 u
•sB
II
z

c
o. s
3 "
g|J $$ s -3 =lss
° '
"jj>

&
o
-22 S a
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

U
This document is copyrighted by the American Psychological Association or one of its allied publishers.

I
1'f • 8 |

'S
c 'S 1
1_<
8

C/) £
1
*^t 8
33 1
"Q •o
c u
c
-2, j?t ^
•a
G |s a
J2 S3
c
1 &

i
Ic 3
Achieveme
knowledge i

"3 Mb 1
CO

c £

£
I
IB-
o e
1 il
J IE

fr
,§
1^
S^ •-* u

1 It ua

"3s
o S O "^ ^s• -
ra u
2 2 u i n —' • ,00

(X,
c <o
5 oo 8•
C 13
^J
S
1
^5
K
•2

p ' .Z 0
B
6 _ "3
U o
6« i
B

.Z U Ol'5 <o.Z U CD ! 0 Ol <csi.Z U CX C Ot O Ot «xZ D Ol

 This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

I
neral ry
ence lEi

8
ii

'-s
s s |
s s
'-a

H H
O1'|.'0.2O Ol g '0.2 U O i
£22
'-S
INTELLIGENCE, PERSONALITY, AND INTERESTS

'-a
&

-. 8 X
il™

'SCra
^ll
ill
til
w ts c
is I
231

a ^s

a| .
.g»|

111
 232 ACKERMAN AND HEGGESTAD

General Findings Similarly, for Extroversion measures, although significantly pos-

itively correlated with the ability traits, the correlations are
Although it is not generally appropriate to aggregate the ta- small, even at the 75th-percentile level. Achievement, however,
hled results across columns (abilities), given the a priori hetero- does show significant positive correlation for Intelligence (G),
geneity of constructs, it is possible to gain insight into the Gf, and Gc, and the 75th-percentile data indicate correlations
general effects found in the meta-analysis by one examining larger than .20 for both G and Gc.
these correlations with a qualitative perspective. With this tack The other frequency distributions are more complex. Al-
in mind, some consistencies readily appear. Personality traits though the mean estimated population correlations for Tradition-
that tend to be positively correlated across ability traits include alism are nonsignificant for G and Gf, the significant heterogene-
Weil-Being, Social Potency, Achievement, Social Closeness, In- ity statistics and the 25th-percentile values suggest that some
tellectence, TIE, Extroversion, and Openness to Experience. combinations of measures and samples may more reliably yield
From a taxonomic view, the positively associated traits fall into negative correlations between Traditionalism and G, and Gf.
two categories: (a) traits associated with the Extroversion cate- Moreover, because Traditionalism bears a lower to higher order
gory from Eysenck and the FFA or with the PEM category from
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

(negative) relationship to Openness to Experience, the relations

This document is copyrighted by the American Psychological Association or one of its allied publishers.

the hierarchical perspective of Tellegen and (b) traits associated of abilities to Traditionalism provide additional support of the
with a broad class of intellectually oriented traits (TIE, Open- presence of relations between broad intellectually oriented per-
ness to Experience, and Intellectence)—see Figure 2 for an sonality traits and broad ability classes.
illustration of the hierarchical relations among these traits. Per- Two other traits, Openness to Experience and TIE, have many
sonality traits thai tend to be negatively correlated across ability fewer coefficients and therefore are not shown in Figure 3.
traits include Stress Reaction, Alienation, lest Anxiety, and Psy- However, the available data suggest that both of these traits have
choticism. With the exception of Test Anxiety, these personality substantial relationships with Gc at least. In reference to Table
traits are associated with Neuroticism/NEM and Psychoticism/ 1, also note that TIE has a significant and substantial relationship
Constraint. The remaining personality traits either have unclear with Ideational Fluency and Knowledge and Achievement—
patterns of correlations (e.g., Conscientiousness and Traditional- in-line with Ackerman's (1994) notion that TIE is closely related
ism) or too little data (e.g., Absorption and Control). to a broad class of verbal abilities, crystallized abilities, and
cultural knowledge. Too few correlations are available to draw
Specific Findings any conclusions about the relations between Openness to Expe-
rience, TIE, and fluid intelligence.
Indications of significant heterogeneity in the estimated popu-
lation correlations suggests that sole examination of the esti-
Summary
mated mean population correlation may obscure effects that
might be attributed to particular choices of sample type, person- The meta-analysis of personality-intellectual ability traits in-
ality measure type, ability measure type, or all three (e.g., see dicates that, first of all, there exist many significant cross-trait
discussions by Hedges & Olkin, 1985; and Rosenthal, 1995). relations. Of the 161 nonempty cells of the 19 X 10 personality
To more closely examine the underlying characteristics of the trait X ability trait matrix, 83 (or 52%) are significantly different
data, we selected five personality traits that either indicate the from zero, with nearly as many significant negative correlations
widest range of correlations or the largest mean correlations. as positive correlations. The personality traits that have ubiqui-
These traits are Achievement, Stress Reaction, Traditionalism, tous negative correlations with ability tend to be from broad
Test Anxiety, and Extroversion. Then we selected the three abil- categories of Neuroticism/NEM and Psychoticism/Constraint,
ity factors for which the greatest number of studies were found, whereas those personality traits that fall under the broad Extro-
namely, General Intelligence, Gf, and Gc. For each of the pair- version/PEM category tend to have positive correlations with
wise personality-ability combinations (5 X 3), frequency distri- intellectual abilities. Test Anxiety, not fitting well within the
butions of corrected correlation coefficients were created and taxonomic personality structure (although substantially corre-
graphed, along with the 25th- and 75th-percentile values of the lated with Neuroticism; see review by Hembree, 1988), also
correlations (and the estimated mean population correlation co- indicates substantial negative correlations across all the ability
efficient). These results are shown in Figure 3. traits. Finally, a trait complex of TIE, Openness to Experience,
From the significant heterogeneity statistics and the frequency and Intellectence shows substantial positive correlations with
distributions shown in Figure 3, it appears that some combina- intellectual abilities, with TIE showing the greatest associations
tions of personality and ability measures and participant samples with Gc and Ideational Fluency factors.
may reliably yield substantial correlations between personality
and ability measures. Most notable among these data are the
Interests and Intellectual Abilities
substantial negative relations between Test Anxiety and the
broad ability measures. Although there is some disagreement
Background
about whether Test Anxiety is best thought of as a personality
trait per se (see the discussion in Test Anxiety), clearly there is Modern interest assessment has nearly as long a history as
a common variance with ability measures. intelligence assessment, with the genesis of interest assessment
Although Stress Reaction indicates significant negative corre- being life insurance salesmen at the Carnegie Institute of Tech-
lations, few of the obtained (corrected) correlations actually nology (Pittsburgh, PA) in 1919 under Yoakum, and Cowdery's
exceed an estimated population correlation greater than .20. interest items for differentiating engineers, lawyers, and physi-
 INTELLIGENCE, PERSONALITY, AND INTERESTS 233

I Estimated
I population
population t 25 Percentile (of corrected correlations)
' J_ 75 percentile (of corrected correlations)
• correlation (p)
General
Achievement Intelligence Gc Gf
10
8 One study =
f 6
4
2
0
-0.6 -0.4
'irh.. I . .if
-0.2 O.| O.q 0.4 0.6 -0.6 -0.4 -0.2 OJO O.| 0.4 0.6 -0.6 -0..
-0.4 -0.2 00
sf
J0.2 0.4 0.6

Stress Reaction
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

24
This document is copyrighted by the American Psychological Association or one of its allied publishers.

20 Two studies =
f, 16
12 ^
8
4
0
tii.
-0.6 -0.4 J.O 0.2 0.4 0.6 -0.6 -0.4 OO 0.2 0.4 0.6
Traditionalism
10 -
One study = • 10 10 i
8 - 8
I
8 I
f 6- fi fi I
4
. iiii,.
4 4 •
2
r> . i f f .1 • 2 •
n
2 •
n sffls.
-0.6 -0.4 _£.2 0£_ 0.2 0.4 0.6 -0.6 -0.4 fo.2 0[0 0.2 0.4 0.6 -0.6 -0.4 -O.| 0.$ 0.2 0.4 0.6
Test Anxiety
10 - One study = • 10 - 10

I
8 fi - H•
f 6-
4
2 .,1
-0.6
i
|).4 fb.2 0.0 0.2 0.4 0.6
6
4
2
•

-0.6 -O.f
•'

-OJ2
i, 0.0 0.2 0.4 0.6
6
4
2 •

-0.6
'

-0.4 "-0 2
\
I
1

0.0 0.2 0.4 0.6

Extroversion
20 -
16 -
1
12 -
8
| 1
4
n. ii • f *
-0.6 -0.4 -0.2 OJO J0.2 0.4 0.6 -0.6 -0.4 -0.2 OJO J.2 0.4 0.6 -0.6 -0.4 -0.2 0|0jp.2 0.4 0.6

Correlation Correlation Correlation

Figure 3. Frequency distributions of corrected correlation coefficients, with 25th-percentile, mean, and
75th-percentile estimates of population correlation coefficient. Gc = Crystallized Intelligence; Gf = Fluid
Intelligence.

cians (e.g., see reviews by Campbell, 1971; and Strong, 1952) review indicates that there are two main reasons for this situa-
and later the general development of the SVIB in 1927. Many tion: (a) Early interest inventories (using or modeled on the
interest measures had been developed up to 1940, however, as SVIB) provided only occupational scores, thus abilities were
with the SVIB, most of these measures were empirically keyed correlated with occupational similarity indices (which them-
and essentially atheoretical. selves are multifaceted); and (b) other early studies used profile
Although there have been many studies of the relations be- analysis (e.g., by investigators separating high and low general
tween interests and intellectual abilities (e.g., see Fryer, 1931), ability respondents and then comparing mean profile patterns).
relatively few provide the kind of correlational data that would As a result of these limitations, our discussion of the interest-
allow for a quantitative meta-analysis procedure to be conducted ability relations below is narrative rather than strictly
(as we did for the personality-ability relations). A historical quantitative.
 234 ACKERMAN AND HEGGESTAD

SVJB (Perception, Number, Verbal, Space, Memory, Induction, and

Reasoning). Again note that this group represents a restriction
In an early study of interest—ability relations with 100 junior of range in ability (as was noted about Thurstone's, 1938, earlier
college students, Segal (1934) selected a set of six "most inde- studies on intelligence). As such, correlations between ability
pendent" interest scores (occupations) from the SVIB (Engi- measures and interest measures were attenuated to an unknown
neering, Medicine, Law, Life Insurance, Personnel Management, degree. (Furthermore, an analysis of gender differences, al-
and Purchasing) and correlated the scores on the American though potentially important and significant in the study, is be-
Council on Education Psychological Exam (ACE, a general in- yond the scope of this article.) In light of the restriction of range
telligence test) and four tests from the Iowa High School Content on the ability measures, the correlations between ability scores
Examination (English Literature, Mathematics, Science, and and interest scales are generally small (only one exceeded r =
History and Social Science). Although there is good reason to .30). However, positive correlations were found for Computa-
expect a restriction of range in talent among this selected group, tional interests and Number ability (Mean r = .30), for Induction
two Strong scores showed salient patterns of correlations with and Reasoning abilities (Mean r = .24) and for Literary interests
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

the ability measures: Engineering interest score with ACE (r =

This document is copyrighted by the American Psychological Association or one of its allied publishers.

and Verbal ability (Mean r = .27); whereas negative correlations

.10), English Literature (r = -.10), Mathematics (r = .49), were found for Artistic interests and Number ability (Mean r —
Science (r = .36), and History and Social Science (r = —.16); -.21) and for Athletic interests and Verbal ability (Mean r —
Purchasing interest score with ACE (r = —.32), English Litera- -.21). The remaining correlations were all under .20.
ture (r = -.43), Mathematics (r = .04), Science (r = -.26), A later study by Triggs (1943) of 234 University of Illinois
and History and Social Science (r — —.26). Segal appeared to students reported correlations between the KPR scales and two
demonstrate a divergence in vocational interests much in-line measures from the ACE (Quantitative and Linguistic). Correla-
with the demands of at least two different occupations. tions were larger in magnitude but generally similar to those
In a war-time (WW2) study of 292 enrollees in introductory reported by Adkins and Kuder (1940). Positive correlations were
engineering subjects across a large age span, Moore (1941) found among Computational interests and Quantitative ability
presented a set of correlations between three SVIB scores and (Mean r — .31) and Literary interests and Linguistic ability
a set of broad ability measures of math, physics, and mechanical (Mean r = .41). Negative correlations were found between So-
knowledge. Sales occupational interest scores correlated in a cial Service interests and both Quantitative ability (Mean r =
range of r = —.37 to —.19, similar to that of Purchasing (r = — .13) and Linguistic ability (Mean r = —.25). Similar patterns
— .14 to —.27), whereas the Engineering interest scores corre- of correlations were reported on a smaller sample of partici-
lated (r = .11 to .35) with the various measures of general pants; again investigators used the KPR but this time with a
ability and math, physics, and mechanical knowledge. Similar series of content-based achievement tests. That is, Scientific
results were also reported by L. Long (1945) in a study of 200 and Computational interests showed positive correlations with
students at the College of the City of New %rk, on a scientific achievement test scores in Mathematics and General Science,
aptitude battery (correlation between Technical/Math occupa- as did Literary interests and English Literature achievement test
tional group scores from the SVIB and scientific aptitude was scores, whereas Social Service interests showed negative corre-
r = .50, whereas the correlation between the Business Contact lations with all of the achievement tests, with the largest negative
occupational group score from the SVIB and scientific aptitude correlations for Math, Science, and English Literature. With
was r = —.37). Aside from the difference in samples and ability respect to the association between general intelligence (the Ohio
measures, the results of these studies are generally consistent State Psychological Test) and the KPR, Phillips and Osborne
with those of Segal (1934). (1949) described a study of 152 University of Georgia students.
One finding from the SVIB, reported by Terman (1954; also They found a positive correlation between general intelligence
see discussions by French & Steffen, 1960; and Strong, 1943, and Literary interests (r = .39) but found negative correlations
1955), is an association between level of intellectual ability and with Mechanical (r = -.18) and Social Service (r = -.24)
breadth of interests; that is, respondents with high levels of interests.
intelligence tend to endorse more interest items across wide Thus, across these different samples and ability tests, there
ranges of domains, whereas lower levels of intelligence are asso- appears to be a robust finding of positive correlations among
ciated with narrower interests or lower, more general levels of Scientific and Computational interests with Mathematical abili-
interest. ties and among Literary interests and Verbal abilities but of
negative correlations between Social Services interests and
Kuder Preference Record Studies many divergent ability measures.

In contrast to the SVIB, the experimental edition of the Kuder

Preference Record (KPR) provided eight preference themes Holland
rather than occupational scores. These themes were Scientific,
Computational, Musical, Artistic, Literary, Persuasive, Athletic, During the 1950s and 1960s, investigators converged on a
and Social Prestige. In a pioneering study of ability-interest structure of occupational interests (e.g., Guilford, Christensen,
associations, with 312 male and 200 female University of Chi- Bond, & Sutton, 1954; Roe, 1956; Roe & Klos, 1969) that has
cago freshmen, Adkins and Kuder (1940) correlated scores on been generally identified with Holland (1959). Holland's initial
the eight themes of the KPR with scores on Thurstone's Tests specification in "A Theory of Vocational Choice" specified
for Primary Mental Abilities, which yielded seven composites six broad vocational orientations. In 1962 and 1963, Holland
 INTELLIGENCE, PERSONALITY, AND INTERESTS 235

modified the terms to "convey more accurately the nature of in tune with intellectual demands (e.g., the Enterprising or Per-
the type" (Holland, 1963, p. 548). The original terms and brief suasive theme is more associated with a desire for intellectual
descriptions follow (with their revised descriptors in challenge than the Realistic or Motoric theme).
parentheses):
1. Motoric orientation (Realistic): These people ' 'enjoy activ- Themes and Ability
ities requiring physical strength, aggressive action, motor coor-
dination and skill" (p. 36). There are relatively few studies that report correlational asso-
2. Intellectual orientation (Investigative): These are "task- ciations between Holland's (1973) themes and intellectual abil-
oriented people who generally prefer to 'think through,' rather ity measures, mainly because most of the research associated
than 'act out,' problems. They have marked needs to organize with Holland's approach is based on profile interpretations
and understand the world" (p. 36). rather than individual theme scores. That is, although Holland's
3. Esthetic orientation (Artistic): These people "prefer indi- theory and instruments also address and assess self-ratings of
rect relations with others. They prefer dealing with environmen- ability, respectively, relatively few investigators have examined
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

tal problems through self expression in artistic media . . . . They associations between interest themes and objective measures of
This document is copyrighted by the American Psychological Association or one of its allied publishers.

resemble persons with an intellectual orientation in their intra- abilities. Nonetheless, we identified five empirical investigations
ceptiveness and lack of sociability" (p. 37). that provided direct correlational assessments of Holland's
4. Supportive orientation (Social): These people "prefer theme-ability relations. The full results of these investigations
teaching or therapeutic roles, which may reflect a desire for are shown in Table 2, and each is discussed in turn.
attention and socialization in a structured, and therefore safe, In a review and archival study of vocational counseling clients
setting. They possess verbal and interpersonal skills" (p. 37). (N = 846) reported by Randahl (1991), correlations between a
5. Persuasive orientation (Enterprising): These people "pre- subsequent edition of the Strong instrument (Strong Campbell
fer to use their verbal skills in situations which provide opportu- Interest Inventory [SCII], which includes Holland's theme
nities for dominating, selling, or leading others . . . . They avoid scores) and the General Aptitude Test Battery [GATE]) were
well-defined language or work situations as well as situations reported. Salient positive correlations were found between Real-
requiring long periods of intellectual effort" (p. 37). istic interests and Spatial ability {r = .34) and between Artistic
6. Conforming orientation (Conventional): These people interests and Verbal ability (r = .28). Modest positive correla-
"prefer structured verbal and numerical activities, and subordi- tions were also found between Investigative (Intellectual) inter-
nate roles. They achieve their goals through conformity" ests and Verbal (r = .22), Numerical (r = .23), and Spatial (r
(P- 37). = .27) abilities.
Holland (1973) also credited a factor analysis by Guilford et In another study of interest—ability associations, with 149
al. (1954) to provide a similar typology. Of the nearly 20 factors female college students, Lowman, Williams, and Leeman (1985)
identified, Guilford et al. found "the well-known factors" of reported multiple correlations between sets of ability measures
interests, namely, "mechanical," "scientific," "aesthetic," and Holland themes (from the Self-Directed Search booklet).
"social welfare," "business," and "clerical" (Guilford et al., Although the multiple correlations may overestimate construct
1954, p. 28). As can be seen from this list, the main Guilford overlap, significant relationships were reported for Investigative
et al.'s factors have a categorization that is strikingly similar to interests and Reasoning/Mechanical abilities (R = .28; where
that offered by Holland (see also Zytowski, 1986). R is the multiple correlation) and for Artistic interests with
Although there has been some controversy about the represen- Musical talent tests (R = .43).
tation of these six broad interest areas (e.g., Gati, 1991), more In a study of 93 college students, Ackerman et al. (1995)
recently investigators have been devoted to validating this typol- administered the Unisex Edition of the American College Testing
ogy of interests within a hexagon or circumplex representation Interest Inventory (UNIACT; see Lamb & Prediger, 1981), along
(e.g., for a meta-analysis, see Tracey & Rounds, 1993). There with Spatial, Verbal, and Math ability tests. With these data,
is little fundamental disagreement in the literature about these several salient correlations were found between abilities and
broad interest themes. interest themes, namely, rAltistic, Vert»i = .37, rRcaiiaic, Math = -38, and
''invesag.tive.Maiii = -34.5 Two other similarly constructed studies
Occupational Level were conducted. The first, by Rolfhus and Ackerman (1996) of

In the original formulation of the theory, Holland (1959) also

5
specified that an individual's occupational level is an additive The results from this study also shed some light on the meaning of
function of intelligence level and self-evaluation level. That is, the general factor of interests (e.g., see Prediger, 1982; Rounds & Tracey,
across the various interest types, an individual's objective intelli- 1993; and Terman, 1954). A composite interest score (summed across
gence and self concept determine the degree of intellectual chal- all six UNIACT scales) shows a substantial correlation with Goff and
lenge or level of achievement within a thematic domain. Later, Ackerman's (1992) TIE measure (.48), somewhat smaller correlations
with Extroversion (.30) and Openness to Experience (.25) from the
Holland (1973) abandoned this formula in favor of a model NEO-PI-R (revised), and with Verbal ability (.21). Correlations be-
profile, that is,' 'the closer a person's resemblance to the person- tween the general interest composite and knowledge scales were all
ality pattern ESAICR [i.e., Enterprising is the highest theme positive and significant, with the largest associations cutting across all
score, followed by Social, and so on], the greater his expected identifiable fields: sociology (.36), education (.36), physics (.35), ge-
vocational aspiration and eventual achievement" (p. 58). In this ometry (.35), music (.33), poetry (.33), geography (.31), agriculture
fashion, Holland suggested that some interest themes are more (.31), and so on.
 236 ACKERMAN AND HEGGESTAD

Table 2
Correlations Between Holland Themes and Intellectual Abilities

Study and ability N Realistic Investigative Artistic Social Enterprising Conventional

Randahl (1991) 846'

Spatial .34* .27* .01 -.13* -.06 -.06
Verbal -.03 .22* .28* -.01 -.17* -.08*
Numerical (computation) .09* .23* -.02 .01 -04 .15*
Lowman et al. (1985) 149*
Reasoning/Mechanical
Knowledge R = .28*
Music R = .43*
Ackerman et al. (1995) 93=
Spatial .24* .13 .01 -.04 -.15 .00
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Math .38* .34* -.20 -.14 -.15 .18

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Verbal .14 .33* .37* -.08 -.32* -.32*

Perceptual Speed .08 .12 -.04 .06 -.05 -.07
Rolfhus & Ackerman (1996) 180=
Spatial .28* .16* .06 -.08 -.16* .01
Mechanical .35* .21* .11 -.21* -.15* .04
Math .27* .14 .11 -.20* .01 .19*
Verbal .23* .20* .24* .03 -.03 -.05
Perceptual Speed .10 .02 .05 .14 .02 .15*
Kanfer et al. (1996) 150s
Spatial .26* .17* .16* .00 -.17* -.12
Math -.04 .02 -.03 -.09 -.21* -.09
Verbal -.12 -.05 .21* -.09 -.31* -.22*
Perceptual Speed .03 .02 -.03 -.10 -.02 .16*

"Vocational assessment clients who completed the Strong Vocational Interest Blank. ""Female college students who completed the Self-Directed
Search booklet. 'College students who completed the Unisex Edition of the American College Testing Interest Inventory.
*/> < .05 (no distinction is made for higher levels of significance, for example, p < .01).

180 college students, found salient correlations between Realis- 4. From Tferman (1954) and Holland (1959), notwithstanding
tic and Investigative themes and most ability measures. Similar 3 above, there is a positive association between level of intellec-
to the Ackerman et al. (1995) results, positive correlations were tual ability (or perceived level) and the depth and breadth of
also found for Artistic interests and Verbal ability. In a second interests.
study of 150 college students, Kanfer et al. (1996) again found 5. From studies with Holland's theme assessment, (a) Spatial
significant correlations between Spatial ability and both Realis- and Math abilities are associated with Realistic and Investigative
tic and Investigative themes but not with Math and Verbal ability, interests (and Math [computation] is associated with Conven-
although Verbal ability was positively correlated with Artistic tional interests), (b) Mechanical ability appears to be more
interests. highly associated with Realistic interests than Investigative inter-
ests, (c) Verbal abilities tend to be most highly correlated with
Summary

From the three broad approaches (SVIB, KPR, and Holland's

Visual Perception
themes), the associations between ability and interests reported (Spatfaf Math
suggest that several sources of overlap exist among these two
broad construct domains (shown in Figure 4). In general,
though, most reported studies "suffered" from restrictions in
range of talent on the ability measures and had interest measures
with low reliability, low validity, or both (e.g., see review by
Cronbach, 1990). As such, the general findings reported may
Numerical/
underestimate construct overlap between interests and abilities:
Perceptual Speed
1. Science and Engineering interests (Investigative and Real- Ni
istic themes) tend to be positively associated with Math, Spatial, Convei
and Mechanical ability scores.
2. Literary interests are positively associated with Verbal
ability and Literature achievement
Enterprising Social
3. Interests in Social Services (which is similar in construct
description to Holland's Supportive/Social theme) tend to be Figure 4. Associations between Holland's interest (regular) themes
negatively correlated with many abilities, especially those in the and intellectual abilities (bold). Solid lines = positive correlations;
Math-Spatial domain. dotted lines = negative correlations.
 INTELLIGENCE, PERSONALITY, AND INTERESTS 237

Artistic and Investigative interests, (d) negative correlations tend ford-Zimmerman Temperament Survey, Hogan Personality In-
to be found between ability measures and Enterprising and be- ventory, CPI, etc.). Interests were measured with either Hol-
tween ability measures and Conventional interests (except for land's Vocational Preference Inventory (VPI) or his Self-
Perceptual Speed/Math computation and Conventional inter- Directed Search booklet. Gottfredson et al. determined that
ests), and (e) negligible or negative correlations tend to be found Social and Enterprising interests were correlated with Extrover-
between Social interests and ability measures. sion, Investigative and Artistic interests were correlated with
Openness to Experience, and Conventional interests were corre-
Interests and Personality lated with Conscientiousness.
Also, in a study of 150 college students who took the MPQ,
Before we continue the discussion of the associations between TIE, and UNIACT (Kanfer, Ackerman, & Heggestad, 1996),
interest orientations and abilities, a functional representation of similar sets of significant correlations were found to those ob-
interest, personality, and ability requires that we identify the tained from the FFA data (Goh & Leong, 1993; Gottfredson et
relations between interests and personality traits. A full-scale al., 1993). Specifically, Well-Being correlated positively with
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

meta-analysis for this domain is beyond the scope of this article, Social interests (r = .28), Social Potency with Social and Enter-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

but a few investigations can be drawn on to gauge trait overlap. prising interests (r = .31 and .48, respectively), and Achieve-
For example, Goh and Leong (1993) examined correlations be- ment with Enterprising interests (r = .27). Both Control and
tween Holland's orientations and H. J. Eysenck's personality Traditionalism showed positive correlations with Conventional
theory, using SCII and EPQ for the respective trait sets. For a interests (r = .20 and .24, respectively), whereas Traditionalism
sample of 119 undergraduate students, cross-correlations were was also negatively correlated with Artistic interests (r = —.24).
computed between the six Holland orientations and the three Absorption was positively related to Artistic interests (r = .55)
trait scales of the EPQ (Extroversion, Neuroticism, and Psy- and Realistic interests (r = .31); whereas Harm-Avoidance
choticism). However, with the exception of TRcaiistic/Neuiocidmi was negatively related to Artistic interests (r = —.26), Realistic
= —.32, no other correlations exceeded .30. A few other corre- interests (r = -.33), and Investigative interests (r = —.23).
lations were significant but not as large: rgrtuacir*y<*ain*ii = .24; Finally, TIE correlated significantly with Investigative (r = .42),
riiiTOtigUJve/Neuroticism = —.26, and rEnlerpri.ias/EMiDvmion = .28. For the Artistic (r = .35), and Social (r = .31) interests.
most part, such data suggest a relatively low-to-moderate associ- From these diverse studies, it is possible to provide a config-
ation between personality traits and interests. They speculated ural summary of the correlations reported between interest and
that one reason for the low associations is because the sample personality measures, as shown in Figure 5.
was made up of college students not older adults, who could 1. Neither Agreeableness/Psychoticism nor Neuroticism/
be expected to have substantial occupational experience, might Stress Reaction correlate appreciably with any of the six Hol-
provide different results. land orientations.
In contrast to the Goh and Leong (1993) study, the investiga- 2. Conscientiousness, Control, and Traditionalism show mod-
tion by Gottfredson, Jones, and Holland (1993) provides greater erate correlations with Conventional interests. Also, Traditional-
support for substantial interest-personality associations (i.e., ism showed moderate (negative) correlations with Artistic
median correlations in the range of r = .20 to .38). They incor- interests.
porated their study of 725 U.S. Navy trainees with a review of 3. Extroversion, Well-Being, and Social Potency show moder-
other investigations, which all used the FFA to assess personality ate to substantial correlations with both Enterprising and Social
traits (e.g., NEO-PI and similar measures from other tests, Guil- interest domains.

Realistic Investigative

Conventiona TIE
Openness
Control
Conscientousness'
Traditionalism^-

Enterprising

Extroversion
Social Potency
Welt-Being

Figure 5. Associations between Holland's interest (regular) themes and personality traits (bold italic).
Solid lines = positive correlations; dotted lines = negative correlations; TIE = typical intellectual
engagement.
 238 ACKERMAN AND HEGGESTAD

4. Openness to Experience and TIE shows moderate to sub- though it is not possible to portray all the significant relations
stantial correlations with Investigative, Artistic, and Social inter- between these three sets of traits into the same configural repre-
ests. Similarly, Absorption shows substantial correlation with sentation, a sense of the various interrelations is shown in Figure
Artistic interests and Realistic interests. 7, where four trait complexes (similar to Snow's "aptitude
5. Harm-Avoidance shows negative correlations with Artis- complexes"; see Snow, 1989; and Snow, Corno, & Jackson,
tic, Realistic, and Investigative interests. 1996) have been drawn that illuminate positive communalities
among personality, interest, and ability traits. We should note
that this is an abstraction and reduction of the data, but this
An Integrated Representation
approach refines the results into a single, generally coherent
It is possible to present integrated representations, based on framework.
the meta-analysis of personality—intelligence relations, reviews The first trait complex shows no positive communality with
of interest—personality relations, interest—ability relations, and ability measures and is made up of a broad social trait complex.
a few empirical studies with simultaneous assessments of per- It includes Social and Enterprising interests, along with Extro-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

sonality, interest, and abilities. However, given the large number version, Social Potency, and Well-Being personality traits. The
This document is copyrighted by the American Psychological Association or one of its allied publishers.

of significant correlations found in the reviews and meta-analy- remaining trait complexes include ability traits. The clerical/
sis, it is difficult (and not particularly useful from a heuristic conventional trait complex includes Perceptual Speed abilities,
perspective) to represent all the significant relations. As such, Conventional interests, and Control, Conscientiousness, and Tra-
we limited the depictions to the largest correlations (i.e., illustra- ditionalism personality traits. The remaining trait complexes
tions are presented with the highest respective communalities overlap to a degree; the third trait complex, science/math, is
among the trait measures, typically above .20) and those that not positively associated substantially with any personality traits
provide the most coherent sets of relationships (i.e., we do not but includes Visual Perception and Math Reasoning abilities
extensively discuss associations that do not appear to have some and Realistic and Investigative interests. The last trait complex,
consistency in the matrix of results). intellectual/cultural, includes abilities of Gc and Ideational Flu-
From an ability-based perspective, personality traits of Open- ency; personality traits of Absorption, TIE, and Openness to
ness to Experience, TIE, Intellectence, and Alienation, as well Experience; and Artistic and Investigative interests.
as Test Anxiety, were found to have substantial communality.
Similarly, Realistic, Investigative, and Artistic interests were Discussion and Conclusions
found to be related to particular abilities. These relationships
are shown in Figure 6. Clearly, the relationships are complex, In this article, we started with a review of intelligence testing
but a pattern does indeed emerge. Test Anxiety has pervasive as a paradigm. From the developments of intelligence assess-
negative associations on abilities, as do Enterprising and Con- ment for children at the beginning of this century, the stage was
ventional interests (with the exception of Conventional interests set to mainly consider intellectual abilities as maximal perfor-
and Perceptual Speed). Positive associations are found with mance and as far removed as possible from experiences and
Openness to Experience, TIE, and Intellectence personality con- background of the examinee. With a focus on these aspects of
structs and Investigative, Realistic, and Artistic interests. Al- intellectual life, intelligence assessment and theory devclop-

Intellectence

3rd order Test Anxiety

i
± Gener
General
..;.•••.;••::;;;"" Intelligeinee
Artistic

Gc
2nd order Gf (Fluid Perceptual Learning Knowledge and Ideational (Crystallized
Intelligence) Speed and Memory Achievement Fluency Intelligence)
Conventional
:
Realistii
•••Enterprising
1st order Math Reasoning Closure
1
Investig;

Alienation

Figure 6. Associations between abilities (bold), interests(regular), and personality (italic) traits. Solid
lines = positive correlations; dotted lines = negative correlations. TIE = typical intellectual engagement.
 INTELLIGENCE, PERSONALITY, AND INTERESTS 239

3. Science/Math

Math Reasoning

visual Perception

Gc
Ideational Fluency

Perceptual Speed
Conventional
Control
onscientousness
Traditionalism
4. Intellectual/Cultural
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

2. Clerical/Conventional / Enterprising
Extroversion
Social Potency
Well-Being

i Social

Figure 7. Trait complexes, including abilities (bold), interests (regular), and personality (italic) traits,
showing positive commonalities. Number categories are trait complexes. Gc — Crystallized Intelligence;
TIE = typical intellectual engagement.

ments over this century have mainly proceeded along a separate That is, abilities, interests, and personality develop in tandem,
path from personality and interest theory and assessment. How- such that ability level and personality dispositions determine the
ever, a consideration of adults, rather than children, and aspects probability of success in a particular task domain, and interests
of Tits' provide a basis to investigate common variance among determine the motivation to attempt the task. Thus, subsequent
measures of intellectual ability, personality traits, and interests. to successful attempts at task performance, interest in the task
We reviewed theories of abilities that provide additional bases domain may increase. Conversely, unsuccessful attempts at task
for an integrated viewpoint, that is, a viewpoint that is much performance may result in a decrement in interest for that do-
more inclusive of other traits than the intelligence-assessment main. Later adult development also appears to be interactive
paradigm. Personality theories and interest theories were re- (abilities, personality, and interests), as demonstrated by Kohn
viewed, and an extensive meta-analysis of personality-ability and Schooler (1973, 1978; see reviews by Schooler, 1987; and
relations was conducted, along with a review of interest-ability Willis & Tbsti-Vasey, 1990). Ultimately, the malleability of indi-
relations. From the meta-analysis, it was determined that many vidual differences in these trait complexes might be investi-
sources of common variance could be found between personality gated—but, given the common variance among the various
and ability measures. Many of these personality-ability rela- traits, it appears that new approaches might best be constructed
tions were pervasive, namely, negative relations between person- as multiply determined interventions—such as an effort that
ality traits falling under the Neuroticism/NEM and Psychot- could be devoted to raise Artistic and Investigative interests,
icism/Constraint factors and Test Anxiety and positive relations expand Openness to Experience and TIE, and provide instruc-
between personality traits of Extroversion and abilities. Other tion in the content area of culture or verbal knowledge. In the
personality—ability relations were more specific, such as the final analysis, it appears that Cronbach (1957) was initially too
relations among Openness to Experience and measures of Gc, restrictive by only suggesting that there are two types of scien-
Gf, and Knowledge and Achievement and among TIE and Gc, tific psychologists, the correlational and the experimental.
Ideational Fluency, and Knowledge and Achievement. Interest Within the correlational discipline, theory has been too restric-
measures also showed substantial positive relations with ability tive along major trait domains. Investigations of higher order
measures, mainly in the domains of Realistic, Investigative, and interactions (e.g., Cronbach, 1975) and trait complexes (see
Artistic interests, whereas Enterprising and Conventional inter- Snow, 1989) may provide for substantially improved under-
ests tend to be negatively associated with abilities. standing of the nature of individual differences in each of these
From this review, we highlighted four trait complexes across traditionally separated domains of cognition, affect, and cona-
the three trait domains: social, clerical/conventional, science/ tion (interests) and the development of intellect across the life
math, and intellectual/cultural. Although it is not possible to span (e.g., see Ackerman, 1996).
determine causal connections with the corpus of data we re- Finally, note the three specific limitations to this investigation.
viewed, the communality among these traits suggests directions First, we only considered linear Pearson product-moment cor-
for future investigations of enabling and moderating influences. relations among ability, personality, and interest trait families.
It seems to be reasonable to propose (as Holland, 1973, did; see Thus, our analysis was insensitive, to the degree that linear
also Sorenson, 1933, 1938) that development of personality- correlations are not sufficiently robust, to any possible nonlinear
interest-intelligence traits proceeds along mutually causal lines. relations that might exist between trait families (e.g., see Ent-
 240 ACKERMAN AND HEGGESTAD

wistle & Cunningham, 1968; and Snow et al., 1996). However, Analysis of data from a large-scale longitudinal study of children
meta-analytic examination of nonlinear hypotheses is impracti- tested at 10-11 and 14-15 years. British Journal of Educational
cal when multiple types of measures and samples are used, Psychology, 53, 374-379. (2)"
mainly because it is necessary to construct models of where *Arwater, L. E. (1992). Beyond cognitive ability: Improving the predic-
tion of performance. Journal of Business and Psychology, 7, 27-44.
inflection points should occur on the particular variables of
*Ayers, J. B., Bashaw, W. L., & Wash, J. A. (1969). A study of the
interest. Other approaches to the corpus of data would be neces-
validity of the Sixteen Personality Factor Questionnaire in predicting
sary for such an evaluation. Second, we have not considered
high school academic achievement. Educational and Psychological
the nature of potential gender differences among the various
Measurement, 29, 479-484.
trait relations, even though significant gender differences are
*Bachman, J. G. (1964). Prediction of academic achievement using the
often found among all three of the trait families under consider- Edwards Need Achievement Scale. Journal of Applied Psychology,
ation. Finally, we reiterate that we limited our discussion to 48, 16-19. (2)
nonpathological samples of participants—evaluation of the rela- *Baehr, M. E., & Orban, J. A. (1989). The role of intellectual abilities
tions between the trait families across the full range of personal- and personality characteristics in determining success in higher-level
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

positions. Journal of Vocational Behavior, 35, 270-287.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

ity trait levels (as well as extremes on ability and interest traits)
would require a different approach (e.g., see H. J. Eysenck, •Bauer, D. H. (1975). The effect of instructions, anxiety, and locus of
1995) or a larger sample of investigations than could be located control on intelligence test scores. Measurement and Evaluation in
in the literature. Guidance, K, 12-19.
*Behrens, L. T, & Vernon, P. E. (1978). Personality correlates of over-
achievement and under-achievement. British Journal of Educational
References Psychology, 48, 290-297.
Binet, A., & Simon, T. (1905). Methodes nouvelles pour le diagnostique
References marked with an asterisk indicate studies included in the
du niveau intellectual des anormaux [New methods for the diagnosis
meta-analysis (numbers in parentheses indicate the number of samples
of the intellectual levels of subnormals] (E. S. Kite, Trans.). In J. J.
contributed greater than one).
Jenkins & D. G. Paterson (Reprint Eds.), Studies in individual differ-
Ackerman, P. L. (1988). Determinants of individual differences during ences: The search for intelligence (pp. 90-96). New ^brk: Appleton-
skill acquisition: Cognitive abilities and information processing. Jour- Century-Crofts. (Reprinted in 1961)
nal of Experimental Psychology: General, 117, 288-318. Binet, A., & Simon, T. (1908). La developpement de ('intelligence chez
Ackerman, P. L. (1994). Intelligence, attention, and learning: Maximal les enfants [The development of intelligence in children] (E. S. Kite,
and typical performance. In D. K. Detterman (Ed.), Current topics Trans.). In J. J. Jenkins & D. G. Paterson (Rprint Eds.), Studies in
in human intelligence. Vol. 4: Theories of intelligence (pp. 1-27). individual differences: The search for intelligence (pp. 96-111). New
Norwood, NJ: Ablex. York: Appleton-Century-Crofts. (Reprinted in 1961)
Ackerman, P. L. (1996). A theory of adult intellectual development: Binet, A., & Simon, T. (1911). La mesure du developpement de I'intelli-
Process, personality, interests, and knowledge. Intelligence, 22, 229- gence chez les jeunes enfants [A method of measuring the develop-
259. ment of the intelligence of young children] (C. H. Town, Trans., 1915;
*Ackerman, P. L., Gernsbacher, M. A., & Kanfer, R. (1996). [Personal- 3rd ed.). Chicago: Chicago Medical Book.
ity, ability, and language comprehension]. Unpublished research data. Block, J. (1995a). A contrarian view of the five-factor approach to
University of Minnesota, Twin Cities, Department of Psychology, personality description. Psychological Bulletin, 117, 187-215.
Minneapolis. Block, J. (1995b). Going beyond the five factors given: Rejoinder to
Ackerman, P. L., & Goff, M. (1994). Typical intellectual engagement Costa and McCrae (1995) and Goldberg and Saucier (1995). Psycho-
and personality: Reply to Rocklin (1994). Journal of Educational logical Bulletin, 117, 226-229.
Psychology, 86, 150-153. •Bolton, B. (1980). Personality (16PF) correlates of WUS scales: A
Ackerman, P. L., & Goff, M. (1995). Intelligence, personality, and inter- replication. Applied Psychological Measurement, 4, 399-401. (2)
ests: Historical review and a model for adult intellect. Unpublished
*Boor, M. (1972). Relationship of test anxiety and academic perfor-
manuscript, University of Minnesota, Twin Cities, Department of Psy-
mance when controlled for intelligence. Journal of Clinical Psychol-
chology, Minneapolis.
ogy, 28, 171-172.
* Ackerman, P. L., & Kanfer, R. (1994). Improving problem-solving and
*Borman, W. C., White, L. A., Pulakos, E. D., & Oppler, S. H. (1991).
decision-making skills under stress: Prediction and training (final
Models of supervisory job performance ratings. Journal of Applied
report to the Office of Naval Research, Contract N00014-91-J-4159).
Psychology, 76, 863-872.
Minneapolis, MN: Authors.
*Brown, M. (1974). Motivational correlates of academic performance.
*Ackerman, P. L., Kanfer, R., & Goff, M. (1995). Cognitive and noncog-
Psychological Reports, 34, 746.
nitive determinants and consequences of complex skill acquisition.
*Bruch, M. A., Juster, H. R., & Kaflowitz, N. G. (1983). Relationships
Journal of Experimental Psychology: Applied, 1, 270-304.
Adkins, D. C., & Kuder, G., F. (1940). The relation of primary mental of cognitive components of test anxiety to test performance: Implica-
abilities to activity preferences. Psychometrika, 5, 251 -262. tions for assessment and treatment. Journal of Counseling Psychology,
Alexander, W. P. (1935). Intelligence, concrete and abstract: A study in 30, 527-536.
differential traits. British Journal of Psychology Monograph, 6(No. *Bruhn, J. G., Bunce, H., Ill, & Greaser, R. C. (1978). Correlations of
19). the Myers-Briggs Type Indicator with other personality and achieve-
*Allen, J., Blanton, P., Johnson-Greene, D., Murphy-Farmer, C., & ment variables. Psychological Reports, 43, 771-776. (2)
Gross, A. (1992). Need for achievement and performance on measures Bushman, B. J. (1994). Vote-counting procedures in meta-analysis. In
of behavioral fluency. Psychological Reports, 71, 471-478. H. Cooper & L. V. Hedges (Eds.), The handbook of research synthesis
*Alpert, R., & Haber, R. (1960). Anxiety in academic achievement (pp. 193-213). New York: Russell Sage Foundation.
situations. Journal of Abnormal and Social Psychology, 61, 207-215. Bushman, B. J., & Wang, M. C. (1995). A procedure for combining
*Anthony, W. S. (1983). The development of extroversion and ability: sample correlation coefficients and vote counts to obtain an estimate
 INTELLIGENCE, PERSONALITY, AND INTERESTS 241

and a confidence interval for the population correlation coefficient. Costa, P. T, Jr., & McCrae, R. R. (1992c). Revised NEO Personality
Psychological Bulletin, 117, 530-546. Inventory and Five-Factor Inventory professional manual. Odessa,
*Butcher, H. J., Ainsworth, M., & Nesbitt, I.E. (1963). Personality FL: Psychological Assessment Resources.
factors and school achievement. British Journal of Educational Psy- Costa, P. T, Jr., & McCrae, R. R. (1995). Solid ground in the wetlands
chology, 33, 276-285. (2) of personality: A reply to Block. Psychological Bulletin, 117, 216-
Butler, J.M., & Fiske, D. W. (1955). Theory and techniques of assess- 220.
ment. Annual Review of Psychology, 6, 327-356. Cronbach, L. J. (1949). Essentials of psychological testing. New York:
*Butterfield, E. C. (1964). Locus of control, test anxiety, reactions to Harper.
frustration and achievement attitudes. Journal of Personality, 32, Cronbach, L. J. (1957). The two disciplines of scientific psychology.
355-370. American Psychologist, 12, 671-684.
*Callard, M. P., & Goodfellow, C. L. (1962). Three experiments using Cronbach, L. J. (1975). Beyond the two disciplines of scientific psychol-
the Junior Maudsley Personality Inventory. British Journal of Educa- ogy. American Psychologist, 30, 116-127.
tional Psychology, 32, 241-251. Cronbach, L. J. (1990). Essentials of psychological testing (5th ed.).
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Campbell, D.P. (1971). Handbook for the Strong Vocational Interest New ^brk: Harper & Row.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Blank. Stanford, CA: Stanford University Press. 'Crookes, T. G., Pearson, P. R., Francis, L. J., & Carter, M. (1981).
Carothers, E E. (1922). Psychological examinations of college students. Extroversion and performance on Raven's matrices in 15-16 year old
Archives of Psychology, 46, 1-82. children: An examination of Anthony's theory of the development of
Carroll, J. B. (1993). Human cognitive abilities: A survey of factor- extroversion. British Journal of Educational Psychology, 51, 109-
analytic studies. New York: Cambridge University Press. 111. (2)
"Caspy, T., Reisler, A., & Mendelson, Y. (1987). MMPI and intelligence *Daniels, B., & Hewitt, J. (1978). Anxiety and classroom examination
correlates: Assessment of relationships via multivariate analyses. performance. Journal of Clinical Psychology, 34, 340-345.
Journal of Clinical Psychology, 43, 415-421. (2) *Day, D. V., & Silverman, S. B. (1989). Personality and job perfor-
Cattell, J. McK. (1890). Mental tests and measurements. Mind, 15, 373- mance: Evidence of incremental validity. Personnel Psychology, 42,
381. 25-36.
Cattell, J. McK., & Farrand, L. (1896). Physical and mental measure- *Deffenbacher, J. L. (1977). Relationship of worry and emotionality to
ments of the students of Columbia University. Psychological Review, performance on the Miller Analogies Test. Journal of Educational
3, 618-648. Psychology, 69, 191-195.
Cattell, R. B. (1946). Description and measurement of personality. Yon- *de Man, A. F., Hall, V., & Stout, D. (1991). Neurotic nucleus and test
kers-on-Hudson, NY: World Book. anxiety. Journal of Psychology, 125, 671-675.
Cattell, R. B. (1949). 16 Personality Factor Questionnaire. Champaign, *Dember, W. R, Naire, E, & Miller, F. J. (1962). Further validation of
1L: Institute for Personality and Ability Testing. the Alpert-Haber Achievement Anxiety Test. Journal of Abnormal
Cattell, R. B. (1987). Intelligence: Its structure, growth, and action. and Social Psychology, 65, 427-428.
Amsterdam: North-Holland. (Reprinted and revised from Abilities: *Denton, J. C., & Taylor, C. W. (1955). A factor analysis of mental
Their structure, growth and action, by R, B. Cattell, 1971, Boston: abilities and personality traits. Psychometrika, 29, 75-81.
Houghton Mifflin) Digman, J. M. (1990). Personality structure: Emergence of the five-
*Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1988). Handbook for factor model. Annual Review of Psychology, 41, 417-440.
the 16PF. Champaign, IL: Institute for Personality and Ability Testing. *Dixon, P. W., Fukuda, N. K., & Berens, A. E. (1971). A factor analysis
(2) of EPPS scales, ability, and achievement measures. Journal of Experi-
*Chatterji, S., & Mukerjee, M. (1983). Relation of intelligence to neu- mental Education, 39, 31 -41.
roticism among school-going children, slum children and the inmates *Dunn, A., & Eliot, J. (1993). Extroversion-introversion and spatial
of a detention home. Psychologia, 26, 223-231. (2) intelligence. Perceptual and Motor Skills, 77, 19-24.
*Chen, H.-G., & Vecchio, R. P. (1992). Nested IF-THEN-ELSE con- *Egan, V. (1989). Links between personality, ability and attitudes in
structs in end-user computing: Personality and aptitude as predictors a low-IQ sample. Personality and Individual Differences, 10, 997-
of programming ability. International Journal of Man-Machine Stud- 1001.
ies, 36, 843-859. *Eno, L. (1978). Predicting achievement and the theory of fluid and
Church, A. T. (1994). Relating the Tellegen and five-factor models of crystallized intelligence. Psychological Reports, 43, 847-852.
personality structure. Journal of Personality and Social Psychology, *Entwistle, N. J., & Cunningham, S. (1968). Neuroticism and school
67, 898-909. attainment—A linear relationship? British Journal of Educational
Church, A. T., & Burke, P. J. (1994). Exploratory and confirmatory tests Psychology, 38, 123-132. (2)
of the Big Five and Tellegen's three- and four-dimensional models. ERIC on CD-ROM [Database]. (1968-1994, March 1995). Washing-
Journal of Personality and Social Psychology, 66, 93-114. ton, DC: U. S. Department of Education, Educational Resources Infor-
•Clarke, M. J. (1985). GCE performance of FE and sixth-form college mation Center [Producer]. Available from: National Information Ser-
students. Educational Research, 27, 169-178. (4) vices Corporation [Distributor].
Conoley, J. C., & Impara, J. (Eds.). (1995 ). The twelfth mental measure- Eysenck, H. J. (1947). Dimensions of personality. London: Routledge &
ments yearbook (O. K. Euros, Series Ed.). Lincoln, NE: Buros Insti- Paul.
tute of Mental Measurement. Eysenck, H. J. (1970). The structure of human personality (3rd ed).
*Cooper, M. (1974). Factor analysis of measures of aptitude, intelli- London: Methuen.
gence, personality, and performance in high school subjects. Journal *Eysenck, H. J. (1971). Relation between intelligence and personality.
of Experimental Education, 42, 7-10. Perceptual and Motor Skills, 32, 637-638.
Costa, P. T., Jr., & McCrae, R. R. (1992a). Four ways five factors are Eysenck, H. J. (1991). Dimensions of personality: 16, 5 or 3? Criteria
basic. Personality and Individual Differences, 13, 653-665. for a taxonomic paradigm. Personality and Individual Differences,
Costa, P. T, Jr., & McCrae, R. R. (1992b). "Four ways five factors are 12, 773-790.
not basic": Reply. Personality and Individual Differences, 13, 861- Eysenck, H. J. (1992). Four ways five factors are not basic. Personality
865. and Individual Differences, 13, 667-673.
 242 ACKERMAN AND HEGGESTAD

Eysenck, H. J. (1994). Personality and intelligence: Psychometric arid *Gough, H. G. (1949). Factors relating to the academic achievement of
experimental approaches. In R. J. Sternberg & P. Ruzgis (Eds.), Per- high school students. Journal of Educational Psychology, 40, 65-78.
sonality and intelligence (pp. 3-31). New "York: Cambridge Univer- Gough, H. G. (1953). A nonintellectual intelligence test. Journal of
sity Press. Consulting Psychology, 17, 242-246.
Eysenck, H. J. (1995). Genius: The natural history of creativity. New Gough, H. G. (1957). Manual for the California Psychological Inven-
York: Cambridge University Press. tory. Palo Alto, CA: Consulting Psychologists Press.
*Eysenck, H. J., & Coulter, T. T. (i972). The personality and attitudes *Gough, H. G. (1987). California Psychological Inventory administra-
of working-class British communists and fascists. Journal of Social tor's guide. Palo Alto, CA: Consulting Psychologists Press. (5)
Psychology, 87, 59-73. *Gough, H. G., & Weiss, D. S. (1981). A nontransformational test of
Eysenck, H. J., & Eysenck, S. B. G. (1969). Personality structure and intellectual competence. Journal of Applied Psychology, 66, 102-
measurement. London: Routledge & Paul 110.
*Eysenck, S. B. G., Russell, T., & Eysenck, H. J. (1970). Extroversion, *Gruber, J. J., & Kirkendall, D. R. (1973). Relationships within and
intelligence, and ability to draw a person. Perceptual and Motor Skills, between the mental and personality domains in disadvantaged high
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

30, 925-926. school students. American Corrective Therapy Journal, 27, 136-140.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

FirstSearch [Database]. (1962-1973, March 1995). Washington, DC: Guilford, J. P., Christensen, P. R., Bond, N. A., Jr., & Sutton, M. A.
American Psychological Association [Producer]. Available from: Sil- (1954). A factor analysis study of human interests. Psychological
verPIatter [Distributor]. Monographs, 68(4, Whole No. 375).
Fiske, D. W., & Butler, J. M. (1963). The experimental conditions for *GuiIford, J. P., Christensen, P. R., Frick, J. W., & Merrifield, P. R.
measuring individual differences. Educational and Psychological (1957). The relations of creative-thinking aptitudes to non-aptitude
Measurement, 23, 249-266. personality Traits (University of Southern California's Psychological
*Flescher, I. (1963). Anxiety and achievement of intellectually gifted Laboratory Rep. No. 20). Los Angeles, CA: University of Southern
and creatively gifted children. Journal of Psychology, 56, 251-268. California Press. (3)
*Frederiksen, N., & Evans, F. R. (1974). Effects of models of creative Gustafsson, J.-E. (1984). A unifying model for the structure of intellec-
performance on ability to formulate hypotheses. Journal of Educa- tual abilities. Intelligence, 8, 179-203.
tional Psychology, 66, 67-82. *Hakstian, A. R., & Cattell, R. B. (1978). An examination of inter-
French, J. L., & Steffen, H. H. (1960). Interest of gifted adolescents. domain relationships among some ability and personality traits. Edu-
Personnel and Guidance Journal, 38, 633-636. cational and Psychological Measurement, 38, 275-290,
*Friis, R. H., & Knox, A. B. (1972). A validity study of scales to *Hasan, Q., & Khan, S. R. (1976). Creativity, intelligence and tempera-
measure need achievement, need affiliation, impulsiveness, and intel- ment: A study of interrelationship. Psychologia: An International
lectuality. Educational and Psychological Measurement, 32, 147- Journal of Psychology in the Orient, 19, 193-199.
154. Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis.
Fryer, D. (1931). The measurement of interests: In relation to human New terk: Academic Press.
adjustment. New "fork: Holt. Hembree, R. (1988). Correlates, causes, effects, and treatment of test
Gallon, F. (1928). Inquiries into human faculty and its development. anxiety. Review of Educational Research, 58, 47-77.
New "Vbrk: Dutton. (Reprinted from 1907 version; original work pub- *Herron, E. W. (1964). Relationship of experimentally aroused achieve-
lished 1883) ment motivation to academic achievement anxiety. Journal of Abnor-
*Gardner, R. W., Jackson, D. N., & Messick, S. J. (1960). Personality mal and Social Psychology, 69, 690-694.
organization in cognitive controls and intellectual abilities. Psycholog- *Hetrick, S. H., Lilly, R. S., & Merrifield, P. R. (1968). Figural creativ-
ical Issues Monograph, S(No. 4), 2. ity, intelligence, and personality in children. Multivariate Behavioral
Gati, I. (1991). The structure of vocational interests. Psychological Research, 3, 173-187.
Bulletin, 109, 309-324. Hogan, R., & Hogan, J. (1995), Hogan Personality Inventory manual.
*Gentry, T. A., Polzine, K. M., & Wakefield, J. A., Jr. (1985). Human Tulsa, OK: Hogan Assessment Systems.
genetic markers associated with variation in intellectual abilities and Holland, J. L. (1959). A theory of vocational choice. Journal of Coun-
personality. Personality and Individual Differences, 6, 111-113. seling Psychology, 6( 1 ) , 35-45.
*Goff, M., & Ackerman, P. L. (1992). Personality-intelligence rela- Holland, J. L. (1962). Some explorations of a theory of vocational
tions: Assessing typical intellectual engagement. Journal of Educa- choice: I. One- and two-year longitudinal studies. Psychological
tional Psychology, 84, 537-552. Monographs: General and Applied, 76, 1-49.
Goh, D. S., & Leong, F. T. L. (1993). The relationship between Hol- Holland, J. L. (1963). Explorations of a theory of vocational choice and
land's theory of vocational interests and Eysenck's model of personal- achievement: II. A four-year prediction study. Psychological Reports,
ity. Personality and Individual Differences, 15, 555-562. 12, 547-594.
Goldberg, L. R. (1971). A historical survey of personality scales and Holland, J. L. (1973). Making vocational choices: A theory of careers.
inventories. In P. Reynolds (Ed.), Advances in psychological assess- Englewood Cliffs, NJ: Prentice Hall.
ment (Vol. 2, pp. 293-336). Palo Alto, CA: Science and Behavior Horn, J. L. (1965). Fluid and crystallized intelligence: A factor analytic
Books. study of the structure among primary mental abilities [Dissertation]
Goldberg, L. R. (1990). An alternative "description of personality": (No. 65-7113). Ann Arbor, MI: University Microfilms International.
The Big-Five factor structure. Journal of Personality and Social Psy- *Hundal, P. S., & Aggarwal, V. (1972). Prediction of academic achieve-
chology, 59, 1216-1229. ment of the higher secondary students using measures of intellectual
Goldberg, L. R., & Saucier, G. (1995). So what do you propose we use and non-intellectual variables. Indian Journal of Applied Psychology,
instead? A reply to Block. Psychological Bulletin, 117, 221-225. 9, 32-37.
Gottfredson, G. D., Jones, E. M., & Holland, J. L. (1993). Personality *Hundal, P. S., & Singh, M. (1971). A factor analytical study of intellec-
and vocational interests: The relation of Holland's six interest dimen- tual and non-intellectual characteristics. Multivariate Behavioral Re-
sions to five robust dimensions of personality. Journal of Counseling search, 6, 503-514.
Psychology, 40, 518-524. Institute for Personality and Ability Testing staff. (1986), Administra-
 INTELLIGENCE, PERSONALITY, AND INTERESTS 243

tor's manual for the 16 Personality Factor Questionnaire. Champaign, and relationship of college women's primary abilities and vocational
IL: Author. interests. Journal of Vocational Behavior, 27, 298-315.
*Ittenbach, R. E, & Harrison, P. L. (1990). Predicting ego-strength from *Lynn, R., &. Gordon, I.E. (1961). The relation of neuroticism and
problem-solving ability of college students. Measurement and Evalua- extroversion to intelligence and educational attainment. British Jour-
tion in Counseling and Development, 23, 128-136. nal of Educational Psychology, 31, 194-203.
*Jensen, A. R. (1973). Personality and scholastic achievement in three •Lynn, R., Hampson, S. L., & Agahi, E. (1989). Genetic and environ-
ethnic groups. British Journal of Educational Psychology, 43, 115- mental mechanisms determining intelligence, neuroticism, extrover-
125. (12) sion and psychoticism: An analysis of Irish siblings. British Journal
*John, O. P., Caspi, A., Robins, R. W., Moffltt, T. E., & Stouthamer- of Psychology, 80, 499-507. (2)
Loeber, M. (1994). The "Little Five": Exploring the nomological *Lynn, R., Hampson, S. L., & Magee, M. (1983). Determinants of
network of the five-factor model of personality in adolescent boys. educational achievement at 16+: Intelligence, personality, home back-
Child Development, 65, 160-178. ground and school. Personality and Individual Differences, 4, 473-
*Johnson, R. C., Nagoshi, C. T, Ahern, F. M., & Wilson, J. R. (1983). 481. (2)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Correlations of measures of personality and intelligence within and *Martel, J., McKelvie, S. J., & Standing, L. (1987). Validity of an
This document is copyrighted by the American Psychological Association or one of its allied publishers.

across generations. Personality and Individual Differences, 4, 331- intuitive personality scale: Personal responsibility as a predictor of
338. (4) academic achievement. Educational and Psychological Measurement,
*Kalmanchey, G. M., & Kozeki, B. (1983). Relation of personality 47, 1153-1163.
dimensions to social and intellectual factors in children. Personality "Matthews, G. (1986). The effects of anxiety on intellectual perfor-
and Individual Differences, 4, 237-243. mance: When and why are they found? Journal of Research in Person-
•Kanfer, R., Ackerman, P. L., & Heggestad, E. D. (1996). Motivational ality, 20, 385-401.
skills and self-regulation for learning: A trait perspective. Learning *McCrae, R. R. (1993-1994). Openness to Experience as a basic di-
and Individual Differences, 8, 185-209. mension of personality. Imagination, Cognition and Personality, 13,
Kohn, M. L., & Schooler, C. (1973). Occupational experience and psy- 39-55.
chological functioning: An assessment of reciprocal effects. American McCrae, R. R. (1994). Openness to Experience: Expanding the bound-
Sociological Review, 38, 97-118. aries of Factor V. European Journal of Personality, 8, 251-272.
Kohn, M. L., & Schooler, C. (1978). The reciprocal effects of the sub- *McCrae, R. R., & Costa, P. T, Jr. (1985). Updating Norman's "ade-
stantive complexity of work and intellectual flexibility: A longitudinal quacy taxonomy": Intelligence and personality dimensions in natural
assessment. American Journal of Sociology, 84, 24-52. language and in questionnaires. Journal of Personality and Social
•Kossowska, M., & Necka, E. (1994). Do it your own way: Cognitive Psychology, 49, 710-721.
strategies, intelligence, and personality. Personality and Individual McCrae, R. R., & Costa, P. T, Jr. (in press). Conceptions and correlates
Differences, 16, 33-46. of Openness to Experience. In R. Hogan, J. A. Johnson, & S. R.
*Krumboltz, J. D., & Farquhar, W. W. (1957). Reliability and validity Briggs (Eds.), Handbook of personality psychology. San Diego: Aca-
of the n Achievement Test. Journal of Consulting Psychology, 21, demic Press.
226-228. (2) McNemar, Q. (1964). Lost: Our intelligence? Why? American Psycholo-
Kuhil, T. S. (1970). The structure of scientific revolutions (2nd ed.). gist. 19, 871-882.
Chicago: Chicago University Press. "Meyers, I.E., & McCaulley, M. H. (1985). Manual: A guide to the
Lamb, R. R., & Prediger, D. J. (1981). The unisex, edition of the ACT development and use of the Myers-Briggs Type Indicator. Palo Alto,
Interest Inventory. Iowa City, IA: American College Testing. CA: Consulting Psychologists Press. (17)
*Lapsley, D. K., & Enright, R. D. (1979). The effects of social desirabil- *Milgram, R. M., & Milgram, N. A. (1977). The effect of test content
ity, intelligence, and milieu on an American validation of the Conser- and context on the anxiety-intelligence relationship. Journal of Ge-
vatism scale. Journal of Social Psychology, 107, 9-14. netic Psychology, 130, 121-127.
•Lessing, E. E., & Barbera, L. (1973). Relationships among scores on 'Mohan, V. (1979). Scores on EPPS and their relation with standard
the California Test of Personality, the IPAT Children's Personality progressive matrices. Indian Psychological Review, 17, 6-8. (2)
Questionnaire, and the IPAT Jr.-Sr. High School Personality Question- Moore, B. V. (1941). Analysis of tests administered to men in engi-
naire. Educational and Psychological Measurement, 33, 985-992. neering defense training courses. Journal of Applied Psychology, 25,
(2) 619-635.
"Ley, P., Spelman, M. S., Davies, A. D. M., & Riley, S. (1966). The *Mueller, J. H., Elser, M. J., & Rollack, D. N. (1993), Test anxiety and
relationships between intelligence, anxiety, neuroticism and extrover- implicit memory. Bulletin of the Psychonomic Society, 31, 531-533.
sion. British Journal of Educational Psychology, 36, 185-191. *Neal, C. M. (1968). Sex differences in personality and reading ability.
Liebert, R., & Morris, L. (1967). Cognitive and emotional components Journal of Reading. 11, 609-614, 633. (2)
of test anxiety: A distinction and some initial data. Psychological "Nye, L. G., & Collins, W. E. (1993). Some personality and aptitude
Reports, 20, 975-978. characteristics of air traffic control specialist trainees. Aviation, Space,
*Lloyd, J., & Barenblatt, L. (1984). Intrinsic intellectuality: Its relations and Environmental Medicine, 64, 711-716.
to social class, intelligence, and achievement. Journal of Personality NTIS Database. (1974-1994, March 1995). Springfield, VA: U. S. De-
and Social Psychology, 46, 646-654. partment of Commerce, National Technical Information Service [Pro-
"Long, H. B. (1970). Relationships of selected personal and social vari- ducer], Available from: Producer.
ables in conforming judgment. Journal of Social Psychology, 81, •Olson, D. R., & Schlottmann, R. S. (1980). Role of model's affect in
177-182. (4) test anxiety. Psychological Reports, 47, 956-958.
Long, L. (1945). Relationship between interests and abilities: A study *Opolot, J. A. (1977). Reliability and validity of Smith's quick measure
of the Strong Vocational Interest Blank and the Zyve Scientific Apti- of achievement motivation scale. British Journal of Social and Clinical
tude Test. Journal of Applied Psychology, 29, 191-198. Psychology, 16, 395-396.
Lorge, I. (1940). Intelligence: Its nature and nurture. The 39th Yearbook •O'Reilly, R. P., & Wightman, L. E. (1971). Improving the identification
39(Pt. I), 275-281. of anxious elementary school children through the use of an adjusted
Lowman, R. L., Williams, R. E., & Leeman, G. E. (1985). The structure anxiety scale. Journal of Educational Measurement, 8, 107-112.
 244 ACKERMAN AND HEGOESTAD

*Orpen, C. (1983). The development of an adjective check-list measure *Sassenrath, J. M. (1967). Anxiety, aptitude, attitude, and achievement.
of managerial need for achievement. Psychology: A Quarterly Journal Psychology in the Schools, 4, 341-346. (2)
of Human Behavior, 20, 38-42. *Satterly, D. J. (1979). Covariation of cognitive styles, intelligence and
•Paul, G. L., & Eriksen, C. W. (1964). Effects of test anxiety on "real- achievement. British Journal of Educational Psychology, 49, 179-
life" examinations. Journal of Personality, 32, 480-494. 181.
Pearson, K. (1906-1907). On the relationship of intelligence to size Schooler, C. (1987). Cognitive effects of complex environments during
and shape of head, and to other physical and mental characters. Biome- the life span: A review and theory. In C. Schooler & K. W. Schaie
trika, 5, 105-146. (Eds.), Cognitive functioning and social structure over the life course
Peterson, J. (1925). Early conceptions and tests of intelligence. Chi- (pp. 24-49). Norwood, NJ: Ablex.
cago: World Book. *Schurr, K. T., Ruble, V. E., & Henriksen, L. W. (1988). Relationships
Phillips, W. S., &. Osbome, R. T. (1949). A note on the relationship of of Myers-Briggs Type Indicator personality characteristics and self-
the Kuder Preference Record scales to college marks, scholastic apti- reported academic problems and skill ratings with Scholastic Aptitude
tude, and other variables. Educational and Psychological Measure- Test scores. Educational and Psychological Measurement, 48, 187-
ment, 9, 331-339. 196.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Pintner, R. (1923). Intelligence testing. New York: Holt. Segal, D. (1934). Differential prediction of scholastic success. School
This document is copyrighted by the American Psychological Association or one of its allied publishers.

*Posey, T. B., & Runyon, 1. C. (1974). Intercorrelations of numerosity and Society, 39, 91-96.
recognition, IQ, and anxiety in college students. Perceptual and Motor *Sepie, A. C., & Keeling, B. (1978). The relationship between types of
Skills, 38, 1054. anxiety and under-achievement in mathematics. Journal of Educa-
Prediger, D. J. (1982). Dimensions underlying Holland's hexagon: Miss- tional Research, 72, 15-19.
ing link between interests and occupations? Journal of Vocational Sharp, S. E. (1899). Individual psychology: A study in psychological
Behavior, 21, 259-287. method. American Journal of Psychology, 10, 329-391.
PsycUT [Database). (1974-1994, March 1995). Washington, DC: Signorella, M. L., & Jamison, W. (1986). Masculinity, femininity, an-
American Psychological Association [Producer]. Available from: Sil- drogyny and cognitive performance: A meta-analysis. Psychological
verPlatter [Distributor]. Bulletin, 100, 207-238.
Randahl, G. J. (1991). A typological analysis of the relations between *Sipps, G. J., Berry, G. W., & Lynch, E. M. (1987). WAIS-R and social
measured vocational interests and abilities. Journal of Vocational intelligence: A test of established assumptions that uses the CP1. Jour-
Behavior, 38, 333-350. nal of Clinical Psychology, 43, 499-504.
*Reid, J. B., Palmer, R. L., Whitlock, J., & Jones, J. (1973). Computer- •Smith, C. P. (1964). Relationships between achievement-related mo-
assisted-instruction performance of student pairs as related to individ- tives and intelligence, performance level, and persistence. Journal of
ual differences. Journal of Educational Psychology, 65, 65-73. Abnormal and Social Psychology, 68, 523-532.
"Richmond, V. P., & Dickson-Markman, R (1985). Validity of the Writ- *Smith, C. P. (1965). The influence on test anxiety scores of stressful
ing Apprehension Test: Two studies. Psychological Reports, 56, 255- versus neutral conditions of lest administration. Educational and Psy-
259. chological Measurement, 25, 135-141.
•Robertson, I. T., & Molloy, K. J. (1982). Cognitive complexity neuroti- Snow, R. E. (1989). Aptitude-treatment interaction as a framework for
cism and research ability. British Journal of Educational Psychology, research on individual differences in learning. In P. L. Ackerman, R. J.
52, 113-118. Sternberg, & R. Glaser (Eds.), Learning and individual differences:
Rocklin, T. (1994). Relation between typical intellectual engagement Advances in theory and research (pp. 13-59). New York: Freeman.
and Openness: Comment on Goff and Ackerman (1992). Journal of Snow, R. E., Corno, L., & Jackson, D., III. (1996). Individual differ-
Educational Psychology, 86, 145-149. ences in affective and conative functions. In D. C. Berliner & R. C.
Roe, A. (1956). The psychology of occupations. New "bbrk: Wiley. Calfee (Eds.), Handbook of educational psychology (pp. 243-310).
Roe, A., & Klos, D. (1969). Occupational classification. Counseling New %rk: Macmillan.
Psychologist, 1, 84-89. Snow, R. E., Kyllonen, P. C., & Marshalek, B. (1984). The topography
*Rolfhus, E. L., & Ackerman, P. L. (1996). Self-report knowledge: At of ability and learning correlations. In R. J. Sternberg (Ed.), Advances
the crossroads of ability, interest, and personality. Journal of Educa- in the psychology of human intelligence (Vol. 2, pp. 47-103). Hills-
tional Psychology, 88, 174-188. dale, NJ: Erlbaum.
Rosenthal, R. (1995). Writing meta-analytic reviews. Psychological Bul- Sorenson, H. (1933). Mental ability over a wide range of adult ages.
letin, 118, 183-192. Journal of Applied Psychology, 17, 729-741.
Rounds, J., & Tracey, T. J. (1993). Prediger's dimensional representation Sorenson, H. (1938). Adult abilities: A study of university extension
of Holland's RIASEC circumplex. Journal of Applied Psychology, students. Minneapolis: University of Minnesota Press.
78, 875-890. »Sperbeck, D. J., Whitbourne, S. K., & Hoyer, W. J. (1986). Age and
*Samuel, W. (1980). Mood and personality correlates of IQ by race Openness to Experience in autobiographical memory. Experimental
and sex of subject. Journal of Personality and Social Psychology, Aging Research, 12, 169-172.
38, 993-1004. (2) Spielberger, C. D. (1972). Anxiety as an emotional state. In C. D. Spiel-
*Sarason, I. G. (1957). Test anxiety, general anxiety, and intellectual berger (Ed.), Anxiety: Current trends in theory and research (Vol. 1,
performance. Journal of Consulting Psychology, 21, 485-490. pp. 23-49). New York: Academic Press.
*Sarason, I. G. (1959). Intellectual and personality correlates of test Spielberger, C. D., Anton, W, & Bedell, J. (1976). The nature and
anxiety. Journal of Abnormal and Social Psychology, 59, 272-275. treatment of test anxiety. In M. Zuckerman & C. D. Spielberger (Eds.),
*Sarason, I. G. (1961). Test anxiety and the intellectual performance of Emotion and anxiety: New concepts, methods, and applications (pp.
college students. Journal of Educational Psychology, 52, 201-206. 317-345). Hillsdale, NJ: Erlbaum.
(2) Sternberg, R. J. (1990). Metaphors of mind: Conceptions of the nature
*Sarason, I. G. (1963). Test anxiety and intellectual performance. Jour- of intelligence. Cambridge, England: Cambridge University Press.
nal of Abnormal and Social Psychology, 66, 73-75. (4) Strong, E. K., Jr. (1943). Vocational interests of men and women. Palo
*Sarason, S. B., & Mandler, G. (1952). Some correlates of test anxiety. Alto, CA: Stanford University Press.
Journal of Abnormal and Social Psychology, 47, 810-817. Strong, E. K., Jr. (1952). Twenty-year follow-up of medical interests.
 INTELLIGENCE, PERSONALITY, AND INTERESTS 245

InL. L. Thurstone (Ed.), Applications of psychology: Essays to honor Wechsler, D. (1950). Cognitive, conative, and non-intellective intelli-
Walter V. Bingham (pp. 111-130). New York: Harper. gence. American Psychologist, 5, 78-83.
Strong, E. K., Jr. (1955). Vocational interests 18 years after college. Weisberg, R. W. (1986). Creativity: Genius and other myths. New York:
Minneapolis: University of Minnesota Press. Freeman.
Tellegen, A. (1982). Brief manual for the Multidimensional Personality Welsh, G. S. (1971). Vocational interests and intelligence in gifted ado-
Questionnaire (MPQ). Minneapolis, MM: Author. lescents. Educational and Psychological Measurement, 31(]), 155—
Tellegen, A. T., & Waller, N. G. (in press). Exploring personality through 164.
test construction: Development of the Multidimensional Personality Welsh, G. S. (1975). Creativity and intelligence: A personality ap-
Questionnaire. In S. R. Briggs & J. M. Cheek (Eds.). Personality proach. Chapel Hill, NC: Institute for Research in Social Science.
measures: Development and evaluation (Vol. 1). Greenwich, CT: JAI Welsh, G. S. ( 1977). Personality correlates of intelligence and creativity
Press. in gifted adolescents. In J. C. Stanley, W. C. George, & C. H. Solano
Terman, L. M. (1924). The mental test as a psychological method. Psy- (Eds.), The gifted and the creative: A fifty-year perspective (pp. 197-
chological Review, 31, 93-117. 221). Baltimore: Johns Hopkins University Press.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Terman, L. M. (1954). Scientists and nonscientists in a group of 800 *Wexner, L. B. (1954). Relationship of intelligence and the nine scales
This document is copyrighted by the American Psychological Association or one of its allied publishers.

gifted men. Psychological Monographs, f5S(7, Whole No. 378), 1- of the Minnesota Multiphasic Personality Inventory. Journal of Social
44. Psychology. 40, 173-176.
Thomdike, E. L. (1920). Intelligence examinations for college entrance. Whipple, G. M. (1914). Manual of mental and physical tests. Pi. 1:
Journal of Educational Research, 1, 329-337. Simpler processes (3rd ed., including supplementary references 1914
Thomdike, E. L. (1921). On a new plan of admitting students at Colum- to 1919). Baltimore: Warwick & York.
bia University. Journal of Educational Research, 4, 95-101. Whipple, G. M. (1922). Intelligence tests in colleges and universities.
Thurstone, L. L. (1919). Mental tests for college entrance. Journal of The 21st yearbook, 21, 253-270.
Educational Psychology, 10, 129-142. *Whitbourne, S. K. (1976). Test anxiety in elderly and young adults.
Thurstone, L. L. (1938). Primary mental abilities. Psychometric Mono- International Journal of Aging and Human Development, 7, 201-
graphs, 1(9), 1-121. 210. (2)
Toops, H. A. (1926). The status of university intelligence tests in 1923- Willis, S. S., & Tosti-Vasey, J. L. (1990). How adult development, intel-
24. Journal of Educational Psychology, 17, 22-36, 110-124. ligence, and motivation affect competence. In S. L. Willis & S. S.
Tracey, T, & Rounds, J. (1993). Evaluating Holland's and Gati's voca- Dubin (Eds.), Maintaining professional competence: Approaches to
tional-interest models: A structural meta-analysis. Psychological Bul- career enhancement, vitality, and success throughout a work life (pp.
letin, 113, 229-246. 64-84). San Francisco: Jossey-Bass.
Triggs, F. O. (1943 ). A study of the relation of Kuder Preference Record •Wilson, R. G., & Lynn, R. (1990). Personality, intelligence compo-
scores to various other measures. Educational and Psychological nents and foreign language attainment. Educational Psychology, 10,
Measurement, 3, 341-354. 57-71.
Tupes, B.C., & Christal, R. E. (1961). Recurrent personality factors Wissler, C. (1901). The correlation of mental and physical tests [Mono-
based on trait ratings (Rep. No. ASD-TR-61 -97). Lackland Air Force graph]. Psychological Review, J6(Suppl. 3), 1-62.
Base, TX: U. S. Air Force, Aeronautical Systems Division, Personnel Wright, L. (1969). Bibliography on human intelligence (Pub. No.
Laboratory. 1839). Washington, DC: U. S. Department of Health, Education, and
Turner, R. G., Willerman, L., & Horn, J. M. (1976). Personality corre- Welfare.
lates of WAIS performance. Journal of Clinical Psychology, 32, 349- Yerkes, R. M., Bridges, J. W., & Hardwick, R. S. (1915). A point scale
354. (2) for measuring mental ability. Baltimore: Warwick & York.
Vemon, P. E. (1950). The structure of human abilities. New York: Wiley. Yoakum, C. S., & Yerkes, R. M. (1920). Mental tests in the American
*Vidler, D. C. (1974). Convergent and divergent thinking, test-anxiety, Army. London: Sidgwick & Jackson.
and curiosity. Journal of Experimental Education, 43, 79-85. *Zeidner, M. (1990). Does test anxiety bias Scholastic Aptitude Test
'Walter, D., Denzler, L. S., & Sarason, I. G. (1964). Anxiety and the performance by gender and sociocultural group? Journal of Personal-
intellectual performance of high school students. Child Development, ity Assessment, 55, 145-160.
35, 917-926. (2) *Zeidner, M. (1991). Test anxiety and aptitude test performance in an
'Waters, C. W. (1974). Multi-dimensional measures of novelty experi- actual college admissions testing situation. Personality and Individual
encing, sensation seeking, and ability: Correlational analysis for male Differences, 12, 101-109.
and female college samples. Psychological Reports, 34, 43-46. (2) Zytowski, D. G. (1986). Comparison of Roe's and Holland's occupa-
*Watson, J. M. (1988). Achievement Anxiety Test: Dimensionality and tional classifications: Diverse ways of knowing. Journal of Counseling
utility. Journal of Educational Psychology, 80, 585—591. Psychology, 33, 479-481.
Webb, E. (1915 ). Character and intelligence. British Journal of Psychol-
ogy Monograph (Suppl. 3), 1-113. Received April 27, 1995
Wechsler, D. (1940). Nonintellective factors in general intelligence [Ab- Revision received June 19, 1996
stract]. Psychological Bulletin, 37, 444-445. Accepted June 20, 1996 •