Psychology of Aesthetics, Creativity, and the Arts © 2013 American Psychological Association

2013, Vol. 7, No. 3, 214-228 1931-3896/13/$12.0O DOI: 10.1037/a0031975

A Multilevel Meta-Analysis of the Relationship

Between Creativity and Schizotypy

Selcuk Acar and Sedat Sen

The University of Georgia

The purpose of this study is to obtain a more coherent picture of the relationship between creativity and
schizotypy. A multilevel approach was used to ensure the statistical independence of multiple effect sizes
from a single study. The mean effect size was r = .07 {SD = .20) based on the analyses from 45 articles
yielding 268 effect sizes. The analyses, including five moderators (i.e., the type of schizotypy, measure
of schizotypy, measure of creativity, content of creativity test, and index of creativity measure)
within a three-level model, indicated that the type of schizotypy was the only significant predictor.
Positive-Impulsive and Unspecified schizotypy each had positive relationships with creativity (r =
.14, 95% CI = .12 and .17), whereas Negative-Disorganized schizotypy was negatively related (r =
- . 0 9 , 95% CI = - . 1 2 , -.06). None of the creativity-related moderators explained the variation. Our
findings revealed that the link between creativity and schizotypy is not uniform and that positive
symptoms represent the intersection of creativity and schizotypy.
,1
Keywords: creativity, schizotypy, meta-analysis, multilevel

The link betweeti psychopathology and creativity has been full-blown and subclinical levels (i.e., psychoticism and schizo-
discussed by researchers in differetit fields (Becker, 1978; Ey- typy) of psychosis with creativity. According to this, creativity has
senck, 1993; Karlsson, 1978; Simotiton, 2005). Some researchers no relationship or a negative relationship with full-blown psycho-
embraced the idea of an intimate association between psychopa- sis, whereas psychoticism and schizotypy can be positively related
thology and creativity (Andreasen, 1987; Lombroso, 1891; to it. Therefore, a closer investigation of the subclinical or milder
Pretitky, 1980, 1989), whereas some others discounted this possi- forms of psychosis is important to the field of creativity.
bility (Brod, 1997). A third perspective acknowledged the com- Schizotypy is a critical construct for understanding the relation-
monalities between the two (i.e., Storr, 1976) but did not suggest ship between creativity and psychopathology. Claridge, Pryor, and
the causal role of psychopathology (Rothenberg, 1979). The in- Watkins (1990) argued that the links between creativity and mental
vestigation of the type of relationship faces at least two challenges. illnesses are most obvious in schizotypy. Kinney et al. (2001)
First, the relationship between psychopathology and creativity is found that peak creativity was observed among those who exhibit
two-way (Runco, 1991), and multiple links exist between the two any two of the schizotypy characteristics and schizotypal person-
(Lauronen et al., 2004). Second, the relationship is not uniform. alities. Also, creativity was lower among those who did not indi-
The association between creativity and mental disorders is at best cate any sign of schizotypy and was lowest among schizophrenics.
fragile and not clearly observed in all types of psychopathologies Theoretical and empirical works underline the construct of schizo-
and all forms of creativity (Lauronen et al, 2004, p. 81). Thus, it typy, and thus, it is the focus of this article. Eugene Bleuler
seems more useful to focus on each psychopathology or trait (1911/1950) used the term "latent schizophrenia" to describe a less
separately. severe and nonpsychotic manifestation of schizophrenia. Rado
The putative link between psychopathology and creativity has (1953) was the first who coined the term "schizotype" which was
actually been investigated in psychosis more than other types of abbreviated from "schizophrenic genotype" to refer to a hereditary
psychopathology (Eysenck, 1993). Other works (Prentky, 1989; disposition to schizophrenia. This term was originally derived
Eysenck, 1994) tend to emphasize the differential relationship of from schizophrenia to describe certain schizophrenia-related be-
haviors. Schizotypy was defined as "a set of behavioral, affective,
and cognitive 'eccentricities'" (Brod, 1997, p. 276) which consti-
tute the foundations of psychotic illnesses. In an earlier work,
This article was published Online First March 4, 2013. Meehl (1962) identified four traits of schizotypy including cogni-
Selcuk Acar and Sedat Sen, Department of Educational Psychology, The tive slippage, interpersonal aversiveness, anhedonia, and ambiva-
University of Georgia. lence. Later, the emphasis changed from anhedonia to personality
We appreciate the valuable contributions of Dr. Rod Dishman and
features and symptoms (e.g., Raine, 1991; Venables, Wilkins,
Samuel Barry.
Mitchell, Raine, & Bailes, 1990). The typical measures of schizo-
Correspondence concerning this article should be addressed to Selcuk
Acar, University of Georgia, 350 Aderhold Hall, Athens, GA 30602. typy involve those specific personality features and symptoms
E-mail: acarse@uga.edu rather than atihedonia.

214
 CREATIVITY AND SCHIZOTYPY 215

Although all psychotic individuals exhibit most of the schizo- thinking can lead to generating a greater number of ideas (thus,
typal characteristics, not all individuals exhibiting schizotypal giving greater potential for creativity), because overinclusive
characteristics are psychotic. The latest description of schizoptypal thinkers possess wider associative networks and have access to
personality in the Diagnostic Statistical Manual of Mental Disor- precociously activated memory and knowledge (Claridge &
ders (DSM-IV-TR) consists of nine symptoms: (a) ideas of refer- Beech, 1995). As a result of this, fiat associative hierarchy has
ence (excluding delusions of reference); (b) odd beliefs or magical been proposed for both creative (Mednick, 1962) and schizo-
thinking that influences behavior and is inconsistent with subcul- phrenic (Mednick, 1958) individuals. For both, high arousal has
tural norms (e.g., superstitiousness, belief in clairvoyance, telepa- been held responsible for a flat associative hierarchy (Martindale,
thy, "sixth sense," or bizarre fantasies or preoccupations; (c) 1975; Mednick, 1958). Based on those findings and theoretical
unusual perceptual experiences, including bodily illusions; (d) odd links, Hasenfus and Magaro (1976) maintained that ideational
thinking and speech (e.g., vague, circumstantial, metaphorical, or fiuency and overinclusion could be similar cognitive processes
stereotyped); (e) suspiciousness or paranoid ideation; (f) inappro- resulting in perception of irrelevant or minimally related concepts
priate or constricted affect; (g) behavior or appearance that is odd as closely associated because of a failure to preserve conceptual
eccentric or peculiar; (h) lack of close ÍTiends or confidants other boundaries (Cameron, 1938). This has been considered a deficit in
than first degree relatives; and (i) excessive social anxiety that schizophrenics whereas it is regarded as a reason for higher cre-
does not diminish with familiarity and tends to be associated with ativity because it (i.e., overinclusive thinking) leads to making
paranoid fears rather than negative judgments about self (Ameri- more and diverse number of associations.
can Psychiatric Association, 2000). The second commonality was suggested by Kris (1952) who
As Claridge (1997) puts it, schizotypy is a dimension of psy- argued that both creative and schizophrenic individuals can expe-
chosis representing "the less deviant bedfellow of 'schizophrenia'" rience regression, albeit in different ways. Creative individuals
(p. 3). The link between the two is conunon genetic loading. Some regress in the service of the ego and schizophrenics usually fail to
studies have shown that individuals with schizotypal personality manage it. 'The regression in the service of the ego can become
disorder (SPD) exhibit the same genotype as those with schizo- controllable at the schizotypal level. The third commonality is
phrenia but only partially (Clementz, Grove, Katsaniz, & Iacono, reduced latent inhibition (Brugger & Graves, 1997; Green &
1991; Kendler et al., 1991). Walker and Gale (1995) put forward Williams, 1999; Weinstein & Graves, 2001). Latent inhibition is
that negative symptoms, including withdrawal and blunting, can be defined as decreased leaming performance when an individual or
observed in schizophrenics. Although there is an obvious overlap- animal is preexposed (but not conditioned) to stimuli (Lubow,
ping, the two are not the same (identical). According to the Diag- 1973). Lubow, Ingberg-Sachs, Zalstein-Orda, and Gewitz (1992)
nostic and Statistical Manual of Mental Disorders (American Psy- and Lipp and Vaitl (1992) found a reduced latent inhibition with
chiatric Association, 20(X)), schizotypal personality disorder is listed higher schizotypy scores, especially in Schizotypy Traits Ques-
on Axis n (personality disorder), whereas schizophrenia is on Axis I. tionnaire (STA). The contribution of reduced inhibition to creative
Positive symptoms are what distinguish SPD from schizophrenia. thinking lies in the fact that seemingly irrelevant information is
Deficits in frontal structure and function were observed in both SPD preserved and used in problem solving (Burch, Hemsley, Pavelis,
and schizophrenia, but deficits in the limbic circuit were evident only & Corr, 2006; Dorfman, Martindale, Gassimova, & Vartanian,
in schizophrenics (Walker & Gale, 1995). 2008). Likewise, Carson, Peterson, and Higgins (2003) found
There is also a tendency to generalize schizotypy to psychosis- lower latent inhibition among lifetime creative achievers than with
proneness as a comprehensive construct, including psychoticism low creative achievers. Psychological correlates of creativity, in-
and anhedonia and schizoidia. They are sometimes even used cluding openness and extroversion, were also related to reduce
interchangeably in the Uterature. Because a focused meta-analytic latent inhibition (Peterson & Carson, 2000; Peterson, Smith, &
study (Acar & Runco, 2012) has already been conducted on the Carson, 2002). Kéri (2011) found that the relationship between
relationship between psychoticism and creativity, psychoticism as creative achievements and unusual experiences did not retain when
measured by the Psychoticism (P) scale of Eysenck Personality latent inhibition was considered.
Questionnaire (EPQ; Eysenck & Eysenck, 1975) and EPQ-R (Ey- The fourth and last crucial link concems the magical and un-
senck & Eysenck, 1991) has been excluded from this study. usual experiences of schizophrenics that allow for vivid imagina-
Therefore, this study is a follow-up meta-analysis of a broader tion and expanded horizons of thought. These are also known as
construct of schizotypy. "psychedelic" experiences (Siegler, Osmond, & Mann, 1971).
Some of the hypothesized links between creativity and schizo- Schizophrenics can be creative when they take advantage of such
typy come from the literature on the relationship between creativ- experiences (Bowers & Ereedman, 1966). Not surprisingly, schizo-
ity and schizophrenia. Researchers have suggested several links phrenics make more unusual and idiosyncratic associations than
between creativity and schizophrenia and, therefore, between cre- does the control group (Buss, 1966; Chapman & Chapman, 1973;
ativity and schizotypy, too. The first link is overinclusive thinking Moran, Mefferd, & Kimble, 1964). However, schizophrenics are
that is defined as a failure to preserve conceptual boundaries which not necessarily creative. Actually, in most instances, they cannot
lead to the incorporation of irrelevant ideas, generation of more be creative, because schizophrenia is involuntary, and such expe-
distant associations, and thinking in a more abstract and less lucid riences become uncontrollable. Therefore, schizotypy, rather than
way (Cameron, 1938; Eysenck, 1993, 1995; Payne, Mattusek, & schizophrenia, can lead to psychedelic experiences, which can be
George, 1959; Weinstein & Graves, 2002). Andreasen and Powers controlled at subchnical levels.
(1975) found that highly creative writers are overinclusive in their More recent works underlined the importance of other factors
thinking. Payne, Caird, and Laverty (1964) reported that overin- that helps to understand the association between creativity and
clusion is frequently observed in schizophrenics. Overinclusive schizotypy. People who had the T/T genotype, a variant of the
216 ACAR AND SEN

neuregulin 1 (NRGl) gene, (SNP8NRG243177/rs6994992), ex- & Broks, 1984); or using those focusing on certain behaviors, such
hibited higher psychosis risk (Kéd, Kiss, & Kelemen, 2009) and as the Revised Anhedonia Scale (Chapman & Chapman,1978;
superior creative achievements and creative-thinking performance Chapman, Chapman, & Raulin, 1976), Perceptual Aberration
(Kéd, 2009). Weinstein and Graves (2002) investigated the pos- (Chapman, Chapman, & Raulin, 1978), and Magical Ideation
sibility that the association of creativity and schizotypy could be Scale (Eckblad & Chapman, 1983). The scope of the measures
related to right hemisphere bias. They found that a response may have an influence on effect sizes. For example, some mea-
critedon requidng dght hemisphere processing is related to indi- sures could yield remarkably higher or lower effect sizes just
vidual differences in both creativity and schizotypy. Folley and because they focus on a particular symptom. Therefore, we in-
Park (2005) reported that highly creative schizotypes prefer to cluded this factor as another moderator.
employ the right prefrontal cortex activation. Considedng the perspectives of the creativity literature, similar
All of these links could be the basis for a relationship between issues related to definition and measurement need to be consid-
schizotypy and creativity, but studies have reported different, even ered. There are numerous ways to measure creativity, including
conflicting, results. Therefore, potential factors underlying the personality-based, self-reports, divergent thinking tests, associa-
variation in the effect size need to be considered to accurately tional tests, and specific tests, such as the Barron—Welsh Art Scale
explain the nature of the relationship. Several theoretically impor- (Barron & Welsh, 1987). Particularly, for this study, they could be
tant moderators, including the dimension or type of schizotypy, the categodzed as personality-based versus performance tests. Theo-
measure of schizotypy, the measure of creativity, the content of retically, personality-based tests can have a higher relationship
creativity test, and the index of creativity measure, will be dis- with schizotypy, which is also considered a personality trait
cussed in the following section. (Raine, Lencz, & Mednick, 1995), compared with other measures.
Schizotypy is not a unidimensional trait. Based on the reviews In a way, this relationship may stem from method vadanee, be-
of several factor analytic studies (Claridge et al., 1996; Gruzelier, cause both use the same method (i.e., self-report). Different indices
1995; Venables, 1995), Brod (1997) argued that schizotypy con- of creativity are reported within some performance tests, specifi-
sists of at least four factors. The first factor has been labeled as cally divergent thinking and word-association tests. Although there
"unusual experiences," "unreality," "perceptual eccentdcity," or are four major indices, including fluency, odginality, flexibility
"positive schizotypy." This profile is prone to magical thinking, and elaboration, only the first two were reported in the studies.
perceptual aberration, hallucinations, hypomania, paranoid ide- Therefore, it would be interesting to see if odginality and fluency
ation, and suspiciousness. The second factor involves withdrawn scores have differential relationships with schizotypy.
schizoid traits, which include social or physical anhedonia and The test content is also important. Creativity tests can be clas-
introversion. The third factor, known as cognitive disorganization, sified as verbal and figurai in terms of content, and they can trigger
consists of the symptoms of social impairment and anxiety, atten- different cognitive processes. Verbal content can have different
tion difficulties, and neuroticism. The fourth factor is impulsive effects on performance, because a major descriptor of schizotypal
nonconformity, which includes the symptoms of psychoticism, personality is "vague, circumstantial, metaphodcal, overelaborate,
extraversion, and impulsiveness and mood-related disinhibition. or stereotyped" language (American Psychiatdc Association,
Of course, those dimensions have some overlapping traits. For 2000, p. 701). Even though this may be seen as a positive thing for
instance, borderline personality is a common trait of both the third unusual thinking, it implies a cognitive impairment and difficulty
and fourth factors, and magical thinking is found in both the first to perceive and use the test matedal as expected.
and third factors. More importantly, positive schizotypy or unusual Finally, the index of creativity that is used can influence the
expedences and impulsive nonconformity are loaded on the same relationship. This actually did explain much of the vadanee in the
factor with extroversion, whereas negative schizotypy or intro- previous meta-analysis on psychoticism (Acar & Runco, 2012). Odg-
vertive anhedonia and cognitive disorganization are loaded on inality and, more specifically, uniqueness scores had significantly
introversion (Claridge et al., 1996). This perspective tallies with higher relationships with the P scale of EPQ. As mentioned above,
that of Prentky (1980), who offered a model suggesting different schizotypy can allow individuals to make unusual associations and to
creative processing in the introverted and extroverted types of give odginal responses. So, a higher correlation with odginality scores
schizotypy. Brod (1997) argued that "the urge for [creative] ex- compared with other indices can be expected for schizotypy, too.
pression must be the outcome of interacting effects, with the more Clearly, there are many potential factors that influence the
extravertive traits of positive schizotypy and impulsive noncon- relationship between creativity and schizotypy. There is no agree-
formity (p. 289)." Another paidng is possible when the anxiety ment about the nature of the relationship between schizotypy and
factor is considered. According to this, cognitive disorganization creativity in the previous empirical reports. Quite a few studies
and positive schizotypy (unusual expedences) represent a high (i.e., O'Reilly, Dunbar, & Bentall, 2001; Poreh, Whitman, & Ross,
anxiety-drive, and negative schizotypy (introverted anhedonia) and 1993; Preti & Vellante, 2007; Rawlings & Georgiou, 2004; Raw-
impulsive nonconformity imply a low-anxiety ddve. The multidi- lings & Toogood, 1997; Schuldberg, French, Stone, & Hebede,
mensional structure of schizotypy is important for this study, 1988) reported positive relationships between schizotypy and cre-
because different factors may have differential links to creativity. ativity. On the other hand, Schuldberg (1990) and Batey and
There are different instruments used to measure schizotypy. One Fumham (2008), for example, reported negative correlations be-
way to classify instruments is whether they target the whole tween the measures of creativity and schizotypy. Given the diver-
construct of schizotypy through instmments, such as the Oxford- sity in the creativity—schizotypy research, a meta-analytical re-
Liverpool Inventory of Feelings and Expedences (O-LIFE; Mason view would be ideal to synthesize the findings and resolve
& Claridge, 2006), Schizotypal Personality Questionnaire (SPQ; inconsistencies. This review can also take different study effects
Raine, 1991), and Schizotypy Traits Questionnaire (STA; Claddge into consideration and identify important factors behind the vari-
 CREATIVITY AND SCHIZOTYPY 217

ation between the individual studies. Thus, it would be possible to 2. Articles that did not use any statistical methods or those
make a data-based and a more generalizable conclusion about the that did not report statistics that allow the calculation of
mad-genius hypothesis rather than the evidence to be taken from an effect size were also excluded.
anecdotal (e.g., Rothenberg, 1990) or biographical accounts and
studies investigating specific populations (e.g., Claridge, Pryor, & 3. Articles on the relationship between creativity and psy-
Watkins, 1990; Post, 1994). To our knowledge, the aforemen- choticism were excluded, because a focused meta-
tioned relationship has never been summarized meta-analytically. analysis is already available (Acar & Runco, 2012).
Additionally, a meta-analysis of the relationship between cre-
ativity and schizotypy as a broader construct than the previous 4. Articles without a control group were excluded.
investigation on psychoticism (Acar & Runco, 2012) would be
5. Indirectly related articles were excluded.
interesting. The analyses of 119 effect sizes from 32 studies
indicated that the relationship can be quite high (r = .50) when The following criteria were used to define the set of articles to
psychoticism is measured by the Eysenck Personality Question- be included in the meta-analysis:
naire and uniqueness is the index of creativity. Otherwise, the
effect size was small (r = .16). Psychoticism is a unidimensional 1. A measure of the relationship between creativity and
construct (Eysenck & Eysenck, 1994; Chico & Ferrando, 1995), schizotypy was required.
whereas schizotypy is clearly multidimensional (Bentall, Claridge,
& Slade, 1989). Because of this, a more complex relationship with 2. Sufficient data for an effect size calculation (correlation
creativity should be expected. Following these purposes, this meta- values, mean and standard deviations, test statistics andp
analysis aims to answer the following questions. values) were required.

1. What is the strength of the relationship between schizo- 3. The search was limited to joumal articles published from
typy and creativity? January 1980 to April 2012.

2. How different is the nature of their relationship from that As a result, 45 articles fulfilled our inclusion criteria. Figure 1
between psychoticism and creativity? provides a flowchart showing which articles were included and
which were excluded along with the specified reasons for their
3. Which moderators explain the variability in the effect exclusion.
sizes?
Data Sources and Search Strategies
4. What conditions are contingent on positive, negative, or
zero correlation? In our search, we focused on published or in-press articles, disser-
tations, book chapters, and unpublished manuscripts. Thus, the term
Method "articles" will be used rather than "studies." Articles reported in the
last three formats were not included in our analyses but were used to
Study Variables find more articles as a step of the secondary search. The studies used
in this meta-analysis were located through a comprehensive literature
Because our main research question is the relationship between
search from January of 1980 through April of 2012 using variants of
creativity and schizotypy, the main outcome measures were the in-
the following search terms: creativity, schizotypy, schizophrenia, di-
struments assessing these two constructs. Thus, we used several
vergent thinking, originality, and fluency.
keywords to identify the articles, which include the outcome variables
needed to answer the target questions. Operational definitions of Several retrieval tools were used to locate target articles. First,
creativity and schizotypy were not explicitly provided in the articles electronic searches were conducted on several sources, including
included in this meta-analysis. However, instruments used in those PsycINFO (EBSCO), ERIC (WebSPIRS), Medline (PubMed), Ac-
studies can give researchers a general idea. In Table 1, the definitions ademic Search Premier (EBSCO), and Social Sciences Citation
and the scope of moderators, including the operational definitions and Index due to the partial overlap among multiple databases (Littell,
measures of creativity and schizotypy, were provided. Creativity was Corcoran, & Pillai, 2008). Second, the reference lists of each
measured with either self-report or personality-based measures relevant study were inspected. Third, hand searches were per-
(26.87%) or performance measures, including DT, associational, or formed in relevant journals, including the Creativity Research
Barron-Welsh tests (65.40%). Schizotypy was measured with gen- Joumal, Joumal of Creative Behavior, Psychology of Aesthetics,
eral (64.18%; i.e., 0-LIFE, SPQ, or STA) or specific measures Creativity, and the Arts, Schizophrenia Bulletin, Schizophrenia
(19.03%), including Anhedonia and Per-Mag scales. Those also Research, Personality and Individual Differences, and Thinking
served as moderators in the analyses. Skills and Creativity.

Inclusion and Exclusion Criteria Coding of Articles

Two independent judges selected the articles by screening the A coding form with basic study information, sample size, and
titles, abstracts, and whole texts considering the following criteria: effect size estimations along with potential moderators was devel-
oped. This form followed the guidelines of the 6th edition of the
1. Articles written in a language other than English were Publication Manual of the American Psychological Association
excluded. (APA, 2010) and Preferred Reporting Items for Systematic Re-
218 ACAR AND SEN

Table 1
Definitions and Scope of Moderators

Gender Male Samples with male groups above 75%

Female Samples with female groups above 75%
Mixed Samples with both males and females
Eminence Eminent Studies with eminent or highly successful individuals in a field of creativity
Noneminent Studies with normal individuals
Domains Arts Either participants were artists or measurement of creativity relied on artistic skills such as Welsh
Figure Preference Test (Welsh, 1949; Barron & Welsh, 1987)
General General population or usual measures of creativity
Measure of Performance Divergent thinking (DT): Alternate Uses, Consequences (Guilford, Christensen, Merrifield, &
creativity Wilson, 1978; Guilford & Hoepfner, 1971), Novel DT (Folley & Park, 2005), Similarities,
Instances, Uses, Pattern Meanings (Wallach & Kogan, 1965); Torrance Test of Creative
Thinking, (TTCT; Torrance, 1970, 1974), Word Huency (Benton & Hamsher, 1989; Goodglass
& Kaplan, 1972; Lezak, 1995; Spreen and Strauss, 1998), Category Fluency (Animals and
Boys' Names; Semel, Wiig, & Secord, 1995; Spreen and Strauss, 1998), Figurai Fluency (Ruff,
1996), Design Fluency (Five Point Test; Regard, Strauss, & Knapp, 1982), Verbal and
Drawing creativity (Miller & Tal, 2007) Letter Fluency, Thurstone Written Fluency Test,
(TWFT; Thurstone & Thurstone, 1962), Utility Test (Wilson, Merrifield, & Guildford, 1969),
Five Points Test (Regard, Strauss, & Knapp, 1982), Comprehensive Ability Battery (Hakstian
& Cattell, 1976)
Associational; WAT (Kent & Rosanoff, 1910; Miller & Chapman, 1983); Word Halo Test
(Armstrong 8i McConaghy, 1977); Remote Associates Test, (RAT; Mednick, 1959; Mednick &
Mednick, 1967), Bridge-the-Associative-Gap, (BAG; Gianotti et al, 2001) Barron-Welsh; B-W
(Welsh & Barron, 1963; Barron & Welsh, 1987); Polygons (Vanderplas & Garvin, 1959);
Welsh Figure Preference Test (Welsh, 1949); and Origence test (Rawlings & Georgiou, 2004)
Other cognitive; Conceptual expansion (Ward, 1994); Constraint of examples (Smith, Ward, &
Schumacher, 1993); Creative imagery (Finke, 1990); Insight problems (Karimi, Windmann,
Güntürkün, &. Abraham, 2007; Metcalfe, 1986; Stemberg & Davidson, 1982)
Creative personality or Adjective Checklist (ACL; Gough & Heilbrun, 1980), Creative Personality Scale (CPS: Gough,
self-reports 1979; Domino, 1970), Biographical Inventory of Creative Behaviours (BICB; Batey, 2007),
Self-rating of creativity (Batey, 2007), Experience of Creativity Questionnaire (ECQ; Nelson &
Rawlings, 2009), Lifetime Creativity Scales (Richards, Kinney, & Merzel, 1988), Independent
Activities Questionnaire (IAQ; Richards, Holland, & Lutz, 1967), Peak Vocational and
Avocational Creativity (Richards, Kinney, Lunde, Benet, & Merzel, 1988), Creativity Styles
Questionnaire-Revised (Kumar Kemmler, & Holman, 1997); How Do You Think (HDYT;
Davis & Subkoviak, 1975)
Measure of General The Oxford-Liverpool Inventory of Feelings and Experiences (O-LIFE; Mason & Claridge, 2006;
schizotypy Mason, Claridge, & Jackson, 1995), Schizotypal Personality Questionnaire (SPQ; Raine, 1991;
Klein, Andresen, & Jahn, 1997), Schizotypal Traits Questionnaire (STA; Claridge & Broks,
1984), Rust Inventory of Schizotypal Cognitions (RISC; Rust, 1988)
Specific Perceptual Aberration (Chapman, Chapman, & Raulin, 1978), Magical Ideation (Eckblad &
Chapman, 1983), Revised Social Anhedonia (Eckblad, Chapman, Chapman, & Mishlove, 1982)
Other Not used
Content of creativity
test Verbal All tests with verbal content including DT, WAT, CPS/ACL and other tests
Figurai All tests with figurai content including DT, B-W, and other tests
Both Composite of both verbal and figurai scores
Indices of creativity Fluency Number of total responses or total score
Originality Number of infrequent ideas including unique scores
Other Flexibility, elaboration and quality as well as none

views and Meta-analyses (PRISMA; Moher, Liberati, Tetzlaff, & Rater Reliability
Altman, 2009). Also, the publication year, type of schizotypy, type of
participant, type of creativity test, domain of creativity, and measures Search and coding procedures were independently conducted by
of creativity and schizotypy were all thoroughly coded. Each variable both authors to ensure the reliability of the results. Twenty percent
in the coding form had different levels. Gender was coded as male, of two forms were randomly selected to estimate reliability. A
female, or mixed. Type of participant was coded as being eminent two-way (effects and raters) intraclass correlation coefficient
versus noneminent. Almost 70% of the participants were university (ICC) for absolute agreement was .94.
students, and the rest consisted of artists, musicians, and writers.
The Calculation of Effect Sizes
There were four types of schizotypy, including negative, positive,
impulsive, and disorganized schizotypy as well as unspecified types. Because our research question is to examine the relationship
Instruments of schizotypy were categorized as general and specific. between creativity and schizotypy and most studies have provided
The domain was coded as general versus artistic. The details of the correlations between the continuous measures of these two con-
categories and definitions are provided in Table 1. structs, correlation was selected as the effect size measure. If
 CREATIVITY AND SCHIZOTYPY 219

57 Studies identified
excluded
6 studies indirectly related

excluded
4 studies provided inadequate statistics

excluded
2 studies without a control group

45 Studies were included

Figure L Elow chart for selection of studies.

correlation was not available, then effect sizes were estimated errors. In the fixed-effects approach, it is assumed that effect sizes
using procedures described by Lipsey and Wilson (2001). As the come fi-om the same distribution and that the population of effect
Pearson's product-moment correlation (r) is not normally distrib- sizes is homogeneous. Thus, sampling error is the only source of
uted, these effect sizes were converted to Fisher's zs in order to difference between the studies. However, in the random-effects
combine the effect sizes properly (Hedges & Olkin, 1985; approach, it is implicitly assumed that each effect size differs from
Rosenthal, 1994). A z-transformed correlation can be back- the population mean by subject-level sampling error and unob-
transformed to a Pearson correlation using an inverse Fisher trans- served random sources, resulting in heterogeneity (Lipsey & Wil-
formation to make the interpretation easier (Lipsey & Wilson, son, 2001). For this reason, the findings of a random-effects model
2001). might be applicable to broader settings, whereas fixed-effects'
results can only be applicable to a specific sample.
The Assessment of Potential Publication Bias A common issue in typical meta-analysis studies is obtaining
multiple effect sizes within a single study. Using multiple effect
Publication bias is a serious problem with meta-analytic studies
sizes in the traditional meta-analysis models may violate the as-
(Rosenthal, 1979). Although it is almost impossible to identify and
sumption of the statistical independence of the data. There are
include every study related to a research question, it should be
three strategies used to overcome this problem in a typical meta-
ensured that the meta-analyst conducted a comprehensive review.
analysis study (Hedges, 2007). The first strategy is to apply the
Fortunately, there are some methods that can be used to examine
same techniques used in the independent effect size situation.
publication bias, such as funnel plot (Light & Pillemer, 1984) and
Ignoring the dependency of the data may not be a problem when
Rosenthal's (1979) fail-safe N. A funnel plot is a graphical method
based on the assumption that the distribution of the effect sizes will multiple effect sizes are small within a study. The second strategy
be normal around the mean effect in the absence of heterogeneity is using dependent effect sizes to obtain a within-study summary
(Littell et al., 2008). The sample size is related to standard errors. and, then, summarizing the independent summaries across studies.
Larger sample sizes yield less error than smaller sample sizes. The third and most elegant strategy is to use multivariate meta-
Funnel plots can be created using effect sizes on the x-axes and analysis techtiiques. However, this strategy is less common than
standard errors or sample sizes on the y-axes. This will give us a the others due to its complexity.
triangle-shaped plot, which is narrower at the top and wider at the We adopted the multilevel modeling approach outlined in Kon-
bottom. In the absence of heterogeneity, effect sizes will be sym- stantopoulos (2011) and Hox (2002) to calculate overall effect size
metrically distributed around the mean effect sizes. estimates and to explain study variances. Multilevel models are
As noted, the second method used in this study is the fail-safe N, particularly useful when the data are nested. Our data are nested
which is the earliest and most commonly used estimate of publi- because we identified multiple effect sizes from a single article in
cation bias (Field & Gillet, 2010). A fail safe N number indicates the majority of the articles included. The use of traditional regres-
the number of new studies that would be needed to increase the p sion models is problematic due to the violation of statistical
value for the meta-analysis to above a desired level (p = .05; independence in this type of data structure. Multilevel models
Rosenthal, 1979). When a meta-analyst calculates a larger fail-safe (MLM) have been designed to deal with this problem. The popu-
N number, s/he can feel confident with the results of the study. larity of the multilevel analysis has increased remarkably in the
psychology literature (de la Torre, Camilli, Vargas, & Vernon,
2007; Dickerson & Kemeny, 2004; Severiens & Dam, 1998; Van
Statistical Analyses
den Bussche, Van den Norrtgate, & Reynvoet, 2009) produced
There are two types of analyses in the traditional meta-analysis: over the last two decades. This is primarily due to its capability of
fixed- and random-effects models (Hedges, 1992; Hedges & Ve- taking variations into account at different levels (Raudenbush &
vea, 1998; Hunter & Schmidt, 2000). The main difference between Bryk, 2002). Multilevel meta-analytic methods provide more flex-
these fixed- and random-effect approaches is the source of error. ibility than traditional methods (Hox, 2002; Hox & Leeuw, 2001).
The former assumes only one error, whereas the latter assumes two Because meta-analysis data can have a multilevel structure (as was
220 ACAR AND SEN

the case with our dataset), multilevel regression can be applied to enabled us to include study characteristics as explanatory variables
meta-analysis with subjects within studies at the first level and on level-1. Five moderators, including the type of schizotypy,
studies at the second level (Bryk & Raudenbush, 1992; Hox, measure of schizotypy, measure of creativity, content of creativity
2002). Unlike in standard multilevel analysis, raw data for all test, and index of creativity measure were included for specific
respondents is not available in multilevel meta-analysis. However, theoretical reasons as discussed in the beginning of this article. These
Bryk and Raudenbush (1992) pointed out that it is possible to covariates were included in the level-1 analyses to explain the article-
apply multilevel analysis to the summary statistics of meta- to-article variation of the reported correlation coefficients. Because all
analysis. They called this specific approach the Level 1 Variance the variables are categorical, the dummy coding was used for each
Known model. variable.
In order to take into account the dependency in the data, two- Restricted maximum likelihood estimation was used to estimate
level models for multivariate meta-analysis have been applied in the parameters. All analyses were conducted using SAS PROC
some studies (Berkey, Anderson, & Hoaglin, 1996; Gleser & MIXED in SAS, Version 9.2. (Konstantopoulos, 2011; Little,
Olkin, 2009). However, it is obvious that a two-level model does Milliken, Stroup, Wolfinger, & Schabenberger, 2007; Singer,
not account for higher levels of nesting in meta-analytic data 1998; van Houwelingen, Arenas, & Stijnen, 2002). In the multi-
(Konstantopoulos, 2011). It should be noted that the between- level meta-analysis, we first estimated the unconditional model
article within-level-3 unit variation and the between-level-3 unit without any covariate. Then, we reestimated the model adding five
variation constitute random variation in the three-level model. This covariates.
displays where the majority of the random variation exists. If the
majority of the variation comes from the third-level variance, then
it should be retained instead of estimating a second-level model
The Homogeneity Test
(Konstantopoulos, 2011). Three-level data structures are also very As a test of homogeneity analyses of effect sizes, Cochran's
common in the literature, such as students nested within classes heterogeneity statistic (Q), which is distributed as a chi-square
and classes nested within schools. Similarly, meta-analytic data in with k — I degrees of freedom, where k = the number of studies,
our study have a three-level structure, with effect sizes nested was calculated. A significant g-test indicates the variability among
within articles, and articles nested within authors. Three-level effect sizes not only due to sampling error but also between study
models are more complicated than two-level models and produce errors (Lipsey & Wilson, 2001). In addition, we also calculated the
an overall mean estimate and two variance component estimates at f statistic (Higgins, Thompson, Deeks, & Altman, 2003) which
Levels 2 and 3. Following the notations of Bryk and Raudenbush indicates the degree of heterogeneity in percentages. Based on the
(1992) and Konstantopoulos (2011), a three-level model can be formula given in Higgins Thompson, Deeks, and Altman (2003),
characterized as the following: an P percentage can be easily calculated using Cochran's hetero-
The first level (i.e., effect size level) can be represented with geneity statistic and the degrees of freedom values. The P ranges
equation 1 in which K,^ is independently and normally distributed from 0% (no heterogeneity) to 100% (high-level heterogeneity;
with a mean of TT,^ and variance of Uj, which is assumed to be Higgins et al., 2003).
known.

- -^ig + (1) Results

In the second level (the between-articles within-level-3-unit As already noted, a total of 268 effect sizes from 45 published
model), the unknown effect-size parameter ir varies around a studies were available for the analyses. The effect size estimates
level-3 unit g mean, namely (rs) ranged from —.42 to .83 with a mean of .07 and a standard
deviation of .20. A stem-and-leaf plot of 268 effect sizes to two
- ßo« + (2) decimal places can be seen in Figure 2. The total number of

where g = 1, . . ., n represents the level-3 units (i.e., authors).

Finally, at the third level, the level-3 unit means vary around an
overall mean y^o. Stem & Leaf

= 7oo (3) -.4 12

-.3 14668
where UQ^ is normally distributed with a mean of zero and -.2 0000011223344568899
between articles variance. Three-level can be represented in a -.1 000111223333445566777888889
-.0 111 12222233333344444444445555556666666788999
single model as follows: .0 000000111122223333444455556666667778888889999
.1 0000111112222222333333444444555566666677788888888899999
Yig = 7oo + «Oí + % + «« (4) .2 0000012222222333344455566666666666778888888899
.3 00113344667789
When there is any additional predictor, the model can be extended .4 0244678
adding estimates for each covariate as follows: .5 5
.6 1
n¡g = ßoo (5) .7 9
.8 3
Z-Transformed correlation coefficients were used as the depen-
dent in this study. The fiexibility of multilevel meta-analysis Figure 2. Stem-and-leaf plot of all effect sizes (rs).
 CREATIVITY AND SCHIZOTYPY 221
subjects in the studies was 39,564 with a median of 1(X) per study Table 2
and a range of 15 to 1,018. Descriptive Statistics of Transformed Effect Sizes
There were 45 level-2 units (articles), and within each article,
the number of effeet sizes ranged from 1 to 25. The descdptive Publication N Mean SD
statistics for the articles can be seen in Table 2. As indicated in the 1 12 0.045 0.228
table, there is considerable vadability within as well as between 2 4 0.023 0.314
articles. 3 4 0.184 0.133
The j2-statistic was found to be significant, (p < .05, x^ = 4 3 0.132 0.099
5 4 -0.028 0.218
1527.27, 267). The P percentage was calculated as 82.5. Based 6 1 -0.435 —
upon these two statistics, this dataset showed heterogeneity. We 7 5 0.306 0.169
investigated the potential publication bias in this study. The funnel 8 2 0.138 0.195
plot is presented in Figure 3. The absence of bias can be claimed 9 1 0.503 —
when this plot looks like a funnel (Begg, 1994). Hence, publication 10 4 0.09 0.09
11 5 0.342 0.153
bias does not seem to be a problem for the present study. We also 12 24 0.104 0.123
found tbat fail-safe N was 573, which revealed a lack of publica- 13 2 0.892 0.253
tion bias. 14 1 0.350
The primary results of three-level analyses estimates are 15 2 0.182 0
16 4 0.208 0.039
given in Tables 3 and 4, which show the fixed coefficients, 17 10 0.163 0.144
variance components, standard errors, and 95% confidence 18 20 -0.027 0.165
intervals (CI). As was the case with tbe multilevel analyses, we 19 25 0.021 0.190
started with an unconditional model, which can also be called 20 4 -0.060 0.060
an intercept-only model. Results of the unconditional model 21 4 -0.135 0.080
22 6 -0.061 0.138
(see Table 3) will be used for the purpose of comparison with 23 2 0.240 0.067
the main model (see Table 4). The overall effect size estimate 24 3 0.312 0.189
was .141 and was significant in the unconditional model. The 25 12 -0.003 0.247
second- and third-level variances were .020 and .033, respec- 26 25 -0.086 0.166
27 1 0.320 —
tively, and were found to be significant. Following the multi- 28 12 0.117 0.134
level research tradition, we continued with a random-effects 29 2 0.072 0.045
model. When five covariates were included in the unconditional 30 1 0.375
model (i.e., main model), the overall effect size estimate was 31 3 0.005 0.269
.230 and was still significant. The variance component esti- 32 1 0.377
33 2 0.321 0.420
mates did not cbange much. The seeond level variance was .012 34 3 0.251 0.112
and was significantly different from zero. The 95% CI was .008 35 4 0.044 0.252
to .016. The third level variance was .032 and was also signif- 36 5 0.181 0.084
icantly different from zero. Tbe 95% CI was .010 to .053. 37 1 -0.040 —
38 12 0.013 0.159
Hence, most of the random variation was between authors and 39 12 0.072 0.080
not between articles. Because tbe level-3 variance is significant, 40 1 0.113 —
we can include it in the computation of the regression estimates. 41 4 0.068 0.106
Tbe type of schizotypy was a significant predictor of the effect 42 4 0.142 0.223
sizes and explained the level-2 or level-3 variance. Tbe mean 43 1 0.266
44 1 1.198
effect sizes for positive-impulsive and unspecified scbizotypy 45 1 0.224 —
were significant (rs = .14, 95% CI = .12, .17,/t= 121; and.11,
95% CI = .06, .16, fc = 71, respectively), although it was
significantly negative for tbe negative-disorganized schizotypy
Discussion
(r = - . 0 9 , 95% CI = - . 1 2 , -.06). However, tbe otber four
covariates were not signiflcant predictors and did not explain Recent accounts (e.g., Silvia & Kaufman, 2010) on creativity and
bigber level variance. mental illness called for a comprehensive meta-analysis in order to
Tbere were two more moderators, eminence and domain, tbat resolve controversies and to provide a more objective perspective.
were not included in tbe main model because of tbe small cells Because of tbe complexity of tbe problem, we chose to focus on a
in tbe categories. Tbe mean effect size values, wbicb are pro- specific construct that could help to illuminate tbis issue. Tbis meta-
vided in Table 5, were remarkably bigber in tbe artistic domain analysis found interesting results about tbe nature of tbe relationsbip
(r = .21, ic = 15, 95% CI = .14 and .29) compared witb tbe between creativity and scbizotypy using multilevel analysis. Tbis is a
general domain (r = .06; Ic = 253, 95% CI = .03 and .09). sound method to analyze tbe nested dataset of tbis meta-analysis,
Likewise, it was higher in tbe studies witb eminent participants because tbis tecbnique ensures tbe assumption of sample indepen-
(r = .20, k = 6, 95% CI = .02 and .36) compared witb dence. Our findings indicated tbat tbe vadation in tbe effect sizes
noneminent participants (r = .06, ic = 262, 95% CI = .04 and between creativity and scbizotypy largely stem from tbe type or
.09). In addition, tbe relationsbip between effect sizes and year dimension of scbizotypy.
of publication was examined. Based on our analysis tbere was As sbown in Table 4, only tbe type of scbizotypy bad inde-
no significant relationsbip between these two (r = .015). pendent main effects in tbe multilevel analyses. In this moder-
222 ACAR AND SEN

Fisher's Z

Figure 3. Funnel Plot of Standard Error by Fisher's Z.

ator, we categorized schizotypy into three major groups: (a) odd speech and behavior, and limited affect (Raine, Lencz, &
positive and impulsive, (b) negative and anhedonia, and (c) Mednick, 1995). The former set of symptoms can help in
unspecified. As we expected, positive and impulsive types had "thinking outside the box" and generating original ideas, which
a higher positive correlation with creativity (r = .14, 95% CI = is consistent with earlier studies (Burch, Pavelis, Hemsley, &
.12 and .17) than the unspecified schizotypy (r = .11, 95% Corr, 2006; O'Reilly, Dunbar, & Bentall, 2001) that have
CI = .06 and .16). The relationship was negative in negative and reported a link between creativity and positive schizotypy
introvcrtive anhedonia types (r = -.09,95% CI = -.12, -.06). It is symptoms. A positive correlation with impulsive nonconfor-
possible to argue a linear trend within those three levels. Although
positive, impulsive, and unspecified schizotypy displayed a positive
correlation with creativity, the relationship turned to negative regard-
Table 4
ing negative and introvertive anhedonia. All three effect sizes are
Main Model
small according to Cohen (1988).
The current meta-analysis indicates that the relationship be- Estimates SE 95% CI
tween creativity and schizotypy is not uniform and that only
some aspects or forms of schizotypy may facilitate creativity, Fixed effects
Intercept 0.230* 0.078 (0.072,0.388)
and other aspects may be counterproductive. Claridge, Pryor, Measure of creativity -0.002 0.040 (-0.097, 0.092)
and Watkins (1990) argued that the personalities of creative Content of creativity test 1 -0.112 0.063 (-0.245,0.022)
individuals overlap with schizotypal traits. Our findings partly Content of creativity test 2 -0.066 0.063 (-0.202, 0.070)
support this hypothesis with the specification that only partic- Index of creativity 1 -0.002 0.033 (-0.072, 0.067)
ular forms of schizotypy can possibly help creativity to emerge. Index of creativity 2 0.047 0.037 (-0.029,0.124)
Type of schizotypy 1 0.108* 0.047 (0.006, 0.209)
The direction and magnitude of the relationship between Type of schizotypy 2 -0.119* 0.049 (-0.223, -0.015)
schizotypy types and creativity underscored that unspecified Schizotypy instrument 1 -0.023 0.043 (-0.129,0.083)
and positive-impulsive schizotypy are symmetrically opposite Schizotypy instrument 2 -0.058 0.041 (-0.160,0.043)
to negative-disorganized schizotypy. The symptoms of both Variance components
Second level 0.012* 0.002 (0.008, 0.016)
types of schizotypy can help to clarify their diverse relation- Third level 0.032* 0.011 (0.010, 0.053)
ships with creativity. Positive-impulsive schizotypy symptoms
include magical ideation, odd beliefs, ideas of reference, and Note. Measure of creativity = a dummy variable coded 1 to indicate
Personality or Self-reports measures and 0 for performance measures (i.e.,
unusual perceptual experiences, whereas negative- DT, B-W, Associational). Content of creativity test 1 = a dummy variable
disorganization symptoms include social isolation and anxiety. coded 1 to Verbal and 0 for both Figurai and Other; Content of creativity
test 2 = a dummy variable coded 1 to Figurai and 0 for both Verbal and
Other. Index of creativity 1 = a dummy variable coded 1 to indicate
Table 3 Fluency and 0 for both Originality and Other; Index of creativity 2 = a
dummy variable coded 1 to indicate Originality and 0 for both Fluency and
Unconditional Model other. Type of schizotypy 1 = a dummy variable coded I to indicate
Positive-Impulsive and 0 for both Negative-Disorganized and Unspecified;
Estimates SE 95% CI Type of schizotypy 2 = a dummy variable coded 1 to indicate Negative-
Disorganized and 0 for both Positive-Impulsive and Unspecified. Schizo-
Fixed effects
typy instrument 1 = a dummy variable coded 1 to indicate Specific
Intercept 0.141* 0.030 (0.080, 0.202)
(Anhedonia, Perceptual Aberration, Magical Ideation) and 0 for both
Variance components
General (O-LIFE, SPQ, STA) and Other; Schizotypy instrument 2 = a
Second level 0.020* 0.003 (0.014, 0.026)
dummy variable coded 1 to indicate General and 0 for both Specific and
Third level 0.033* 0.011 (0.011,0.054) Other.
'p < .05. 'p < .05.
 CREATTVITY AND SCHIZOTYPY 223

Table 5
Number of Studies and Categories, and Univariate Analyses of Moderators

k Mean r 95% CI P Heterogeneity

Gender
Male 10 .08 - . 0 1 , .17 .10 Homogeneous
Female 27 .08 - . 0 1 , .17 .09 Heterogeneous
Mixed 231 .07 .04, .09 <.OO1 Heterogeneous
Eminence
Eminent 6 .20 .02, .36 .03 Heterogeneous
Noneminent 262 .06 .04, .09 <.OO1 Heterogeneous
Domains
Arts 15 .21 .14,.29 <.OO1 Heterogeneous
General 253 .06 .03, .09 <.OO1 Heterogeneous
Measure of creativity
Self-report, personality 72 .06 .01, .11 .03 Heterogeneous
Performance measures 196 .07 .05, .09 <.OO1 Heterogeneous
Type of creativity test
Verbal 200 .04 .01, .07 .004 Heterogeneous
Figurai 54 .13 .08, .17 .001 Heterogeneous
Both or none 14 .25 .12, .37 <.OO1 Heterogeneous
Indices of creativity
Fluency 84 .06 .01,.11 .01 Heterogeneous
Originality 43 .09 .01, .16 .02 Heterogeneous
Other 141 .07 .04, .11 <.OO1 Heterogeneous
Instrument of schizotypy
General 172 .08 .05, .11 <.OO1 Heterogeneous
Specific 51 .01 - . 0 5 , .07 .085 Heterogeneous
None or other 45 .11 .04, .17 .001 Heterogeneous
Type of schizotypy
Positive & impulsive 121 .14 .12, .17 <.0Ol Heterogeneous
Disorganized & negative 76 -.09 -.12, - . 0 6 <.OO1 Heterogeneous
Unspecified 71 .11 .06, .16 <.0Ol Heterogeneous

mity was also linked to the absence of self-censorship and the thology. Our findings indicated that this is true even within the
inclination to express inappropriate ideas rather than creativity construct of schizotypy.
(Brod, 1997). These results are also interesting in terms of the earlier
Our findings can be interpreted under the light of the factor findings on the P scale (Acar & Runco, 2012). The mean effect
analytic studies on schizotypy. Chapman, Chapman, Kwapil, size between creativity and the P scale was .16, which is close
Eckblad, and Zinser (1994) reported that measures that fall to the mean effect size with a positive-impulsive schizotypy
under the category of "positive trait schizotypy" predict psy- {r = .14). P and positive-impulsive schizotypy seem to have the
chosis rather than schizophrenia whereas Physical Anhedonia same strength of relationship with creativity and are distinc-
Scale did not predict psychosis or psychosis proneness. This tively different from negative-introverted anhedonia types. This
difference can explain the discrepancy in the mean effect sizes. situation could be explained with the possibility that P and
Positive-negative schizotypy dimensions were found in both positive schizotypy are closely related constructs. It should be
Lewandowski et al. (2006) and Kwapil, Barrantes-Vidal, and noted that none of the categories of moderators had as high
Silvia (2008) studies. The latter also reported that negative mean effect size of r = .50 as obtained between creativity and
schizotypy scores were positively correlated with introversion. P, when P was measured with EPQ-R and uniqueness was the
As mentioned in the beginning of this article, Claridge et al. index of creativity.
(1996) suggested that four factors of schizotypy could be re- Further studies could follow up this one by focusing on schizo-
duced into two dimensions: extroversion and introversion. Spe- phrenia. Claridge and Beech (1995) argued that creativity and
cifically, positive and impulsive schizotypy loaded on extro- schizophrenia are linked through schizotypy. In this way, we could
version, and negative and introvertive anhedonia loaded on compare the findings of this study and the previous one with
introversion. Our analyses indicated that the parts of schizotypy subclinical levels. This would be a meta-analytic test of the find-
that loaded on extroversion are conducive to creativity, whereas ings by the Kinney et al. (2001) study, who found the least
the parts that loaded on introversion are potentially detrimental. creativity in schizophrenics and the highest from those who exhibit
These results are striking, because it is not possible to gener- only a few symptoms.
alize schizotypy in its relationship with creativity. This finding
Two of the moderators (i.e., eminence and domain) were not
also justified this and other focused meta-analyses that inves-
included in the multilevel regression analyses because of the small
tigated specific links between creativity and particular forms or
sample sizes in some categories. Inclusion of those could allow the
levels of psychopathologies. Lauronen et al. (2004) suggested
testing of a few more hypotheses and could influence the regres-
that there are multiple links between creativity and psychopa-
sion model. The current results from a few studies revealed a
224 ACAR AND SEN

stronger relationship with the studies that included an eminent Benton, A. L., & Hamsher. K. (1989). Multilingual aphasia examination.
sample and were in the artistic domain. The higher mean effect Iowa City, Iowa: AJA Associates.
sizes from the eminent group of participants could be explained Berkey, C. S., Anderson, J. J., & Hoaglin, D. C. (1996). Multiple-outcome
with the salience effect. Kasof (1999) argued that creative behavior meta-analysis of clinical trials. Statistics in Medicine, 15, 537-557.
is salient because of the inherent novelty of creativity. Symptoms doi:10.1002/(SICI)1097-0258(19960315)15:5<537::AID-SIM176>3.0
of schizotypy or psychopathology become more visible in highly .C0;2-S
Bleuler, E. (1911/1950). Dementia praecox or the group of schizophrenias.
successful or eminent people's behaviors than those of others. If a
(J. Zinkin, Trans.). New York, NY: Intemational Universities Press.
higher effect size is not because of the fundamental attribution
Bowers, M. B., & Freedman, D. X. (1966). Psychedelic experiences in
error led by the salience effect, one could argue that eminent
acute psychosis. Archives of General Psychiatry, 15, 240-248. doi:
people are those who can take advantage of their schizotypal traits
lO.lOOl/archpsyc.1966.01730150016003
for the sake of creativity. The higher mean effect size in the artistic Brod, J. H. (1997). Creativity and schizotypy. In G. Claridge (Ed.), Schizo-
domain is also intriguing. Burch, Pavelis, et al. (2006) argued that typy: Implications for illness and health (pp. 274-298). Oxford, UK:
higher rates of "taboo" responses (Rawlings, 1984) in visual artists Oxford University Press.
can lead to higher creativity. They suggested that visual artists Brugger, P., & Graves, R. E. (1997). Testing vs. believing hypotheses:
have original ideas and the will to express them, even if they are Magical ideation in the judgment of contingencies. Cognitive Neuropsy-
socially unacceptable. The former refers to creativity and the latter chiatry, 2, 251-272. doi:10.1080/135468097396270
is found among the characteristics of schizotypy. More could be Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models:
said about the type of relationship between creativity and schizo- Applications and data analysis methods. Newbury Park, CA: Sage.
typy if more studies were conducted using the data from eminent •Burch, G. S. J., Hemsley, D. R., Pavelis, C , & Corr, P. J. (2006).
samples and from different domains. Personality, creativity, and latent inhibition. European Journal of Per-
sonality, 20, 107-122. doi:10.1002/per.572
'Burch, G. S. J., Pavelis, C , Hemsley, D. R., & Corr, P. J. (2006).
References Schizotypy and creativity in visual artists. British Journal of Psychology,
97, 177-190. doi:10.1348/000712605X60030
Note: References with asterisk (*) indicate studies included to
Buss, A. H. (1966). Psychopathology. New York, NY: Wiley.
meta-analysis.
Cameron, N. (1938). Reasoning, regression and communication in schizo-
'Abraham, A., & Windmann, S. (2008). Selective information processing phrenics. Psychological Monographs, 50, 1-33
advantages in creative cognition as a function of schizotypy. Creativity Carson, S. H., Peterson, J. B., & Higgins, D. M. (2003). Decreased latent
Research Journal, 20, 1-6. doi:10.1080/10400410701839819 inhibition is associated with increased creative achievement in high-
Acar, S., & Runco, M. A. (2012). Psychoticism and creativity: A meta- functioning individuals. Journal of Personality and Social Psychology,
analytic review. Psychology of Aesthetics, Creativity and the Arts, 6, 85, 499-506. doi:10.1037/0022-3514.85.3.499
341-350. doi:10.1037/a0027497 Chapman, L. J., & Chapman, J. P. (1973). Disordered thought in schizo-
American Psychiatric Association. (2000). Diagnostic and statistical man- phrenia. New York, NY: Appleton-Century-Crofts.
ual of mental disorders (4th ed., text rev.). Washington, DC: Author. Chapman, L. J., & Chapman, J. P. (1978). The Revised Physical Anhedonia
American Psychological Association. (2010). Publication manual of the Scaie. (Unpublished test). University of Wisconsin, Madison.
American Psychological Association (6th ed.). Washington, DC: Author. Chapman, L. J., Chapman, J. P., Kwapil, T. R., Eckblad, M., & Zinser,
Andreasen, N. C. (1987). Creativity and mental illness: Prevalence rates in M. C. (1994). Putatively psychosis-prone subjects 10 years later. Journal
writers and their first-degree relatives. The American Journal of Psychi- of Abnormal Psychology, 103, 171-183.
atry, 144, 1288-1292. Chapman, L. J., Chapman, J. P., & Raulin, M. L. (1976). Scales for
Andreasen, N. C , & Powers, P. S. (1975). Creativity and psychosis. physical and social anhedonia. Journal of Abnormal Psychology, 85,
Archives of General Psychiatry, 32, 70-73. doi:10.1001/archpsyc.l975 374-382. doi:i0.1037/0021-843X.85.4.374
.01760190072008
Chapman, L. J.. Chapman, J. P., & Raulin, M. L. (1978). Body-image
Armstrong, M. S., & McConaghy, N. (1977). Allusive thinking, the word
aberration in schizophrenia. Journal of Abnormal Psychology, 87, 399-
halo and verbosity. Psychological Medicine, 7, 439-445. doi:10.1017/
407. doi:10.1037/0021-843X.87.4.399
S0033291700004414
Chico, E., & Ferrando, P. J. (1995). A psychometric evaluation of the
Barron, F., & Welsh, G. W. (1987). Barron-Welsh Art Scale. Menio Park,
revised P scale in delinquent and non-delinquent Spanish samples.
CA: Mind Garden.
Personality and Individual Differences, 18, 331-337.
Batey, M. (2007). A psychometric investigation of everyday creativity.
Claridge, G. (1997). Schizotypy: Theoretical background and issues. In G.
(Unpublished doctoral thesis). University of London, London, UK.
Claridge (Ed.), Schizotypy: Implications for illness and health (pp.
'Batey. M., & Furnham, A. (2008). The relationship between measures of
creativity and schizotypy. Personality and Individual Differences, 45, 3-18). Oxford, UK: Oxford University Press.
816-821. doi:10.1016/j.paid.2008.08.014 Claridge, G., & Beech, T. (1995). Fully and quasi-dimensional construc-
'Batey, M., & Furnham, A. (2009). The relationship between creativity, tions of schizotypy. In A. Raine, T. Lencz, & S. A. Mednick (Eds.),
schizotypy, and intelligence. Individual Differences Research, 7, 272- Schizotypal personality (pp. 192-216). New York, NY: Cambridge
284. University Press. doi:10.1017/CBO9780511759031.010
Becker, G. (1978). The mad genius controversy: A study in the sociology 'Claridge, G., & Blakey, S. (2009). Schizotypy and affective temperament:
of deviance. Beverly Hills, CA: Sage. Relationships with divergent thinking and creativity styles. Personality
Begg, C. B. (1994). Publication bias. In H. Cooper & L. Hedges (Eds.), and Individual Differences, 46, 820-826. doi:10.1016/j.paid.2009.01
Handbook of research synthesis (pp. 399-409). New York, NY: Sage .015
Publication. Claridge, G., & Broks, P. (1984). Schizotypy and hemispheric function. I.
Bentall, R. P., Claridge, G. S., & Slade, P. D. (1989). The multidimensional Theoretical considerations and the measurement of schizotypy. Person-
nature of schizotypal traits: A factor analytic study with normal subjects. ality and Individual Differences, 5, 633-648. doi: 10.1016/0191-
British journal of Clinical Psychiatry, 28, 363-375. 8869(84)90111-9
 CREATIVITY AND SCHIZOTYPY 225

Claridge, G., McCreery, C , Mason, O., Bentall, R., Boyle, G., & Slade, P. Eysenck, H. } . , & Eysenck, S. B. G. (1994). Manual of the Eysenck
(1996). The factor structure of schizotypal traits: A large replication Personality Questionnaire: Comprising the EPQ-Revised (EPQ-R) and
study. British Joumal of Clinical Psychology, 35. 103-115. doi: 10.1111/ EPQ-R Short Scale. San Diego, CA: EdITS.
j.2044-8260.1996.tb01166.x Field, A. P., & Gillet, R. (2010). How to do a meta-analysis. British
"Claridge, G., & McDonald, A. (2009). An investigation into the relation- Joumal of Mathematical and Statistical Psychology, 63, 665-694. doi:
ships between convergent and divergent thinking, schizotypy, and au- 10.1348/000711010X502733
tistic traits. Personality and Individual Differences, 46, 794-799. doi: Finke, R. A. (1990). Creative imagery: Discoveries and inventions in
10.1016/j.paid.2009.01.018
visualization. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Claridge, G., Pryor, R., & Watkins, G. (1990). Sounds from the bell Jar:
•Fisher, J. E., Heller, W., & Miller, G. A. (2007). Semantic associations,
Ten psychotic authors. London, UK: Macmillan Press.
lateralized frontal function, and context maintenance in schizotypy.
Clementz, B. A., Grove, W. M., Katsanis, J., & Iacono, W. G. (1991). The
Neuropsychologia, 45, 663-672. doi:10.1016/j.neuropsychologia.2006
psychometric detection of schizotypy: Perceptual aberration and physi-
.07.017
cal anhedonia in relatives of schizophrenics. Joumal of Abnormal Psy-
•FoUey, B. S., & Park, S. (2005). Verbal creativity and schizotypal per-
chology, 100, 607-612. doi:10.1037/0021-843X.100.4.607
sonality in relation to prefrontal hemispheric laterality: A behavioral and
Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
near-infrared optical imaging study. Schizophrenia Research, 80, 271-
f2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.
282. doi:10.1016/j.schres.2005.06.016
"Cox, A. J., & Leon, J. L. (1999). Negative schizotypal traits in the relation
of creativity to psychopathology. Creativity Research Joumal, 12, 25- *Giaiiotti, L. R. R., Mohr, C , Pizzagalli, D., Lehmann, D., & Brugger, P.
36. doi:10.1207/sl5326934crjl201_4 (2001). Associative information processing and paranormal belief. Psy-
Davis, G. A., & Subkoviak, M. J. (1975). Multidimensional analysis of a chiatry and Clinical Neurosciences, 55, 595-603. doi : 10.1046/j. 1440-
personality based test of creative potential. Joumal of Educational 1819.2001.00911.x
Measurement, 12, 37-43. "Gibson, C , Folley, B. S., & Park, S. (2009). Enhanced divergent thinking
de la Torre, J., Camilli, G., Vargas, S., & Vemon, R. F. (2007). Illustration and creativity in musicians: A behavioral and near-infrared spectros-
of a multilevel model for meta-analysis. Measurement and Evaluation in copy. Brain and Cognition, 69, 162-169. doi:10.1016/j.bandc.2008.07
Counseling and Development, 40, 169-180. .009
Dickerson, S. S., & Kemeny, M. E. (2004). Acute Stressors and cortisol Gleser, L. J., & Olkin, I. (2009). Stochastically dependent effect sizes. In
responses: A theoretical integration and synthesis of laboratory research. H. Cooper, L. V. Hedges, & J. C. Valentine (Eds.), The handbook of
Psychological Bulletin, 130, 355-391. doi:10.1037/0033-2909.130.3 research synthesis and meta-analysis (pp. 357-376). New York, NY:
.355 Russell Sage.
*Dinn, W. M., Harris, C. L., Aycicegi, A., Greene, P., & Andover, M. S. Goodglass, H., & Kaplan, E. (1972). The assessment of aphasia and
(2002). Positive and negative schizotypy in a student sample: Neuro- related disorders. Philadelphia, PA: Lea & Febiger.
cognitive and clinical correlates. Schizophrenia Research, 56, 171-185. Gough, H. (1979). A creative personality scale for the Adjective Check
doi: 10.1016/S0920-9964(01 )00230-4 List. Joumal of Personality and Social Psychology, 37, 1398-1407.
Domino, G. (1970). Identification of potentially creative persons from the Gough, H. G., & Heilbmn, A. B. (1980). The Adjective Check List Manual.
Adjective Check List. Joumal of Consulting and Clinical Measurement, Palo Alto, CA: Consulting Psychologists Press.
12, 37-43. *Green, M. J., & Williams, L. M. (1999). Schizotypy and creativity as
Dorfman, L., Martindale, C , Gassimova, V., & Vartanian, O. (2008). effects of reduced cognitive inhibition. Personality and Individual Dif-
Creativity and speed of information processing: A double dissociation ferences, 27, 263-276. doi:10.1016/S0191-8869(98)00238-4
involving elementary versus inhibitory cognitive tasks. Personality and Gruzelier, J. H. (1995). Syndromes of schizotypy: Patterns of cognitive
Individual Differences, 44, 1382-1390. doi:10.1016/j.paid.2007.12.006 asymmetry, arousal and gender. In A. Raine, T. Lencz, & S. A. Mednick
•Duchene, A., Graves, R. E., & Brugger, P. (1998). Schizotypal thinking (Eds.), Schizotypal personality (pp. 329-352). Cambridge, UK: Univer-
and associative processing: A response commonality analysis of verbal
sity Press. doi:10.1017/CBO9780511759031.015
fluency. Joumal of Psychiatry and Neuroscience, 23, 56-60.
Guilford, J. P., Christensen, P., Merrifield, P., & Wilson, R. (1978).
Eckblad, M., & Chapman, L. J. (1983). Magical ideation as an indicator of
Altemate uses (Form B, Form C). Orange CA: Sheridan Psychological
schizotypy. Joumal of Consulting and Clinical Psychology, 51, 215-
Services Inc.
225. doi:I0.1037/0022-006X.51.2.215
Guilford, J. P., & Hoepfner, R. (1971). The analysis of intelligence. New
Eckblad, M. L , Chapman, L. J., Chapman, J. P., & Mishlove, M. (1982).
York, NY: McGraw-Hill.
The revised social anhedonia scales. Unpublished manuscript.
Hakstian, A. R., & Cattell, R. B. (1976). Manual for the Comprehensive
*Eisenman, R. (1990). Creativity, preference for complexity, and physical
and mental illness. Creativity Research Joumal, 3, 231-236. doi: Ability Battery (CAB). Champaign, IL: Institute for Personality and
10.1080/10400419009534355 Ability Testing.
Eysenck, H. J. (1993). Creativity and personality: Suggestions for a theory. Hasenfus, N., & Magaro, P. (1976). Creativity and schizophrenia: An
Psychological Inquiry, 4, 147-178. doi:10.1207/sl5327965pli0403_l equality of empirical constructs. The British Joumal of Psychiatry, 129,
Eysenck, H. J. (1994). The measurement of creativity. In M. Boden (Ed.), 346-349. doi:10.1192/bjp.l29.4.346
Dimensions of creativity (pp. 199-242). Cambridge, MA: The MIT Hedges, L. V. (1992). Meta-analysis. Joumal of Educational Statistics, 17,
Press. 279-296. doi:10.2307/l 165125
Hysenck, H. i. (1995). Genius: The natural history of human creativity. Hedges, L. V. (2007). Meta-analysis. In C. R. Rao & S. Sinharay (Eds.),
New York, NY: Cambridge University Press. doi:10.10I7/ Handbook of Statistics (Vol. 26, pp. 919-953), Amsterdam, the Neth-
CBO9780511752247 erlands: Elsevier.
Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual of the Eysenck Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis.
Personality Questionnaire (adult and junior). London, UK: Hodder & Orlando, FL: Academic Press.
Stoughton. Hedges, L. V., & Vevea, J. L. (1998). Fixed- and random-effects models in
Eysenck, H. J., & Eysenck, S. B. G. (1991). Manual of the Eysenck meta-analysis. Psychological Methods, 3, 486-504. doi: 10.1037/1082-
Personality Scales (EPS Adult). London, UK: Hodder & Stoughton. 989X.3.4.486
226 ACAR AND SEN

Higgins, J. P., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Kumar, V. K., Kemmler, D., & Holman, E. R. (1997). The Creativity
Measudng inconsistency in meta-analyses. British Medical Journal, Styles Questionnaire-Revised. Creativity Research Journal, 10, 51-58.
327, 557-560. doi:10.1136/bmj.327.7414.557 doi: 10.1207/s 15326934crj 1001 _6
*Hod, H., Nagamine, M., Soshi, T., Okabe, S., Kim, Y., & Kunugi, H. Kwapil, T. R., Barrantes-Vidal, N., & Silvia, P. J. (2008). The dimensional
(2008). Schizotypal traits in healthy women predict prefrontal activation stmcture of the Wisconsin schizotypy scales: Factor identification and
pattems dudng a verbal fluency task: A near-infrared spectroscopy construct validity. Schizophrenia Bulletin, 34, 444-457. doi:10.1093/
study. Neuropsychobiology, 57, 61-69. doi: 10.1159/000129669 schbul/sbmO98
Hox, J. J. (2002). Multilevel analysis: Techniques and applications. Mah- 'Lack, S. A., Kumar, V. K., & Arevalo, S. (2003). Fantasy proneness,
wah, NJ: Lawrence Erlbaum. creative capacity, and styles of creativity. Perceptual and Motor Skills,
Hox, J. J., & de Leeuw, E. D. (2001). Multilevel models for meta-analysis. 96, 19-24. doi:10.2466/pms.2003.96.1.19
In N. Duan & S. Reise (Eds.), Multilevel modeling: Methodological Lauronen, E., Veijola, J., Isohanni, I., Jones, P. B., Nieminen, P., &
advances, issues, and applications (pp. 90-111). Mahwah, NJ: Law- Isohanni, M. (2004). Links between creativity and mental disorder.
rence Erlbaum Associates. Psychiatry, 67, 81-98.
Hunter, J. E., & Schmidt, F. L. (2000). Fixed effects vs. random effects Lewandowski, K. E., Barrantes-Vidal, N., Nelson-Gray, R. O., Clancy, C ,
meta-analysis models: Implications for cumulative research knowledge. Kepley, H. O., & Kwapil, T. R. (2006). Anxiety and depression symp-
Intemational Journal of Selection and Assessment, 8, 275-292. doi: toms in psychometrically identified schizotypy. Schizophrenia Re-
10.1111/1468-2389.00156 search, 83, 225-235.
*Karimi, Z., Windmann, S., Güntürkün, O., & Abraham, A. (2007). Insight Lezak, M. (1995). Neuropsychological assessment. Oxford, UK: Oxford
problem solving in individuals with high versus low schizotypy. Journal University Press.
of Research in Personality, 41, 473-480. doi:10.1016/j.jrp.2006.03.008 Light, R. J., & Pillemer, D. B. (1984). Summing up: The science of
Karlsson, J. L. (1978). Inheritance of creative intelligence. Chicago, IL: reviewing research. Cambddge, MA: Harvard University Press.
Nelson-Hall. Lipp, O. V., & Vaitl, D. (1992). Latent inhibition in human Pavlovian
differential conditioning: Effect of additional stimulation after preexpo-
Kasof, J. (1999). Attribution and creativity. In M. A. Runco & S. R.
sure and relation to schizotypal traits. Personality and Individual Dif-
Pdtzker (Eds.), Encyclopedia of creativity (pp. 147-156). San Diego,
ferences, 13, 1003-1012. doi:10.1016/0191-8869(92)90133-A
CA: Academic Press.
Lipsey, M. W., & Wilson, D. B. (2001 ). Practical meta-analysis. Thousand
Kendler, K. S., Ochs, A. L., Gorman, A. M., Hewitt, J. K., Ross, D. E., &
Oaks, CA: Sage.
Mirsky, A. F. (1991). The stmcture of schizotypy: A pilot multi-trait
Littell, J. H., Corcoran, J., & Pillai, V. (2008). Systematic reviews and
twin study. Psychiatry Research, 36, 19-36. doi: 10.1016/0165-
meta-analysis. Oxford, UK: Oxford University Press. doi:10.1093/
1781(91)90114-5
acprof:oso/9780195326543.001.0001
Kent, G. H., & Rosanoff, A. J. (1910). A study of association in insanity.
Little, R. C , Milliken, G. A., Stroup, W. W., Wolfinger, R. D., &
American Journal of Insanity, 67, 37-96; 317-390.
Schabenberger, O. (2007). SAS for mixed models (2nd ed.). Cary, NC:
Kéd, S. (2009). Genes for psychosis and creativity: A promoter polymor-
SAS Institute Inc.
phism of the Neuregulin 1 Gene is related to creativity in people with
Lombroso, C. (1891). Man of genius. London, UK: Walter Scott.
high intellectual achievement. Psychological Science, 20, 1070-1073.
Lubow, R. E. (1973). Latent inhibition. Psychological Bulletin, 79, 398-
doi:10.1111/j.l467-9280.2009.02398.x
407. doi:10.1037/h0034425
'Kéd, S. (2011). Solitary minds and social capital: Latent inhibition,
Lubow, R. E., Ingberg-Sachs, Y., Zalstein-Orda, N., & Gewirtz, J. (1992).
general intellectual functions and social network size predict creative
Latent inhibition in low and high "psychotic-prone" normal subjects.
achievements. Psychology of Aesthetics, Creativity, and the Arts, 5,
Personality and Individual Differences, 13, 563-572. doi:10.1016/0191-
215-221. doi:10.1037/a0022000
8869(92)90197-W
Kéd, S., Kiss, I., & Kelemen, O. (2009). Effects of a neuregulin 1 vadant 'Ludwig, A. M. (1992). Creative achievement and psychopathology: Com-
on conversion to schizophrenia and schizophreniform disorder in people padson among professions. American Journal of Psychotherapy, 46,
at high dsk for psychosis. Molecular Psychiatry, 14, 118-119. doi: 330-356.
10.1038/mp.2008.1 'Ludwig, A. M. (1994). Mental illness and creative activity in female
Kinney, D. K., Richards, R., Lowing, P. A., LeBlanc, D., Zimbalist, M. E., wdters. The American Journal of Psychiatry, 151, 1650-1656.
& Harlan, P. (2001). Creativity in offspring of schizophrenics and *MacPherson, J. S., & Kelly, S. W. (2011). Creativity and positive schizo-
controls. Creativity Research Journal, 13, 17-25. doi:10.1207/ typy influence the conflict between science and religion. Personality and
S15326934CRJI301_3 Individual Differences, 50, 446-450. doi:10.1016/j.paid.2010.11.002
Klein, C , Andresen, B., & Jahn, T. (1997). Erfassung der schizotypen Martindale, C. (1975). What makes creative people different. Psychology
Persoenlichkeit nach DSM-HI-R: Psychometdsche Eigenschaften einer Today 9, 44-50.
autodsierten deutschsprachigen Uebersetzung des "Schizotypal Person- Mason, O., & Claddge, G. (2006). The Oxford-Liverpool Inventory of
ality Questionnaire" (SPQ) von Raine [Assessment of schizotypal per- Feelings and Expedences (O-LIFE): Further descdption and extended
sonality according to DSM-III-Psychometric properties of an authodzed norms. Schizophrenia Research, 82, 203-211. doi:10.1016/j.schres.2005
German translation of "Schizotypal Personality Questionnaire" by .12.845
Raine]. Diagnostica, 43, 347-369. Mason, O., Claridge, G., & Jackson, M. (1995). New scales for the
Konstantopoulos, S. (2011). Fixed effects and vadanee components esti- assessment of schizotypy. Personality and Individual Differences, 18,
mation in three-level meta-analysis? Research Synthesis Methods, 2, 7-13. doi:10.1016/0191-8869(94)00132-C
61-76. doi:10.1002/jrsm.35 Mednick, S. A. (1958). A leaming theory approach to research in schizo-
Kds, E. (1952). Psychoanalytic explorations in art. New York, NY: phrenia. Psychological Bulletin, 55, 316-327. doi:10.1037/h0040425
Intemational Universities Press. Mednick, S. A. (1959). The remote associates test. Boston, MA: Houghton
'Krysanski, V., & Ferraro, F. R. (2007). Creativity, unique designs, and Mifflin.
schizotypy in a non-clinical sample. Psychological Reports, 101, 273- Mednick, S. A. (1962). The associative basis of the creative process.
274. Psychological Review, 69, 220-232. doi:10.1037/h0048850
 CREATTVITY AND SCHIZOTYPY 227

Mednick, S. A., & Mednick, M. T. (1967). Examiner's manual: Remote Rado, S. (1953). Dynamics and classification of disordered behavior.
Associates Test. Boston, MA: Houghton Mifflin. American Joumal of Psychiatry, UO, 406-416.
Meehl, P. E. (1962). Schizotaxia, schizotypy, schizophrenia. American Raine, A. (1991). The SPQ: A scale for the assessment of schizotypal
Psychologist, 17, 827-838. doi:10.1037/h0041029 personality based on DSM-III-R criteria. Schizophrenia Bulletin, 17,
Metcalfe, J. (1986). Premonitions of insight predict impending error. 555-564. doi: 10.1093/schbul/l7.4.555
Joumal of Experimental Psychology: Leaming, Memory, and Cogni- Raine, A., Lencz, T., & Mednick, S. A. (Eds.). (1995). Schizotypal per-
tion, 12, 623-634. doi:10.1037/0278-7393.12.4.623 sonality. Cambridge, UK: Cambridge University Press. doi:10.1017/
Miller, E. N., & Chapman, L. J. (1983). Continued word association in CBO9780511759031
hypothetically psychosis-prone college students. Joumal of Abnormal Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models (2nd
Psychology, 92, 468-478. doi:10.1037/0021-843X.92.4.468 ed.). London, UK: Sage.
*MilIer, G. F., & Tal, I. (2007). Schizotypy versus intelligence and open- Rawlings, D. (1984). The correlation of EPQ psychoticism with two
ness as predictors of creativity. Schizophrenia Research, 93, 317-324. behavioural measures of impulsivity. Personality and Individual Differ-
doi:10.1016/j.schres.2007.02.007 ences, 5, 591-594. doi:10.1016/0191-8869(84)90034-5
Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred *Rawlings, D., & Georgiou, G. (2004). Relating the components of figure
Reporting Items for Systematic Reviews and Meta-Analyses: The preference to the components of hypomania. Creativity Research Jour-
PRISMA Statement. Annals of Internal Medicine, 151, 264-269. nal, 16, 49-57. doi:10.1207/sl5326934crjl601_5
*Mohr, C , Graves, R. E., Gianotti, L. R., Pizzagalli, D., & Brugger, P. •Rawlings, D., & Locamini, A. (2008). Dimensional schizotypy, autism,
(2001). Loose but normal: A semantic association study. Joumal and unusual word associations in artists and scientists. Journal of Re-
of Psycholinguistic Research, 30, 475-483. doi:10.1023/A: search in Personality, 42, 465-471. doi:10.1016/j.jip.2007.06.005
1010461429079
*Rawlings, D., & Toogood, A. (1997). Using a "taboo response" measure
Moran, L. J., Mefferd, R. B., Jr., & Kimble, J. P., Jr. (1964). Idiodynamic
to examine the relationship between divergent thinking and psychoti-
sets in word association. Psychological Monographs: General and Ap-
cism. Personality and Individual Differences, 22, 61-68. doi:10.1016/
plied, 78, 1-22. doi:10.1037/h0093833
S0191-8869(96)00177-8
Nelson, B., & Rawlings, D. (2009). Experience of Creativity Questionnaire
*Rawlings, D., Twomey, F., Bums, E., & Morris, S. (1998). Personality,
(ECQ). Creativity Research Journal, 21, 43-53. doi: 10.1080/
creativity, and aesthetic preference: Comparing psychoticism, sensation
10400410802633442
seeking, schizotypy, and openness to experience. Empirical Studies of
'Nelson, B., & Rawlings, D. (2010). Relating schizotypy and personality to
the Arts, 16, 153-178. doi:10.2190/8GVJ-ERL3-HYUM-EH88
the phenomenology of creativity. Schizophrenia Bulletin, 36, 388-399.
Regard, M., Strauss, F., & Knapp, P. (1982). Children's production of
doi: 10.1093/schbul/sbnO98
verbal and nonverbal ñuency tasks. Perceptual and Motor Skills, 55,
O'Reilly, T., Dunbar, R., & Bentall, R. P. (2001). Schizotypy and creativ-
839-844. doi: 10.2466/pms. 1982.55.3.839
ity: An evolutionary connection? Personality and Individual Differ-
Richards, J. M., Jr., Holland, J. L., & Lutz, S. W. (1967). Prediction of
ences, 31, 1067-1078. doi:10.1016/S0191-8869(00)00204-X
student accomplishment in college. Joumal of Educational Psychology,
Payne, R. W., Caird, W. K., & Laverty, S. G. (1964). Ovednclusive
58, 343-355.
thinking and delusions in schizophrenic patients. Joumal of Abnormal
and Social Psychology, 68, 562-566. doi:10.1037/h004Il 17 Richards, R., Kinney, D. K., Lunde, I., Benet, M., & Merzel, A. P. (1988).
Payne, R. W., Mattusek, P., & George, E. I. (1959). An experimental study Creativity in manic depressives, cyclothymes, their normal relatives, and
of schizophrenic thought disorder. Joumal of Mental Science, 105, control subjects. Joumal of Abnormal Psychology, 97, 281-288. doi:
627-652. 10.1037/0021-843X.97.3.281
Peterson, J. B., & Carson, C. (2000). Latent inhibition and openness to Richards, R., Kinney, D. K., & Merzel, A. P. C. (1988). Assessing
experience in a high-achieving student population. Personality everyday creativity: Characteristics of the Lifetime Creativity Scales and
and Individual Differences, 28, 323-332. doi:10.1016/S0191- validation with three large samples. Joumal of Personality and Social
8869(99)00101-4 Psychology, 54, 476-485. doi: 10.1037/0022-3514.54.3.476
Peterson, J. B., Smith, K., & Carson, S. (2002). Openness and extraversion Rosenthal, R. (1979). The file drawer problem and tolerance for null
are associated with reduced latent inhibition: Replication and commen- results. Psychological Bulletin, 86, 638-641. doi:10.1037/0033-2909.86
tary. Personality and Individual Differences, 33, 1137-1147. doi: .3.638
10.1016/SO191 -8869(02)00004-1 Rosenthal, R. (1994). Parametric measures of effect size. In H. Cooper &
*Poreh, A., Whitman, D., & Ross, T. R. (1993). Creative thinking abilities L. V. Hedges (Eds.), Handbook of research synthesis (pp. 231-244).
and hemispheric asymmetry in schizotypal. college students. Current New York, NY: Russell Sage Foundation.
Psychology: Research and Reviews, 12, 344-352. doi: 10.1007/ Rothenberg, A. (1979). The emerging goddess: The creative process in art,
BF02686814 science, and other fields. Chicago, IL: University of Chicago Press.
Post, F. (1994). Creativity and psychopathology: A study of 291 world- Rothenberg, A. (1990). Creativity and madness. Baltimore, MD: The John
famous men. The British Journal of Psychiatry, 165, 22-34. doi: Hopkins University Press.
10.1192/bjp.l65.I.22 Ruff, R. M. (1996). Ruff Figurai Fluency Test (RFFT) Examiner's manual.
Prentky, R. A. (1980). Creativity and psychopathology: A neurocognitive Odessa, FL: Psychological Assessment Resources, Inc.
perspective. New York, NY: Praeger. Runco, M. A. (1991). Divergent thinking. Norwood, NJ: Ablex.
Prentky, R. A. (1989). Creativity and psychopathology: Gamboling at the Rust, J. (1988). The Rust Inventory of Schizotypal Cognitions (RISC).
seat of madness. In J. A. Glover, R. R. Ronning, & C. R. Reynolds Schizophrenia Bulletin, 14, 317-322. doi:10.1093/schbul/14.2.317
(Eds.), Handbook of creativity: Assessment, research, and theory (pp. *Rust, J., Golombok, S., & Abram, M. (1989). Creativity and schizotypal
243-269). New York, NY: Plenum. thinking. The Journal of Genetic Psychology, 150, 225-227. doi:
*Preti, A., & Vellante, M. (2007). Creativity and psychopathology: Higher 10.1080/00221325.1989.9914593
rates of psychosis proneness and nonright-handedness among creative *SchuIdberg, D. (1990). Schizotypal and hypomanic traits, creativity, and
artists compared to same age and gender peers. Joumal of Nervous and psychological health. Creativity Research Joumal, 3, 219-231. doi:
Mental Disease, 195, 837-845. doi:10.1097/NMD.0b013e3181568180 10.1080/10400419009534354
228 ACAR AND SEN

'Schuldberg, D. (2000-2001). Six subclinical spectrum traits in normal Van den Bussche, E., Van den Noortgate, W., & Reynvoet, B. (2009).
creativity. Creativity Research Journal, 13, 5-16. doi:10.1207/ Mechanisms of masked priming: A meta-analysis. Psychological Bulle-
S15326934CRJ1301_2 tin, 135. 452-477. doi:10.1037/a0015329
*Schuldberg, D., French, C , Stone, B. L., & Heberle, J. (1988). Creativity Vanderplas, J. M., & Garvin, E. A. (1959). The association value of
and schizotypal traits: Creativity test scores, perceptual aberration, mag- random shapes. Joumal of Experimental Psychology, 57, 147-154.
ical ideation, and impulsive nonconformity. Journal of Nervous and van Houwelingen, H. C , Arends, L. R., & Stijnen, T. (2002). Advanced
Mental Disease, 176, 648-657. doi:10.1097/00005053-198811000- methods in meta-analysis: Multivariate approach and meta-regression.
00002 Statistics in Medicine, 21, 589-624. doi:10.1002/sim.l040
Semel, E., Wiig, E. H., & Secord, W. A. (1995). Clinical evaluation of Venables, P. H. (1995). Schizotypal status as a developmental stage in
language fundamentals (CELF) Examiner's manual. Chicago, IL: Psy- studies of risk for schizophrenia. In A. Raine, T. Lencz, & S. A. Mednick
chological Corporation Harcourt Brace & Co. (Eds.), Schizotypal personality (pp. 107-131). Cambridge, UK: Univer-
Severiens, S., & Dam, G. T. (1998). A multilevel meta-analysis of gender sity Press. doi:10.1017/CBO9780511759031.007
differences in learning orientations. British Joumal of Educational Psy- Venables, P. H., Wilkins, S., Mitchell, D. A., Raine, A., & Bailes, K.
chology, 68, 595-608. doi:10.1111/j.2044-8279.1998.tb01315.x (1990). A scale for the measurement of schizotypy. Personality
Siegler, M., Osmond, H., & Mann, H. (1972). Laing's models of madness. and Individual Differences, 11, 481-495. doi:10.1016/0191-
In R. Boyers & R. Orrill (Eds.), Laing and anti-psychiatry (pp. 99-122). 8869(90)90061-U
Harmondsworth, UK: Penguin. Walker, E., & Gale, S. (1995). Neurodevelopmental processes in schizo-
Silvia, P. J., & Kaufman, J. C. (2010). Creativity and mental illness. In J. C. phrenia and schizotypal personality disorder. In A. Raine & S. Med-
Kaufman & R. J. Stemberg (Eds.), Cambridge handbook of creativity nick (Eds.), Schizotypal personality disorder (pp. 56-75). London, UK:
(pp. 381-394). New York, NY: Cambridge University Press. Cambridge University Press. doi:10.1017/CBO9780511759031
Simonton, D. K. (2005). Are genius and madness related? Contemporary .005
answers to an ancient question. Psychiatric Times, 22, 21-23. Wallach, M. A., & Kogan, N. (1965). Modes of thinking in young children.
Singer, J. D. (1998). Using SAS PROC MIXED to fit multilevel growth New York: Holt, Rinehart, & Winston.
models, hierarchical models, and individual growth models. Journal of *Ward, P. B., McConaghy, N., & Catts, S. V. (1991). Word association and
Educational and Behavioral Statistics, 23, 323-355. measures of psychotic-proneness in university students. Personality
Smith, S. M., Ward, T. B., & Schumacher, J. S. (1993). Constraining and Individual Differences, 12, 473-480. doi:10.1016/0191-
effects of examples in a creative generation task. Memory & Cognition, 8869(9 l)90065-J
21, 837-845. Ward, T. B. (1994). Structtired imagination: The role of category stmcture
Spreen, O., & Strauss, E. (1998). A compendium of neuropsychological in exemplar generation. Cognitive Psychology, 27, 1-40.
tests: Administration, norms, and commentary. New York, NY: Oxford 'Weinstein, S., & Graves, R. E. (2001). Creativity, schizotypy, and later-
University Press. ality. Cognitive Neuropsychiatry, 6, 131-146. doi:10.1080/
Stemberg, R. J., & Davidson, J. E. (1982). The mind of the puzzler. 13546800042000098
Psychology Today, 16, 37-44. "Weinstein, S., & Graves, R. E. (2002). Are creativity and schizotypy
'Stoneham, A. C. S., & Coughtrey, A. E. (2009). The role of schizotypy products of a right hemisphere bias? Brain and Cognition, 49, 138-151.
and creativity in a group problem-solving task. Personality and Individ- doi:10.1006/brcg.2001.1493
ual Differences, 46, 827-831. doi:10.10I6/j.paid.2009.01.014 Welsh, G. S. (1949). Welsh Figure Preference Test. Palo Alto, CA:
Storr, A. (1976). 77!« dynamics of creation. Harmondsworth, UK: Pelican. Consulting Psychologists Press.
'Suzuki, A., & Usher, M. (2009). Individual differences in language Welsh, G. S., & Barron, F. (1963). Barron-Welsh Art Scale. Palo Alto, CA:
latéralisation, schizotypy and the remote-associate task. Personality and Consulting Psychologists Press.
Individual Differences, 46, 622-626. doi:10.1016/j.paid.2009.01.006 Wilson, R. C , Merrifield, P. R., & Guilford, J. P. (1969). Scoring guide for
Thurstone, L. L., & Thurstone, T. G. (1962). Primary mental abilities. the Utility Test. Form A. Beverly Hills, CA: Sheridan Psychological
Chicago, IL: Science Research Associates. Services.
Torrance. E. P. (1970). Torrance tests of creative thinking: Directions 'Wuthrich, V. M., & Bates, T. C. (2001). Schizotypy and latent inhibition:
manual and scoring guide. Princeton, NJ: Personnel Press. Non-linear linkage between psychometric and cognitive markers. Per-
Torrance, E. P. (1974). The Torrance Tests of Creative Thinking: Norms— sonality and Individual Differences, 30, 783-798. doi:10.1016/S0191-
Technical manual. Bensenville, IL: Scholastic Testing Service. 8869(00)00071-4
'Tsakanikos, E., & Claridge, G. (2005). More words, less words: Verbal
fluency as a function of "positive" and "negative" schizotypy. Person- Received June 20, 2012
ality and Individual Differences, 39, 705-713. doi:10.1016/j.paid.2005 Revision received December 13, 2012
.02.019 Accepted December 31, 2012 •
Copyright of Psychology of Aesthetics, Creativity & the Arts is the property of American
Psychological Association and its content may not be copied or emailed to multiple sites or
posted to a listserv without the copyright holder's express written permission. However, users
may print, download, or email articles for individual use.