Intelligence xxx (xxxx) xxx

Contents lists available at ScienceDirect

Intelligence
journal homepage: www.elsevier.com/locate/intell

The future of intelligence: The role of specific abilities

Thomas R. Coyle a, *, Samuel Greiff b
a
University of Texas at San Antonio, Department of Psychology, 1 UTSA Circle, San Antonio, TX 78249, USA
b
University of Luxembourg, Department of Behavioural and Cognitive Sciences, L-4366 Esch-sur-Alzette, Luxembourg

A R T I C L E I N F O A B S T R A C T

Editor: Dr. Richard Haier This contribution to the Special Issue on the Future of Intelligence focuses on specific abilities. Specific abilities
are distinct abilities (e.g., math, verbal, spatial) that differ conceptually and empirically from general intelligence
Keywords: (g; variance common to tests). Two specific abilities are discussed: (a) ability tilt, defined as an ability pattern
Ability tilt based on differences in math and verbal scores on standardized tests (e.g., SAT and ACT), yielding math tilt
Complex problem solving (CPS)
(math > verbal) and verbal tilt (verbal > tilt); and (b) complex problem solving (CPS), defined as the ability to
General intelligence (g)
adapt to novel and dynamically changing environments. Although g is generally a strong predictor of life out­
SAT
Armed Services Vocational Aptitude Battery comes, tilt and CPS predict criteria beyond g. Math tilt predicts science, technology, engineering, and math
(ASVAB) (STEM) criteria (e.g., jobs, majors, aptitudes), whereas verbal tilt predicts humanities criteria. Similarly, CPS
Program for International Student Assessment abilities (e.g., knowledge acquisition and knowledge application) predict academic achievement and job per­
(PISA) formance after controlling for g. CPS and tilt can be integrated into theories of intelligence such as Cattell-Horn-
Carroll theory and investment theories. Future research should consider (a) reporting norms for tilt and CPS in
testing programs (e.g., SAT and ACT); (b) using tilt and CPS to reduce disparate impact between groups; (c) using
computerized testing to track tilt and CPS over time; and (d) integrating tilt and CPS into trait complexes
involving personality traits, vocational interests, and specific abilities.

This contribution to the Special Issue on the Future of Intelligence with predictive power beyond g. The final section will discuss future
will focus on the predictive power of specific abilities and their impli­ research on specific abilities. Topics include reporting specific abilities
cations for intelligence research. Specific abilities are distinct abilities in testing programs (e.g., SAT, ACT, PISA); examining group differences
(e.g., math, verbal, spatial) that differ conceptually and empirically from in ability tilt and complex problem solving; tracking specific abilities as
general intelligence (g; variance common to tests), which largely ex­ a function of instruction; and examining trait complexes involving
plains the predictive power of tests (Jensen, 1998, pp. 270–305). specific abilities, personality traits, and vocational interests.
Although the predictive power of specific abilities is hotly debated, some
specific abilities robustly predict outcomes beyond g (for a review, see 1. Distinction between g and specific abilities
Coyle, 2018a; see also, Coyle, 2014; Coyle, 2018b) and can be as
important as g in predicting criteria (Kell & Lang, 2017; Lang & Kell, A key distinction in intelligence research is between g (general in­
2020). Here, we discuss two specific abilities. The first is ability tilt, telligence) and specific abilities. g arises mainly from positive correla­
defined as an ability pattern based on differences in math and verbal tions among tests (i.e., positive manifold), indicating that people who do
scores on standardized tests (e.g., SAT and ACT), yielding math tilt well on one test generally do well on other tests. The g loading of a test is
(math > verbal) and verbal tilt (verbal > math). The second is complex related to its predictive power (Coyle, 2018a, p. 2). In general, tests with
problem solving, defined as the ability to adapt to novel and dynami­ higher g loadings (high correlations with g) have strong predictive power
cally changing tasks that measure specific abilities (e.g., knowledge for school and work criteria (e.g., school grades or supervisor ratings),
acquisition and knowledge application). even though this prediction is far from perfect (e.g., Coyle, 2018a). In
The article is divided into three sections. The first will discuss the addition, statistically removing g from tests usually lowers (or even
distinction between g and specific abilities. The second section will neutralizes) their predictive power, suggesting that the predictive power
discuss ability tilt and complex problem solving, two specific abilities of tests is partly attributable to g.

* Corresponding author.
E-mail addresses: [email protected] (T.R. Coyle), [email protected] (S. Greiff).

https://doi.org/10.1016/j.intell.2021.101549
Received 10 March 2021; Received in revised form 17 May 2021; Accepted 17 May 2021
0160-2896/© 2021 Elsevier Inc. All rights reserved.

Please cite this article as: Thomas R. Coyle, Samuel Greiff, Intelligence, https://doi.org/10.1016/j.intell.2021.101549
T.R. Coyle and S. Greiff Intelligence xxx (xxxx) xxx

In contrast to g, specific abilities are distinct abilities that are unre­ complementary domains (math) but inhibits abilities in competing do­
lated or weakly related to g. Specific abilities include test specific vari­ mains (verbal). Consistent with investment theories, math tilt correlates
ances (e.g., verbal, math, spatial) or other abilities (e.g., problem positively with STEM criteria (e.g., STEM GPAs) but negatively with
solving) that are conceptually distinct from g. It has been claimed that humanities criteria (e.g., humanities GPAs). In contrast, verbal tilt
specific abilities have limited or negligible predictive power compared shows the opposite pattern. Because time is limited, investment in one
to g. Consistent with this claim, Thorndike (1984) claimed that domain (math) comes at the expense of competing domains (verbal),
compared to g, specific abilities of IQ tests (obtained after removing g) producing negative relations between competing abilities.
accounted for about five times less variance in academic performance More recent research has examined another type of tilt, tech tilt
(80–90% vs. 10–20% for g vs. specific abilities, respectively). In a similar (Coyle, 2019, 2020, 2021). Tech tilt refers to a pattern of specific abil­
vein, other research has found that the predictive power of specific ities based on within subject differences in technical abilities (e.g.,
abilities decreases after controlling for g and correcting for measurement mechanical, electrical, automotive) and academic abilities (math or
error (e.g., Schmidt, 2017). verbal). These differences yield tech tilt (tech > academic) and academic
Although the predictive power of g is firmly established (e.g., Ree & tilt (academic > tech). Coyle (2019, 2020, 2021) measured tech tilt and
Carretta, 2002), claims that “nothing matters beyond g” may be pre­ academic tilt using the Armed Services Vocational Aptitude Battery
mature (cf. Ree & Earles, 1991; Ree, Earles, & Teachout, 1994). (ASVAB), a selection test used by the U.S. military. Tech tilt and aca­
Compared to g, specific abilities have received less attention, and less is demic tilt showed a domain-specific pattern similar to ability tilt. Tech
known about their predictive power. The situation is different in per­ tilt predicted STEM criteria (jobs and college majors), while academic
sonality research, which has examined the predictive power of a gen­ tilt predicted humanities criteria, with the effects replicating after
eral factor of personality (i.e., variance common to personality traits), removing g. The results were attributed to complementarity between
personality traits (e.g., Conscientiousness), and personality facets tech tilt and STEM, which involve technical skills (e.g., mechanical
(Condon et al., 2020; Mõttus et al., 2020). Following the example of reasoning and electrical knowledge), versus academic tilt and human­
personality research, intelligence research should pay more attention to ities, which involve non-technical verbal skills. In addition, tech tilt
specific abilities, their predictive power, and their relation to other predicted relatively high math ability but relatively low verbal ability
variables (e.g., vocational interests and personality traits). (based on the SAT and ACT), whereas academic tilt showed the opposite
pattern. The results again support investment theories, which assume
2. Two key specific abilities: ability tilt and complex problem that investment in one ability (technical) comes at the expense of
solving competing abilities (verbal), producing negative relations between
competing abilities.
Although the predictive power of specific abilities is hotly debated Research on tilt highlights the predictive power of specific abilities
(e.g., Kell & Lang, 2018), several specific abilities predict outcomes beyond g. A similar example, from a different theoretical tradition, is
robustly. Two examples from different lines of research are ability tilt (for complex problem solving (CPS). Research on CPS emerged due to crit­
a review, see Coyle, 2018a) and complex problem solving (for an over­ icism of tests of g that were allegedly too far detached from real-world
view, see Herde, Wüstenberg, & Greiff, 2016). Ability tilt refers to a problems (Greiff & Wüstenberg, 2014). Indeed, early studies showed
pattern of specific abilities based on within subject differences in math small correlations between CPS and measures of g (e.g., Putz-Osterloh,
and verbal scores on standardized tests such as the SAT and ACT, two 1981), which were quickly traced back to measurement error in CPS
widely used college admissions tests. These differences yield math tilt measures (Greiff, Wüstenberg, & Funke, 2012).
(math > verbal), indicating math strength, and verbal tilt (verbal > Recent research on CPS has advanced in measurement and theoret­
math), indicating verbal strength. Complex problem solving (CPS) ical understanding. First, new tools that focus on scalability and mea­
captures how individuals adapt to novel and dynamically changing surement quality have been developed, whereas the original
tasks. Tests of CPS typically measure knowledge acquisition (i.e., desideratum of face validity (e.g., resemblance to complex real-world
knowledge about the problem’s structure; Scherer, Greiff, & Hautamäki, problems) has been largely dropped (Greiff, Fischer, Stadler, & Wüs­
2015) and knowledge application (i.e., actually solving the problem; tenberg, 2015; Kröner, Plass, & Leutner, 2005; Neubert, Kretzschmar,
Molnar, Greiff, & Csapo, 2013), two specific abilities that predict school Wüstenberg, & Greiff, 2015). Following these developments, CPS was
and work outcomes (Danner, Hagemann, Schankin, Hager, & Funke, added as a new domain to the 2012 Program for International Student
2011; Greiff et al., 2013). Assessment (PISA), which measures educational achievement across the
Ability tilt on the SAT and ACT has been examined using the National world (Greiff, Holt, & Funke, 2013). Second, theories of CPS have been
Longitudinal Survey of Youth (NLSY), a large and representative sample refined. In the early years, differences between g and CPS were the
of U.S. students (Coyle, 2018a; see also, Coyle, 2016, 2018b, 2019, theoretical focus. In recent studies, CPS has been integrated into over­
2020; Coyle & Pillow, 2008; Coyle, Purcell, Snyder, & Richmond, 2014; arching models of the human intellect such as Cattell-Horn-Carroll
Coyle, Snyder, & Richmond, 2015). Two key findings have emerged. (CHC) theory (McGrew, 2009; see also, Wüstenberg, Greiff, & Funke,
First, ability tilt is unrelated to g and, second, ability tilt robustly predicts 2012). At the same time, advances in CPS theory and research go beyond
diverse criteria in STEM (science, technology, engineering, math) and conceptualizations of g to include dynamic changes of the environment,
humanities (e.g., history, English, foreign languages) fields. The criteria intransparency, and interrelatedness (Greiff et al., 2014).
include jobs, college majors, test scores (based on other tests), and col­ Like ability tilt, CPS predicts diverse outcomes beyond g. The pre­
lege grades (e.g., grade point average, GPA). In general, math tilt pre­ dictive power of CPS has been attributed to specific abilities (e.g.,
dicts STEM criteria, whereas verbal tilt predicts humanities criteria. knowledge acquisition and knowledge application), which facilitate
Moreover, the predictive power of ability tilt has been replicated with dynamic problem solving. The predictive power of CPS has been
tests other than the SAT and ACT (e.g., PSAT and Armed Services demonstrated for job performance (e.g., Danner et al., 2011; Mainert,
Vocational Aptitude Battery); with samples other than the NLSY (e.g., Niepel, Murphy, & Greiff, 2019) and academic performance (Greiff &
Study of Mathematically Precocious Youth); and with ability tilt based Fischer, 2013; Greiff, Wüstenberg, et al., 2013) after removing variance
on manifest variables (e.g., test scores) and latent variables (e.g., verbal from g (Lotz, Scherer, Greiff, & Sparfeldt, 2017; Wüstenberg et al., 2012)
or math abilities) (e.g., Coyle, 2018a). or working memory (Schweizer, Wüstenberg, & Greiff, 2013).
The results of ability tilt research are consistent with investment CPS theory and research often focus on dynamic problem solving in
theories (Cattell, 1987, pp. 138–146; see also, Coyle, 2016, 2019, 2020; educational contexts. CPS has been characterized as a stratum II (broad)
Coyle et al., 2015). Investment theories assume that investment (time ability in CHC theory, while specific CPS abilities (e.g., knowledge
and effort) in a particular domain (STEM) boosts abilities in acquisition and knowledge application) may be considered stratum I

2
T.R. Coyle and S. Greiff Intelligence xxx (xxxx) xxx

(narrow) abilities. In developmental research, CPS has been character­ and ACT). As a result, students with high spatial abilities may not be
ized as a higher-order thinking skill that develops with reasoning identified by college assessments or given adequate support in school
(Frischkorn, Greiff, & Wüstenberg, 2014) and working memory (Greiff (cf. Lakin & Wai, 2020; see also, Gohm, Humphreys, & Yao, 1998).
et al., 2015). CPS has been integrated into cascade theories of devel­ Testing programs that consider both g and specific abilities can profit
opment, which focus on the development of basic abilities (e.g., pro­ from two great psychometric traditions (cf. Humphreys, Lubinski, &
cessing speed) and higher order thinking skills (e.g., causal reasoning) in Yao, 1993, pp. 250–252). One tradition emphasizes ability level and the
early childhood and adolescence (Galotti, 2011). predictive value of g and specific abilities. Another emphasizes the
The predictive power of ability tilt and CPS may seem paradoxical in pattern of specific abilities (strengths and weaknesses) and the selection
light of claims that specific abilities have negligible predictive power of environmental niches (e.g., STEM vs. humanities). The two traditions
(cf. Jensen, 1998, pp. 270–305; see also, Coyle, 2018a, p. 2). However, describe different aspects of cognitive ability that predict distinct but
there is no necessary contradiction. Consistent with their domain- complementary outcomes. Ability level predicts performance within a
specific nature, specific abilities robustly predict criteria in specific niche, whereas ability pattern (e.g., math tilt or verbal tilt) predicts
domains (e.g., STEM and humanities) and sometimes even in general choice of niche (e.g., STEM or humanities).
domains (e.g., Greiff, Wüstenberg, et al., 2013; Lang & Kell, 2020). In
contrast, g predicts criteria in specific domains and, importantly,
composite criteria (averaged across domains) such as academic 3.2. Ensuring fairness: the example of race and ethnic differences
achievement in different courses or job performance in different
occupations. Future research could compare the predictive power of specific
abilities for different groups such as race or ethnic groups. For example,
3. Future research on specific abilities: an agenda Whites and Blacks in the U.S. show mean differences in test scores on IQ
tests and scholastic aptitude tests (e.g., SAT and ACT). The reasons for
Up to this point, we have argued that intelligence research should these differences are heavily disputed and there is an ongoing scientific
consider specific abilities that account for more than g – just as per­ debate surrounding this topic, which is beyond the scope of this article.
sonality research considers facets of personality that account for more However, as a practical matter, test score differences between groups
than the general factor of personality. In this last section, we outline a that differ on a sociopolitical variable such as race can produce disparate
research agenda for specific abilities. impact by reducing school or work opportunities (e.g., Nisbett et al.,
2012). Further, school and work opportunities may depend on g-loaded
3.1. Looking at specific abilities from a normative and criterion-oriented skills or admission criteria, exacerbating existing societal disparities (e.
angle g., Flynn, 2008).
Given that group differences in test scores can produce disparate
Standardized tests (e.g., SAT and ACT) provide information on the impact, it would be desirable to identify specific abilities that minimize
ability level of individuals compared to a reference population. Future group differences yet retain predictive power. Ability tilt and CPS have
testing programs could examine specific abilities using normative the potential to combine these desirable features. For example, ability
criteria (e.g., where individuals stand in the reference population) and tilt on the SAT and ACT shows relatively small White-Black differences,
mastery criteria (e.g., whether individuals can solve a particular prob­ yet predicts diverse criteria (e.g., jobs, college majors, GPAs), with math
lem). Mastery criteria might be easier to develop for specific abilities tilt predicting STEM criteria and verbal tilt predicting humanities
(versus g), in part because specific abilities have more limited scope and criteria (Coyle, 2016). Similarly, tech tilt on the ASVAB shows negligible
therefore may be more easily specified. In fact, it is standard practice in White-Black differences, yet also predicts diverse outcomes, with tech
educational research to use mastery criteria to compare students when tilt predicting STEM outcomes and academic tilt predicting humanities
evaluating learning goals and progress. outcomes (Coyle, 2021). Both tech tilt and ability tilt robustly predict
Ability tilt could be incorporated into testing programs by reporting performance criteria (e.g., aptitude test scores) and preference criteria
tilt scores for standardized tests (e.g., SAT or ACT). Following prior (e.g., job choice and college majors), suggesting that the predictive
research (Lubinski, 2016; see also, Lubinski, Webb, Morelock, & Ben­ power of tilt is not an artifact of a particular criterion (Coyle, 2018b).
bow, 2001; Park, Lubinski, & Benbow, 2007), tilt scores could be The low or negligible g loadings of tilt and CPS make them good
computed by taking the difference in math and verbal scores on stan­ targets to reduce disparate impact and societal disparities. One appli­
dardized tests, and computing percentiles based on the normative dis­ cation would be to use tilt scores, CPS scores, and other measures of
tribution of tilt scores. The percentiles would indicate the degree of math specific abilities for personnel selection or admissions in educational
tilt (math > verbal) or verbal tilt (verbal > math) for an individual contexts. Compared to g-loaded tests that reflect societal disparities
(relative to the normative sample). Similarly, CPS could be integrated (sometimes referred to as “institutional racism”), specific abilities may
into testing programs by linking a person’s standing on a normative reduce group disparities and maintain predictive power in key domains
scale to a description of a problem he or she is likely to solve from a such as STEM and humanities, which have been linked to educational
mastery perspective. Such approaches have been adopted by interna­ and occupational outcomes (Coyle, 2018a, 2018b).
tional large-scale assessments such as PISA (e.g., Greiff, Wüstenberg, Although specific abilities may reduce disparate impact, such abili­
et al., 2013) and could be used to target specific abilities in need of ties should complement (not replace) other measures of cognitive ability
intervention. Finally, specific abilities could be used for college admis­ that predict educational and occupational criteria (e.g., Schmidt &
sions or counseling. For example, students with math tilt (math > ver­ Hunter, 1998). This is because specific abilities and other measures of
bal) might consider college majors in STEM (e.g., physics and ability may predict different criteria. For example, whereas measures of
engineering), whereas students with verbal tilt (verbal > math) might general ability (g) typically predict educational and occupational level
consider humanities fields (e.g., law and journalism). (based on occupational or subject matter complexity), specific abilities
Future research might incorporate spatial abilities into existing such as math or verbal tilt typically predict educational or occupational
testing programs (e.g., SAT, ACT, PSAT), which often focus on academic category (e.g., STEM or humanities) (Coyle, 2018a, 2018b). Although
abilities (math and verbal). Spatial abilities are distinct from academic specific abilities have been linked to educational or occupational cate­
abilities and have been linked to STEM achievement and creativity (e.g., gory, specific abilities have also been found to predict g-loaded criteria
patents, inventions, refereed publications) (Kell, Lubinski, Benbow, & such as aptitude test scores, even after removing g (e.g., Coyle, 2018a).
Steiger, 2013; Wai, Lubinski, & Benbow, 2009). Spatial abilities receive Such findings suggest that specific abilities measure “more than g” and
scant attention in school curricula and college assessments (e.g., SAT may reduce disparities related to g.

3
T.R. Coyle and S. Greiff Intelligence xxx (xxxx) xxx

3.3. Improving specific abilities and tracking them over time 1997, p. 239; see also, von Stumm & Ackerman, 2013). In addition, trait
complexes could include non-academic abilities such as spatial, me­
Future research could develop interventions to improve specific chanical, or technical abilities. Such abilities are associated with prac­
abilities via direct instruction. For example, STEM instruction and tical and conventional interests and often receive short shrift in school
problem solving might boost math tilt, whereas humanities instruction curricula and college assessments (cf. Lakin & Wai, 2020; see also, Gohm
might boost verbal tilt. Similarly, metacognitive instruction and in­ et al., 1998).
structions on self-regulated learning might improve CPS performance by
facilitating the understanding of causal models. In addition, CPS and tilt 4. Conclusion
interventions could use computer adaptive instruction to link instruc­
tional support with ability level, presenting items of optimal difficulty to This contribution to the Special Issue on the Future of Intelligence
facilitate the acquisition of specific abilities. Computerized instruction focused on specific abilities such as ability tilt and CPS. Specific abilities
could also track the g and non-g loadings of specific abilities over time. have been found to predict diverse criteria (e.g., jobs, majors, GPAs),
Increases in the non-g loadings (and decreases in g loadings) of specific even after removing g. Although g is generally a strong predictor of life
abilities could be used to measure completeness of training, with in­ outcomes, specific abilities provide incremental validity beyond g and
creases in non-g loadings of specific abilities indicating that the acqui­ hold promise for reducing disparate impact and overreliance on g-loaded
sition of specific abilities is approaching saturation. measures. To realize this promise, specific abilities could be incorpo­
Interventions might also consider technical (e.g., mechanical and rated into testing programs (e.g., SAT, ACT, PISA) and trait complexes
electrical) and spatial abilities, two abilities that have been linked to involving vocational interests and personality traits (Ackerman & Heg­
STEM criteria (e.g., Lakin & Wai, 2020; Wai et al., 2009). Spatial abil­ gestad, 1997; von Stumm & Ackerman, 2013; see also, Ackerman,
ities, in particular, have been found to improve following different types 2003).
of spatial training (e.g., video games, semester long courses, repeated
testing) (Uttal et al., 2013). Such training may boost specific abilities
and also boost g (e.g., IQ), which correlates positively with educational Acknowledgement
duration (e.g., Ritchie & Tucker-Drob, 2018). Although educational
duration is related to IQ improvements, improvements from more spe­ T. R. Coyle was supported by a grant from the National Science
cific training or interventions may fade over time (e.g., Protzko, 2015, Foundation’s Interdisciplinary Behavioral and Social Science Research
2016). Competition (IBSS-L 1620457).

3.4. Long term agenda: nomological network of specific abilities References

A long-term agenda is to identify a nomological network of specific Ackerman, P. L. (2003). Aptitude complexes and trait complexes. Educational
abilities, including their observable measurements and relations with Psychologist, 38, 85–93. https://doi.org/10.1207/S15326985EP3802_3.
Ackerman, P. L., & Heggestad, E. D. (1997). Intelligence, personality, and interests:
each other and external criteria (e.g., personality and interests). The Evidence for overlapping traits. Psychological Bulletin, 121, 219–245. https://doi.
nomological network could be modeled after the Cattell-Horn-Carroll org/10.1037/0033-2909.121.2.219.
(CHC) model of intelligence (McGrew, 2009). The CHC model de­ Cattell, R. B. (1987). Intelligence: Its structure, growth and action. New York, NY: North-
Holland.
scribes cognitive abilities as a hierarchy, with g at the top, followed by Condon, D., Wood, D., Mõttus, R., Booth, T., Costantini, G., Greiff, S., … Zimmermann, J.
broad (e.g., fluid reasoning) and narrow abilities (e.g., induction and (2020). Bottom up construction of a personality taxonomy. European Journal of
deductive reasoning), which are related to each other and to g. Psychological Assessment, 36, 923–934. https://doi.org/10.1027/1015-5759/
a000626.
Following the CHC model, the non-g components of specific abilities Corno, L., Cronbach, L. J., Kupermintz, H., Lohman, D. F., Mandinach, D., Porteus, A. W.,
could be characterized as either broad or narrow. Narrow abilities might & Talbert, J. E. (2002). Remaking the concept of aptitude: Extending the legacy of
represent non-g components of specific tests such as the SAT or ACT, or Richard E. Snow. Mahwah, NJ: Erlbaum.
Coyle, T. R. (2014). Predictive validity of non-g residuals of tests: More than g. Journal of
non-g components of cognitive processes such as knowledge acquisition Intelligence, 2, 21–25. https://doi.org/10.3390/jintelligence2010021.
in CPS. Broad abilities might represent non-g factors based on multiple Coyle, T. R. (2016). Ability tilt for whites and blacks: Support for differentiation and
tests of a specific ability, such as the non-g components of multiple math investment theories. Intelligence, 56, 28–34. https://doi.org/10.1016/j.
intell.2016.02.002.
or verbal tests in the ASVAB (e.g., Coyle, 2018b). Relations among non-g
Coyle, T. R. (2018a). Non-g factors predict educational and occupational criteria: More
components of specific abilities may differ from the g-loaded abilities in than g. Journal of Intelligence, 6, 1–15. https://doi.org/10.3390/
CHC models. Whereas relations among CHC abilities are positive jintelligence6030043.
(reflecting positive manifold), relations among non-g components of Coyle, T. R. (2018b). Non-g residuals of group factors predict ability tilt, college majors,
and jobs: A non-g nexus. Intelligence, 67, 19–25. https://doi.org/10.1016/j.
specific abilities may be positive, negative, or nonexistent. Based on intell.2017.12.003.
investment theories, positive relations are expected for complementary Coyle, T. R. (2019). Tech tilt predicts jobs, college majors, and specific abilities: Support
abilities such as math tilt and STEM criteria. In contrast, negative re­ for investment theories. Intelligence, 75, 33–40. https://doi.org/10.1016/j.
intell.2019.04.002.
lations are expected for competing abilities such as math tilt and hu­ Coyle, T. R. (2020). Sex differences in tech tilt: Support for investment theories.
manities criteria (for a review, see Coyle, 2018a). Intelligence, 80, 1–13. https://doi.org/10.1016/j.intell.2020.101437.
The nomological network of specific abilities might include trait Coyle, T. R. (2021). White-Black differences in tech tilt: Support for Spearman’s law and
investment theories. Intelligence, 84, 1–9. https://doi.org/10.1016/j.
complexes, defined as clusters of personality traits, vocational interests, intell.2020.101504.
and specific abilities (Ackerman & Heggestad, 1997; von Stumm & Coyle, T. R., & Pillow, D. R. (2008). SAT and ACT predict college GPA after removing g.
Ackerman, 2013; see also, Ackerman, 2003; Corno et al., 2002). Trait Intelligence, 36, 719–729. https://doi.org/10.1016/j.intell.2008.05.001.
Coyle, T. R., Purcell, J. M., Snyder, A. C., & Richmond, M. C. (2014). Ability tilt on the
complexes involving tilt, CPS, and other specific abilities might be SAT and ACT predicts specific abilities and college majors. Intelligence, 46, 18–24.
associated with grit (e.g., persistence in acquiring a specific ability) or https://doi.org/10.1016/j.intell.2014.04.008.
self-concept, both of which could facilitate the acquisition of specific Coyle, T. R., Snyder, A. C., & Richmond, M. C. (2015). Sex differences in ability tilt:
Support for investment theory. Intelligence, 50, 209–220. https://doi.org/10.1016/j.
abilities (Mustafic, Niepel, & Greiff, 2017). Trait complexes might also
intell.2015.04.012.
be associated with vocational interests in STEM and humanities. For Danner, D., Hagemann, D., Schankin, A., Hager, M., & Funke, J. (2011). Beyond IQ: A
example, math tilt may form a trait complex with investigative interests, latent state trait analysis of general intelligence, dynamic decision making, and
which are associated with STEM preferences. In contrast, verbal tilt may implicit learning. Intelligence, 39, 323–334. https://doi.org/10.1016/j.
intell.2011.06.004.
form a trait complex with artistic interests and openness, which are Flynn, J. R. (2008). Still a question of Black vs White? New Scientist, 199, 48–50. https://
associated with humanities preferences (cf. Ackerman & Heggestad, doi.org/10.1016/S0262-4079(08)62253-8.

4
T.R. Coyle and S. Greiff Intelligence xxx (xxxx) xxx

Frischkorn, G., Greiff, S., & Wüstenberg, S. (2014). The development of complex problem McGrew, K. S. (2009). CHC theory and the human cognitive abilities project: Standing on
solving: A latent growth curve analysis. Journal of Educational Psychology, 106, the shoulders of the giants of psychometric intelligence research. Intelligence, 37,
1004–1020. https://doi.org/10.1037/a0037114. 1–10. https://doi.org/10.1016/j.intell.2008.08.004.
Galotti, K. M. (2011). Cognitive development. Thousand Oaks, CA: SAGE. Molnar, G., Greiff, S., & Csapo, B. (2013). Inductive reasoning, domain specific and
Gohm, C. L., Humphreys, L. G., & Yao, G. (1998). Underachievement among spatially complex problem solving: Relations and development. Thinking Skills and Creativity,
gifted students. American Educational Research Journal, 35, 515–531. https://doi. 9, 35–45. https://doi.org/10.1016/j.tsc.2013.03.002.
org/10.3102/00028312035003515. Mõttus, R., Wood, D., Condon, D., Back, M., Baumert, A., Costantini, G., …
Greiff, S., & Fischer, A. (2013). Der Nutzen einer Komplexen Problemlösekompetenz: Zimmermann, J. (2020). Descriptive, predictive, and explanatory personality
Theoretische Überlegungen und empirische Befunde [Usefulness of complex problem research: Different goals, different approaches, but a shared need to move beyond
solving competency: Theoretical considerations and empirical results]. Zeitschrift für the Big Few traits. European Journal of Personality, 34, 1175–1201. https://doi.org/
Pädagogische Psychologie, 27, 1–13. https://doi.org/10.1024/1010-0652/a000086. 10.1002/per.2311.
Greiff, S., Fischer, A., Stadler, M., & Wüstenberg, S. (2015). Assessing complex problem Mustafic, M., Niepel, C., & Greiff, S. (2017). Assimilation and contrast effects in the
solving skills with multiple complex systems. Thinking & Reasoning, 21, 356–382. formation of problem-solving self-concept. Learning and Individual Differences, 54,
https://doi.org/10.1080/13546783.2014.989263. 82–91. https://doi.org/10.1016/j.lindif.2017.01.006.
Greiff, S., Holt, D. V., & Funke, J. (2013). Perspectives on problem solving in cognitive Neubert, J. C., Kretzschmar, A., Wüstenberg, S., & Greiff, S. (2015). Extending the
research and educational assessment: Analytical, interactive, and collaborative assessment of complex problem solving to finite state automata: Embracing
problem solving. The Journal of Problem Solving, 5, 71–91. https://doi.org/10.7771/ heterogeneity. European Journal of Psychological Assessment, 31, 181–194. https://
1932-6246.1153. doi.org/10.1027/1015-5759/a000224.
Greiff, S., & Wüstenberg, S. (2014). Assessment with microworlds: Factor structure, Nisbett, R. E., Aronson, J., Blair, C., Dickens, W., Flynn, J., Halpern, D. F., &
invariance, and latent mean comparison of the MicroDYN test. European Journal of Turkheimer, E. (2012). Intelligence: New findings and theoretical developments.
Psychological Assessment, 30, 1–11. https://doi.org/10.1027/1015-5759/a000194. American Psychologist, 67, 130–159. https://doi.org/10.1037/a0026699.
Greiff, S., Wüstenberg, S., Csapo, B., Demetriou, A., Hautamäki, J., Graesser, A. C., & Park, G., Lubinski, D., & Benbow, C. P. (2007). Contrasting intellectual patterns predict
Martin, R. (2014). Domain-general problem solving skills and education in the 21st creativity in the arts and sciences: Tracking intellectually precocious youth over 25
century. Educational Research Review, 13, 74–83. https://doi.org/10.1016/j. years. Psychological Science, 18, 948–952. https://doi.org/10.1111/j.1467-
edurev.2014.10.002. 9280.2007.02007.x.
Greiff, S., Wüstenberg, S., & Funke, J. (2012). Dynamic problem solving. A new Protzko, J. (2015). The environment in raising early intelligence: A meta-analysis of the
measurement perspective. Applied Psychological Measurement, 36, 189–213. https:// fadeout effect. Intelligence, 53, 202–210. https://doi.org/10.1016/j.
doi.org/10.1177/0146621612439620. intell.2015.10.006.
Greiff, S., Wüstenberg, S., Götz, T., Vainikainen, M. P., Hautamäki, J., & Bornstein, M. H. Protzko, J. (2016). Does the raising IQ-raising g distinction explain the fadeout effect?
(2015). A longitudinal study of higher-order thinking skills: Working memory and Intelligence, 56, 65–71. https://doi.org/10.1016/j.intell.2016.02.008.
fluid reasoning in childhood enhance complex problem solving in adolescence. Putz-Osterloh, W. (1981). Über die Beziehung zwischen Testintelligenz und
Frontiers in Psychology, 6, 1060. https://doi.org/10.3389/fpsyg.2015.01060. Problemlöseerfolg [On the relation between test intelligence and problem solving
Greiff, S., Wüstenberg, S., Molnar, G., Fischer, A., Funke, J., & Csapo, B. (2013). Complex success]. Zeitschrift für Psychologie, 189, 79–100.
problem solving in educational settings—Something beyond g: Concept, assessment, Ree, H. M., & Carretta, T. R. (2002). g2K. Human Performance, 15, 3–24. https://doi.org/
measurement invariance, and construct validity. Journal of Educational Psychology, 10.1207/s15327043hup1501&02_02.
105, 364–379. https://doi.org/10.1037/a0031856. Ree, M. J., & Earles, J. A. (1991). Predicting training success: Not much more than g.
Herde, C. N., Wüstenberg, S., & Greiff, S. (2016). Assessment of complex problem Personnel Psychology, 44, 321–332. https://doi.org/10.1111/j.1744-6570.1991.
solving: What we know and what we don’t know. Applied Measurement in Education, tb00961.x.
29, 265–277. https://doi.org/10.1080/08957347.2016.1209208. Ree, M. J., Earles, J. A., & Teachout, M. S. (1994). Predicting job performance: Not much
Humphreys, L. G., Lubinski, D., & Yao, G. (1993). Utility of predicting group membership more than g. Journal of Applied Psychology, 79, 518–524. https://doi.org/10.1037/
and the role of spatial visualization in becoming an engineer, physical scientist, or 0021-9010.79.4.518.
artist. Journal of Applied Psychology, 78, 250–261. https://doi.org/10.1037/0021- Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve
9010.78.2.250. intelligence? A meta-analysis. Psychological Science, 29, 1358–1369. https://doi.org/
Jensen, A. R. (1998). The g factor: The science of mental ability. Westport, CT: Praeger. 10.1177/0956797618774253.
Kell, H., & Lang, J. W. B. (2017). Specific abilities in the workplace: More important than Scherer, R., Greiff, S., & Hautamäki, J. (2015). Exploring the relation between time on
g? Journal of Intelligence, 5, 1–18. https://doi.org/10.3390/jintelligence5020013. task and ability in complex problem solving. Intelligence, 48, 37–50. https://doi.org/
Kell, H. J., & Lang, J. W. B. (2018). The great debate: General ability and specific abilities 10.1016/j.intell.2014.10.003.
in the prediction of important outcomes. Journal of Intelligence, 6, 1–8. https://doi. Schmidt, F. L. (2017). Beyond questionable research methods: The role of omitted
org/10.3390/jintelligence6030039. relevant research in the credibility of research. Archives of Scientific Psychology, 5,
Kell, H. J., Lubinski, D., Benbow, C. P., & Steiger, J. H. (2013). Creativity and technical 32–41. https://doi.org/10.1037/arc0000033.
innovation: Spatial ability’s unique role. Psychological Science, 24, 1831–1836. Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods in
https://doi.org/10.1177/0956797613478615. personnel psychology: Practical and theoretical implications of 85 years of research
Kröner, S., Plass, J. L., & Leutner, D. (2005). Intelligence assessment with computer findings. Psychological Bulletin, 124, 262–274. https://doi.org/10.1037/0033-
simulations. Intelligence, 33, 347–368. https://doi.org/10.1016/j.intell.2005.03.002. 2909.124.2.262.
Lakin, J. M., & Wai, J. (2020). Spatially gifted, academically inconvenienced: Spatially Schweizer, F., Wüstenberg, S., & Greiff, S. (2013). Validity of the MicroDYN approach:
talented students experience less academic engagement and more behavioural issues Complex problem solving predicts school grades beyond working memory capacity.
than other talented students. British Journal of Educational Psychology, 90, Learning and Individual Differences, 24, 42–52. https://doi.org/10.1016/j.
1015–1038. https://doi.org/10.1111/bjep.12343. lindif.2012.12.011.
Lang, J. W. B., & Kell, H. J. (2020). General mental ability and specific abilities: Their von Stumm, S., & Ackerman, P. L. (2013). Investment and intelligence: A review and
relative importance for extrinsic career success. Journal of Applied Psychology, 105, meta-analysis. Psychological Bulletin, 139, 841–869. https://doi.org/10.1037/
1047–1061. https://doi.org/10.1037/apl0000472. a0030746.
Lotz, C., Scherer, R., Greiff, S., & Sparfeldt, J. (2017). Intelligence in action. Effective Thorndike, R. L. (1984). Intelligence and information processing: The mind and the computer.
strategic behaviors while solving complex problems. Intelligence, 64, 98–112. Bloomington, IN: Center on Evaluation, Development, and Research.
https://doi.org/10.1016/j.intell.2017.08.002. Uttal, D. H., Meadow, N. G., Elizabeth, T., Hand, L. L., Alden, A. R., Christopher, W., &
Lubinski, D. (2016). From Terman to today: A century of findings on intellectual Newcombe, N. S. (2013). The malleability of spatial skills: A meta-analysis of
precocity. Review of Educational Research, 86, 900–944. https://doi.org/10.3102/ training studies. Psychological Bulletin, 139, 352–402. https://doi.org/10.1037/
0034654316675476. a0028446.
Lubinski, D., Webb, R. M., Morelock, M. J., & Benbow, C. P. (2001). Top 1 in 10,000: A Wai, J., Lubinski, D., & Benbow, C. P. (2009). Spatial ability for STEM domains: Aligning
10-year follow-up of the profoundly gifted. Journal of Applied Psychology, 86, over 50 years of cumulative psychological knowledge solidifies its importance.
718–729. https://doi.org/10.1037/0021-9010.86.4.718. Journal of Educational Psychology, 101, 817–835. https://doi.org/10.1037/
Mainert, J., Niepel, C., Murphy, K., & Greiff, S. (2019). The incremental contribution of a0016127.
complex problem solving skills to the prediction of job level, job complexity, and Wüstenberg, S., Greiff, S., & Funke, J. (2012). Complex problem solving – More than
salary. Journal of Business and Psychology, 34, 825–845. https://doi.org/10.1007/ reasoning? Intelligence, 40, 1–14. https://doi.org/10.1016/j.intell.2011.11.003.
s10869-018-9561-x.