“Genetic & Environmental Influences on the Phenotypic Associations between Intelligence, Personality, & Creative Achievement in the Arts and Sciences”, 2018-07 (; backlinks):
- Intelligence and Openness contribute to creative achievement in the arts and sciences.
Creativity in the arts and sciences is influenced by genes and unique environment.
Artistic but not scientific creativity is also influenced by shared environment.
There is a genetic overlap between openness and creativity in the arts and sciences.
Most genetic influences on intelligence are also involved in scientific creativity.
Several studies suggest a different effect of intelligence and personality on creative achievement in the arts and sciences. There is also research showing that all these variables are influenced by both genes and environmental factors. The aim of this study was to move further and investigate whether the relative influence of genes and environment on the associations between personality, intelligence, and creative achievement differs between the arts and sciences.
Measures of intelligence (Wiener Matrizen Test), personality traits (BFI-44), and creative achievement (Creative Achievement Questionnaire) were obtained from a twin cohort. The sample size differed between measures, ranging between 6,606 and 9,537 individuals (1,349 and 2,250 complete twin pairs).
Firstly, we performed several phenotypic analyses.
These analyses collectively showed that intelligence and the personality trait ‘openness to experience’ were the only traits which contributed statistically-significantly to achievement, in either creative domain. Intelligence showed a stronger association with science than with art (non-linear and linear form, respectively), while relations between openness and achievement showed the opposite pattern.
Secondly, we performed genetic modeling.
Univariate analyses showed artistic creative achievement to be the only variable statistically-significantly influenced by shared environment. Individual differences in the remaining traits could be accounted for by additive genetic effects and non-shared environment.
Results from two trivariate analyses, which included intelligence, openness, and creative achievement in either the arts or sciences, indicated a substantial and fairly equal genetic overlap between openness and achievement in the two creative domains. Genes associated with intelligence however, played a statistically-significantly greater role in scientific achievement than in artistic achievement. In fact, the majority of genetic influences on intelligence were also involved in scientific creative achievement. There was also an overlap of unique environmental influences between intelligence and scientific creative achievement that was not present between intelligence and artistic creative achievement.
…The univariate genetic modeling showed no dominant genetic effects on BFIO or MAXSCI, i.e. both these variables could be fit with AE-models, which was also the case for WMT. The heritability, i.e. the proportion of variance explained by genetic factors, were estimated to AMAX_SCI = 0.68 (CI: 0.60; 0.75), ABFI_O = 0.57 (CI: 0.53; 0.60) and AWMT = 0.59 (CI: 0.55; 0.63). For MAXART, the ACE-model showed the best fit, with AMAX_ART = 0.37 (CI: 0.13; 0.63) and CMAX_ART = 0.32 (CI: 0.06; 0.53). The complete results of the univariate genetic modeling can be found in Table S6 in the Supplementary material.
The model fitting results from the trivariate ACE Cholesky decompositions are summarized in Table 9: There were no statistically-significant differences between the ACE and AE models for either creative domain, i.e. there were no statistically-significant effects of shared environment in the trivariate analysis on any of the variables of interest. The only variable for which such an effect would have been expected based on the univariate modeling was MAXART, but it would seem that power was too low to render it statistically-significant in the trivariate analysis. The estimated AE models are illustrated in Figure 2, Figure 3. Estimates for the full ACE models can be found in Figure S1 & Figure S2 in the Supplement. The heritability estimates for the variables in the AE models were AMAX_ART = 0.69 (CI: 0.63; 74), AMAX_SCI = 0.67 (CI: 0.59; 0.74), ABFI_O = 0.57 (CI: 0.53; 0.61) and AWMT = 0.59 (CI: 0.55; 0.63) (calculated by adding the squared genetic pathways for each variable, see Figure 2, Figure 3). The heritability estimate for MAXART was notably higher in the trivariate AE model than in the trivariate or univariate ACE models (0.37), since all variance explained by C in the latter models was picked up by A in the former.
For both domains of creative achievement, most of the covariance with intelligence (artistic: 89%, scientific: 88%) and Openness (artistic: 82%, scientific: 85%) could be explained by shared genetic influences. A substantial proportion of the total genetic variance was shared between Openness and artistic creative achievement (63%), and between Openness and scientific creative achievement (59%). Unique environmental influences were also shared between Openness and artistic (24%) as well as scientific (17%) creative achievement.
The main difference between the two creative domains/traits appeared to be with respect to the genetic and unique environmental correlations with intelligence. Out of the total variance in intelligence and artistic creative achievement explained by genetic factors, 33% was shared between the two traits. Between intelligence and scientific creative achievement, 58% of the genetic variance was shared, which was statistically-significantly higher than for the arts. As for the total variance accounted for by unique environmental factors, 14% was shared between intelligence and scientific creative achievement, while no such influences were shared between intelligence and artistic achievement.
