Biological essentialism, the belief that human attributes are determined by biology, is a core component of essentialist thinking. Previous studies have shown that individual differences in essentialist thinking are associated with heuristic thinking, cognitive ability and style, conservative values, and prejudice. None, however, have examined whether biological essentialism is itself heritable, or the extent to which familial aggregation explains associations with core correlates.
In order to do this, we analyzed data from a genetically informative sample of families with twins in Australia (n = 2,103), as well as general population samples from the UK (n = 501) and the US (n = 500).
Genetic factors had little influence in individual differences in biological essentialism or in its relationship with heuristic thinking. Conservative values were genetically correlated with cognitive styles (ie. need for closure and heuristic thinking).
These findings support a bigger role of genes in explaining the relationship between cognitive processes and moral reasoning and ideology than they do the association between cognitive processes and essentialist thinking.
[Keywords: biological essentialism, motivated cognition, heuristics, moral foundations, need for closure, twin study]
…Biological basis of behavior scale: Endorsement of biological essentialism was assessed using the biological basis of behavior scale (2006). The scale is composed of 8 items (eg. “The kind of person someone is can be largely attributed to their genetic inheritance”), evaluated on a 6-point Likert scale that range from Strongly disagree to Strongly agree with no neutral option…The 2 factors extracted for the ‘biological basis of behavior scale’ separated positively-worded items (4 indicators) and negatively-worded items (4 indicators). We interpret the first factor as ‘degree to which someone believes that there are kinds of people and these are genetically determined’ and we will refer to this factor as Genetic Essentialism. Items in the second factor referred to the ‘degree to which someone rejects the idea of any genetic influence on any human trait’ and we will refer to this factor as Genetic Indeterminism.
…Twin Models: At the univariate level, the 7 constructs showed some degree of genetic influence, with cognitive reflection test scores showing the strongest genetic influence (VA = 59%) and genetic essentialism showing the weakest (VA = 17%). The low genetic influence on genetic essentialism along with no influence of shared-environment suggests that individual differences in these beliefs (within the current population) are mostly due to the participants’ unique experiences.
In general, the genetic correlations were moderate to high while the environmental correlations were low, except Need For Predictability & Need For Order (see Table 4). We found statistically-significant genetic and environmental correlations between Genetic Essentialism and (1) the need for predictability; (2) conservative values; and (3) prejudice, which points towards shared sources of genetic and environmental variation, which could also be consistent with a causal relationship between a genetic tendency towards essentialist thinking and those variables (De Moor et al 200816ya). These correlations were positive, indicating that factors associated with higher endorsement of Genetic Essentialism are associated with more conservative values and prejudice in the context of mental health genetics.
We also found statistically-significant genetic and environmental correlations between conservative values and Need For Predictability, and a statistically-significant genetic correlation between conservative values and cognitive reflection test scores. This correlation was negative, indicating that tendency to commit heuristic errors and conservative values might share genetic pathways, and that the factors associated with committing less heuristic errors would be associated with endorsing less conservative values. Genetic bivariate models showed good fit: RMSEA values were between 0.00 and 0.01, TLI values between 0.97 and 1, and the LRT did not show differences between the saturated model and the fitted model. variance components, correlations, and percentage of variance explained by each component for the 9 bivariate comparisons are also presented in Table 4.