Contents lists available at ScienceDirect Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid Is general intelligence responsible for differences in individual reliability in personality measures? David Navarro-González, Pere Joan Ferrando, Andreu Vigil-Colet ⁎ Universitat Rovira i Virgili, Research Center for Behavior Assessment, Spain ARTICLE INFO Keywords: Person reliability Response bias Personality differentiation hypotheses ABSTRACT One possible hypothesis for personality differentiation is the higher reliability of high-ability individuals in typical response measures. This differential reliability has been explained as resulting from different verbal abilities as a consequence of the difficulties that low-ability individuals have in understanding items, or as the effect of response bias, or due to higher precision in the answers of high-ability individuals. The lack of an estimation of individual reliability has made it difficult to test these hypotheses. However, recent psychometric advances have made it possible to measure person reliability and thus address the issue. The present study analyses the relationships between person reliability measures and the response bias of different personality measures in measurements of intelligence in a sample of 532 adolescents. The results show that person reliability is more closely related to general intelligence than to specific abilities and that the results for low-ability in- dividuals cannot be explained by verbal deficits or by higher levels of acquiescence or social desirability. The differential reliability of measures across ability levels therefore seems to be related to higher levels of trait- edness in high-ability individuals, i.e. traits are represented in them with greater strength and clarity. 1. Introduction The potential interactions between intelligence and personality measures are a subject that has generated considerable controversy for many decades. These interactions do not refer directly to the relation- ships between personality and intelligence, but rather to a series of problems related to (a) the extent to which intelligence levels affect the factorial structure of personality measures or the relationships between personality dimensions, and (b) the possibility that the level of differ- entiation of abilities may depend on certain personality dimensions. The issue summarized above was first reported by Shure and Rogers (1963), who found that the factor structure of personality scales dif- fered as a function of individual levels of intelligence, and Eysenck and White (1964), who found a different factor structure of intelligence depending on individual levels of neuroticism. These types of result were later integrated into the personality differentiation hypothesis (PDH) framework developed by Brand, Egan, and Deary (1994). The PDH suggests that people with a higher level of ability have a more differentiated personality structure because they have more freedom to develop their personality, and this, results in greater distinction be- tween them. If this hypothesis is true, then certain outcomes can be predicted when analysing the interactions between measures of per- sonality and measures of ability. First we can expect a lack of factorial invariance when assessing the structure of personality measures across different intelligence levels, insofar as fewer dimensions will be needed to describe the personality structure of less intelligent individuals. Second, high-ability individuals will show greater variability in per- sonality measures than low-ability individuals. Finally, we can expect a lack of invariance of ability measures across levels of personality due to different relationships between ability measures across levels of dif- ferent personality dimensions such as neuroticism. The above predictions have generated a considerable amount of research over the last 30 years, but so far the evidence in favour of the PDH is inconsistent. With respect to the first issue mentioned, certain studies have detected a lack of invariance in personality measures across intelligence levels (Allik, Laidra, Realo, & Pullmann, 2004; Mclarnon & Carswell, 2013) or different correlations between person- ality measures across ability levels (Austin et al., 2002). Others, how- ever, have reported that personality remains essentially invariant (De Fruyt, Aluja, García, Rolland, & Jung, 2006; Waiyavutti, Johnson, & Deary, 2012) or that the correlations between personality measures were equal across ability levels (Austin, Deary, & Gibson, 1997). With regard to the second prediction, different authors have re- ported an increased variance of personality scores among high-ability individuals, but only for some of the personality dimensions analysed. Austin et al. (1997), for instance, reported this effect only for openness https://doi.org/10.1016/j.paid.2018.03.034 Received 31 January 2018; Received in revised form 16 March 2018; Accepted 17 March 2018 ⁎ Corresponding author at: Universitat Rovira i Virgili, Departament de Psicologia, Crtra. Valls s/n, 43007 Tarragona, Spain. E-mail address: andreu.vigil@urv.cat (A. Vigil-Colet). Personality and Individual Differences 130 (2018) 1–5 0191-8869/ © 2018 Elsevier Ltd. All rights reserved. T