Personality similarity in unrelated look-alike pairs: Addressing a twin study challenge
Highlights
► Personality similarity is negligible in unrelated look-alikes. ► MZ twins’ personality similarity is most likely associated with their shared genes. ► MZ twins are alike due to their shared genes, not due to their similar treatment by others.
Introduction
Twin studies have proliferated, addressing the origins of intelligence (Johnson, Bouchard, & McGue, 2007), personality (Jang, Livesley, Ando, et al., 2006) and voting behaviors (Fowler, Baker, & Dawes, 2008). Greater resemblance between monozygotic (MZ) than dizygotic (DZ) co-twins demonstrates genetic influence on virtually all measured traits (Segal, 2012). However, MZ co-twin resemblance is never perfect, a finding largely explained by non-shared environmental influences (events producing differences between relatives). Modest contributions to behavior from shared environmental effects (events producing similarities between relatives) have been found for childhood IQ (Segal, McGuire, Havlena, Gill, & Hershberger, 2007), juvenile delinquency (Rowe, 1994) and vocational interests (Betsworth et al., 1994). However, a significant finding over the last three decades is that shared environments have little effect on most behavioral traits measured during adulthood (Plomin, 2011).
A recurrent criticism is that MZ twins are alike behaviorally because people treat them alike, due to their matched physical appearance (Joseph, 2001; Palmer, 2011). An opportunity to revisit the question of whether physical resemblance eventuates in personality and self-esteem similarity was offered by unrelated individuals who look physically alike (U-LAs).
Twin studies have demonstrated genetic influence on personality traits. The Multidimensional Personality Questionnaire (MPQ) scales yielded a mean heritability of .46 (ri = .33–.56), based on 74 pairs of MZ twins raised apart (MZA) (Bouchard, 2007); the MZA intraclass correlation directly estimates heritability. A remarkable finding is that MZA twins are as similar as MZT twins across most personality traits, showing that personality similarity resides mostly in shared genes, not shared environments (Bouchard, 2007). An MPQ study of 165 MZT and 352 DZT twin pairs yielded a mean heritability estimate of .44 (.35–.54) (Finkel & McGue, 1997), nearly identical to the .46 value based on MZA twins.
Twin studies have also examined the origins of personality traits comprising the five-factor model (Riemann, Angleitner, & Strelau, 1997; Yamagata et al., 2006). The Big Five traits include Neuroticism, Extraversion, Agreeableness, Openness and Conscientiousness. Model-fitting procedures using data from adult twin pairs have yielded trait heritabilities ranging from .42 to .56 (Riemann et al., 1997). Reared apart twin studies have also found genetic influence on the Big Five personality traits. Analyses of various subsets of twins from the Minnesota Study of Twins Reared Apart and the Swedish Adoption Twin Study of Aging yielded mean MZA correlations of .28 to .56 and mean DZA correlations of .11 to .18 across the five factors (Bouchard, 1993).
Heritable effects on self-esteem have been demonstrated by studies of adolescent and adult twins and siblings. McGuire et al. (1999), using the Harter Self-Perception Profile, found that genetic influence rose from .16 to .60 between the ages of 13.6 and 16.2 years. Kamakura, Ando, and Ono (2007), using a Japanese translation of the Rosenberg Self-Esteem Scale (RSES), showed that genetic influence increased from .31 to .49 between the ages of 19.8 and 21.1 years. This increase was explained by the possibility that genetic effects underlying different facets of self-concept become more evident as young people acquire greater control over their environment. Environmental events unique to each pair member, possibly associated with peer networks and/or social activities, also appear to be important.
A longitudinal study of self-esteem among young and middle-aged adult female twins from the Virginia Twin Registry, also using the RSES, yielded heritabilities of .40 at time 1 and .36 at time 2 (16 months later). A repeated measurement model estimated that the heritability of self-esteem was .52 (Roy, Neale, & Kendler, 1995). A subsequent self-esteem study from the Virginia group reported heritabilities of .32 for females and .29 for males (Kendler, Gardner, & Prescott, 1998), with negligible shared environmental effects.
In summary, self-esteem appears to have meaningful genetic effects that vary across the life span. The question of what is inherited remains, but relevant factors may reside in personality and temperamental traits. Another relevant body of research, the reported associations between physical attractiveness and personality traits, is reviewed below.
A sizable literature supports the existence of stereotypes linking greater physical attractiveness with more desirable social and personality traits (Dion et al., 1972, Langlois and Stephan, 1981). The reverse association, namely that more favorably depicted individuals are perceived as more physically attractive than those depicted unfavorably, has also been demonstrated (Gross & Crofton, 1977). A recent meta-analytic study also linked greater physical attractiveness with more favorable personality attributes in children and adults, but could not examine a possible causal link between treatment by others and personality (Langlois et al., 2000).
A somewhat different picture emerges in studies using self-ratings of personality and others’ perceptions of attractiveness. A meta-analysis of seventy-eight experiments discerned no meaningful relationships between physical attractiveness and self-rated personality traits, such as sociability and dominance (Feingold, 1992). In contrast, self-rated physical attractiveness correlated positively, but modestly, with extraversion (r = .25) and self-esteem (r = .32). A more recent study found that the effects of the Big Five personality traits on social status were independent of attractiveness (Anderson, John, Keltner, & Kring, 2001). Other investigators identified facial symmetry as a possible correlate of personality, especially for Openness, Extraversion and Agreeableness, but did not find the predicted positive associations between them (Fink, Neave, Manning, & Grammer, 2005).
This work must be reconciled with the twin studies cited above because genetic explanations of MZ twins’ behavioral similarity are still challenged. Despite evidence against causal connections between treatment and personality, critics argue that similar looking people are treated alike, such that MZ twins’ similar treatment—not their similar genes—explains their behavioral resemblance.
Nisbett (2009) recently faulted genetic findings from MZA twin studies, asserting that they are alike behaviorally “because they look so much alike or have other characteristics in common that tend to elicit the same sorts of behavior from other people” (p. 27). Based on this belief, critics reason that MZ twins’ genetic identity has little effect on their similar behaviors, relative to their appearance.
A number of studies have addressed this misconception and have found it wanting. Loehlin and Nichols (1976) showed that MZ twins treated alike were not more similar in personality than those treated differently. Plomin, Willerman, and Loehlin (1976) found that frequently confused twins were rated less similar behaviorally by their parents than twins who looked less alike. Rowe, Clapp, and Wallis (1987) showed that twins remain alike in personality even after controlling for their degree of physical attractiveness. Twins’ behavioral resemblance is also consistent with their true, rather than perceived, zygosity (Goodman and Stevenson, 1989, Kendler et al., 1993). Recent studies have generally concurred with these findings (Cronk, Slutske, Madden, Bucholz, & Heath, 2002). However, despite the supporting evidence, misconceptions persist, a situation warranting new attempts at resolution.
U-LAs parallel MZA twins due to their matched appearance and separate rearing, but lack a genetic link. If, as twin research critics have argued, physical appearance triggers certain classes of treatment by others, then the personality similarity of U-LAs should approach that of MZ twins, or at least DZ twins. The U-LA vs. DZ contrast may be especially revealing because DZ twins look less alike then U-LAs, but share 50% of their genes, on average. Alternatively, if shared genes underlie MZ co-twins’ matched behaviors, then U-LAs should show little or no personality or self-esteem resemblance.
Section snippets
Participant sample
U-LA pairs were identified initially by the French Canadian portrait photographer, François Brunelle. Brunelle has been creating black and white photographs of these rare dyads for many years as part of his “I’m Not a Look-Alike!” project. U-LAs are identified through the media and via his web site where potential participants can register (Brunelle, 2012). Photographs of two U-LA pairs are displayed in Fig. 1A and B.
In winter 2010, questionnaire booklets (described below) were mailed to
Analysis
The resemblance of the U-LA pairs was assessed by intraclass correlations, which express the proportion of shared variance, together with 95% confidence intervals. The sample size offered limited power, so an initial decision was made to set the statistical significance level at p < .10. However, the results were virtually the same at the more stringent level of p < .05, so that value was applied throughout the present study. All personality and self-esteem data were age- and sex-corrected prior to
PfPI
Age and sex showed negligible to modest size correlations with the Big Five personality scales with none reaching statistical significance (age: r = −.08 to .23; sex: r = −.26 to .16, n [individuals] = 48). Within-pair age differences showed little to modest association with the intra-pair personality score differences (r = −.14 to .31), but none were statistically significant. The interval between pair members’ meeting and study participation also showed little relationship with their personality
Discussion
The present study began by identifying a common criticism of twin research, namely that MZ co-twins’ matched physical appearance is responsible for their behavioral resemblance, due to their similar treatment by others. The lack of statistical evidence that physically similar people match in personality and self-esteem challenges that claim. If matched physical appearance contributes significantly to the behavioral resemblance between two people, then the U-LAs should have been more alike than
Conclusions
Two conclusions can be drawn from the present study. First, it appears that MZ co-twins’ personality similarity mostly reflects their shared genes. Second, reactive gene-environment correlation may best explain MZ co-twins’ similar treatment by others.
A theoretically ideal, albeit practically difficult, analysis of physical and behavioral similarity would involve “U-LATs”--unrelated look-alike, same-age individuals raised together. Virtual twins (VTs or same-age unrelated individuals), composed
Acknowledgements
François Brunelle, Laura Brunelle, Scott Hershberger, Jamie Graham, Jamie Munson and William Marelich are gratefully acknowledged.
References (41)
- et al.
Genetic and environmental influences on vocational interests assessed using adoptive and biological families and twins reared apart and together
Journal of Vocational Behavior
(1994) - et al.
Emotional and behavioral problems among female twins: An evaluation of the equal environments assumption
Journal of the American Academy of Child and Adolescent Psychiatry
(2002) - et al.
Facial symmetry and the ‘big-five’ personality factors
Personality and Individual Differences
(2005) - et al.
Behavioral genetics of the higher-order factors of the Big Five
Personality and Individual Differences
(2006) - et al.
Genetic and environmental influences on the Verbal-Perceptual-Image Rotation (VPR) model of the structure of mental abilities in the Minnesota Study of Twins Reared Apart
Intelligence
(2007) - et al.
Genetic and environmental effects of stability and change in self-esteem during adolescence
Personality and Individual Differences
(2007) - et al.
Intellectual similarity of virtual twin pairs: Developmental trends
Personality and Individual Differences
(2007) - et al.
What virtual twins reveal about general intelligence and other behaviors
Personality and Individual Differences
(2012) - et al.
Who attains social status? Effect of personality and physical attractiveness in social groups
Journal of Personality and Social Psychology
(2001) - et al.
Social ecology of similarity: Big schools, small schools and social relationships
Group Processes and Intergroup Relations
(2011)
Genetic and environmental influences on adult personality: Evaluating the evidence
Genes and human psychological traits
Sources of human psychological differences: The Minnesota Study of Twins Reared Apart
Science
What is beautiful is good
Journal of Personality and Social Psychology
Good-looking people are not what we think
Psychological Bulletin
Sex differences and nonadditivity in heritability of the Multidimensional Personality Questionnaire scales
Journal of Personality and Social Psychology
Genetic variation in political participation
American Political Science Review
Parental criticism and warmth toward unrecognized monozygotic twins
Behavioral and Brain Sciences
What is good is beautiful
Sociometry
Separated twins and the genetics of personality differences: A critique
American Journal of Psychology
Cited by (13)
The genetics of investment biases
2014, Journal of Financial EconomicsCitation Excerpt :One concern has been that the matched physical appearance of identical twins results in more similar treatment by those who are a part of these individuals׳ environments, in the end causing more similar outcomes. Using a novel research design, Segal (2013) studies unrelated look-alike individuals, and finds that their correlations for personality measures are much lower than for identical twins, suggesting that identical twins׳ similarity with respect to personality mostly reflects similarity in their genes, and not similar treatments by others. Finally, a concern relevant for this paper is that inheritances that twins receive from their parents might be more similar (for example, with respect to the asset composition) for identical than fraternal twins.
"hey, This is not like me!" Convergent validity and personal validation of computerized personality reports
2013, Revue europeenne de psychologie appliqueeCitation Excerpt :The Dutch manual (De Fruyt & Rolland, 2013) reports a series of validity studies demonstrating that specific PfPI traits predict a broad range of work performance indices, including task (quality and quantity), contextual and adaptive performance, underscoring its claim as a work-relevant personality inventory. Although the PfPI has been primarily developed for personnel selection and career counseling purposes, it has been also used in scientific research as an indicator of general traits (e.g., Segal, 2013). The PfPI is online-administered, scored and further provides a computerized personality descriptive text report in addition to the raw and normative scale scores.
Unrelated look-alikes: Replicated study of personality similarity and qualitative findings on social relatedness
2013, Personality and Individual DifferencesCitation Excerpt :Conscientiousness showed the highest correlation in both the replication (ri = .18) and the original analysis (ri = .29), while the sizes of the other four correlations varied across the two analyses, most likely reflecting small sample fluctuation. Most importantly, the mean correlation of −.03 was nearly identical to the mean correlation of −.05 reported by Segal (2013), confirming the lack of personality similarity in U-LA pairs. These correlations were also substantially below those reported previously for MZA (ri = .53), MZT twin pairs (ri = .48,), DZA (ri = .15) and DZT (ri = .20) twin pairs (Bouchard, 1993).
Evidence for the scarr-rowe effect on genetic expressivity in a large u.s. sample
2018, Twin Research and Human GeneticsReared-Apart Chinese Twins: Chance Discovery/Twin-Based Research: Twin Study of Media Use; Twin Relations over the Life Span; Breast-Feeding Opposite-Sex Twins/Print and Online Media: Twins in Fashion; Second Twin Pair Born to Tennis Star; Twin Primes; Twin Pandas
2017, Twin Research and Human Genetics