Developmental Psychology1993. Vol. 29, No. 1,96-109Copyright 1993 by ihe American Psychological Association, Inc.0012-1649/93/S3.00Personality Stability and Change in Early Adulthood:A Behavioral Genetic AnalysisMatt McGue, Steven Bacon, and David T. LykkenSeventy-nine monozygotic and 48 same-sex dizygotic twin pairs completed the MultidimensionalPersonality Questionnaire twice, averaging 20 years of age at first and 30 years at second testing.There were significant mean decreases in measures of Negative Emotionality (NE), increases inmeasures of C o n s t r a i n t (CO), but no significant mean changes for measures of Positive Emotional-ity (PE). Variance decreased for measures of NE but remained stable for measures of PE and CO.Biometrical analyses revealed that (a) NE variance reduction was due to diminishing genetic influ-ences, (b) personality stability was due largely to genetic factors, and (c) although some evidence forgenetic influence on personality change was observed, change was determined largely by environ-mental factors. It is concluded that the stable core of personality is strongly associated with geneticfactors but that personality change largely reflects environmental factors.There is a growing consensus among personality psycholo-gists that personality stabilizes in adulthood (e.g., McCrae &Costa, 1990). Nonetheless, adolescence, the period leading upto adulthood, has long been recognized as a psychologicallyturbulent age (Hall, 1904; Rutter, Graham, Chadwick, & Uhle,1976). Thus, for example, the rates of violent behavior (Daly &Wilson, 1988), criminal activity (Gottfredson & Hirschi, 1990),alcohol and drug abuse (Johnston, O'Malley, & Bachman,1985), as well as other antisocial behavior all peak before age 25years. Yet, the majority of troubled youths do not go on to haveserious problems in adulthood (e.g., Robins, 1966). The attain-ment of physical maturity is accompanied by significant psy-chological change (Siegel, 1982) a nd increasing levels of psycho-logical adjustment (Hathaway & Monachesi, 1953). Th e precisemechanisms that underlie these psychological changes, how-ever, remain to be fully determined.Only a few longitudinal studies have investigated personalitychanges during the transition from late adolescence to earlyadulthood, yielding a limited and at times inconsistent set offindings. For example, Stevens and Truss (1985) administeredthe Edwards Personal Preference Survey (EPPS) to two samplesof college alumni, one 12 years and the other 20 years afterinitial completion of the inventory while in college. Stevens andTruss reported mean increases—consistent across the two sam-ples—for Achievement, Autonomy, and Dominance; consis-tent mean decreases in Abasement and Affiliation; but no con-sistent changes in Aggression or the other EPPS scales. BlockMatt McGue, Steven Bacon, and David T. Lykken, Department ofPsychology, University of Minnesota. Steven Bacon is now at the Vet-erans Administration Medical Center, San Diego, California.This research is supported in part by U.S. Public Health ServiceGrants AG06886, MH73860, and DA05147. We acknowledge thehelpful comments of Thomas J. Bouchard, Jr., Crista Carmichael,Auke Tellegen, and four anonymous reviewers of an earlier version ofthis article.Correspondence concerning this article should be addressed to MattMcGue, Department of Psychology, N218 Elliott Hall, 75 East RiverRoad, University of Minnesota, Minneapolis, Minnesota 55455.(1971), in the Berkeley longitudinal studies, concluded that, formale subjects, interviewer-rated expressed hostility increasedwhile rebelliousness decreased as the sample aged from seniorhigh school to early adulthood. Stein, Newcomb, and Bentler(1986) reported significant mean increases in Law Abidance,Congeniality, Diligence, Generosity, and Leadership from theearly teen years to the early twenties in a large and broadlyrepresentative school-based sample.The evidence concerning the stability of individual differ-ences in personality from late adolescence to early adulthood issimilarly inconsistent. Thus, for example, Stein et al. (1986)reported 4-year stability coefficients from the late teens to earlytwenties that generally fell in the range of .5 to .6, leading themto conclude that individual differences crystallize with advanc-ing age. Stevens and Truss (1985), however, reported 12- and20-year stability coefficients for the EPPS that generally fell inthe .2 to .4 range, suggesting only modest stability of individualdifferences. The apparently inconsistent pattern of resultscould be caused by different sampling strategies (college stu-dents vs. representative samples), different assessment proce-dures (e.g., self-ratings vs. ratings by others), and very probablyalso the use of different measuring devices. In the present study,a community-based sample completed an omnibus self-reportpersonality inventory—the Multidimensional Personality Ques-tionnaire (MPQ; Tellegen, 1982)—first in late adolescence(average age of 20 years) and then again 10 years later when inearly adulthood. The MPQ has excellent psychometric proper-ties and includes validity scales to screen out likely invalid pro-files. Of particular significance to the present study is theMPQ's hierarchical factor structure, which provides a basis forassessing consistency of results within the study.The psychological changes that accompany the transitionfrom late adolescence to early adulthood have typically beenattributed to environmental factors; for example, to "culture"(e.g., Stein et al., 1986), to societal expectations (e.g., Dubow,Huesman, & Eron, 1987), and so on. An alternative, but lessfrequently considered, possibility is that biological factors con-tribute to personality changes. The field of developmental be-96PERSONALITY STABILITY AND CHANGE97havioral genetics seeks to identify and characterize genetic andenvironmental influences on behavioral stability and change(Plomin, 1986). The present study uses a standard yet powerfulbehavioral genetic design, a longitudinal twin study, to investi-gate genetic and environmental influences on personality sta-bility and change in the transition from late adolescence toearly adulthood.Loehlin (1992) reviews longitudinal behavioral genetic stud-ies of normal between-person variation in personality and con-cludes that although there are significant genetic contributionsto personality change in childhood, personality change in lateadolescence and early