T wo observations suggest a genetic similarity among close friends: (1) the observed similarity among friends in certain traits and behaviors; (2) the evidence that many of these traits and behaviors have a genetic component. This hypothesis is tested using monozygotic (MZ) twins, dizygotic (DZ) twins, full biological siblings, and their best friends from the National Longitudinal Study of Adolescent Health (Add Health). Whether the resemblance between MZ twins' friends is larger than that between friends of DZ twins and full siblings is investigated with respect to 4 individual traits or behaviors: grade point average (GPA), AHPVT (the version Peabody Picture Vocabulary Test [PPVT] designed for Add Health), aggressive behavior, and depression. Our hypothesis is supported in analyses among same-sex friends (not among different-sex friends) for GPA, aggressive behavior, and AHPVT.
HypothesisEmpirically, two sets of well-established findings suggest the hypothesis of a genetic similarity among close friends, at least with respect to certain attitudes, traits, and behaviors: (1) friends share certain attitudes, traits and behaviors (e.g., Bearman & Bruckner, 2000;Berelson et al., 1954;Duncan et al., 1972;Matsueda, 1982;Matsueda & Heimer, 1987); (2) many of these attitudes, traits, and behaviors have been shown to be partially genetic (e.g., Plomin et al., 2003;Plomin et al., 2001;Rodgers et al., 1999). These findings point to the possibility of a genetic similarity among best friends because they are similar in certain characteristics and because the source of the similarity is partially genetic.A hypothesis of genetic similarity shared by spouses and friends (Rushton, 1989) was suggested to expand Hamilton's (1964) theory of inclusive fitness. Hamilton's theory was put forward as an explanation for altruism: individuals who perform altruistic acts towards genetic relatives are more efficient in spreading their genes (inclusive fitness) than individuals who only attempt to spread their genes through their own offspring (individual fitness). A number of studies have tested the hypothesis of genetic similarity by showing that spouses and close friends are more similar than randomly paired individuals and that the friends and spouses of more genetically related individuals (e.g., monozygotic [MZ] twins) are more similar to each other than are those of less genetically related individuals (e.g., dizygotic [DZ] twins; Baker & Daniels, 1990;Daniels & Plomin, 1985;Iervolino et al., 2002;Manke et al., 1995;Rowe & Osgood, 1984;Rushton & Bons, 2005).In this analysis, we test the hypothesis of genetic similarity shared by best friends among adolescents using MZ twins, DZ twins, and full siblings as well as their best friends collected by the National Longitudinal Study of Adolescent Health (Add Health; Harris et al., 2003). Our analysis makes at least three contributions. First, we use data on both genetically related individuals and their friends directly. In almost all previous empirical tests, friends data are reported b...