It is important to examine the magnitude of twin correlations across methods of assessment, as heritability estimates may vary depending on who is rating the subject. This is especially important given the age trends found for heritable influences in Fig. 8.1, since different methods of assessment tend to be employed for different age groups. As previously discussed, studies of younger subjects rely on parent or teacher ratings, while self-report methods are typically used in studies of adults and often adolescents. The twin correlations are summarized in Fig. 8.2, according to laboratory observation, self-reports, teacher ratings, and parent/caregiver ratings. Indeed, twin correlations—and thus the estimates of genetic and environmental influences on aggressive behavior—do appear to vary across method of assessment. According to the twin correlations summarized in Fig. 8.2, the heritability of aggressive behavior based on laboratory observation is approximately 32% [h2: 2(rMZ–rDZ)=2(0.27–0.11)], dominant genetic effect accounts for approximately 10% [d2: 2(rMZ–rDZ)=2(0.27)–4(0.11)], and the nonshared environment accounts for the remaining 58% of the variance. The heritability of aggressive behavior based on self-reports is 40% and the nonshared environment accounts for the
