The mixed or multilevel model for behavior genetic analysis.

摘要: We propose the mixed model or multilevel as a general alternative approach to existing behavior genetic analysis—an correlation analysis, DeFries-Fulker and structural equation modeling. The handles readily families of behavioral data, which include paired sibling data (e.g., pairs MZ DZ twins) clustered family more than two biological siblings) special cases. Not only can have siblings, it also contain multiple types siblings pair twins, full sibling, half sibling). In contrast traditional approaches, is insensitive order in cluster. apply our large, nationally representative sample collected recently by Add Health Study. demonstrate through several applications using both complex data: conventional variance decomposition analysis interactions between environmental influences, possible basis for friendship selection. compare results from model, Pearson's model.