among them can affect only the ERP values of individual twins, rather than the pair-wise similarity in ERP values of twin pairs. Determining such effects are not, accordingly, our primary goal. Instead, we target interactions that influence intrapair resemblance, and, of these, we particularly focus on the two-way interactions that test whether intrapair similarities in the twins' ERP values change with their alcohol use. These interactions are the central focus of our analyses. Three-way interactions test whether genetic influences (indexed by MZ vs. DZ comparisons) modulate pairwise differences associated with alcohol use; these are of obvious interest, as well. But because we selected twin pairs based on their intrapair similarities and differences for adolescent alcohol-related problems, and because larger differences are less often observed among MZ twin pairs, the three-way interactions in these results must be interpreted with caution and regarded as suggestive only, pending confirmation in a fully representative twin sample.
