All analyses, including the calculation of the mean externalizing scores and the genetic analyses, were conducted using the statistical program SPSS Statistics 17 (PASW Statistics Release 17.0.2). We ran standard linear regression analyses to test for an association between mean externalizing score and each SNP from CHRM2. Our coding system reflected an additive genetic model, with each of the SNPs coded 0, 1, or 2 according to the number of copies of the major (i.e., most common) allele. We first tested main-effects models separately for each measured polymorphism and for parental monitoring. We then tested for interaction effects between each SNP and parental monitoring. Sex was included as a covariate in all analyses.
