“…Even when data are gathered that provide information both about non‐drug variables during pregnancy and the postnatal environment, the interpretation of the role that these variables play in influencing the teratogenic outcomes is complex and rarely definitive. Much has been written about both the need and the strategies for assessing and controlling a wide gamut of potentially pre‐ and postnatal confounding variables in prospective longitudinal assessments of drug‐exposed offspring (e.g., Bellinger, 2000; Fried, 1998; Jacobson & Jacobson, 1996; Kilby & Ashgar, 1991; Wetherington et al., 1996). However, even with this emphasis, it is important to keep in mind that the frame of reference for behavioral teratology is not that the focus teratogen is the best predictor of the dependent variable being assessed, but rather that the drug of primary interest continues to have predictive power after appropriate statistical control for other significant predictors (Streissguth, Barr, Sampson, Darby, & Martin, 1989).…”