Can minimally trained observers provide valid global ratings?

From a cost-benefit perspective, the validity of minimally trained (MT) coders' ratings was examined individually and relative to ratings by coders who had undergone "gold standard" (GS) training in the rating of maternal overreactive discipline. The ratings of MT and GS coders were strongly associated (r = .72). MT ratings were also correlated with maternal self-reported overreactive discipline (r = .39), laxness (r = .46), depressive symptoms (r = .32), and observed child misbehavior (r = .40). These correlations were of similar magnitude to correlations based on GS ratings. MT ratings were more strongly associated with GS ratings than with child misbehavior. Thus, their ratings demonstrated convergent, concurrent, and discriminant validity. Implications and limitations of these findings are discussed. ((c) 2006 APA, all rights reserved).