Results of this study reveal that well-trained observers report significantly lower levels of cognitive activity on all measures than raters who had no training or only moderate training. Consequently, interpretations of cognitive components of environmental and teacher evaluation scales must be questioned unless generated by trained student raters. (Author)