Do situational variables act as substantive causes of relationships between individual difference variables? Two large-scale tests of ''common cause'' models

James, Demaree, Mulaik, and Ladd (1992) proposed that situational variables may act as substantive (''common'') causes of relationships between individual difference variables as well as statistical artifacts (i.e., measurement unreliability) associated with these variables, thus invalidating assumptions of current validity generalization/meta-analysis procedures. In this investigation, we report the results of two large-scale studies designed to test hypothesized relationships derived from two ''common cause'' models. Study 1 examines relationships between store-level organizational climate variables and employee satisfaction and performance variables for 33,097 sales personnel in 537 retail stores. Study 2 investigates relationships between store-level situational constraints and customer service perception and shopping variables for 31,611 customers from 564 retail stores. The results of these studies did not support the proposition that situational variables act as substantive causes of correlations among the respective employee and customer variables or the variances and reliabilities of these variables. The implications of these findings for meta-analyses in applied psychology as well as the generalizability of the findings are discussed.