Assessing the influence of confounding subject variables in neuropsychological research in alcoholism and related disorders.

With the recent emphasis on clinical research there has developed a greater sophistication in identifying and uncovering possible confounding subject variables. The investigator searching for etiologies and underlying relationships is accordingly increasingly faced with the question of how to deal with, or control numerous subject variables which may covary with the condition or disorder under study. The present paper describes some of these variables and the problems their presence raises for the conduct of clinical research. Although most of the examples described such as childhood hyperactivity, head injury, psychopathology, and physical disorders of the subjects are specifically relevant to neuropsychological research in alcoholism, many of them should be pertinent also for related areas of inquiry. The difficulties and problems in controlling, ruling out and taking such variables into account, as well as guidelines and strategies for their treatment and management, are considered. The question of the generality of findings which arises when so many subject variables are controlled is discussed. It is concluded that we are just on the threshold of identifying and clarifying the behavioral effects of numerous subject variables, conditions and disorders. The implications of this situation for the validity of our current research findings are discussed.