Theoretical formulation in social epidemiology increasingly posits the existence of modifying effects on the relationship between key risk factors and disease onset or course. Unfortunately, there is conceptual and methodological disagreement over how the parameters in models that contain modifier effects should be estimated and interpreted. These disagreements are reviewed and guidelines are offered for researchers who work with models of this sort. (June 1982)
Journal of Health and Social Behavior
1982
Cleary PD and Kessler RC
http://www.jstor.org/discover/10.2307/2136512?uid=3739256&sid=21101999149803