At the risk of causing readers to wonder about the state of my mental health, I admit to hearing statements such as these knocking about in my head - statements gleaned from many interactions with Bob Rosenthai during my tenure as a graduate student at Harvard. I now make these comments, and similar others, to my own students and colleagues. Bob served as a mentor to me in the domain of quantitative methods, and I continue to be heavily influenced by his approach to research: by his delineation of the types of controls needed to reduce experimenter effects and other artifacts in behavioral research, by his zest for plunging into new areas of methodology and statistics, and by his concern for external as well as internal validity in empirical research.

In the present chapter, I would like to take up the issue of experimenter effects - and experimenter expectancies in particular - in the context of an applied area of research that heavily utilizes nonlaboratory, field interviewing. Specifically, I focus on the nature and manner by which researchers survey mental health in the community, and on difficulties and rewards encountered in such field interviewing when we attempt to limit the many ways in which researchers can inadvertently influence individuals' responses.