21 - The volunteer problem revisited

Summary

This chapter is a retrospective look at an early aspect of Bob Rosenthal's work and its implications for present-day research. The theme of this particular work was the characteristics of subjects who volunteer for research participation and the effects associated with their perception of the experimenter's expectations. For a number of years, Bob and I collaborated in research and reviews on this topic. Our initial aim was to discover in what ways volunteers were not representative of the population under investigation. We also sought to discover whether subjects who volunteered were more likely to acquiesce to cues they perceived as reflecting the experimenter's expectancies. Our investigation of these problems began in the 1960s (e.g., Rosenthal, 1965; Rosnow & Rosenthal, 1966) and culminated in the following decade with several syntheses (e.g., Rosnow & Rosenthal, 1976), including our monograph entitled The Volunteer Subject (Rosenthal & Rosnow, 1975).

I begin by focusing on the significance of the topic. I then summarize what we learned about the characteristics of volunteer subjects and turn to a sample of studies (by ourselves and others) in order to show the way in which volunteer status affected a variety of experimental outcomes. Next, I summarize our recommendations for lessening volunteer bias by increasing subject participation. In reviewing the characteristics of volunteers and the factors that stimulate participation, I also mention the degree of confidence that we believed was warranted by each set of conclusions. Finally, I allude to the delicate balance that exists between artifacts and ethics, a topic that has more recently occupied our attention (Rosenthal & Rosnow, 1984b, 1991; Rosnow, 1990; Suls & Rosnow, 1981).