9 Human Subjects

Abstract The evidence presented up to this point that the expectancy of the experimenter may in part determine the results of his research has been at least somewhat equivocal. Some of the evidence has been anecdotal. Some has required the untenable assumption that the expectancy of the experimenter, and not some correlated variable, had led to the effects observed. The most clear-cut evidence for the effects of the experimenter's expectancy, therefore, must come from experiments in which experimenters are given different expectancies. Of the studies examined, that by Stanton and Baker (1942) comes closest to meeting this requirement of the experimental induction of an expectancy. That study does require, however, the assumption that experimenters will expect the subjects to answer correctly the items being presented. The same assumption is required to interpret the case of Clever Hans as an experiment in expectancy effects. This chapter describes experiments that appear to be fairly straightforward tests of the hypothesis of the effects of the experimenter's expectancy on his research results.