The Experimenter's Expectancy Effect: A Meta-experiment

Abstract

The claim that the outcome of an experiment may be determined by what its experimenter expects to obtain was empirically assessed with a meta-experiment. Three groups of experimenters were asked to conduct Rosenthal & Fode's (1963a) photorating task under two conditions which jointly satisfied the formal requirement of an experiment. The three groups of experimenters were given different information about the expected outcome. There was no evidence of experimenter's expectancy effect when it was properly defined in terms of the difference between the two conditions. Some issues raised by Rosenthal & Fode's (1963a) study, in its capacity as evidence for the experimenter's expectancy effect in particular, are examined. Also discussed are a few metatheoretical issues, as well as some pedagogical implications, of accepting uncritically experimenter's expectancy effect.