THE EFFECTS OF TASK AMBIGUITY AND EXPECTANCY CONTROL GROUPS ON THE EXPERIMENTER BIAS EFFECT.

The present study was designed to investigate the effects of task ambiguity and expectancy control groups on the Experimenter Bias Effect (EBE). The results indicated that the main effects of Induced Expectancy, Ambiguity, and Actual State (actual pre-established score differences on the test protocols), as well as the interaction of Ambiguity X Induced Expectancy and Ambiguity X Actual State, were all significant at the .01 level. The Ambiguity X Induced Expectancy interaction indicated that Induced Expectancy had an effect only when subjects were given ambiguous criteria with which to score the test protocols. The Ambiguity X Actual State interaction indicated that the high ambiguity condition resulted in inaccurate scoring, while low and medium ambiguity did not. Finally, the main effect of Induced Expectancy was small when compared to the Actual State effect, suggesting that some of the experimenter bias effects in the literature may have been obtained because basically unscorable stimuli were used. [ABSTRACT FROM AUTHOR]