Demand Characteristics and three conceptions of the frequently deceived subject

Experimentally naive undergraduates participated either in 1 of 5 group-administered attitude-change experiments or in all 5. Of those who took part in 5, some experienced the experiments in 1 sequence while others experienced them in the reverse sequence. The aim of the design was to hold constant the experiments, to vary the frequency of previous deceptions and debriefings, and to see if Ss with a longer experimental history would seek to infirm hypotheses, confirm them, or disregard them and obey only experimental instructions. While experimental history affected global attitudes towards experiments, it did not affect attitude, incidental learning, or task performance in Metaexperiment I. An attempt was made in Metaexperiment II to manipulate the Ss' suspicion in a context where the E's hypothesis could be readily guessed. The experiment pointed to 2 kinds of experimental history which may induce bias, showed that experimental performance was not biased in the condition of greatest presumed suspicion, and demonstrated that experiencing deception and knowing of deception (without experiencing it) are not functionally equivalent.