Probing Suspicion Among Participants in Deception Research

Deception is a popular and useful technique for examining a variety of important social psychological phenomena. Indeed, deception has proved to be indispensable in studying and establishing causality within a variety of important human social phenomena (e.g., aggression, prejudice, obedience, conformity, impression management). Upwards of 81% of studies published in the top social psychological journals use deception in their procedures (Adair, Dushenko, & Lindsay, 1985). One typical aspect of deception experiments is the postexperimental inquiry in which experimenters, using interviews or paper-and-pencil questionnaires, probe whether participants were deceived and whether they truly believed the ostensible purpose of the study. "ihe goal of the inquiry is to identify and perhaps eliminate data from participants whose behavior stems from an awareness of the deception or hypotheses rather than a natural response to the manipulations. Participants who identify the deception may supply data that reflect a response to their knowledge of the deception rather than a response to the manipulated variables, thus providing an inadequate test of the experimental hypotheses. Researchers using deception techniques obviously make an assumption that participants will report any and all suspicions about the experiment and that the experimenter can ferret out instances in which participants have uncovered or guessed the nature of the deception and the experimental hypothesis. However, a recent event in our laboratory prompted us to pause and reevaluate the effectiveness of established procedures (see Greenberg & Folger, 1988) for detecting suspicion of deception among research participants. Eight introductory psychology students were recruited to participate in a pilot study examining the effect of practice task feedback on subsequent performance. Because of a last minute cancellation, we recruited a graduate student to step in and pose as a participant. Fortunately, the s e v e n real participants in the study were unaware that the eighth participant was a graduate student who knew of the deception and the purpose of the experiment. During the study, the experimenter was required to leave the room for a brief period. Although they were admonished not to, the real participants began to discuss the experiment among themselves during the experimenter's absence. By comparing feedback they had individually received, the participants uncovered the deception and realized that their individual feedback was false. When the experimenter returned, no participant revealed his or her ill-found knowledge but, instead, continued to participate as though nothing had happened. Because we were piloting new procedures, the experimenter took great care during the postexperimental inquiry to assess whether participants were suspicious of the procedures and aware of any deception. After announcing that the experiment involved deception, the experimenter asked participants on three separate occasions whether they were suspicious of the procedures or anything that happened in the experiment. Although the participants were clearly aware of the deception, none divulged this knowledge. Why did participants not divulge their suspicions of the procedures and their awareness of the deception? It is possible that participants did not report their suspicions because to do so would reveal that they had blatantly disobeyed the experimenter's instructions not to interact during his brief absence. Participants may also have fallen into the role of "good subject" (Orne, 1962), withholding any information they felt might undermine the experiment, infirm the experimental hypotheses, or otherwise cause problems for the experimenter. Furthermore, participants may have refrained from voicing their suspicions to avoid appearing foolish should the knowledge they gained illicitly about the deception be in error. Perhaps most troubling was the inability of a highly trained experimenter to identify suspicious participants. Not only did participants not divulge their knowledge of the deception when questioned, but their responses to manipulation check questions completed before the questioning provided little evidence that they were aware of the deception. Six of the seven participants responded to the manipulation check items in a manner that was consistent with the cover story and feedback yet was inconsistent with the knowledge they had gained during their discussion. Perhaps participants are conscious that it is problematic for them to know more about an experiment than they are told and that such knowledge may invalidate their data or the experiment (Ome, 1962). Of course, our seven participants do not represent all experimental participants. Nevertheless, their behavior suggests that, e v e n when pressed, participants cannot be counted on to reveal knowledge they may acquire or suspicions they may develop about the deception or experimental hypotheses. Our serendipitous discovery has several implications for investigators using deception in their research. First, admonishing participants not to discuss the experiment among themselves in the experimenter's absence is clearly inadequate. Experimenters using deception should make it a point never to leave unsupervised any participants who have knowledge that, when shared, could lead to discovery of the deception or hypothesis. In general, it is problematic to leave participants unsupervised in any setting where discussion might alter responses to stimulus material. Second, investigators need to evaluate more carefully the procedures they use to assess suspicion among research participants. Merely having par-