Deception in Research

Why to Employ Deception

The choice to use deception as part of a research design is based, at times, on the need to avoid biased responses by participants. Demand characteristics [Page 360]in an experimental design are features of the experimental setting that have the potential to create biases toward certain behaviors or responses by participants independent of the true effect being investigated.

In some cases, demand characteristics take the form of participants identifying the true nature of the hypotheses being tested. In such cases, participants may alter their behavior in response to what they perceive as the experimenter’s expectations instead of responding naturally based on what they are experiencing. One such case involves the good subject effect. The good subject effect refers to the tendency of participants who identify the purpose of an experiment responding in a manner consistent with what they anticipate the experimenter desires from them. Research using appeals involving attitudes considered pro-social may produce outcomes associated with the target of the message appeal rather than the specific communication strategy used to generate the outcome. Another example involves the negativistic subject effect. In this example, participants behave in ways that they believe will provide results inconsistent with those predicted by the researchers. A person may communicate in a way to appear as a good person rather than communicate in a way that reflects the actual belief.

These biases serve as potential confounds for experimental manipulations within the research design. That is, biased responses serve as an alternative explanation for participants’ behavior in the study. As such, they are a threat to the internal validity of the study. Results found in support of the researcher’s hypotheses may be explained in an alternative manner. In this circumstance, deception could be used to prevent the participant from recognizing the purpose of the experiment and therefore prevent biased responses that directly correspond to experimenters’ predictions. One example involves research about cheating whereby a confederate is used that actually does the cheating on some outcome from which the participant benefits and the examination is whether the participant is honest about the cheating. The use of a confederate involves deception because the respondent may not respond truthfully if he or she were aware that the other participant is actually a confederate.

Deception is also necessary to generate participant responses to circumstances that could not otherwise be feasibly or ethically introduced. At times, researchers may want to examine phenomena that could not reasonably be studied without deceiving the participants because it would be impossible or unethical to create the true experiences in which the researcher is interested. Such circumstances would include instances where the true experience would create unreasonable risks for potential harm to the participants or where it is unfeasible to try to create the true experience in a laboratory setting. Instead, the participants may be led to believe they are experiencing something that is not, in fact, truly occurring. The Stanley Milgram investigations involving obedience to authority used a deceptive set of communications that led participants to believe that their actions injured another person. No such injury took place, but understanding the response to a communicative event was argued to necessitate the deception. The next section considers different ways a researcher may create false impressions in participants as part of the research design.

...