The process should be conducted by researchers who are qualified to approach the debriefing in a manner that supports subjects in expressing any thoughts or feelings they may have about being deceived and can be appropriately responsive to their reactions. The APA outlines three basic requirements for debriefing. Debriefing sessions should mitigate the potential harm of deception by explaining the rationale for the deception. Participants should be given a clear and informative explanation for the design of the study and the methods used, and they should have the opportunity to ask questions.
Dehoaxing is the process of convincing subjects who have been deceived as part of a research study that they have in fact been deceived. The purpose of dehoaxing is to prevent possible future harm to the subject. For example, subjects may be given false pretest scores in order to test the effect of these scores on subsequent tests of motivation levels. If subjects believe that the false scores represent their true abilities, their level of self-esteem may become jeopardized.
In cases such as these, simply informing the subjects that they were deceived and that the pretest scores were false may not be sufficient. In addition to informing the subjects, some form of demonstration may be needed to convince subjects that they were deceived and thereby diminish the undesirable effects of the study [14]. Since subjects may feel a range of emotions at different intervals about being deceived, a process for continuous or staged debriefing may be needed; however, this is usually only done for greater than minimal risk studies.
If a study requiring debriefing will run over several days or weeks, subjects who have completed the study might tell others about it. If they have been debriefed and thus know the real purpose of the study activities, they might share that information with prospective subjects, thus compromising the scientific validity of the study.
Under these circumstances, the IRB may consider a delayed debriefing based upon the level of risk to subjects and the justification for delay. There are several strategies to handle a delayed debriefing. Provided that the delay is not extensive and an in-person debriefing is not necessary to assess and address potential harms, debriefing information can be sent via e-mail or mail.
The IRB will consider the length of a proposed delay in relation to the other study details. If names and contact information are not collected, researchers could:. Give subjects a URL where they can get debriefing information after a particular date upon which the information will be available. Have each subject self-address an envelope before they leave the study session for the purpose of receiving post-study debriefing information.
Depending on the details of the study and the level of risk, an immediate in-person debriefing may be necessary to minimize risk, even if it jeopardizes future enrollment. There are certain circumstances under which the IRB may waive the requirement for debriefing when a study involves deception, such as when the debriefing regarding deception may cause more harm than the deception itself.
Research involving deception could fall into any of the three review levels exempt, expedited, or full board depending on the specifics of the study. Our research actually focuses on the development of "status hierarchies" in small groups.
In many small groups such as project teams, ad hoc committees, or juries, some people tend to "take charge" more than others.
However, the process by which these small group hierarchies develop is not well understood. In this study, we are attempting to understand what happens when two members of a group disagree as to who should take charge. To try and obtain unbiased or natural reactions, we had to give you some false information at the beginning of the study. We informed you that, based on your scores on the tests from the prescreening packet, we had determined that you were the most suited to lead the group in the group task, and we told you that you were the only member in the group who received this information.
But in fact, we gave this same information to one other group member, i. Hu and colleagues showed that the instruction to selectively speed up deceptive answers along with a short training substantially altered the pattern of response times such that truthful and deceptive responses became indistinguishable Hu et al.
However, it cannot be generalized from these results that merely emphasizing that an examination aims at detecting deception necessarily reduces lie detection efficacy. By contrast, a study using a variant of the differentiation-of-deception paradigm in conjunction with functional magnetic resonance imaging revealed larger differences between deceptive and truthful answers in the neural activation of different brain areas when participants believed that a lie-detector was activated Sip et al.
In line with the majority of neuroimaging studies in this domain Gamer, , activity in the right inferior frontal gyrus was also modulated by deception. This region was frequently supposed to reflect the recruitment of response inhibition processes.
However, temporary disruption of the inferior frontal gyrus by means of continuous theta-burst stimulation did not significantly alter the pattern of behavioral responses in a variant of the differentiation-of-deception paradigm Verschuere et al. These results thus question the frequently assumed functional role of the inferior frontal gyrus in deception.
Besides exploring specific cues of deceptive behavior in highly standardized situations and with highly standardized interrogation techniques, it also seems interesting to examine deception in more naturalistic settings. For example, Spence and colleagues asked participants to provide relatively unrestricted honest and deceptive accounts of their opinion regarding social issues.
For these accounts, speech parameters were extracted and the authors found a significantly reduced speech rate along with increased response latency during deception compared with truth-telling Spence et al. In a similar vein, Duran and colleagues examined movement dynamics accompanying deceptive and truthful accounts. Instead of searching for specific discrete cues such as the rise of a brow, they examined the whole time course of movements and provided preliminary evidence for unique dynamic signatures of deception in these kinetic variables Duran et al.
Finally, Mackinger and Jonas explored determinants of deception in advisor-client interactions and provided evidence for the use of explicit and implicit strategic deceptive behavior in advisors aiming to receive an incentive.
Research on deception has a long tradition in psychology and related fields. On the one hand, the drive for detecting deception has inspired research, teaching, and application over many decades. On the other hand, research on deception as a process or phenomenon is characterized by manifold interactions with other areas of psychological research such as attention, memory, executive control, or motor behavior.
It remains to be debated whether deception and its detection should be studied as a key topic which entails addressing these other fields, or rather as a particular, illustrative manifestation of them.
We regard the present Research Topic as clearly underlining the scientific benefits arising from the broad and multidisciplinary perspective that characterizes deception research today and we hope that it will enrich and inspire future research in this domain. Agosta, S. The autobiographical IAT: a review.
Ambach, W. Face and voice as social stimuli enhance differential physiological responding in a concealed information test. Ben-Shakhar, G. Current research and potential applications of the concealed information test: an overview. Duran, N. Exploring the movement dynamics of deception. Furedy, J. Differentiation of deception as a psychological process: a psychophysiological approach. Psychophysiology 25, — Pubmed Abstract Pubmed Full Text. Gamer, M. Verschuere, G.
Ben-Shakhar, and E. Meijer Cambridge: University Press , 90— CrossRef Full Text. P amplitudes in the concealed information test are less affected by depth of processing than electrodermal responses. Ganis, G.
Concealed semantic and episodic autobiographical memory electrified. Hu, X. A repeated lie becomes a truth? The effect of intentional control and training on deception.
Jang, K. Effects of the combination of P3-based GKT and reality monitoring on deceptive classification. Leue, A. Lykken, D. The GSR in the detection of guilt. It is possible that researchers with stronger feelings about deception both in favor and in opposition were more likely to complete the survey than their more ambivalent colleagues.
Second, we did not define deception for our respondents. There are multiple types of experimental deception e. The attitudes of psychologists and economists are broadly consistent with policies in their respective fields, but this result alone does not provide evidence for or against the use of deception. Therefore, we now turn to Study 2, in which we tested the relationship between past experience and present suspicion, and between present suspicion and behavior in common economic tasks.
Six hundred thirty-six people women, men participated in this study. All laboratory-based experimental sessions had four, eight, or 12 mixed-sex participants divided into groups of four. In both the economics and psychology laboratories, participants sat at semiprivate computer terminals; they were not informed which individuals in the session were assigned to their four-member interaction group.
The AMT experimental sessions each had four participants who interacted in real time online as in the laboratory-based session, but because the AMT participants participated remotely, they never saw the other participants in their group.
The AMT participants did not interact with any experimenter—all their instructions were written and presented by the computer program. The in-lab participants interacted with two experimenters during the study to minimize any effects of expectancy or experimenter demand.
Experimenter A gave participants the written consent form, directed participants to the computer laboratory where participants could choose their own computer terminal , and answered any preexperiment questions. Participants traded their labeled card for the matching envelope. Neither experimenter was aware of which participant was assigned which role in the study, and neither experimenter could see any of the specific decisions that participants made.
During the study, participants engaged in four interactions in fixed order: a welfare trade-off allocation task, a dictator game, an ultimatum game, and a welfare trade-off allocation estimation. The within-experiment written instructions explained all four parts of the study before participants started any of the interactions, but participants were not told which role they would have in each interaction until the time of that interaction. Participants were informed that they would interact with the members of their group in real time but that they would never interact with the same participant in more than one task.
In-lab experimental sessions were advertised as being 30 min in duration, including the consenting, payment, and debriefing, and generally took between 15 and 25 min.
The AMT experimental sessions were advertised as being between 15 and 40 min in duration due to the variable time it took for four participants to log into the online waiting room and did not exceed this time. The order of these decisions was randomized.
Participants next completed a dictator game DG. Participants next completed an ultimatum game UG. In addition to changing roles, the participants who had been paired together in the DG were not paired in the UG. Participants were informed that they were paired with a new partner. In the final task, each participant learned the amount that an anonymous other participant from their interaction group allocated as either a dictator in the DG or a proposer in the UG.
Given this information, participants were asked to guess how that anonymous other participant made each of the 12 trade-off allocation decisions during the initial trade-off allocation task. Participants completed demographic questions followed by a postexperimental questionnaire that assessed their suspicion of having been deceived in the study. They then answered a free-response question about whether they believed their suspicion level had influenced their behavior see the supplemental information, Appendix 2.
Once all participants in a session had completed the questionnaire, the in-lab participants were directed by Experimenter A to take the labeled card at their computer terminal to Experimenter B, who gave participants their payment envelope and additional debriefing information.
AMT participants were immediately directed to online debriefing information after all participants in their session had completed the questionnaire, and they were paid their bonus and base pay within 7 days via the AMT interface. We coded suspicion in four variables.
First, spontaneous suspicion was coded from the responses to the first four free-response questions of the postexperiment questionnaire, which did not explicitly mention deception. Three independent raters coded these responses, with disagreements settled by majority rule.
Any participant who was coded as suspicious in at least one of these questions was coded as spontaneously suspicious , or otherwise as not spontaneously suspicious. Finally, we constructed a suspicion type variable that coded participants as being either spontaneously suspicious , suspicious only when prompted , or not suspicious. The final free-response question regarding the influence of suspicion on behavior in the task was also coded by the same three independent coders for containing the sentiment my suspicion had no influence on my behavior , my suspicion made me less generous , my suspicion made me more generous , and other, hard to interpret sentiment.
Previous experience being deceived or participating in a deception-tolerant subject pool did not predict suspicion in the present study. We tested this in two ways. First, because AMT is a single pool in which some researchers use deceptive methods and some do not, we tested whether individual differences in prior history of being deceived predicted suspicion of being deceived in the present study.
Second, we tested whether suspicion rates differ across subject pools as would be expected if experience in deception-tolerant subject pools e. Figure 2A—C presents the rates of prompted and spontaneous suspicion in each of the five pools. Pool differences in suspicion of deception. A Rates of spontaneous suspicion coded from nonleading free-response text. B Rates of prompted suspicion in response to an explicit forced choice question. C Combined suspicion codes, depicting all spontaneous suspicion, suspicion only after prompting, and the remaining nonsuspicious participants.
Participants who expressed both spontaneous and prompted suspicion are categorized as spontaneously suspicious. These pool-level effects do not show a clear positive relationship between the permission of deception and participant suspicion.
When prompted, psychology participants were more suspicious than economics participants, who were more suspicious, in turn, than AMT participants. Note that the majority of participants who expressed suspicion when prompted did not do so spontaneously see Fig.
Furthermore, there were no pool effects on spontaneous suspicion. That over a quarter of economics participants were suspicious when asked, despite explicit pool-level instruction that deception is banned, and the fact that AMT participants are commonly deceived but were the cohort least likely to express suspicion when prompted, argues against the hypothesis that banning deception maintains the naivety of the subject pool. Each game behavior was modeled with a set of five stepwise regressions Tables 1 , 2 , 3 , 4 and 5.
At the first step of each regression, dummy variables coding pool were added in order to test and control for pool differences in each behavior. Models 2—4 individually added each of the three suspicion measures spontaneous suspicion, prompted suspicion and suspicion confidence one at a time. Model 5 added the three suspicion measures simultaneously.
Because these models do not control experiment-wise alpha, it is reasonable to suspect that this single significant result out of 15 tests in Models 2—4 may be a false positive. To test this question more holistically across the full dataset, we proceeded with a within-study meta-analytic approach.
For each of the three suspicion measures, we calculated the effect size of suspicion for each game behavior in each pool, for every aggregation by pool and game behavior, and then in total for the full dataset. Although two of 25 such comparisons exceed the test-wise alpha for the equivalent t test, they do so in opposite directions, and every aggregation of these results by pool or measure suggests that the underlying effect of suspicion is nonexistent aggregations by pool and in total based on the subject-wise means of z -scored game behaviors.
Forest plot of the effect of prompted suspicion of deception. Psychology pool participants are plotted in blue; economics pool participants are plotted in red; Amazon Mechanical Turk participants are plotted in green; cross-pool subtotals and totals are plotted in black.
We highlight here the results investigating the effect of prompted suspicion because it more closely matches how many prior studies operationalize suspicion.
Furthermore, because more participants expressed suspicion when prompted, these tests are higher-powered. However, we also completed the same analysis for spontaneous suspicion and suspicion confidence, and those results Figs. S 1 and S 2 provide convergent evidence that the true effect of suspicion of deception is likely null. Note that we have moderate power to detect even small effects of suspicion here. To assess the evidential weight that should be given to this null result, we calculated Bayes factors for each game behavior measure and the subject-wise aggregate i.
That is, for the aggregate total scores, these data support the null hypothesis of negligible differences over 30 times more than they do the alternative hypothesis of substantial differences. Therefore, the present results provide strong evidence against the existence of even a very small effect of suspicion on behavior in the tasks used. Posterior probabilities of the null hypothesis, as a function of priors and the weight of the evidence.
This figure plots posterior probabilities of the null hypothesis that suspicion has a negligible effect on behavior with various effect sizes as the cutoff for a negligible effect versus the alternative hypothesis that suspicion does affect behavior, as a function of prior probability and the weight of the evidence the Bayes factor. The figure is based on subject-wise aggregation of the effect of prompted suspicion bottom, Total row of Fig.
As can be seen, even if one thinks it is likely that prompted suspicion has an appreciable effect—that is, has a low prior belief in the null hypothesis—the weight of the present evidence argues strongly that suspicion that has no appreciable effect on behavior.
Interestingly, participants themselves expected this null effect. In free response to a probe asking how they believed their suspicion of being deceived influenced their behavior, In contrast to the suspicion measures, behavior in all five tasks did differ by pool Tables 1 , 2 , 3 , 4 and 5.
For example, as compared to participants in the economics pools, participants in the psychology pools were more generous in the dictator game [ Participants in the psychology pools also offered more in the ultimatum game The effect of pool on behavior may be due to differential self-selection into these pools Frank et al. These pool effects on behavior show that this study was indeed capable of revealing differences, further strengthening our confidence in the null effects of suspicion.
The results of our survey of researchers Study 1 suggest that attitudes toward banning deception vary as expected by field, with economists supporting banning deceiving study participants more so than psychologists, but the results of our behavioral study Study 2 do not support the public goods argument for banning deception. Recall that the logic of banning deception to preserve a public good rests on the assumptions that experience in a subject pool that permits deception will foster participant suspicion, and that suspicion will affect participant behavior in future studies.
We found that experience in a pool that permits deception or in deceptive experiments did not reliably lead to greater suspicion, and that suspicion was surprisingly widespread in economics subject pools despite the deception ban. Furthermore, suspicious participants did not behave differently from credulous participants. In other words, we did not find evidence for the effectiveness or necessity of banning deceptive research methodologies.
Of course, the present null results only extend to the measures included. We deliberately used common partner-based experimental economics tasks because suspicion of the use of sham partners a common form of deception may be expected to affect behavior in precisely these tasks e. We found no effect of suspicion in these tasks, but our data cannot speak to the effect suspicion of deception may have in other tasks.
For example, psychological realism is an important consideration in experimental design, and to the extent that suspicion of deception undermines psychological realism, it may indeed affect behavior.
Future research may continue to test different possible pathways by which suspicion may affect behavior in different possible domains. A second limitation of our behavioral study concerns self-selection into the subject pools we used. Our AMT and economics pools consisted entirely of self-selected participants, and there is evidence that the experience of deception affects the choice to participate in future research e. Therefore, the null effects of suspicion observed here may be due in part to the persistence of participants who are either less prone to or less affected by suspicion than the participants who left those subject pools.
The convergent results from the psychology subject pools suggest that self-selection is unlikely to have had a major effect on the results, however, since psychology subject pools including those used here generally compel undergraduates to participate in studies in exchange for course credit though note that all the participants in Study 2 received pay, and that the Harvard psychology subject pool contains some self-selected community members in addition to students.
Therefore, we expect self-selection in response to previous experiences of deception to have had less of an effect on the results from psychology subject pools, which were broadly consistent with the results from the economics subject pools and AMT. The increasing unification of the social sciences and advent of powerful shared research platforms like Amazon Mechanical Turk benefit the entire scientific community.
However, these developments may create friction between fields e. Fortunately, a growing body of research from both economics and psychology is assessing the effects of deceptive methodologies on the research enterprise e. Across five subject pools and four common tasks, we did not find evidence of increased suspicion among participants previously exposed to deceptive pools or studies, and behavior did not differ significantly between suspicious and credulous participants, undercutting the pragmatic logic of the deception ban.
If participant suspicion is not primarily dependent on prior exposure to deception, and suspicion does not reliably influence participant behavior, it may be that the benefits of the judicious use of deception could outweigh its costs see, e. Although the economists we surveyed had more negative views of deception than psychologists and were more likely to endorse banning deceptive methodologies across fields, we also found variance and ambivalence in their attitudes Fig.
We hope that further interdisciplinary research will continue to reveal what role, if any, different types of deception should have in the social science toolbox.
Alberti, F. Studying deception without deceiving participants: An experiment of deception experiments. Journal of Economic Behavior and Organization , 93 , — Article Google Scholar. Barrera, D. Much ado about deception: Consequences of deceiving research participants in the social sciences. Sociological Methods and Research , 41 , — Bonetti, S. Experimental economics and deception.
0コメント