A Defense of Deception on Scientific Grounds

debate is emerging concerning the useof deception in social science research(especially when it employs experimen-tal methods), driven primarily by the relative-ly recent move by many economists intoexperimental work. These economists general-ly argue that deception should be banned.Deception includes a variety of practices insocial science research and is not limited toexperimental research. Most often it involvesproviding limited information about the truepurpose of the research, omitting informationin the instructions to subjects not deemed cen-tral to the study but important to the researcheffort, or giving a Ocover storyO for the studythat does not reveal the actual topic of study.Regardless of type, all deception used insocial science research must be approvedthrough protocol review by InstitutionalReview Boards (IRBs) in university settings.As political scientist Wilson (2007:5)notes in a recent essay: OIn economics alldeception is forbidden. Reviewers are quiteadamant on the point and a paper with anydeception will be rejected.O Thus, if the econ-omists prevail with journals, funding agen-cies, and universities in their efforts to ban theuse of deception, the impact on our capacity toconduct some kinds of research on importantintellectual and scientific issues will be largeand potentially very negative. Given spaceconstraints we comment on only a few of therelevant issues posed by this debate (e.g.Hertwig and Ortmann 2001) and the aggres-sive stance of most experimental economists(see also Eckel 2007).Hertwig and Ortmann (2001:397) argue(as do almost all experimental economists)that deception should generally not be used insocial science experimentation not on ethicalgrounds, but on pragmatic grounds.Deception, it is argued, produces a contami-nated subject pool for their use. They contendthat this public goods problem should besolved by strong sanctions, not by semi-volun-tary mechanisms such as IRB review.Although they acknowledge the need for suchreview, they believe it to be inadequate and toopermissive even by APA (AmericanPsychological Association) standards. Whilesocial psychologists and experimental sociol-ogists generally agree that deception shouldbe a last resort and is sometimes unnecessary,there are underlying issues relevant to thisdebate that need to be discussed. In the spaceallowed we briefly outline only some of ourconcerns, providing just a few examples of themajor theoretical problems at stake.As a rule we agree that deception shouldbe limited by necessity, but we do not agreethat the use of deception should be sanctionedby nonpublication of papers employing it,nonfunding of proposals for research requir-ing deception, or any other heavy sanction thatimposes one disciplineOs priorities on others.Important insights into human behavior anddecision-making have come from award-win-ning experiments (some involving deception)that revealed things about ourselves we maynot have wished to know, such as our tenden-cy toward obedience to authority, which takesmany forms and continues to have devastatingeffects on human history.The need for deception under certain cir-cumstances depends in part on oneOs view ofwhat is driving human decision-making andbehavior. Many experimental economists