Chapter 10.  Accuracy: Historical, Political, and Conceptual Objections

Abstract

This chapter reviews many of the reasons why accuracy is a controversial topic in social psychology.  The historical interest in accuracy is briefly reviewed, as are some older reviews from the 1950s that, for 30 years, essentially killed accuracy research.  Modern theoretical and political sources of objections to accuracy research are identified.  These objections are themselves critically evaluated, and, in general, are found to be more flawed than the accuracy research to which they object.  Although some criticisms of accuracy research are well-justified, none justify the claims (that have sometimes appeared) suggesting that accuracy research cannot or should not be conducted.

EXCERPT:
    Consider also the following from Fiske (1998, Page 381 (commenting on McCauley, et al’s [1995] chapter on stereotype accuracy):
Moreover, they differ from the present review in their conclusions, which do not follow from their premises:  If two resumes are otherwise equivalent, it is permissible to use stereotypes associated with group membership as a factor in hiring choice, if group membership has previously predicted success on the job.  (In this they evidently disagree with U.S. civil rights law).”
“Permissible”!?  We (I am one of the “et al”) made claims about accuracy, not about “permissibility.”  These are clear political rationales for quashing accuracy research.  Fiske does not criticize accuracy research as failing to demonstrate that relying on stereotypes increases accuracy of predictions.  Instead, she criticizes such research on grounds of the “permissibility” of relying on stereotypes.  People in power make decisions about permissibility.  I have never seen permissibility presented as a criterion for establishing the accuracy of judgments and it is not included as one in this book.
Once the permissibility criterion is established, however, it has liberated Fiske (1998) to completely ignore our claim that relying on stereotypes sometimes increases the accuracy of judgments.  Instead, she relies on two classic logical fallacies to make her point.  First, she makes the “red herring” type of argument – being that she does not even attempt to refute the point that relying on stereotypes sometimes does increase accuracy, she changes the subject – to the politics of permissibility. She claims we disagree with civil rights law, even though our paper never discussed civil rights law or any other law.  This also takes advantage of a second classic logical fallacy: the ad hominem attack.  After all, who disagrees with civil rights laws except bigots?  And we can’t believe anything a bigot says, can we?  By implicitly insinuating political issues into the topic, Fiske’s quote is a masterful piece of misdirection, denigration and dismissal that is likely to appear compellingly convincing to many of those sympathetic to her goals.  But it is irrelevant to helping figure out whether and when relying on a stereotype increases or reduces accuracy.
Regardless of who agrees or disagrees with civil rights law, this is a fundamentally political, not scientific, rationale for quashing accuracy work.  Fiske (1998) presents neither evidence nor argument that relying on group membership necessarily reduces the accuracy of perceivers’ judgments of individual targets.  I submit that that is because she cannot do so, although to fully understand why, you will have to read Chapter 18 (if you are generally interested in the issue of accuracy and inaccuracy in stereotypes, you should read Chapters 15-19).