Chapter 16. What Constitutes Evidence of Stereotype Accuracy?

Abstract

     This chapter lays the groundwork for understanding Chapter 17, which reviews every high quality scientific investigation of the (in)accuracy of stereotypes that I could find.  Chapter 16 facilitates understanding Chapter 17 by: 1) Discussing how understanding accuracy requires first understanding three different levels of analysis (population, small group, person perception), that (in)accuracy at one level rarely provides any information about (in)accuracy at any of the other levels, and that researchers committed to a view of stereotypes as inaccurate frequently do indeed confound these levels; 2) reviewing the earliest research suggesting that stereotypes are not necessarily inaccurate, including the “kernel of truth” hypothesis, and the fact that this empirical evidence apparently had no effect on the dominant view in the social sciences that stereotypes were inaccurate; 3) discussing the quality standards that studies had to meet in order to be included in the Chapter 17 review; 4) discussing the main two types of stereotype (in)accuracy, discrepancies from perfection and correspondence with real differences; and 5) discussing how these two types of (in)accuracy can occur at either the level of consensual stereotypes (what most people generally believe) or personal stereotypes (what a particular individual believes). Understanding these issues are necessary for understanding the review of empirical research on stereotype accuracy in the next chapter.

EXCERPTS:
    A common reaction to research demonstrating stereotype accuracy is “yes but…”  A “yes but” occurs whenever incontrovertible evidence of stereotype accuracy is presented, and a person steeped in traditions viewing stereotypes as unmitigated evils responds with “Yes, but what about ______?”  (You can fill in the blank with your preferred objection to stereotype accuracy research, if you have one; if not, suffice it to say that some common “yes buts” include: “Yes, but what about self-fulfilling prophecies?” “Yes, but what about biased evaluations,” “Yes, but what about information-seeking biases?” “Yes, but what about stereotype threat?” etc.).  Of course, each “yes but” can be addressed on its own merits.  Chapter 1’s discussion of stereotype threat was essentially a refutation of a “yes but”; and Chapters 6-9 should refute most of the “yes buts” arising out of all the expectancy-confirmation literature addressed in Chapters 4 and 5.
In general, people engage in the “Yes, but…” tactic in response to evidence of stereotype accuracy when they: 1) are committed to a position emphasizing stereotype inaccuracy, bias, or irrationality; 2) are unwilling or unable to refute the evidence demonstrating rationality, reasonableness, or accuracy; so that 3) they (perhaps somewhat defensively) attempt to “limit the damage” by acknowledging the existence of some degree of stereotype accuracy (this is the “yes” in “Yes, but…”), and then returning as quickly as possible to the evidence of bias with which they are generally much more familiar or comfortable (this is the “but…” in “Yes, but…”).
Some “yes buts,” however, seem so reasonable, and are so widely believed to negate any possibility of stereotype accuracy that it is worth spending some time on them.  One such common “yes but” is the following: “Yes, but even a stereotype that is, in some sense, “accurate” as a description of a group mean, will not apply to most members of the stereotyped group, because hardly anyone falls on the mean. Therefore, even such stereotypes are inaccurate most of the time.” Let’s examine this more closely.  Variations on this “yes but” summarizes a class of criticisms of notion of stereotype accuracy that has periodically appeared in the social psychological literature (e.g., Allport, 1954; APA, 1990; Fiske, 1998; Hamilton, et al, 1990; Nelson, 2002; Stangor, 1995):
…
Claims suggesting that stereotypes are inaccurate because they do not apply to all individual members of a group (Allport, 1954/1979; APA, 1990; Fiske, 1998; Hamilton, et al, 1990; Nelson, 2002; Stangor, 1995) are both true and false.  The claim that stereotypes cannot possibly apply to all individual members of a group is completely true.  The suggestion that this renders stereotypes inaccurate is, however, unjustified because it confounds levels of analysis (population and either small group, individual, or both).  A claim about a population cannot be evaluated against the characteristics of an individual, or even small groups of individuals.  Consistency between the level of the perception and the level of the criterion must be maintained when assessing accuracy by comparing: beliefs about populations (stereotypes) to characteristics of those population groups; beliefs about differences between small groups of individuals to the actual differences between those small groups of individuals.
…
The Fate of the Early Hints and Whispers That Stereotypes Are Not Always Inaccurate
Nearly all of this early scholarship pointed in the same direction – that, frequently, many stereotypes had at least some degree of accuracy.  Nonetheless, this research had no effect whatsoever on the conception of stereotypes as irrational reflections of bigotry in popular culture.  Even more startling is that, until the 1990s, it had a nearly equally nonexistent effect on most social scientific views of stereotypes (see, e.g., the discussion of stereotypes that appear in almost any graduate or undergraduate psychology text from the 1980s and early 1990s and in prestigious and influential Handbook chapters, Annual Review chapters, and the like)…
This is not just my opinion.  Prominent psychological articles and texts from the 1970s, 1980s, and early 1990s are peppered with claims emphasizing the unjustified or inaccurate nature of stereotypes (e.g., Fiske & Taylor, 1984, 1990; Jones, 1986; Pinderhughes, 1989; Snyder, 1984; Snyder, et al., 1977).   As Chapters 5 and 15 documented, they are also quite common today.  These claims were typically made without citation of articles demonstrating inaccuracy, which reflects the widespread agreement that one could take for granted as fact that stereotypes were inaccurate.  Just as one need not cite evidence to support the claim that “the sky is blue” one needed no research citation.