Implicit measures reveal prejudice that explicit self-report instruments cannot capture

Evidence

Recent advances in social-structural theories of prejudice address how seemingly benevolent forms of prejudice undermine women’s socioeconomic progress. These approaches are supported by new methodologies, including response latency techniques, such as the Implicit Association Test (IAT; Greenwald, McGhee, & Schwartz, 1998) that assess reactions that may be unconscious and automatic, and thus are not susceptible to impression management strategies. Because this task measures the time it takes people to make decisions about pairs of stimuli (e.g., men and agentic traits) rather than directly asking people their opinions, the IAT (along with similar methods) has the advantage of overcoming people’s inability (or unwillingness) to report their attitudes and beliefs (see also Fiske, ch. 3 this volume; Batson & Stocks, ch. 25 this volume). Thus, implicit measures provide unique insights into the psychology of prejudice beyond the direct, explicit measures that were the tools of social psychology when The Nature of Prejudice was written. — John F. Dovidio, Peter Glick, Laurie A. Rudman, On the Nature of Prejudice, 759

Fairly high “immunity” to desirable responding is another advantage of the IAT, compared with explicit self-report measures and some behavioral indexes of racial prejudice. That is, some people who underestimate their racial prejudice on self-report scales or who fake pleasant social interactions cannot easily manipulate the time it takes to categorize pleasant and unpleasant terms and names of Whites or Blacks during an IAT. Even if a few participants do categorize terms in the IAT too quickly (indicating being in a hurry to finish) or too slowly (indicating inattentiveness or not understanding the instructions), the participants with outlying reaction times can be identified and removed. — Wade C. Rowatt and Lewis M. Franklin, Christian Orthodoxy, Religious…, 29

Categorization

• XREF: Connects to the broader skepticism about IAT predictive validity in later meta-analyses and replication debates; worth cross-checking against current IAT criticism. — connects to: Implicit prejudice measures are not clearly more valid than survey measures — Christopher Tarman and David O. Sears, The Conceptualization and Measurement…, 223; Explicit and implicit attitudes operate on different levels and need different measures. — Wade C. Rowatt and Lewis M. Franklin, Christian Orthodoxy, Religious…, 14; The implicit association test measures bias via response latency on concept-sorting tasks. — Wade C. Rowatt and Lewis M. Franklin, Christian Orthodoxy, Religious…, 25 (+4 more)
• DEFINE: This names one of the reader’s core research axes directly: self-reported conviction diverging from measured bias — here the mechanism by which the IAT bypasses desirable responding.
• corroborated: 3 independent passages support this claim — John F. Dovidio, Peter Glick, Laurie A. Rudman, On the Nature of Prejudice: Fifty Years…, 759; Wade C. Rowatt and Lewis M. Franklin, Christian Orthodoxy, Religious…, 29