Evidence locker · triangulate the cluster
Prejudice
Prejudice is a pre-judgment about a person or group that persists even when specific evidence is limited, mixed, or contradictory, often operating through automatic cognitive processes that can conflict with consciously held egalitarian beliefs.
Devine stereotype automatic activation studies
— · n?
Greenwald et al. Implicit Association Test
d=1.13 · n≈26
Bertrand and Mullainathan resume discrimination study
50% · n≈5000
★ Resume discrimination in Boston and C…
field incident · 2004
★ Starbucks implicit bias training afte…
field incident · 2018
★ Physician implicit bias and thromboly…
field incident · 2007
The claim
Gordon Allport's 1954 book 'The Nature of Prejudice' established the modern scientific study of prejudice.
Reading the scatter. ★ chits are real-world incidents (vivid, but n = 1, so they sit low on strength). The cluster is split: the best-powered field evidence lands upper-right, while several lab paradigms sit lower and weaker. Read the construct paradigm by paradigm.
Exhibit index
Devine stereotype automatic activation studies. Both high- and low-prejudice subjects showed automatic stereotype activation when conscious monitoring was precluded. Only low-prejudice subjects inhibited stereotype-congruent thoughts and replaced them with egalitarian thoughts in a controlled thought-listing task.
Patricia G. Devine, 1989 · Three studies with undergraduate participants
Greenwald et al. Implicit Association Test. The IAT was sensitive to expected individual differences in evaluative associations and to consciously disavowed evaluative differences (Black + pleasant vs. White + pleasant for self-described unprejudiced White subjects). Effect sizes for IAT measures were larger than for explicit measures on socially sensitive topics.
Greenwald, McGhee, & Schwartz, 1998 · Experiment 3: n = 26 White undergraduate students
Bertrand and Mullainathan resume discrimination study. White names received 50 percent more callbacks for interviews. A White name yielded as many more callbacks as an additional eight years of experience. The racial gap was uniform across occupation, industry, and employer size. Federal contractors and Equal Opportunity employers discriminated as much as other employers.
Marianne Bertrand & Sendhil Mullainathan, 2004 · Nearly 5,000 resumes sent in response to over 1,300 employment ads
Why the shortcut forms
Prejudice operates through a dual-process mechanism (Devine, 1989). Stereotypes associated with social groups are stored in semantic memory and automatically activated when encountering group members or symbols. This automatic activation is equally strong in high- and low-prejudice individuals because it reflects cultural knowledge. What distinguishes low-prejudice individuals is the engagement of controlled processes to inhibit activated stereotypes and replace them with egalitarian responses. Recent predictive brain accounts frame implicit stereotypes as learned associations arising from normal brain functioning rather than individual cognitive biases.
Some researchers argue the IAT measures cultural knowledge rather than personal prejudice (Karpinski & Hilton, 2001). The predictive validity of the IAT for behavior has been contested (Oswald et al., 2013), though a 2009 meta-analysis by Greenwald et al. supported it. The malleability of implicit bias suggests it may be more situational than trait-like.
Replication
Mixed
The IAT produces large and reliable effects across populations and has been replicated thousands of times. However, its predictive validity for actual discriminatory behavior is debated. Some meta-analyses support predictive validity (Greenwald et al., 2009) while others question it (Oswald et al., 2013). Devine's core findings on automatic stereotype activation have been widely replicated, though the precise boundary conditions continue to be studied. Implicit bias interventions show short-term effects but limited long-term change.
The IAT has been criticized for conflating cultural knowledge with personal prejudice. Some researchers question whether implicit measures predict behavior above and beyond explicit measures. The construct of implicit prejudice continues to be refined, and the field has moved toward more nuanced models of intergroup bias.
What to log instead
Engage in perspective-taking and seek individuating information about outgroup members. Implement structured decision-making processes that require evaluating specific qualifications rather than relying on overall impressions. Exposure to counter-stereotypic exemplars can reduce automatic bias.
Devine et al. (2012) showed that a prejudice habit-breaking intervention produced long-term reductions in implicit race bias through strategies like stereotype replacement, perspective-taking, and increasing concern about discrimination.
Catch it in the act
Watch for situations where judgments about people seem to be influenced by group membership rather than individual characteristics. In hiring, note if resumes with certain names receive different callback rates. Be aware that even people who consciously reject prejudice may show automatic stereotype activation.
Adjacent lockers
- Implicit BiaschildImplicit bias refers to the automatic activation component of prejudice, typically measured by the IAT. It is a more specific cognitive construct within the broader phenomenon of prejudice.
- StereotypingsiblingStereotyping is the cognitive component of prejudice - the beliefs about traits characteristic of a group. Prejudice adds an evaluative (positive or negative) component to stereotypic beliefs.
- Confirmation Biasmechanistically-linkedOnce prejudiced attitudes are formed, confirmation bias leads individuals to seek and remember information that confirms their pre-existing beliefs about outgroups.