Bias Map and Stereotype Content Model — Lecture Notes (Comprehensive)

Stereotype Content Model and Bias Map — Comprehensive Notes

• Core framing from last lecture

  • Stereotype Content Model (SCM): two core dimensions shape group stereotypes and prejudice
  • Warmth: perceived intent toward the in-group or other groups (axis of benevolence, friendliness)
  • Competence: perceived ability or capability to act on that intent (status, capability axis)
  • These two dimensions predict broad patterns of emotion and behavior toward social groups
  • Warmth and competence combine to produce four basic stereotype quadrants

• The Bias Map (Behaviors from Intergroup Affect and Stereotypes Map)

  • Key idea: discrimination can be analyzed along two domains:
  • Directness of behavior: active vs. passive (direct vs. indirect)
  • Outcome valence: harm vs. facilitation (negative vs. positive impact on the out-group)
  • Formalization:
  • Let the two axes be
    • Warmth (W): judgments of intentional positivity toward the out-group
    • Competence (C): judgments of ability to act on those intentions
  • Let Directness D ∈ {Direct, Indirect (Passive)} and Outcome O ∈ {Harm, Facilitation}
  • The bias map predicts how stereotypes map onto behavior across these two dimensions
  • In this account, high warmth leads to more active facilitation; low warmth leads to more harm
  • High competence leads to facilitation of the out-group’s goals; low competence leads to more harm in some forms
  • The combination of warmth and competence predicts a stable pattern of behaviors toward a group across the four quadrants

• Two key axes to keep in mind

  • Warmth axis: judgment of behavioral intention toward the out-group (toward “us” or the in-group)
  • Competence axis: judgment of the out-group’s ability to act on its intentions
  • These two axes together predict both emotional responses and the likelihood of different discriminatory behaviors

• Mapping from warmth/competence to behaviors and emotions

  • Quadrant logic (two-by-two with four quadrants):
  • Let $W ext{ be High or Low},\, C ext{ be High or Low}.$ Then the following mappings hold:
    • If $W= ext{High}, C= ext{High}$ → Emotion: Admiration; Behaviors: active facilitation + passive facilitation
    • If $W= ext{High}, C= ext{Low}$ → Emotion: Pity; Behaviors: active facilitation + passive harm
    • If $W= ext{Low}, C= ext{High}$ → Emotion: Envy; Behaviors: passive facilitation + active harm
    • If $W= ext{Low}, C= ext{Low}$ → Emotion: Contempt; Behaviors: active harm + passive harm
  • Conceptual interpretation of the two behavioral axes:
  • Directness of action: direct vs indirect (passive) actions toward the out-group
  • Outcome: whether actions are harmful or facilitative for the out-group
  • Example groups discussed
  • Elderly: High warmth, Low competence
    • Emotional reaction: pity
    • Behaviors: active facilitation (help when possible) and passive harm (neglect, patronizing, not listening)
  • High competence, Low warmth groups (e.g., Jews, some Asian groups in discussion):
    • Emotional reaction: envy
    • Behaviors: passive facilitation (cooperation when it serves self-interest) and active harm (directly harmful actions)
  • Low warmth, Low competence groups (e.g., homeless people):
    • Emotional reaction: contempt
    • Behaviors: both active and passive harm (direct hostility and neglect)
  • High warmth, High competence groups (admired):
    • Emotional reaction: admiration
    • Behaviors: both active facilitation and passive facilitation (cooperation and supportive acts)

• Definitions of the four behavior types

  • Active Harm (AH)
  • Definition: behaviors that are intentionally directed toward an out-group with the goal of harming or causing negative experiences
  • Examples: hate crimes, attacking or threatening, making others feel unsafe, verbal/sexual harassment, bullying
  • Passive Harm (PH)
  • Definition: demeaning or distancing out-groups by diminishing social worth through exclusion, ignoring, or neglect
  • Institutional examples: disrespecting needs, limiting access to education, housing, health care; patronizing; telling others what’s best for them
  • Active Facilitation (AF)
  • Definition: proactively helping or aiding out-group members in reaching their goals
  • Examples: protecting group members, helping secure safety and security, charitable giving, anti-discrimination policies, overtly supporting out-group goals
  • Passive Facilitation (PF)
  • Definition: acting in ways that benefit the out-group as a byproduct of pursuing one’s own goals (not the out-group’s goals)
  • Examples: hiring an out-group member because they are perceived as competent; tacit cooperation with an out-group to achieve self-interested ends (e.g., trade with oppressive regimes when it serves one’s interests)

• Measurement and empirical approach to discrimination in the bias map

  • Factor-analytic structure for self-report items measuring discrimination
  • Four factors corresponding to the four behavioral domains:
    • Active Harm (AH)
    • Passive Harm (PH)
    • Active Facilitation (AF)
    • Passive Facilitation (PF)
  • Item examples and loadings (illustrative):
  • AH items load on Factor 1 with relatively strong loadings (e.g., loadings around $0.3-0.5$); items include threats, intimidation, harassment, making one feel unsafe
  • PF items cluster on Factor 4 (distinct from AH/PH) and reflect cooperative interactions that are self-serving rather than group-serving
  • PF items include cooperative actions that benefit oneself but also coincide with interacting with the out-group (e.g., working with, associating with an out-group member when convenient)
  • PH items form a separate factor (Factor 2), reflecting neglect, patronizing attitudes, and exclusion
  • Significance: passive harm can be as consequential as active harm, and sometimes more predictive of long-term well-being and identity threat

• Regional data and group patterns (New Zealand focus)

  • Ethnic stereotypes in NZ (representative sample in prior years):
  • Asian NZers: high competence, low warmth
  • Pacific NZers: high warmth, low competence
  • Pākehā/European NZers: high warmth, high competence
  • Māori: moderate on both warmth and competence
  • Focus on Asian NZers (as a case study):
  • Asian men reported more experience of active harm than Asian women and than European women/men
  • Passive facilitation: Asian men reported higher levels of PF (e.g., being cooperated with only when it serves others’ interests)
  • Gender and group patterns noted in this dataset
  • Women generally report higher discrimination on gender lines across ethnic groups
  • Asian men show notably higher AH and PF relative to other subgroups in this sample
  • Implications for interpretation
  • The bias map is most useful for ethnic/racial groups and age/gender comparisons, with caveats about how intersectionality may alter patterns
  • Some stereotypes (e.g., terrorist stereotypes) fit less well with SCM; the model best explains widely observed patterns in ethnicity/racial groups

• Example data visuals and patterns from New Zealand data (described)

  • Asian vs. Pacific vs. Māori vs. European/NZ (with gender splits)
  • Asian NZers: consistently highest perceived discrimination on ethnicity, with Asian men typically higher than Asian women
  • Pacific NZers: lower relative levels than Asian groups; pattern relatively stable across gender
  • Māori: moderate-high on both warmth and competence; discrimination patterns show gender differences with Māori women showing some of the highest discrimination in gender-domain analyses
  • Pākehā/European: highest discrimination increases among men over time in some analyses; gender differences show European men experiencing more discrimination than European women

• Time-series and longitudinal work ( honors thesis by Van )

  • Research design and data scale
  • Data from an honors thesis modeled by nonlinear regression (intercept and slope framework; intercept fixed at 2010)
  • Sample size: ~ $N \approx 70{,}000$ individuals across 14+ years
  • Outcome: self-reported discrimination on the basis of ethnicity (not limited to active harm; these items load more on active harm domain historically, but cover broad discrimination experiences)
  • Time frame and axes
  • Time axis: $t \in [2010, 2023]$
  • Y-axis: discrimination rate (self-reports) over time
  • Key patterns and group differences
  • Asian men consistently report the highest level of discrimination across the period
  • Asian women also high but typically lower than Asian men
  • Pacific groups show consistently lower levels than Asians, with relatively small gender gaps
  • Māori: men higher than women in many years; patterns converge in some years but gender gaps persist
  • Pākehā/European men show rising discrimination in some years, with a polarization pattern where a subset experiences large increases while others stay flat
  • Temporal peaks and potential drivers
  • Peak around 2010 (post-global financial crisis) possibly tied to job/status competition and rising threats perception
  • Peak around the COVID era (late 2010s/early 2020s) likely tied to global insecurity and policy responses
  • Interpretive takeaway
  • Longitudinal data reveal persistent group differences and highlight how macro events (economic shocks, COVID) relate to perceived discrimination

• Questions, clarifications, and theoretical notes from the lecture

  • About the fit of the bias map
  • The bias map aligns well with ethnic/racial group stereotypes; less clear for groups like terrorists where the stereotype profile is less commensurate with the two dimensions
  • The model is grounded in the idea that status (competence) and warmth guide emotions (pity, envy, admiration, contempt) which in turn guide discriminatory behaviors
  • Distinctions between fear and envy
  • Fear is not a core explicit component of the bias map; envy often better captures the affect toward high-competence, low-warmth groups
  • Some fear-like responses may arise in threat-driven prejudice (context dependent) but not formalized as a primary axis in SCM
  • Relationship to broader theories
  • The model can be integrated with social dominance orientation (SDO) and right-wing authoritarianism (RWA) predictions about out-group threats and status dynamics
  • Evolutionary framing is acknowledged by some researchers (e.g., Susan Fiske) as compatible with person-perception theories
  • Intersections and limitations
  • Anthropological and structural accounts show that intersecting identities (gender, homelessness, disability, etc.) can modify or intensify the patterns predicted by warmth/competence stereotypes
  • The dataset examples illustrate that overlapping identities do exist and can amplify or alter discrimination patterns (e.g., Asian men in NZ data)

• Gender, stereotypes, and in-group/out-group dynamics

  • Gender differences in stereotypes are robust in the NZ data
  • Women are reported as more discriminated against on gender lines across ethnic groups
  • Asian men show specific vulnerabilities in the discrimination data (AH and PF higher than other subgroups in the discussed sample)
  • Theoretical note on “double jeopardy” and related concepts
  • Double Jeopardy: overlapping identities (e.g., woman + minority group) can produce multiplicative rather than additive disadvantages in some contexts
  • Evidence for double jeopardy is context-dependent and not uniformly observed across NZ data
  • Practical implications for research and policy
  • Targeted anti-discrimination policies need to account for intersectional patterns (e.g., gendered dimensions within ethnic groups)
  • Measurement should assess both active and passive forms of harm and both forms of facilitation to capture a complete discrimination profile

• Cross-cutting implications and future directions

  • Overlapping identities and homelessness
  • Students asked about how overlapping identities (e.g., ethnicity + homelessness) change discrimination patterns
  • The lecturer suggested deferring to Shiloh (homelessness expert) for domain-specific data, while noting broad population dynamics are the focus of much of the current work
  • Practical relevance
  • Understanding the bias map helps explain why certain groups receive nurturing or protection in some contexts while facing overt hostility in others
  • It also highlights the hidden (passive) forms of discrimination that can be silently debilitating

• Quick recap of key takeaways for exam readiness

  • The Bias Map/SCM links warmth and competence to four stereotype quadrants and predictable patterns of emotions (admiration, pity, envy, contempt) and behaviors (AH, PH, AF, PF)
  • Active vs. passive and harm vs. facilitation are the two behavioral dimensions that organize discriminatory actions
  • Different groups cluster into distinct quadrants (elderly: high warmth/low competence; Asians: high competence/low warmth; Pacific: high warmth/low competence; Europeans: high on both; homeless: low on both)
  • Active harm and passive harm are easier to observe in some contexts, but passive forms can be equally consequential and harder to detect
  • Measurement relies on factor analysis showing four distinct clusters corresponding to AH, PH, AF, PF
  • NZ data illustrate robust gender and ethnic group differences, with Asian men often most affected by AH and PF; trends show peaks around economic crises and COVID; time-series analysis reveals persistent intergroup disparities
  • Overlaps of identities matter and can intensify experiences of discrimination; careful interpretation is needed to account for intersectionality
  • The model is particularly informative for ethnic/racial discrimination but may be less applicable to groups that do not fit cleanly into two-dimensional warmth/competence profiles

• Final reflection

  • The lecture emphasizes the importance of considering both overt and covert forms of bias, the emotional substrates of prejudice, and the ways in which stereotypes translate into concrete behaviors
  • It also highlights the value of longitudinal, cross-sectional, and intersectional data in understanding how discrimination operates in real-world settings and across different demographics