Intergroup Conflict Theory
Literature seems to identify predictors of intergroup conflict as falling into the following categories: 1) contextual (environmental predictors: social norms, history of intergroup hostility, opportunity for intergroup contact), 2) perceived threat (symbolic, realistic, self, honor, obstacle, economic), 3) affect (anger, anxiety, hatred, contempt, disgust, fear), 4) motivation (defense, retaliation, thrill, mission, self-enhancing), and 5) individual characteristics (social identification, pride in group identity, stereotype endorsement, aggressive personality). Following protocol for conducting a systematic review and analysis developed by Johns Hopkins University’s Cochrane Collaboration (Higgins & Green, 2008), we seek to find all available research addressing predictors of aggression toward outgroup members. Specifically, we will include research that empirically examines factors increasing the likelihood (i.e., predictors) that an individual will intentionally attempt to engage in behavior to cause harm to another (i.e., aggression; can be material, physical, social, or psychological harm) who s/he perceives as possessing a demographic identity (e.g., religious, ethnic, sexual) distinct from one’s own (i.e., not ingroup member). This question is limited because we are not considering buffers and we are casting a wider net by examining aggression as the outcome variable, not violence. Yet, it is important to note that much bias crime against a person (e.g., threats) or property (e.g., vandalism) is not violence, per se. Further, evidence supports that research on aggressive behavior can be extended to understand violence, as violence is an extreme aggression sub-type (Huesmann & Kirwil, 2007).
The next step in the systematic review plan includes developing study selection criteria. We will conduct 2 systematic literature analyses: 1) correlational, and 2) experimental and quasi-experimental. To be included, studies must meet 4 criteria; they must: 1) measure aggressive behavior (self-report, collateral report, or behavioral observation; not including aggressive thoughts or affect); 2) test one or more factor(s) other than group identity (e.g., context, threat, motive) contributing to increasing aggressive behavior; 3) delineate that different group identities of aggressor and target are salient; and 4) use pre-existing (not lab created) group identities. However, we will consider lab-created groups (e.g., Minimal Group Paradigm; Otten, 2016) if studies of naturally-occurring groups are rare, as may be the case for experimental designs.
The third step requires search strategy development, accomplished through: 1) identifying sources for published (e.g., Scopus, PsycINFO, SocIndex) and unpublished material (e.g., conference and dissertation abstracts); 2) identifying keywords for research question variables (e.g., intergroup conflict, aggression, bias or hate crime) and related variables such as theories (e.g., realistic group conflict theory, integrated threat theory), or predictor categories (e.g., motivation, threat, emotion, personality); 3) identifying exemplar articles (i.e., articles that meet all inclusion criteria) for citation searches, keyword collection, and “find similar” searches; 4) developing quality metrics (e.g., sample size and power, scale reliability, manipulation checks, effect sizes); and 5) developing exclusion criteria (e.g., small sample studies with low power or poor scale reliability [α ≥ .60], or those that fail to include manipulation checks if published after 2000).
The fourth step is to identify what data is to be gleaned; we are looking to identify: 1) groups examined (e.g., migrants vs. natives), 2) methodology (archival and correlational vs. experimental and quasi-experimental), 3) sample size and type (e.g., community or college), 4) presence and frequency of different theoretical applications (e.g., number of studies using integrated threat theory), 5) label overlap in predictors examined (e.g., number of studies testing symbolic threats), 6) operationalization overlap of predictors and outcomes (e.g., number of studies using the same instruments), 7) commonalities in predictors of different labels (e.g., economic challenge = realistic threat), 8) predictors using I2AM framework to identify themes, and 9) indicators of relationship strength between identified predictors (e.g., realistic threat perception) and aggressive behavior (e.g., self-reported past behavior, observed aggressive acts). We will include significant and non-significant coefficients to determine evidence weight in our meta-analysis. Each of these steps will be posted on Open Science Framework (OSF) during analysis, and updated with inclusion and exclusion logs, search protocol, and data files post-analysis. Two coders will be assigned per article.
As such, anticipated contributions of this year include: 1) a comprehensive compilation, review, and integration of an interdisciplinary body of literature relevant to understanding intergroup aggression, of which bias crimes is a subtype; 2) qualitative evidence regarding most commonly tested predictors (and measurement thereof) of intergroup aggression; 3) quantitative evidence regarding predictive utility of variables for understanding intergroup aggression; and 4) a refined theory (I2AM) with specific variables identified within categories based on empirical evidence.
The next step in the systematic review plan includes developing study selection criteria. We will conduct 2 systematic literature analyses: 1) correlational, and 2) experimental and quasi-experimental. To be included, studies must meet 4 criteria; they must: 1) measure aggressive behavior (self-report, collateral report, or behavioral observation; not including aggressive thoughts or affect); 2) test one or more factor(s) other than group identity (e.g., context, threat, motive) contributing to increasing aggressive behavior; 3) delineate that different group identities of aggressor and target are salient; and 4) use pre-existing (not lab created) group identities. However, we will consider lab-created groups (e.g., Minimal Group Paradigm; Otten, 2016) if studies of naturally-occurring groups are rare, as may be the case for experimental designs.
The third step requires search strategy development, accomplished through: 1) identifying sources for published (e.g., Scopus, PsycINFO, SocIndex) and unpublished material (e.g., conference and dissertation abstracts); 2) identifying keywords for research question variables (e.g., intergroup conflict, aggression, bias or hate crime) and related variables such as theories (e.g., realistic group conflict theory, integrated threat theory), or predictor categories (e.g., motivation, threat, emotion, personality); 3) identifying exemplar articles (i.e., articles that meet all inclusion criteria) for citation searches, keyword collection, and “find similar” searches; 4) developing quality metrics (e.g., sample size and power, scale reliability, manipulation checks, effect sizes); and 5) developing exclusion criteria (e.g., small sample studies with low power or poor scale reliability [α ≥ .60], or those that fail to include manipulation checks if published after 2000).
The fourth step is to identify what data is to be gleaned; we are looking to identify: 1) groups examined (e.g., migrants vs. natives), 2) methodology (archival and correlational vs. experimental and quasi-experimental), 3) sample size and type (e.g., community or college), 4) presence and frequency of different theoretical applications (e.g., number of studies using integrated threat theory), 5) label overlap in predictors examined (e.g., number of studies testing symbolic threats), 6) operationalization overlap of predictors and outcomes (e.g., number of studies using the same instruments), 7) commonalities in predictors of different labels (e.g., economic challenge = realistic threat), 8) predictors using I2AM framework to identify themes, and 9) indicators of relationship strength between identified predictors (e.g., realistic threat perception) and aggressive behavior (e.g., self-reported past behavior, observed aggressive acts). We will include significant and non-significant coefficients to determine evidence weight in our meta-analysis. Each of these steps will be posted on Open Science Framework (OSF) during analysis, and updated with inclusion and exclusion logs, search protocol, and data files post-analysis. Two coders will be assigned per article.
As such, anticipated contributions of this year include: 1) a comprehensive compilation, review, and integration of an interdisciplinary body of literature relevant to understanding intergroup aggression, of which bias crimes is a subtype; 2) qualitative evidence regarding most commonly tested predictors (and measurement thereof) of intergroup aggression; 3) quantitative evidence regarding predictive utility of variables for understanding intergroup aggression; and 4) a refined theory (I2AM) with specific variables identified within categories based on empirical evidence.
