ABSTRACT
In-group identification has been suggested to consist of two-dimensions (group based self-definition and self-investment) that hierarchically relate to five lower order components (individual self-stereotyping, in-group homogeneity, satisfaction, solidarity, and centrality). The goal of the present research was to test the generalizability of the two-dimensions-five-components structure of in-group identification across identities with which people show converging and diverging group based self-definition and self-investment. We manipulated the mean level and the linear correlational strength of the two identification dimensions by asking participants to indicate in-groups to which respective identification criteria apply. Confirmatory factor analyses showed that the two-dimensions-five-components model of in-group identification fits both converging and diverging identification patterns better than alternative models, indicating generalizability of the model across various identification patterns.
Disclosure statement
No potential conflict of interest was reported by the authors.
Supplementary material
Supplementary data for this article can be accessed here.
Notes
1. “Traditional” students are often characterized by higher education entrance qualification, enrolling full time shortly after finishing secondary school, and either not working or working part time during the school year. “Nontraditional” students at the FernUniversität in Hagen are characterized by diverse educational backgrounds (e.g., 40% already completed a university-degree, 5% have no formal higher education entrance qualifications), diverse socio-demographic characteristics (e.g., a wide range of age groups, with the largest group being between 29 to 35 years of age), and heterogeneous living conditions (e.g., 80% work full time or part time; 10% live permanently outside of Germany) (FernUniversität in Hagen, Citation2017).
2. A reasonable fit of the model to the data is indicated when the χ2 statistic is nonsignificant, RMSEA ≤ .06, sRMR ≤ .08, CFI and TLI ≥ .95 (Hu & Bentler, Citation1999). The χ2 statistics reached significance (as it was the case in the original validation studies of Leach et al., Citation2008). However, because the χ2 statistic is sensitive to sample size and might prompt to reject a model it is not as appropriate as the alternative fit indices that adjust for the effect of sample size (RMSEA, sRMR, CFI, TLI; Hu & Bentler, Citation1999; Schermelleh-Engel, Moosbrugger, & Müller, Citation2003).
3. The AIC and SABIC of Model 4 cannot be used to compare the quality of the alternative models since this model only considers the items which are supposed to measure the self-investment dimension and excludes items that refer to the self-definition components.