PurposeThis study used a series of multi-group confirmatory factor analyses (MG-CFAs) to test a fundamental assumption in the procedural justice literature that the public judgments of procedural justice are invariant across groups including the split between those who have had recent interactions with the police and those lacking that experience. MethodsA public survey from three Ukrainian cities (N = 4005) provided the data for separate MG-CFAs to test configural, metric, scalar, and residual invariance in a procedural justice scale across the paired groups that differed by age, gender, education, and whether they had recent contact with the police. ResultsThe MG-CFAs found that procedural justice was strongly invariant across the demographic variables while also not finding a significant difference in procedural justice between those with and without recent police interactions. ConclusionsPolice scholars cannot know whether findings of differences in procedural justice across groups or relationships between procedural justice and other variables reflect measurement bias in the sample or real differences unless they have first established measurement invariance. This study's finding of strong invariance across the tested groups serves as a critical step in promoting a valid, standardized measure of police procedural justice.
Read full abstract