Affiliation:
1. Decision Research Center Oregon Research Institute Springfield Oregon USA
2. Department of Political Science, School of Public and Global Affairs, and College of Law and Institute for the Study of the Environment, Sustainability, and Energy Northern Illinois University DeKalb Illinois USA
Abstract
AbstractObjectiveEfforts to measure cultural identities in survey research rely on self‐reported ethnic, racial, and national identities. We test how survey operationalization of grid–group cultural theory (CT) influences the classification of individuals’ (sub)cultural identities.MethodsA national online sample of Americans (n = 697 for current analyses) rated items from CT indices, CT statements, and cultural cognition theory (CCT) indices in a 2016–2017 panel survey. Individuals were classified as identifying with a culture if they supported it (e.g., rating it above the scale mean or median, or in the top 35 percent of the scale distribution) or agreed with each item constituting the scale (the “midpoint method” introduced here).ResultsDifferent classification methods and cultural measures yield different proportions of support of cultural biases, yielding statistically significant differences despite most people being similarly classified. Survey measures can unequivocally assign a minority of people to a single cultural identity, with a majority so classified only if one does not require the individual to support only one bias.ConclusionsUsing a short, conceptually valid measure of culture with the novel midpoint method seems best for CT survey researchers but should have implications more broadly in cultural identity research and social science efforts to classify individuals.