What does it mean to be “presidential”? A multiple-group confirmatory factor analysis of the invariance of indicators in a unidimensional scale

Abstract: U.S. presidential candidates aspire to be perceived as "presidential." Political communication researchers, political scientists, pollsters, campaign consultants, and media pundits speculate about who is "presidential" and "unpresidential." No prior research, however, has empirically measured and validated a presidential image construct. Based on data collected prior to the 2020 U.S. presidential election (N = 618 registered likely voters), we tested congeneric models of presidentiality. Results indicated (1) a six-item, single-factor model fit well when single group confirmatory factor analyses (CFA) were performed separately on ratings of Donald Trump (Study 1) and Joe Biden (Study 2), (2) multiple-group CFA indicated configural and metric invariance (generalizability across candidates) of the model, and (3) presidential factor scores were strongly associated with the candidate participants planned to vote for in the election. Theoretical implications are discussed.