BEYOND TRADITIONAL SCALES: DEVELOPING A SELF-CALIBRATING MEASURE OF HOMONEGATIVITY

Abstract

The article presents the development and validation of a novel measure of homonegativity based on a self-calibrating scoring method. Unlike traditional Likert-type scales, this approach applies Multiple Correspondence Analysis (MCA) to derive optimal quantifications for categorical responses, allowing the data themselves to determine the numerical weights of response options.

Starting from an initial pool of eight items, the instrument was administered to a large Italian sample (N = 2,088 adults). After data screening and selection procedures, a refined dataset was analyzed using Correspondence analysis. The results supported a strong unidimensional structure, as evidenced by the dominance of the first eigenvalue and the presence of a clear Guttman-like pattern, indicating that a single latent dimension captures most of the variance. SEM models were used to confirm the unidimensional nature of the scale (Cronbach’s alpha = .78). Correlations with Adorno’s authoritarianism scale, as well as comparisons across religion, education, interest in the research and sexual orientation provided support for construct validity.

Overall, the study demonstrates that the self-calibrating method offers a flexible and data-driven alternative to traditional scoring procedures, improving measurement precision while maintaining conceptual coherence. The resulting scale provides a robust tool for assessing homonegativity and may be extended to other domains involving ordinal or categorical psychological data.