Toward a Robust Assessment of Construct Differentiation: A Preliminary Evidence for Convergent Validity of Repertory Grid Differentiation Indices
| Autoři | |
|---|---|
| Rok publikování | 2026 |
| Druh | Článek v odborném periodiku |
| Časopis / Zdroj | JOURNAL OF CONSTRUCTIVIST PSYCHOLOGY |
| Fakulta / Pracoviště MU | |
| Citace | |
| www | https://www.tandfonline.com/doi/full/10.1080/10720537.2025.2471607 |
| Doi | https://doi.org/10.1080/10720537.2025.2471607 |
| Klíčová slova | Cognitive complexity; convergent validity; differentiation; repertorygrid technique; semistructured interview |
| Přiložené soubory | |
| Popis | Differentiation of personal constructs as a component of cognitive complexity is an important topic in research on personal construct theory and practice. The article addresses the validity of repertory grid differentiation indices such as intensity, percentage of variance accounted for by the first factor, Bieri’s index, and the bootstrap clustering index. To determine their validity, we compared them with an independent index of differentiation based on an analysis of semistructured interviews. We used data from two previous studies that had focused on construction processes related to conspiracy beliefs in the context of the COVID-19 pandemic (Study 1) and the context of the war in Ukraine (Study 2). In Study 1, intensity, percentage of variance accounted for by the first factor, and Bieri’s index correlated moderately with the text-analysis index; the bootstrap clustering index did not correlate with the text-analysis index. In Study 2, the repertory grid indices correlated with the text-analysis index lower and insignificantly, but in the predicted direction. In merged data, intensity, and percentage of variance accounted for by the first factor exhibited low significant correlations with the text-analysis index. We argue that in Study 2, the correlations could be lowered due to a lower reliability of the text coding for differentiation and due to a broader scope of the interview. In conclusion, the studies report preliminary evidence of the validity of percentage of variance accounted for by the first factor and intensity. |