ABSTRACT
A conceptual framework for examining different facets of temporal stability is proposed. Empirical studies, involving the temporal stability of 200 specific acts subsumed by eight dispositional categories and assessed over time with two data sources in a sample of 130 subjects, are used to illustrate the merits and deficiencies of this framework. Results suggest moderate stability of single acts, as well as robust increments in obtained stability when multiple-act composites or act trends are employed as units of analysis. Individually computed stability coefficients, not dependent on other sample members for meaning, were found to be moderate and similar in magnitude to traditionally computed correlations. Variations in stability magnitudes were reliable across data sources. Three hypothesized moderators of stability differences were operationalized. Eight tests of each hypothesis provided no support for the context specificity or the motive inference hypotheses and moderate support for the cross-method agreement hypothesis. Implications for the conceptual framework and for alternative facets of temporal stability are discussed.
