| Sumario: | Data from the organizational research literature were reanalyzed to compare and contrast classic and current methods for analyzing construct validity to establish a basis for assessing the validity of measures used in organizational research. The results indicate that the Campbell and Fiske (1959) criteria are too limited an approach, particularly in their assumptions, diagnostic information, and power. The more modern method of confirmatory factor analysis (CFA) overcomes most of the limitations inherent in the Campbell and Fiske method, although their are 2 potential shortcomings: the confounding of random error with measure-specific variance, and the inability to test for interactions between traits and methods. Alternative methods for analyzing construct validity, which go farther than currently used procedures in enhancing the ability to ascertain the validity of certain research methods, are discussed.
|
|---|