Kristine Y. Hogarty
University of South Florida
Abstract: The explication and empirical testing of theories are critical components of research in any field. Despite the long history of science, the extent to which theories are supported or contradicted by the results of empirical research remains ill defined. Meehl (1997) has proposed an index of corroboration (C) that may provide a standardized weans of expressing the extent to which empirical research supports or contradicts n theory. The index is the product of a theory’s precision of prediction and the extent to which observed data are close to those predictions. Large values of C are expected from strong theories making tight, accurate predictions. Small values should result from (a) weak theories /linking weak predictions (regardless of their accuracy), Or (b) strong theories that are not accurate.
Simulation methods were employed to evaluate the sampling behavior of C. Factors ill the research design included the precision of prediction, degree of congruence between known population parameters and the theoretical prediction, sample size, psychometric reliability, and the influence of a confounding variable. The results suggest that precision of prediction is far more influential in the value of C than is the accuracy of prediction. As anticipated, less reliable measures yielded smaller values of C. An uncontrolled extraneous variable resulted in biased C values, but the direction of bias could not be anticipated. Surprisingly, sample size evidenced negligible influence on the average value of C, although sampling error was reduced with larger samples.
Citation: Hogarty, K. Y. (2000). Risky predictions and damn strange coincidences: An initial consideration of Meehl’s Index of Corroboration. Florida Journal of Educational Research, 40(2), 76-99.
Download: Hogarty.401.pdf (148 downloads)