Suchergebnisse

The failure to understand the mathematical structure of causal models, the sense in which such models represent causal assumptions, & the indirect nature of the test of such models seem to be responsible for much of the misunderstanding of what causal models are & for certain misconceptions about how causal models cannot be sci'ly useful. An attempt is made to clarify some of these matters in the reply to J. Gibbs "Causation and Theory Construction," Social Science Quarterly, 1972, 52, 4, Mar, 815 -- 826, pointing out that the "purist" empiricism expressed in this paper stands in contrast to the causal reasoning usefully applied to his own work. REJOINDER, by J. Gibbs criticizes H. Costner & H. Blalock for not specifically defining the concept of causation. With causal assumptions remaining obscure, when they test a causal model it will always be difficult to decipher those assumptions. When emphasizing that "goodness of fit" is crucial, the authors fail to address 3 important & related questions: (1) what is the criterion for a "good" fit?, (2) how can such criterion avoid arbitrary distinctions?, & (3) in what sense is "goodness of fit" divorced from the magnitude of r's. In reference to the status integration theory, they say nothing of its merits (as to testability range, or predictive accuracy), & criticize Gibbs for refraining from stating a theory in causal language; ironically enough, Costner & Blalock merely allude to their thoughts about causation. Modified AA.