Chapter 1: Causal Inference in Randomized and Non-Randomized Studies: The Definition, Identification, and Estimation of Causal Parameters

Michael E. Sobel

Causal inference in randomized and non-randomized studies: The definition, identification, and estimation of causal parameters The distinction between causation and association has figured prominently in science and philosophy for several hundred years at least, and, more recently, in statistical science as well, indeed, since Galton, Pearson and, Yule developed the theory of correlation. Statisticians have pioneered two approaches to causal inference that have proven influential in the natural and behavioral sciences. The oldest dates back to Yule (1896), who wrote extensively about ‘illusory’ correlations, by which he meant correlations that should not be endowed with a causal interpretation. To distinguish between the illusory and non-illusory correlations, Yule invented partial correlation to ‘control’ for the influence of a common factor, arguing in context that because the relationship between pauperism and out relief did not vanish when ‘controlling’ for poverty, this relationship could be deemed causal. A half century later, philosophers, psychologists ...

Users without subscription are not able to see the full content on this title. Please, subscribe or login to access all content on this website.

• Buy The Book