Cronbach to the Future: Methodological Expertise and Validity Theory in Educational Policy since the 1950s

Alexander Kindel, Princeton University

How much should statistical findings be trusted? What conclusions can we synthesize from hundreds or thousands of statistical claims? In modern America, as the production of statistical knowledge became a routinized element of public administration, questions like these became important foci for intellectual debate and scientific innovation. This paper narrates one case of mid-century innovation in statistical reasoning—validity theory—as it played out through the work and career of Lee J. Cronbach, a psychometrician working in educational policy evaluation between the 1950s and the 1990s. Cronbach’s career trajectory, published scholarship, and personal correspondence illustrates how the development of a national research economy created opportunities for inventive statistical thought. Faced with unruly research objects and a growing volume of published science, quantitatively savvy researchers like Cronbach developed new techniques for shoring up the objectivity of statistical knowledge production. The distinctive qualities of this new kind of methodological expertise yield important lessons for historians and sociologists of quantification, statistics, and the administrative state.

No extended abstract or paper available

 Presented in Session 188. Expertise III: The Politics of Policy Ideas