History of randomization (take 2)

Last year I wrote a blog post about the history of random design after it came up in a couple of different settings. I have now written a more exhaustive paper on the history of randomized assignment (with a special focus on the social sciences), where I use the term to refer to any randomization by the researcher that is carried out in order to equate across observations (in expectation) any factors outside experimental control. In addition to several obscure but interesting examples, some but not all of which were familiar to this literature, the main novelty of the paper is the observation that randomized control trials arose in multiple fields almost simultaneously in the mid to late 1920s.…

Continue reading