Inference on treatment effect in a pretest–posttest study is a routine objective in medicine, public health, and other fields, and a number of approaches have been advocated. Typically, subjects are randomized to two treatments, the response is measured at baseline and a prespecified...