Abstract
An alternative method of analysis for randomized pretest-posttest experiments is proposed. It yields treatment contrasts identical to those from ordinary analys...