Dilemma of the Untreated Control Group in Applied Research: A Proposed Solution
Among the various impediments to controlled research in applied settings are the difficulties attendant upon the withholding or withdrawal of treatment that might prove beneficial to subjects and the inability of researchers to present simultaneously treatment conditions to all subjects. The multiple pretest-posttest control group design is introduced as an efficient and effective remedy to the problems with special focus upon controls for internal and external validity threats. Several variations of this basic strategy, including the alternating model, the staggered model, and the multiple pretest-posttest replications design, are illustrated and discussed. Finally, the arrangement of factors for statistical analysis is presented for both the case of analysis of covariance with posttest scores and analysis of variance using corrected gain scores.