ABSTRACT

Selection can interact with all the other threats to internal validity. For instance, consider selection-history interaction. One threat in the design is maturation. Perhaps the participants matured during the period between the pretest and post-test, and the increase is due to maturation and not to the treatment. Instrumentation, as it is sometimes called, is also a threat. Another threat is testing, which is defined as the effects of the pretest on the performance exhibited on the post-test. Statistical regression is a threat that occurs only if participants are selected on the basis of their extreme scores. Because the selection of participants for the two groups was not at random, they may be systematically subjected to different life experiences. All threats to internal validity can be overcome by using a true experimental design, in which participants are assigned at random to experimental and control conditions.