Advisability of Pretest Designs in Psychological Research

The advantages, disadvantages, and appropriateness of pretest designs vis-a-vis posttest-only investigations are reviewed from a pragmatic perspective. Generally, pretests provide information on pretreatment states of nature and allow analysis of change scores, but they also tend to introduce main effects and interactions as well as difficulties in analysis which preclude an unambiguous interpretation of treatment effects. These problems are illustrated with data from a field experiment conducted in the manner of a Solomon (1949) design. The results showed that three alternative analyses of pretested subjects yielded conclusions which conflicted with those obtained in both an after-only group analysis and a factorial analysis of the complete Solomon design. It was concluded that pretesting may introduce more interpretative problems than it resolves and that the cost-efficient posttest-only design may be adequate in many cases.