Statistical Procedures Used in Pretest-Posttest Control Group Design: A Review of Papers in Five Iranian Journals

Abstract

The pretest-posttest control group design is one of the most widely used quantitative experimental design models for evaluating the efficacy of programs, treatments, and interventions. Despite the prevalence and utility of this research design, best practices for data analytical procedures are not clearly defined. Invalid results decrease the chance of generalization. Given that Iranian Journals are interested in publishing pretest-posttest control group design studies, it is important to denote the accuracy of them. The aim of the current study is to explore the correct procedure for using ANCOVA in pretest-posttest control group designs to mitigate the potential limitations of this approach. This study explores the use of ANCOVA in pretest-posttest control group design. It has been done by analyzing data from experimental studies published in five Iranian journals indexed in PubMed or Scopus between 2011 and 2018. The results indicate that among the 280 published experimental studies in these journals, 53 papers (18.9 percent) used ANCOVA as the statistical test in pretest-posttest studies. The power of the test represents the probability of detecting differences between the groups being compared when such differences exist. Our analysis concludes that ANCOVA, which runs a multiple linear regression, is a suitable method for comparing and examining pretest-posttest study designs. Implications of this study have potential utility for researchers employing the use of pretest-posttest control group designs in various fields in and outside of Iran.