This chapter explores the viability of meta-analysis as a research tool for helping career and technical educational and organizational professionals make decisions. Following many of the same steps involved in the basic research process, meta-analysis provides a means for reconciling contradictory quantitative results from multiple studies, thereby generating a conclusive answer. However, meta-analysis is subject to many forms of bias and can pose practical problems. Meta-analysis has been used to study many issues in administration and management. From this chapter, educational and organizational professionals can determine if it is an appropriate tool to help them make decisions about specific challenges that they face.
History of Meta-Analysis
Prior to the establishment of meta-analysis, researchers often summarized findings of various studies by grouping together similar aspects of the research. For example, if a researcher was reviewing literature regarding the effect of a reading program intervention, he or she may have stated that 15 of 25 studies showed that reading programs led to increases in student ACT scores. However, this method did not take into account the diverse designs and quality of the various research studies that were being summarized or the discrepancies in study results (Gay, Mills, & Airasian, 2009; Rubin & Babbie, 2008). Furthermore, the summarization approach resulted in subjective research, often varying from one researcher to the next due to inconsistent ways of selecting and analyzing such research (Gay, Mills, & Airasian). Introduced in the early 20th century (Moncrieff, 1998; Rosenthal & DiMatteo, 2001), meta-analysis initially was used primarily in clinical medical research (Nijkamp & Pepping, 1998). However, in the 1970’s, Glass, McGaw, and Smith (1981) adapted this method of research for the social sciences, and over the past 30 years, it has been widely applied to quantitative research in education, psychology (personal and industrial), criminology, and other social sciences (Bangert-Drowns & Lawrence, 1991; Duvall & Tweedie, 2000; Mann, 1994; Robey & Dalebout, 1998).
Key Terms in this Chapter
Meta-Analysis Independent Variables: Study characteristics can be thought of as the independent variable.
Test For Homogeneity: Meta-analytical homogeneity tests include the Q statistic, credibility intervals, and chi-square tests for homogeneity.
Criteria for Inclusion of Studies: General categories for study eligibility criteria include distinguishing features, research respondents, key variables, research designs, or time frame.
Effect Size (ES): A numerical way of expressing the strength or magnitude of a reported relationship.
Coding: A researcher includes factors that have had possible moderating effects on the study’s results.
Meta-Analysis Dependent Variables: Dependant variables are a study’s empirical findings.
Problem Selection: An appropriate problem for meta-analysis balances the two extremes of enough has been written about it, but not too much.
Meta-Analysis: A quantitative technique for combining results of multiple studies with similar hypotheses to clarify findings.