An Exploration of Bias in Meta-Analysis: The Case of Technology Integration Research in Higher Education
Journal of Computing in Higher Education Volume 26, Number 3, ISSN 1042-1726
This article contains a second-order meta-analysis and an exploration of bias in the technology integration literature in higher education. Thirteen meta-analyses, dated from 2000 to 2014 were selected to be included based on the questions asked and the presence of adequate statistical information to conduct a quantitative synthesis. The weighted random effects average was g [superscript ++] = 0.393, p <0.000. The article goes on to report an assessment of the methodological quality of the thirteen studies based on Cooper's ("Research synthesis and meta-analysis: a step-by-step approach." Sage, Thousand Oaks, 2010) seven stages in the development of a meta-analysis. Two meta-analyses were found to have five out of seven stages where methodological flaws could potentially create biased results. Five meta-analyses contained two flawed stages and one contained one flawed stage. Four of the stages where methodological flaws can create bias are described in detail. The final section attempts to determine how much influence the methodological flaws exerted on the results of the second-order meta-analysis.
Bernard, R.M., Borokhovski, E., Schmid, R.F. & Tamim, R.M. (2014). An Exploration of Bias in Meta-Analysis: The Case of Technology Integration Research in Higher Education. Journal of Computing in Higher Education, 26(3), 183-209.