The Need for Meta-Analytic Thinking in Educational Technology Research

The present journal article formatted dissertation assessed the extent of meta-analytic thinking currently used educational technology research. In the first study, the author examined the journals, Computers & Education, International Journal of Computer-Supported Collaborative Learning, British Journal of Educational Technology, Australasian Journal of Educational Technology, and Educational Technology Research and Development, between 2012 and 2013 to offer empirical evidence of the field?s current status with regard to reporting results using meta-analytic thinking. These articles represented a total of 32,131 research methods and statistical techniques recorded from 1,171 articles. Findings point to little change in how educational technology researchers conduct investigations. Quantitative methods continue to dominate the field as a whole. Most authors reported the type of sampling procedure used in their investigations. Few researchers reported score reliability estimates using their own data. Findings also suggest few authors report informationally-adequate statistics. One area of concern is the tendency to report a mean without the SD about the mean. Another area of concern is the lack of reporting correlation matrices with accompanying means and standard deviations or covariance matrices.

In the second study, the author conducted a meta-analysis to offer a glimpse of where the field could go once researchers begin to think meta-analytically. The author cumulated findings from nine studies which used the Technology Acceptance Model (TAM) to explain undergraduate students? acceptance of online learning. The author used meta-analytic structural equation modeling and multiple-group analysis to test four path models. The meta-analytic findings suggest the TAM is not a valid theoretical model to explain undergraduates? acceptance of college online courses. The multiple group analysis emphasized that the parameter estimates between studies resulted in statistically different findings, suggesting the findings across studies are not replicable.