TY - JOUR
AU - Busse, Thomas V.
AB - LETTERS Meta-Analysis of Research: A Rejoinder to Glass Glass (November, 1976) has called Turning to the second point, we find 1. Locate all available studies on a attention to an increasingly significant fault with computing the average effect topic, including dissertations, research problem: how to interpret the findings size across studies varying greatly in reports, reports to foundations, etc. from many individual studies, when the methodological sophistication. The ef­ 2. Eliminate from consideration studies are diverse with respect to popu­ fect sizes obtained in small, poorly de­ studies with severe methodological in­ lations sampled, instruments used, signed studies are likely to be mislead­ adequacies. Results from these studies variables measured, methodological ing for two reasons. First, the use of are likely to be more misleading than designs employed, and findings ob­ small samples may produce spuriously helpful. tained. Our attempts to summarize the large treatment effects by chance alone. 3. Divide the remaining studies into research on creativity (Mansfield & Second, the effect size for any treat­ two categories: those that are reason­ Busse, in press) encounter the same ment or program is probably related to ably well designed and those that have problem but lead us to take 
TI - Letters: Meta-Analysis of Research: A Rejoinder to Glass
JF - Educational Researcher
DO - 10.3102/0013189X006009003
DA - 1977-10-01
UR - https://www.deepdyve.com/lp/sage/letters-meta-analysis-of-research-a-rejoinder-to-glass-OR8R0WgMkD
SP - 3
EP - 3
VL - 6
IS - 9
DP - DeepDyve
ER -