TY - JOUR
AU1 - J. Light, Richard
AU2 - B. Pillemer, David
AB - To date, six books on meta-analysis have appeared. The tative methods and how best to match these methods with first by Glass, McGaw, and Smith (1981) was reviewed in Con­ the purpose of the review. The dangers of losing critical infor­ temporary Education Review (Hedges & Olkin, 1982). The mation by aggregating across studies in search of "average current review deals with four other books. A sixth book results" are addressed. Then the authors emphasize examin­ (Hedges & Olkin, 1985) has appeared recently. The four books ing the actual variation in quantitative indices before aggre­ reviewed here differ considerably with respect to the amount gation and describe several techniques for facilitating such of statistical methodology discussed. We review them in the examination. order of increasing statistical complexity. Simple frequency distributions are shown to be useful for some real examples. Funnel diagrams, in which the effect size or quantitative index is plotted against sample size, are Summing Up: The Science of Reviewing Research introduced. Because sampling variance is a function of sam­ Richard J. Light and David B. Pillemer ple size, the characteristic shape of this plot is that of an in­ Harvard University Press: Cambridge, MA, 1984, xüi+191 pp. 
TI - Summing Up: The Science of Reviewing Research Harvard University Press: Cambridge, MA, 1984, xiii+191 pp
JF - Educational Researcher
DO - 10.3102/0013189X015008016
DA - 1986-10-01
UR - https://www.deepdyve.com/lp/sage/summing-up-the-science-of-reviewing-research-harvard-university-press-YTI9zQUhRY
SP - 16
EP - 17
VL - 15
IS - 8
DP - DeepDyve
ER -