河野万里奈唱哭:急需翻译！！高手 进！！！

Of course much has been done since. In 1993 Lipsey and Wilson claimed– in their review of 302 meta-analyses (nearly 10,000 individual studies) – that current meta-analysis methodology shows larger effects of psychological, educational and behavioral treatment, than could be found before. The 80 articles that specifically analyze studies of experimental versus control treatments in education show an overall mean effect size of 0.34 (difference of means, devided by standard deviation; 0.20 is called ‘small’ and 0.50 ‘medium’ [Cohen, 1977]). A closer look however complicates the interpretation of their results. Nine meta-analyses explicitly deal with higher (college) education methods and have achievement as a dependent variable. Of these, the three reviews with the largest effect sizes seem somewhat contradictory. Two ‘Keller plan-method’ reviews report that personalized instruction leads to higher achievement than traditional education (0.49 and 0.52 effect size), whereas a review on cooperative learning methods reports thatcooperation yields better achievement than individual instruction (0.62 effect size).