Meta-analysis Of Researches On The Relationship Between Anxiety And Students' Attainment In The Past Decade In China

Meta-analysis, known as"analysis of analysis", was first proposed by Glass in 1976. It's a kind of systematic review. Meta-analysis includes a procedure of statistical technology which deals with a large amount of quantitative results and produces an"effect size"as the final result.The relationship between anxiety and students'attainment is one of the hottest topics in educational and psychological circles in recent years. Literature retrieval from 1997 to 2006 proves that there are many primary researches on this topic, but the research results vary differently. The largest effect size is 0.18, and the smallest effect size is -0.68.So, the results contradict with one another.This thesis uses meta-analysis to re-analyze 99 effect sizes, which are from 24 primary researches on the relationship between anxiety and students'attainment. The result of this meta-analysis shows that r equals -0.21149, and p equals 0.00000. The 95% confidence interval of r is -0.221480 to -0.201446. The research result supports the popular opinion on the relationship between anxiety and students'attainment.As the overall effect size is heterogeneous, we have taken measures to detect the moderating variables. And the results show that there are two significant moderating variables, i.e. anxiety instrument and the sources of effect size. It means that different anxiety instrument and effect size sources are reasons for the heterogeneity. So, we should take more consideration of the anxiety instrument and the sources of effect size when doing this kind of meta-analysis thereafter. And the meta-analysis results show that the level of the grade, the type of the article, the year of publication and the subject being tested are not moderating variables.We could take measures to deepen the research from the following two respects. First, classify the researches on the relationship between anxiety and students'attainment, such as doing the meta-analysis depending on the anxiety instrument. Second, enlarge the number of the research.