Shrinkage Procedures on Meta-Analysis

• Main Authors: Jhan, Shu-Wun, 詹淑雯
• Other Authors: Hsieh, Jin-Jian
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/31472294597250434340

碩士 === 國立中正大學 === 數理統計研究所 === 100 === Meta analysis is an approach to combine results from different studies. In this article, we would like to compare the efficiency between the ordinary method and the shrinkage approach. We would consider four shrinkage approaches in meta analysis. Via large simulation studies, we find there is no approach dominates others for all situations, and conclude that when the variances of each study are similar, we should adopt a shrinkage approach for meta analysis; when the variances of each study are dispersed, we should consider the ordinary approach. Finally, we consider the children's development data by Hedges and Pigott (2004) for illustrations.