Meta分析中标准化均差效应量的计算

Calculation of standardized mean difference effect size in Meta-analysis

背景:大多数情况下,在进行两组比较研究的Meta分析时,纳入Meta分析的各研究文献结果变量的操作差异较大,不同的研究结果间不能进行直接的比较,给Meta分析带来挑战。目的:深入挖掘原始研究的信息方法,增强Meta分析结果的效能。方法:根据原始研究提供的结局变量最常见的近10种情形介绍了标准化均差效应量的计算方法。结果与结论:在进行两组比较研究Meta分析时需要综合分析各种可能情况,特别是当原始研究文献提供的信息与研究者已掌握的公式存在着距离时需要尽可能挖掘到所需要的信息,避免信息的流失。

BACKGROUND： In most Meta-analysis of two-group comparison study, the outcome variables of original research literature are quite different between different researches, so outcome variables cannot be directly compared, limiting Meta-analysis. OBJEC＇~tVE： To investigate methods for mining the information of original research and enhance the performance of Meta-analysis. METHODS： According to the outcome variables of the original research, 10 common standardized mean difference effect size calculation methods were described. RESULTS AND CONCLUSION： Before conducting Meta-analysis of two-group comparison study, we should consider all kinds of possible situations, especially when there were some distances between the original research literature information and those we have grasped. We have to dig information as much as possible to avoid information loss.