a University of North Carolina Greensboro.
Multivariate Behav Res. 2013 Jul;48(4):592-618. doi: 10.1080/00273171.2013.802978.
This article investigates an effect size (MLM ES) and its variance for cluster randomized trials based on parameter estimates from multilevel modeling analysis. Accuracy and precision of MLM ES were evaluated using Monte Carlo simulation methods and compared with the performance of an effect size, computed from summary statistics, proposed by Hedges (2007; Hedges' dB ). Simulation results indicated that MLM ES had acceptable accuracy in all conditions, also demonstrating efficiency and consistency. With small sample sizes, MLM ES did not suffer from the same negative bias as Hedges' dB due to overestimation of between-cluster variance. With large sample sizes, MLM ES and Hedges' dB were comparable for accuracy and efficiency. Both MLM ES and Hedges' dB showed considerable bias in some conditions when cluster sizes were unequal. An illustrative example using real data was provided.
本文基于多层次建模分析的参数估计，研究了群组随机试验的效应量（MLM ES）及其方差。采用蒙特卡罗模拟方法评估了 MLM ES 的准确性和精密度，并与 Hedges（2007；Hedges' dB）提出的基于汇总统计量计算的效应量的性能进行了比较。模拟结果表明，在所有条件下，MLM ES 的准确性均令人满意，同时还具有效率和一致性。在小样本量的情况下，由于高估了组间方差，MLM ES 不会像 Hedges' dB 那样受到同样的负偏差影响。在大样本量的情况下，MLM ES 和 Hedges' dB 在准确性和效率方面相当。当簇大小不等时，MLM ES 和 Hedges' dB 在某些情况下都存在相当大的偏差。提供了一个使用真实数据的说明性示例。
