Schwarzer Guido, Antes Gerd, Schumacher Martin
Institute of Medical Biometry and Medical Informatics, University of Freiburg, Freiburg, Germany.
Stat Med. 2007 Feb 20;26(4):721-33. doi: 10.1002/sim.2588.
A new test for the detection of publication bias in meta-analysis with sparse binary data is proposed. The test statistic is based on observed and expected cell frequencies, and the variance of the observed cell frequencies. These quantities are utilized in a rank correlation test. Type I error rate and power of the test are evaluated in simulations; results are compared to those of two other commonly used test procedures. Sample sizes were generated according to findings in a survey of eight German medical journals. Simulation results indicate that, in contrast to existing test procedures, the new test holds the prescribed significance level when data are sparse. However, the power of all tests is low in many situations of practical importance.
本文提出了一种用于检测具有稀疏二元数据的荟萃分析中发表偏倚的新检验方法。检验统计量基于观察到的和预期的单元格频率以及观察到的单元格频率的方差。这些量被用于秩相关检验。通过模拟评估了检验的I型错误率和检验功效；并将结果与其他两种常用检验程序的结果进行了比较。样本量是根据对八本德国医学期刊的调查结果生成的。模拟结果表明，与现有检验程序不同，当数据稀疏时，新检验保持规定的显著性水平。然而，在许多实际重要的情况下，所有检验的功效都很低。
