Estimating heterogeneity variance in meta-analysis

• 成果类型: Article
• 署名作者: Rukhin, Andrew L.
• 署名单位: National Institute of Standards & Technology (NIST) - USA
• 刊物名称: JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY
• ISSN/ISSBN: 1369-7412
• DOI: 10.1111/j.1467-9868.2012.01047.x
• 发表日期: 2013
• 页码: 451-469
• 关键词: random-effects model; difference; inference

摘要

. Several new estimators of the between-study variability in a heterogeneous random effects meta-analysis model are derived. One is the unbiased statistic which is locally optimal for small values of the parameter. Others are Bayes procedures within a class of quadratic statistics for a diffuse prior with different choices of the prior mean. These estimators are compared with the DerSimonianLaird procedure and the Hedges statistic in particular via the quadratic risk of the treatment effect estimator. A Monte Carlo study supports the usage of confidence intervals derived from the new estimators.