Parast Layla, McCaffrey Daniel F, Burgette Lane F, de la Guardia Fernando Hoces, Golinelli Daniela, Miles Jeremy N V, Griffin Beth Ann
RAND Corporation, 1776 Main Street, Santa Monica, CA, 90403 USA.
Educational Testing Service, 660 Rosedale Road Princeton, NJ 08541 USA.
Health Serv Outcomes Res Methodol. 2017 Dec;17(3-4):175-197. doi: 10.1007/s10742-016-0168-2. Epub 2016 Dec 26.
While propensity score weighting has been shown to reduce bias in treatment effect estimation when selection bias is present, it has also been shown that such weighting can perform poorly if the estimated propensity score weights are highly variable. Various approaches have been proposed which can reduce the variability of the weights and the risk of poor performance, particularly those based on machine learning methods. In this study, we closely examine approaches to fine-tune one machine learning technique (generalized boosted models [GBM]) to select propensity scores that seek to optimize the variance-bias trade-off that is inherent in most propensity score analyses. Specifically, we propose and evaluate three approaches for selecting the optimal number of trees for the GBM in the twang package in R. Normally, the twang package in R iteratively selects the optimal number of trees as that which maximizes balance between the treatment groups being considered. Because the selected number of trees may lead to highly variable propensity score weights, we examine alternative ways to tune the number of trees used in the estimation of propensity score weights such that we sacrifice some balance on the pre-treatment covariates in exchange for less variable weights. We use simulation studies to illustrate these methods and to describe the potential advantages and disadvantages of each method. We apply these methods to two case studies: one examining the effect of dog ownership on the owner's general health using data from a large, population-based survey in California, and a second investigating the relationship between abstinence and a long-term economic outcome among a sample of high-risk youth.
虽然倾向得分加权已被证明在存在选择偏倚时可减少治疗效果估计中的偏差，但也已表明，如果估计的倾向得分权重变化很大，这种加权可能表现不佳。已经提出了各种方法来减少权重的变异性和表现不佳的风险，特别是那些基于机器学习方法的方法。在本研究中，我们仔细研究了微调一种机器学习技术（广义增强模型[GBM]）以选择倾向得分的方法，这些方法旨在优化大多数倾向得分分析中固有的方差-偏差权衡。具体而言，我们提出并评估了三种为R语言中twang包的GBM选择最优树数量的方法。通常，R语言中的twang包会迭代选择最优树数量，即能使所考虑的治疗组之间达到最大平衡的树数量。由于所选的树数量可能导致倾向得分权重变化很大，我们研究了调整用于估计倾向得分权重的树数量的替代方法，以便我们在牺牲一些治疗前协变量平衡的情况下换取权重变异性更小的结果。我们使用模拟研究来说明这些方法，并描述每种方法的潜在优缺点。我们将这些方法应用于两个案例研究：一个使用来自加利福尼亚州一项大型基于人群的调查数据，研究养狗对主人总体健康的影响；另一个调查高危青年样本中禁欲与长期经济结果之间的关系。
