Missing value imputation in longitudinal measures of alcohol consumption.
Missing value imputation in longitudinal measures of alcohol consumption.
复制标题
DOI：
10.1002/mpr.330
复制
发表时间：
2011-03
影响因子：
3.1
通讯作者：
Wicki, Matthias
中科院分区：
文献类型：
--
作者：
Grittner, Ulrike;Gmel, Gerhard;Ripatti, Samuli;Bloomfield, Kim;Wicki, Matthias
纵向研究中的过度投入可能导致偏倚结果。这项研究的动机是出乎意料的观察，酒精消费量下降，尽管增加了可用性，这可能是由于大量饮酒者的样本磨损。已经提出了几种插补方法，但很少在酒精消费的纵向研究中进行比较。消费水平测量的插补在计算上特别具有挑战性，因为酒精消费是一个半连续变量（饮酒者的二分饮酒状态和连续量），以及连续部分数据的非正态性。数据来自于丹麦的一项纵向研究，该研究有四个波（2003-2006年），基线时有1771名个体。比较了缺失数据的五种技术：最后值结转（LVCF）用作单一插补方法，Hotdeck、Heckman建模、链式方程多元插补（MICE）和贝叶斯方法用作多重插补方法。使用预测平均值匹配来解释非正态性，其中插补相似病例的“真实的”观察值，而不是插补回归估计值。还通过模拟数据集对方法进行了比较。模拟结果表明，贝叶斯方法产生了最无偏的估计插补。尽管供应量增加，但消费水平没有增加，这一结论仍然没有改变。
Attrition in longitudinal studies can lead to biased results. The study is motivated by the unexpected observation that alcohol consumption decreased despite of increased availability, which may be due to sample attrition of heavy drinkers. Several imputation methods have been proposed, but rarely compared in longitudinal studies of alcohol consumption. The imputation of consumption level measurements is computationally particularly challenging due to alcohol consumption being a semi-continuous variable (dichotomous drinking status and continuous volume among drinkers), and the non-normality of data in the continuous part. Data come from a longitudinal study in Denmark with four waves (2003–2006) and 1771 individuals at baseline. Five techniques for missing data are compared: Last value carried forward (LVCF) was used as a single, and Hotdeck, Heckman modelling, multivariate imputation by chained equations (MICE), and a Bayesian approach as multiple imputation methods. Predictive mean matching was used to account for non-normality, where instead of imputing regression estimates, “real” observed values from similar cases are imputed. Methods were also compared by means of a simulated dataset. The simulation showed that the Bayesian approach yielded the most unbiased estimates for imputation. The finding of no increase in consumption levels despite a higher availability remained unaltered.
登录
查看更多内容
影响因子：
2
作者：
Paddock, Susan M.;Ebener, Patricla
通讯作者：
Ebener, Patricla
影响因子：
2
作者：
Little, Roderick J.;Yosef, Matheos;Harlow, Sioban D.
通讯作者：
Harlow, Sioban D.
影响因子：
6
作者：
Makela, Pia;Bloomfield, Kim;Room, Robin
通讯作者：
Room, Robin
影响因子：
2
作者：
Kaciroti, Niko A.;Schork, M. Anthony;Julius, Stevo
通讯作者：
Julius, Stevo
影响因子：
6
作者：
Barnes, Sunni A.;Larsen, Michael D.;Decker, Paul A.
通讯作者：
Decker, Paul A.