Goudy W J
J Gerontol. 1985 May;40(3):358-67. doi: 10.1093/geronj/40.3.358.
The attrition of respondents in panel studies of older people can create bias in data analysis. Changes in the examination of attrition are suggested. Bivariate and multivariate tests comparing those eligible to respond and those actually continuing in the panel are reported for five waves of data from the Retirement History Study. The results indicate that variable relationships in the data examined are relatively free of the bias that can be created by respondent attrition. Thus, attrition from samples of older respondents probably is less of a factor than suggested by previous studies reporting on frequency distributions.
在针对老年人的面板研究中，受访者的流失可能会在数据分析中造成偏差。文中提出了关于流失情况考察的变化。报告了对退休历史研究五轮数据中符合回应条件者与实际继续参与面板者进行比较的双变量和多变量检验。结果表明，所检验数据中的变量关系相对不受受访者流失可能造成的偏差影响。因此，与以往报告频率分布的研究相比，老年受访者样本中的流失可能不那么重要。
