Bye B V, Gallicchio S J
Soc Secur Bull. 1993 Fall;56(3):75-87.
Since 1987 the Social Security Administration (SSA) has published a special set of tabulations on SSA program recipients in the Annual Statistical Supplement to the Social Security Bulletin using data derived from the Census Bureau's Survey of Income and Program Participation (SIPP). Estimates of sampling errors pertaining to these tabulations were derived from the 1984 SIPP panel. This article provides updated sampling error estimates for the 1990 SIPP panel to be used in conjunction with the SIPP-based tabulations provided in the Annual Statistical Supplement for 1992 and 1993. The computational approach is essentially the same as that used in the earlier analysis. Sampling variances are estimated by half-sample replication using the pseudo stratum and half-sample codes available on SIPP public use data files. Generalized tables of standard errors are provided for all SSA program participants. An appendix provides detailed specifications about the calculations. In order that it be self-contained, this article repeats much of the methodological exposition in the previous article that appeared in the October 1988 issue of the Social Security Bulletin.
自1987年以来，社会保障管理局（SSA）利用人口普查局的收入与项目参与调查（SIPP）得出的数据，在《社会保障公告》年度统计增刊中发布了一组关于SSA项目受益人的特殊表格。与这些表格相关的抽样误差估计值来自1984年的SIPP样本。本文提供了1990年SIPP样本的更新抽样误差估计值，以便与1992年和1993年年度统计增刊中基于SIPP的表格一起使用。计算方法与早期分析中使用的方法基本相同。抽样方差通过使用SIPP公共使用数据文件中可用的伪层和半样本代码进行半样本重复估计。为所有SSA项目参与者提供了标准误差的通用表格。附录提供了关于计算的详细说明。为了做到内容完整，本文重复了1988年10月《社会保障公告》上一篇文章中的许多方法阐述。
