An assessment of the reliability of WFS data.
An assessment of the reliability of WFS data.
复制标题
对 WFS 数据可靠性的评估。
DOI：
--
复制
发表时间：
1980
期刊：
影响因子：
--
通讯作者：
Marckwardt Am
中科院分区：
文献类型：
--
作者：
Omuircheartaigh Ca;Marckwardt Am
在评估世界生育率调查数据的可靠性时，大部分实质性分析是以斐济和秘鲁的数据为依据的。印度尼西亚和孟加拉国的结果被包括在内，以说明在不同的设置中发现的可靠性范围。使用秘鲁和莱索托调查设计的一些特殊特征来说明将总方差划分为3个分量：抽样方差、简单响应方差和访谈引起的相关响应方差。5个（秘鲁、莱索托、斐济、印度尼西亚、孟加拉国）重新访谈研究有某些共同特点。对在基本类似的调查条件下进行和处理的两次访谈（斐济和秘鲁）的个人一级答复进行的比较表明，答复的变异程度相当大，其程度与答复者的某些背景特征有关。不稳定的程度之间有很大的变化与态度的问题和有关事件的日期在相对遥远的过去往往是特别不稳定的问题的变量。不一致性指数表示观测值的变异性在多大程度上是由简单响应方差引起的。
In assessment of the reliability of World Fertility Survey (WFS) data the bulk of the substantive analysis is based on data from Fiji and Peru. Results from Indonesia and Bangladesh are included to illustrate the range of reliability found in different settings. Some special features of the survey design in Peru and Lesotho are used to illustrate the partitioning of the total variance into 3 components: sampling variance simple response variance and the correlated response variance due to interviews. There were certain common traits to the 5 (Peru Lesotho Fiji Indonesia Bangladesh) reinterview studies considered. Comparison of individual level responses from the 2 interviews (Fiji and Peru) conducted and processed under essentially similar survey conditions indicated a considerable degree of response variability the extent of which is associated with certain background characteristics of the respondent. The degree of instability varied considerably between variables with attitudinal questions and questions concerning dates of events in the relatively distant past tending to be particularly unstable. The index of inconsistence indicated the extent to which the variability in the observations is due to the simple response variance.