Separating interviewer and sampling-point effects
Separating interviewer and sampling-point effects
复制标题
分离采访者和采样点效应
DOI：
--
复制
发表时间：
2003
期刊：
影响因子：
--
通讯作者：
F. Kreuter
中科院分区：
文献类型：
--
作者：
R. Schnell;F. Kreuter
摘要：在实证社会科学研究中使用的数据，尤其是在面对面调查中使用的数据，通常是以多阶段整群抽样的形式收集的。所选择的簇的相对同源性可能导致在抽样阶段的设计结果。面试者倾向于在抽样点内进一步同质化回答。这里提出的研究是为了区分这两个来源而设计的。多水平模型被用来分离面试者选择和采样点选择。尽管人们会认为“对附近犯罪的恐惧”问题的设计应该是由于空间同质性，但事实证明，对于大多数项目，面试者在同质化的问题上所占的份额远远大于空间聚类。
Inter- viewer Behavior, Survey Sampling Abstract: The data used in empirical social-science research, especially in face-to-face surveys, are often collected in multistage cluster samples. The relative homo- geneity of the clusters selected in this could lead to design eects at the sampling stage. Interview- ers tend to further homogenize answers within the sampling points. The study presented here was de- signed to separate the two sources. Multilevel mod- els had been used to separate interviewer eects and sampling-point eects. Even though one would as- sume that a design eect in questions of "fear of crime in the neighborhood" should be due to spa- tial homogeneity it turned out that, for most of the items, the interviewer takes a far greater share of the homogenized eect than the spatial clustering does.
影响因子：
5
作者：
Ohidul Siddiqui;D. Hedeker;B. Flay;F. Hu
通讯作者：
Ohidul Siddiqui;D. Hedeker;B. Flay;F. Hu