Language:
English
Pages:
1 Online-Ressource (16 pages)
Parallel Title:
Erscheint auch als Etang, Alvin Interviewer Design Effects in Household Surveys: Evidence from Sudan
Keywords:
Household Survey
;
Interview Design
;
Interviewer Effects
;
Poverty Reduction
;
Sample Variance
;
Survey Variance
;
Computer Assisted Personal Interviewing (capi)
;
Household Survey Accuracy
;
Interviewer Effects
;
Measurement Error
;
Mixed Effects Regression
;
Survey Interviewer
Abstract:
Interviewer design effects occur when data collected by the same interviewer is more similar than data collected by different interviewers. Design effects inflate survey variance and reduce the precision of estimates. Using household survey data collected via computer assisted personal interviewing (CAPI) in Sudan this paper employs a two-level mixed effects regression model to identify interviewer design effects for key variables. The study finds mean interviewer design effect values of 7 with a maximum of 16, implying a significant loss of precision. Recommendations to mitigate interviewer design effects include simplifying questions, sound survey implementation practices, and utilizing multi-way cluster robust standard errors to account for both area and interviewer clustering during data analysis
DOI:
10.1596/1813-9450-10242
