China fertility survey 2017: design and implementation

This paper examines the design and the process used to carry out the China Fertility Survey 2017, a national representative survey that collected data on fertility desire, childbearing behavior, the use of childbearing services, and the determinants of childbearing behavior. The sampling method adopted was three-stage stratified probabilities proportional to size (PPS), and survey implementation made use of Computer Assisted Personal Interviewing (CAPI). CAPI played a significant role in survey design, last-stage sampling, interviewer training, face-to-face interviews, and questionnaire review and quality control. The survey results were compared with relevant data in the Integrated Management Information System for Population and Family Planning to check consistency. Ex post facto weighting was applied to correct sample structure bias. The process used to acquire accurate personal information is summarized. Suggestions based on consideration of sampling frame distortion by population mobility and other factors are put forward in the hope of improving similar sampling surveys in the future.