Missing Data in the Health Interview Survey 1997 in Belgium
Health surveys, as other types of population surveys, inevitably face the problem of incomplete data, which may influence the final results and therefore should receive careful attention, starting from the design phase. Also, an investigation of the influence on the results should be undertaken. The Belgian Health Interview Survey (HIS) was conducted in 1997. In this paper the methods used in design and conduct of the HIS to reduce incompletness are described and an analysis of the influence of the missing data on results of the survey is presented. Some aspects of such a survey are difficult... Mehr ...
| Verfasser: | |
|---|---|
| Dokumenttyp: | Artikel |
| Erscheinungsdatum: | 1999 |
| Schlagwörter: | Missing data / Epidemiology / Surveys / beta-binomial model / multiple imputation / health survey |
| Sprache: | Englisch |
| Permalink: | https://search.fid-benelux.de/Record/base-26993747 |
| Datenquelle: | BASE; Originalkatalog |
| Powered By: | BASE |
| Link(s) : | http://hdl.handle.net/1942/361 |
Health surveys, as other types of population surveys, inevitably face the problem of incomplete data, which may influence the final results and therefore should receive careful attention, starting from the design phase. Also, an investigation of the influence on the results should be undertaken. The Belgian Health Interview Survey (HIS) was conducted in 1997. In this paper the methods used in design and conduct of the HIS to reduce incompletness are described and an analysis of the influence of the missing data on results of the survey is presented. Some aspects of such a survey are difficult (e.g. household participation is difficult to study), and therefore we carefully discuss the advantages and possible limitations of such efforts.