CitationCao, Honggao & Hill, Daniel H. (2005). Active versus passive sample attrition: the Health and Retirement Study. EconPapers Working Paper No. 0505006.
AbstractThis paper investigates sample attrition in the Health and Retirement Study (HRS). We compare attrition behavior in two of the HRS cohorts: original HRS cohort and AHEAD cohort. We distinguish attrition due to death (passive attrition) from attrition due to other causes (active attrition), examining potential effects of different attrition modes on the representativeness of the remaining samples. This distinction is justified based on a specification test on a multinomial logistic regression model. Among other results from the study are differences between passive and active attritors in a set of demographic, economic, and health measures, and a finding that active attrition occurring in the HRS is perhaps not selective and, thus, is statistically ignorable.
Reference TypeJournal Article
Journal TitleEconPapers Working Paper No. 0505006
Hill, Daniel H.