Abstract
The Household Component of the National Medical Expenditure Survey (NMES) was designed to provide unbiased national estimates of the health care utilization, expenditures, sources of payment and health insurance coverage for the U.S. civilian noninstitutionalized population. The same panel of households were contacted four times over the course of the survey to obtain data on their health care experience for 1987. Furthermore, an additional screening interview was conducted in the fall of 1986 to permit the oversampling of population subgroups of particular health policy interest: blacks, Hispanics, the poor and near poor, the elderly, and the functionally impaired. Due to the panel design of the survey, the NMES experienced two distinct forms of nonresponse: complete nonresponse and partial nonresponse, where sample participants did not respond for their entire period of sample eligibility. In this paper, the characteristics of the NMES part-year respondents are contrasted with those of the complete respondents. The investigation includes an analysis of the effects of alternative nonresponse adjustment strategies for compensating for partial nonresponse.
Get full access to this article
View all access options for this article.