Studying people living in non-private households: results of a large pilot study in the Netherlands (THESIS VERSION)

Abstract

Residents of non-private households are currently excluded from most general social surveys. Their inability to answer survey questions and difficulties in approaching them are viewed as obstacles to survey this population. However, excluding them can potentially bias survey estimates. These two claims are investigated in this study. Firstly, we evaluate the opportunities to study residents of elderly and nursing homes, two major groups belonging to the non-private population. Secondly, we assess possible differences between the key characteristics of these two groups and those of comparable residents of private households to explore the potential bias of excluding them from survey research. To do so, a pilot study is conducted at Statistics Netherlands. In this pilot study, 537 responses are collected from residents of elderly and nursing homes. We find that with an adjusted approaching strategy and questionnaire, conducting survey research among residents of elderly and nursing homes is certainly feasible in terms of response rates and data quality. An examination of the potential bias caused by excluding non-private households from general nation-wide social surveys reveals substantial differences between non-private and private households on almost all key survey variables.