Which Questions in the Health and Retirement Study are Used by Researchers? Evidence from Academic Journals, 2006-2009

Since 2002, the average number of questions asked per respondent in the Health and Retirement Study (HRS) has risen by 39 percent, from 413 to 581. Yet there is little or no understanding of which questions, or how many in total, should be included and more importantly, maintained in longitudinal surveys. In this paper, we propose a simple approach to assessing the value of survey questions: journal citation counts. A sample of journal articles and book chapters published in 2006-09 (N = 206) is used to document which questions, and categories of questions, were used most and least frequently. A disproportionate number of published articles used a relatively small number of questions regarding health, wealth, income, and employment. By contrast, several categories of questions were rarely used, and many specific questions were never used. This evidence-based approach to measuring the value of survey questions can have applications for other surveys beyond the HRS.

method of Questioning/survey Methods/Methodology/journal citation counts/journal citation counts/survey questions

25200