Journal of Official Statistics, Vol.20, No.2, 2004. pp. 357377

Current Issue
Personal Reference Library (PRL)
Personal Page

Exploring Confidentiality Issues Related to Dependent Interviewing: Preliminary Findings

Several surveys employ a panel design in which respondents are interviewed at multiple points in time (“waves”) over the course of several months, or even years, in order to investigate the dynamics of certain life events. Many of these surveys use some form of “dependent” interviewing in which information gathered in one wave is carried over into subsequent waves in an attempt to reduce repetitiveness and burden, aid recall, reduce spurious change, and generally provide a sense of continuity over the life of the survey. Recent efforts to improve the Survey of Income and Program Participation (SIPP) have begun to focus on understanding and improving these dependent interviewing techniques. A related research effort has recently begun at the U.S. Census Bureau (which administers the SIPP) regarding its new “Respondent Identification Policy” or RIP. This policy, implemented in 1998, is designed to ensure that responses are not shared among individuals within a household unless the respondent has given consent for this. The implementation of dependent interviewing techniques and the new RIP policy prompted the current research. The main goals were to develop and assess a RIP item – a request for consent to share the respondent’s data with other household members during later waves – and to explore respondents’ reactions and concerns about dependent interviewing in general. In the service of those goals, the current research had several components. First, respondent debriefings were conducted following a Wave 1 interview to explore: (1) their reactions to the RIP item; and (2) their attitudes toward confidentiality and dependent interviewing. The debriefing findings were used to further refine the RIP item, and then the original and the new RIP item were evaluated in a Wave 2 follow-up debriefing. In addition, Wave 2 cognitive interviews were conducted to assess respondents’ overall reactions to dependent interviewing. Results from these exploratory inquiries indicate that some respondents expressed concerns about sharing information with children and about sharing financial information more generally. In general, respondents reacted positively to dependent interviewing techniques and most had no privacy or confidentiality concerns. Among the limitations of the research are: (1) its qualitative nature (testing was conducted on small numbers of respondents not selected at random, and semi-structured interviewing techniques – versus standardized interviewing – were used), (2) the survey context (topics within the SIPP include income, earnings and health insurance; other topics of a more or less sensitive nature may not yield the same results) and (3) the face-to-face lab setting, which may promote more trust and rapport-building than would be found in an actual field test.

Dependent interviewing, confidentiality, privacy, respondent debriefing, cognitive interviewing

Copyright Statistics Sweden 1996-2018.  Open Access
ISSN 0282-423X
Created and Maintained by OKS Group