Reducing Person Non-Response in Conversational Interviewing: The Redesign of the Survey of Income and Program Participation

Rachael Walsh, U.S. Census Bureau
Daniel Perez-Lopez, U.S. Census Bureau

The Survey of Income and Program Participation is currently undergoing redesign to incorporate an Event History Calendar (SIPP-EHC). Both the 2010 and 2011 field tests had substantially higher person non-response and interview length than the 2008 SIPP. This research uses Kaplan-Meier survival curves to identify the point of saturation in a household interview. Cox regression analysis assessed the effect of interview length on person non-response relative to both household and respondent level characteristics. A saturation point was seen in the 2010 SIPP-EHC at approximately one hour and 50 minutes of interviewing and after two hours of interviewing in a household in the 2011 SIPP-EHC. In the 2011 SIPP-EHC, home ownership, interviewer experience with the SIPP-EHC, and the gender, race, and interview length of the householder were found to significantly impact person non-response rates. The implications of the two hour threshold, relevant respondent characteristics, and suggestions for survey improvement, are discussed.