Project Talent: Weighting Adjustments Comparison for Nonresponse and Tracking Loss in a Follow-Up Survey 50 Years Later

• Activity Number: 288
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Danielle Battle*+
• Companies: American Institutes for Research
• Keywords: Tracking ; Non-response ; Weighting Adjustment ; CHAID ; Longitudinal

Abstract

All longitudinal studies face a common challenge of both locating study participants and having participants respond to future waves. One challenge is ensuring that adjustments can be made to the weights to correct for any non-response and tracking loss. This paper provides results from the 2011-12 Project Talent Follow-up Pilot Study (PTPS12), a pilot test developed to assess the feasibility of finding and reengaging a representative random subsample of Project Talent participants who had not been contacted in 37 to 51 years. To adjust the weights for non-response and tracking loss a CHAID (Chi-Squared Automatic Interaction Detector) Analysis was used to develop the weighting cells. This paper will focus on weighting two different methods for weighting the PTPS12. The first treats those not tracked and non-respondents as the same group and consider the impact on bias analysis. The second method first adjusts for tracking loss and then adjusts for non-response separately and again considers the impact of the bias analysis. A comparison of the two methods will be used to determine which of the two has a lower variance and a smaller bias.