Overcoming the Challenges of Computing a Single Analysis Weight When a State-Level Component Is Incorporated into a Traditional National-Level Study

• Activity Number: 250 - Weighting and Variance Estimation in Complex Samples
• Type: Contributed
• Date/Time: Tuesday, August 9, 2022 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Ruby Johnson* and Peter Siegel and Kimberly Janda
• Companies: RTI International and RTI International and RTI International
• Keywords: weighting; weight adjustments; poststratification; state-level data; NPSAS

Abstract

When a national-level study updates its design to include a state-level component, the historical methodology of the study must be revised to accommodate the new design including revisions to the weighting process. The 2017–2018 National Postsecondary Student Aid Study, Administrative Collection (NPSAS:18-AC) was the first NPSAS study to include a state-level component for undergraduates in all 50 states, Washington, D.C., and Puerto Rico.

We use our experience with NPSAS:18-AC to illustrate the challenge of adapting established weighting methods to a new study design. We describe our process for computing a single undergraduate weight that can be used for analyses at both the national and state levels. We discuss the evolution of the NPSAS:18-AC weighting approach which included both model-based adjustments using the SUDAAN WTADJUST procedure and weighting-class adjustments. We evaluate the final undergraduate analysis weights by comparing NPSAS:18-AC estimates with prior NPSAS estimates to show how the new single weight produces national-level estimates that are comparable to prior estimates and yet has the ability to produce state-level estimates.