HSR&D Citation Abstract
Search | Search by Center | Search by Source | Keywords in Title
A Randomized Controlled Trial of Special Mailing Procedures on Survey Response Rates: UPS vs USPS Priority Mail.
Hagel Campbell EM, Bangerter AK, Erbes CR, Polusny MA. A Randomized Controlled Trial of Special Mailing Procedures on Survey Response Rates: UPS vs USPS Priority Mail. Poster session presented at: VA HSR&D / QUERI National Meeting; 2015 Jul 8; Philadelphia, PA.
Objectives: To optimize response rates, VA HSRandD researchers using mailed survey research methodology typically employ a protocol involving repeated mailings to non-responders. This experiment examined the effect of two different types of special mailing procedures United Parcel Service (UPS) versus United States Postal Service (USPS) Priority mail on the response rate of participants to a final survey mailing.
Methods: Best practices for mailed survey methodology, as outlined by Dillman (2007), include procedures for implementing a series of follow-up contacts with non-responders. Typically, the initial survey mailing is followed by a postcard thank you/reminder, a replacement survey that appears similar to the initial mailing, and a final contact sent using special mailing procedures. The final survey mailing should appear different from the others and preferably sent by registered mail to emphasize its importance. A total of 341 National Guard Soldiers and partners (all non-respondents to previous efforts to elicit their response to a mailed survey) who were scheduled to receive a final survey mailing were randomized to receive either a UPS mailing (n = 171) or a USPS Priority mailing (n = 170). UPS mailings were sent in brown envelopes and delivered to the individual's door step. USPS mailings were white in color and delivered to the individual's mailbox.
Results: Across Soldiers and partners, those randomized to UPS had a higher response rate than those randomized to USPS Priority mail (38% vs 27% p = 0.05, respectively). When survey response was examined separately for Soldiers and partners, the differences in response rates remained in the Soldiers (40% vs 26% p = 0.02, respectively); while partners' response rates did not show a difference (28% vs 32% p = 0.74, respectively).
Conclusions: Use of UPS for implementing the final special mailing procedures of the Dillman protocol resulted in a 14% higher response rate for Veterans than US Postal Service Priority mail.
Impact: This study provided information that can be used by survey researchers to optimize overall response rates, which is essential to conducting high quality survey research.
Subject: Research methods: New Statistical Methods, Research Designs, and Data Tools