Employing Collaborative Survey Development and Mixed-Mode Strategies to Facilitate Data Collection Among Those Living in Rural Communities

Abstract

Purpose

Among those living in rural areas, knowledge regarding best community-based strategies to prevent negative health outcomes, including suicide, remains limited. A survey was collaboratively developed regarding individual- and community-based suicide risk and protective factors, and Veterans and non-Veterans living in rural communities were outreached. Mixed-mode strategies were employed to improve response rates and reduce non-response bias.

Methods

Recruitment letters were mailed to a random address-based sample of 45,000 households in rural Mississippi (MS), New Mexico (NM), and Oregon (OR) to achieve 3,600 responses assuming an 8% yield rate. Recruitment prioritized survey completion by Veterans. Letters encouraged survey completion online, however, paper-and-pencil surveys were also mailed. Enhanced recruitment efforts to increase yield included weekly emails, telephone reminders, increased incentives, letters to those who started but did not complete, and Spanish language paper-and-pencil surveys to members of NM households.

Findings

Completed surveys totaled 3,864 across the three communities; an 8.6% yield rate; resulting in a 21.7% American Association for Public Opinion Research (AAPOR) response rate. Approximately, 77% of surveys were completed online and 23% by paper-and-pencil, respectively. The distribution of completion modes was similar across communities. Increasing email frequency and post-incentives were most effective for boosting survey completion.

Conclusions

Input from rural communities is needed to facilitate innovative, community-driven suicide-prevention. Mixed-mode strategies enabled successful recruitment to obtain vital information regarding community-specific suicide risk and protective factors. Using paper-based methods may be important in rural areas to account for limited internet infrastructure and accommodate traditional communication preferences.