Strategies for Recruiting Diffuse Practitioner Groups: Lessons Learned from a National Survey with Mobile Crisis Workers

Background:

Social workers occupy a wide array of positions in a broad range of settings. Understanding what constitutes a representative sample and recruiting a representative sample among such diffuse practitioner groups presents significant challenges; sampling frames may not exist nor contain accessible lists of practitioners. This poster aims to assess strategies and convey lessons learned from a national survey focused on one sub-group of service providers, mobile crisis workers. The strategies and lessons presented have application to other “hard to reach” professional groups that social work researchers commonly study.

Methods:

The study through which we illustrate strategies and draw lessons learned used an experimental vignette survey design to understand factors that impact decision-making among mobile crisis workers. The focal population was mobile crisis workers and the inclusion criteria was the active engagement in direct service provision. Based on sample size estimation, we sought 360 participants. In the absence of a pre-existing sampling frame, we used an availability sampling approach. Recruitment sources included social media, university field placement offices, state and municipal service providers, professional groups, a previously constructed list of national crisis intervention organizations, and professional contacts within the field. To keep track of recruitment, a database was created consisting of individuals and organizations known or believed to operate within the focal population. The dates contacted, responses to contact, and follow-ups for each were documented. Participants were offered a $20 Amazon gift card upon survey completion.

Results:

The study team spent 60 hours over 6 weeks completing recruitment activities. The recruitment process yielded 369 eligible participants. Concerning sample composition, participants were predominantly white (71.5%), female (73.9%), and master’s educated (46.6%). Participants were from 17 states and reported working in urban (33.6%), rural (27.6%) and suburban (35.5%) environments. The sample demographically aligns with expectations, but Crisis Intervention Team workers (45.2%) and supervisors (16.8%) may be overrepresented within the target population. Recruitment sources, the pre-existing database of crisis organizations, one professional group’s listserve (CIT International), and personal contacts known by the researchers yielded the greatest response.

Conclusion:

The recruitment approach facilitated enrollment of a sufficiently sized sample in a relatively short period that in many respects likely reflects the target population. Successful strategies included the construction of a master database for tracking, which was critical to keeping recruitment organized and efficient, intentionally seeking sources across states, and drawing on pre-existing relationships. Challenges included the lack of a sampling frame that could be enumerated for probability sampling, lack of standardized crisis systems across states that could act as sources, limited accessibility to contact information within known or discovered organizations for frontline workers, and high response from a single professional group. Future research should consider the adoption of quota sampling to avoid pitfalls such as the overrepresentation of administrative staff or participants from any single organization or professional group.