Best practices for enrolling young women in online research studies
Online recruitment is an innovative and cost-effective way to reach potential participants for research studies. Through social media advertisements, researchers can recruit from a large and diverse population. However, without the face-to-face interactions and personal follow-ups that traditional recruitment methods employ, online recruitment efforts may have difficulty building the trust and credibility needed to get sufficient numbers of eligible potential study participants to complete the enrollment process. In order to recruit the desired sample size online, researchers should recruit broadly to account for drop-off throughout the enrollment process, find ways to streamline the enrollment process, and use targeted recruitment strategies.
In 2016, Child Trends began a partnership with Healthy Teen Network to recruit more than 2,300 primarily Black and Latinx women (ages 18 to 20) into a rigorous three-year evaluation study of Pulse, a web-based teen pregnancy prevention program. All recruitment and enrollment took place online, primarily via smartphones. Participants were recruited primarily through Instagram, Facebook, and Twitter ads over 15 months from 2016 to 2019. Of nearly 20,000 potential participants who clicked on a recruitment ad and arrived at the study website, about 12 percent were eligible study participants who completed the enrollment process.
Based on the lessons learned in the Pulse study, we offer researchers some tips for online enrollment to recruit the desired sample size:
- Plan to recruit many more participants than you expect to have in your final study sample. About 12 percent of initial visits to the Pulse study page resulted in eligible evaluation participants who completed the enrollment process. Enrolling in an online study requires participants to complete numerous steps—clicking on a recruitment ad, completing a screener, consenting to the study, and/or completing a survey, to name a few. Researchers should expect participants to drop off throughout the process and should, accordingly, cast a wide net during initial recruitment to obtain the final desired sample size.
- Streamline the screening process. In the Pulse study, the largest drop-off in participation occurred in the initial recruitment stage: completing the screener. Originally, the screener had one question per page, requiring participants to click through several pages to complete it. This resulted in a drop-off of approximately 90 percent at the beginning of the study. To improve these rates, the study team reformatted the screener so that all questions appeared on one page, which dramatically improved the completion rate. However, even after this streamlining, only about half of all participants finished the eight-question screener (54%).
- Use targeted advertisements and recruitment strategies. One of the main reasons for drop-off in the enrollment process was participant eligibility. For example, participants in the Pulse study had to be female, 18 to 20 years old, not currently pregnant or trying to become pregnant, have a smartphone, live in the United States, and speak English. Only 54 percent of individuals who clicked on a recruitment ad completed the screener, and only 30 percent were eligible to participate in the study. To attract the intended study demographic(s) and improve eligibility, researchers should carefully craft the language of social media advertisements or offer targeted incentives. The Pulse study found that participants were more likely to click on ads that emphasized an incentive for study participation (e.g., $25 gift card) and ads that featured relevant events (e.g., New Year’s resolutions). Targeted recruitment strategies, such as heavily advertising in zip codes with higher populations of the targeted demographic, can improve the number of eligible recruits.
- Reduce enrollment fatigue. Sharp drop-offs also occurred during the consent and enrollment steps of the Pulse study (only 3,073 of the 5,870 eligible participants completed these steps), perhaps due to participant fatigue. Participants who reached these steps had already clicked on a recruitment ad and completed the screener; being asked to consent and then prompted to fill out an enrollment form may have discouraged participants from completing the process. Streamlining recruitment and enrollment—by consolidating information where possible and minimizing the number of required steps—can greatly improve participant completion rates. For example, screeners, enrollment forms, and surveys should all be completed on one page, and consent could be included within the screener or enrollment forms.