Domain-specific participant recruitment exceeds the application of "Targeted" advertisement from common online advertising platforms.

Abstract

Introduction: Insufficient sample sizes threatened the fidelity of the primary research trials. Even if the research group recruits a sufficient sample size, the sample may lack diversity, reducing the generalizability of the results of the study. Evaluating the effectiveness of online advertising platforms (e.g., Facebook & Google Ads) versus traditional recruitment methods (e.g., flyers, clinical participation) is essential.

Methods: Patients were recruited through email, electronic direct message, paper advertisements, and word-of-mouth advertisement (traditional) or through Google Ads and Facebook Ads (advertising) for a longitudinal study on monitoring COVID-19 using wearable devices. Participants were asked to wear a smart watch-like wearable device for ∼ 24 hours per day and complete daily surveys.

Results: The initiation conversion rate (ICR, impressions to pre-screen ratio) was better for traditional recruitment (24.14) than for Google Ads, 28.47 ([0.80, 0.88]; p << 0.001). The consent conversion rate (CCR, impressions to consent ratio) was also higher for traditional recruitment (66.54) than for Google Ads, 2961.20 ([0.015, 0.030]; p << 0.001). Participants recruited through recommendations or by paper flier were more likely to participate initially (Χ² = 23.65; p < 0.005). Clinical recruitment led to more self-reporting white participants, while other methods yielded great diversity (Χ² = 231.47; p << 0.001).

Conclusions: While Google Ads target users based on keywords, they do not necessarily improve participation. However, our findings are based on a single study with specific recruitment strategies and participant demographics. Further research is needed to assess the generalizability of these findings across different study designs and populations.