Piloting Using CloudResearch for Collecting High-Quality Audio Recordings from a Clinical Population

Background: CloudResearch is a popular platform for online research, but its potential for audio data collection remains relatively unexplored. This pilot study aimed to assess the feasibility and effectiveness of using CloudResearch for gathering audio data. Specifically, as a test case, we sought to capture the perceptions of recently suicidal individuals about their future selves.Methods: We recruited participants (N = 167) with varying levels of suicidal thoughts and/or behaviors via CloudResearch. Participants were instructed to record responses to 15 prompts about their future selves using Vocaroo, a user-friendly online voice recording tool. Bonus payments were provided for high-quality recordings.Results: Of the participants who attempted the recording task, 98% (N = 164) successfully recorded and submitted their audio responses, demonstrating the feasibility of using CloudResearch for high-quality audio data collection. The recordings were substantive, with an average total duration of 15 minutes and 40 seconds (median: 14 minutes and 8 seconds).Conclusions: Our findings suggest that CloudResearch can be effectively employed for audio data collection, enabling novel applications such as qualitative analysis and natural language processing. However, further refinement and consideration of identified limitations and areas for improvement are necessary to fully leverage the method's potential.