Conducting real-time interactive experiments online: A guide for researchers

The COVID-19 pandemic has accelerated the shift towards online behavioral experiments and the increased adoption of crowdsourcing platforms for participant recruitment. While conducting online studies has become more common, carrying out real-time interactive experiments online presents unique challenges. We provide an overview of existing tools for conducting interactive online experiments and discuss key considerations for setting up and running interactive online studies. We make special mention of the setup combination of studies programmed in oTree and carried out with Prolific samples, which has become popular in experimental and behavioral economics and finance research. Using a case study of a large multi-round experiment involving synchronous decision-making among groups of participants recruited via Prolific, we examine critical factors such as sample availability, participant arrival speed, traffic management, as well as dropout risks and mitigation approaches. We discuss strategies researchers can adopt and suggest areas for further investigation.