Reproducible Esp Testing

Zoltán Kekecs, a Ph.D. holder in Behavioural Science and an Assistant Professor at the ELTE Institute of Psychology, joins us today. His research centers around mind-body medicine interventions used in medicine, and the mechanisms underlying the effects of these interventions. Zoltán joins us to discuss his research on replicating research findings.

Zoltán began by discussing the current problem with biomedicine and social science research — the low replication rates of research findings. He stated how low replicability affects the trustworthiness of published papers. He also mentioned reasons for the low replicability. One such reason is incomplete information in published studies. In such cases, other researchers need to contact the authors to reproduce their work.

Zoltán delved into the approach his team deploys when replicating a previous research finding. He particularly discussed the process for interacting with psychologist Daryl Bem to test the famous Bem experiment. He explained what the Bem experiment was and the debatable findings of the experiment. Zoltán used a Delphi survey of both proponents and opponents of Bem’s hypothesis to facilitate a consensus about the specifics of the study design before data collection started. For example, he explains how this consensus survey led to a hybrid hypothesis testing framework to be implemented in the replication study, and how the panel members came to an agreement on the interpretation of the possible outcomes of the project.

Zoltán extensively talked about the process of data collection for his study. He listed the three key efforts made for ensuring the replicability of his study. He also discussed other steps taken to avoid artificial bias during the experiments.

Concluding, Zoltán disclosed his results after reproducing the Bem experiment using his approach. He concluded that Bem’s findings could not be replicated given the discrepancy between both results. He also discussed a project in the works that seeks to develop a trustless survey platform for conducting highly credible and transparent research for social scientists. You can reach Zoltán via email at kekecs.zoltan@ppk.elte.hu, and follow him on Twitter @kekecs_zoltan.