Improving Results Reporting in RCT Registries: Evidence from Interviews with Registrants
Abstract
The complete and accurate registration of trials is a critical component of credible science. Comprehensive registries provide a record of trials conducted in a given field – regardless of the studies’ publication – thereby helping combat publication bias. Thanks in part to requirements by journals and funders, the social sciences have seen a marked increase in registration over time.
Yet the benefits of registries cannot be realized if researchers know which trials were started but have incomplete information on post-trial outcomes (trial completion, results, etc.). Failure to update registrations with post-trial information is a challenge in both medicine, where government-sponsored scientists comply less than 50% of the time with results-reporting requirements, and the social sciences. In 2021, only 29% of registrations in the AEA RCT Registry had any post-trial fields completed one year post-trial end date.
This study explores barriers to updating AEA RCT Registry trials and interventions to increase post-trial reporting. We conducted qualitative interviews with 48 registrants randomly chosen from 5,752 trials registered before May, 2022 and at least one year past their trial end date. Sampled registrants completed a self-guided survey and follow-up semi-structured interview covering attitudes towards research transparency, registration, updating, and the design of messages to increase updates. Results informed the design of an RCT that aimed to identify methods to increase registry post-trial update rates.
We identify preferences for messaging about publication bias and nudges with pre-filled information and fields to be completed. Respondents also identify journal requirements and peer norms as the most likely incentives to increase reporting. Absence of perceived benefits and of staff/own time are found to be primary constraints to reporting. Respondents identify the importance of building a culture of transparency, but also see a gap between personal importance and the perceived importance to the economics field.