Investigating the occurrence of reporting bias and publication bias in registered reports with the use of p-curves.
Publication bias and reporting bias are major problems in academic research. Registered reports aim to prevent both these biases. In the proposed study, we aim to check how successful they are in doing so. We want to do this by comparing p-curves of registered reports with those of traditional publications. The p-curve is formed from the distribution of statistically significant p-values. The skewness of the p-curve (good materials at p-curve.com) is a measure of how much evidential value a set of studies contain, or, in other words, how likely it is to be biased.
Selective outcome reporting (SOR) in randomized controlled trials (RCTs) is deduced from outcome changes between trial registries and published reports, favoring statistically significant findings. Though limited, examination of SOR in psychotherapy trials indicates a pervasive problem. We examined the prevalence of selective reporting across a complete cohort of RCTs of psychotherapies for depression, as well as its influence on the estimation of treatment effects. Our findings showed that prospective registration was implemented in only 40% of trials of psychotherapies for depression. Among prospectively registered trials, 25% displayed discrepancies between registration and publications, and 17% were favoring statistical significance. Trials with selective reporting were associated with considerably larger effectiveness.
Many universities in the Netherlands have set up programmes to facilitate the transition to Open Science. However, there is considerable variability in the definition and scope of Open Science between these programmes. At the VU, an Open Science programme is being set up that has a broad scope. Sander will discuss the plans for the VU and the values underlying the programme.
A criminological approach to research misconduct
Concerns over research misconduct (RM) have been voiced in a wide range of sciences and reflections and analysis on the topic have surged in the last decade. Its causes, processes and consequences; as well as the effectiveness of prevention and sanctioning measures have been widely debated and analyzed. However, most of the times Criminology, the social science that studies crime and deviance and social control mechanisms, has been absent from the debate. In this meeting, results on RM from a criminological perspective will be briefly presented, both regarding individual and organizational causes, as well as about the effectiveness of training in preventing RM and promoting responsible conduct of research. Arguments will be made about the need to produce multi and interdisciplinary research that includes criminological scholarship.