On June 7th, the Department of Psychology hosted a symposium on robust research and open science practices organized by lab member Máté Gyurkovics. The goal of the event was to provide an overview of current statistical and methodological practices and how they could be improved to ensure more reliable findings.
It is becoming more and more widely accepted that many common practices in the scientific world can lead to results of questionable reliability. These include creating hypotheses once data has already been collected and analysed, dropping certain conditions and variables from the analyses, or removing participants for no a priori reason. These questionable research practices (QRPs) are at the heart of the crisis psychology (and other disciplines) are currently facing, as it has been found that many findings in the field are hard or even impossible to reproduce.
The speakers –Dr Chris Chambers, Dr Kate Button, Dr Lewis Halsey, Dr Verena Heise, Dr Hannah Hobson, Dr Marcus Munafo – discussed different initiatives (e.g., journals accepting reports before data is collected and analysed, committing to publishing it regardless of the findings) and practical tips (e.g., creating a detailed analysis plan before data collection) to combat these QRPs, and emphasized the importance of transparency in science which can be improved by making raw data and codes used for analyses available to other scientists. These new practices are worth investing the time and effort as they can guarantee more robust and reliable results, and thus, could ultimately benefit science and society as a whole.
We thank everyone who attended for their interest, and the speakers for their wonderful talks. You can find the slides that were presented and audio recordings of the talks at the following URL: http://mate-gyurkovics.postgrad.shef.ac.uk/robustresearch/