Meeting with the next generation of psychologists in the North West to talk about reproducibility and open research, May/June 2019

Over the last month or so I’ve been fortunate to have given invited presentations to early career psychologists at the universities of Keele, Staffordshire and Lancaster. The combined Keele/Staffordshire Psychology Postgraduate Research Conference gave me the opportunity to provide some (recent) historical context to the current ‘replication crisis’ in the biosciences. A combination of questionable research practices (including p-hacking, and hypothesising after results are known), publication bias (not being able to easily publish null results), and the academic incentive structure (publish or perish!) has created a situation where many published discoveries are likely wrong (that is, the reported findings cannot be replicated and are probably false positives). In response to this, many researchers are now adopting open research practices. These include pre-registration of hypotheses, registered reports (where research papers are unconditionally accepted for publication before data collection begins), reproducible workflows, open data and open code via resources such as OSF and GitHub.

After my talk, there was a lively discussion with lots of questions from early career researchers about how they might start adopting open research practices and how they might go about learning R. As most psychology undergrads are taught statistics using SPSS, and have little or no coding experience, it’s clear that this is a real challenge. The University of Glasgow provides one possible way forward as their Psychology UG programme involves students learning and coding in R right from the get go. So, it can be done! There was also some discussion about whether the academic incentive structure is beginning to change in a way that will support engagement with open research practices. Given that open research has recently been highlighted as important by REF, we hope that it is beginning to change. You can access the slides from my talk on GitHub.

Several weeks later I was at Lancaster to speak at their ESRC-funded methods workshop on “Future proofing your research” organised by Lancaster’s Open Science Working Group, led by Dermot Lynott. This was a full-day event and featured talks by Kirstie Whitaker (from The Alan Turing Institute) on the Turing Way; Lisa deBruine (from the University of Glasgow) on the Psychological Science Accelerator for multi-lab collaborations; Priya Silverstein (from Lancaster University) on her experiences with the Many Babies multi-lab study; Louise Connell (from Lancaster University) on some of the challenges – and possible solutions – around sharing big data; Josh Sendall and Louise Tripp (Lancaster University) on Open Publishing; and then myself on reproducible data visualisations. Chris Chambers (Cardiff University) joined us for a Q&A on registered reports. Again, there were lots of great questions from the audience (who had travelled from many corners of the UK) with lots of enthusiasm to start engaging with open research practices and interest in working on multi-lab projects. Oh, and the pizza we had for lunch was amazing (thanks Dermot!). Access the (reproducible) slides from my Lancaster talk.

It was the first time I had used the Xaringan markdown template in R to create my slides – given my talk was on reproducibility I thought creating reproducible slides was the least I could do.

Keele and Lancaster Universities are part of our North West Open Research Hub (along with MMU, Sheffield, and Chester) – given the (relative) closeness of these institutions, the Hub provides a great opportunity for the open research communities in each of these places to meet and to share best practice, and run collaborative workshops and events centred on reproducible and open research. If you’re in the North West and want to keep up to date with our open research activities (or are just interested in what we’re up to), please feel free to sign up to our listserv.

Andrew Stewart
Fellow of the Software Sustainability Institute and Senior Lecturer

My interests include open research and reproducibility, data science, experimental psychology, and psycholinguistics. I am a fellow of the Software Sustainability Institute.