I was recently interviewed by Amy Orben, Sophia Crüwell, and Sam Parsons for their excellent podcast ‘ReproducibiliTea’. We talked a little bit about myself before getting into some more interesting topics: open science, educational psychology, how the two interact (or how they don’t), and the developing nuance in understanding around open science practices. The link for the podcast is here and below are links to papers I mentioned/additional resources.

ReproducibiliTea links

I mentioned 2 blog posts which altered my thinking around preregistration and gave me, I believe, some much needed nuance [zotpressInText item=”{8R2SVDHL},{YEJCNNCG}”]. These blog posts argue that preregistration does not add any value in the areas of theory development and cognitive modelling. However, Sophia Crüwell linked me to a recent article by herself and others which argues that cognitive modelling would greatly benefit from preregistration [zotpressInText item=”{A45HTJM4}”].

Tim van der Zee was another person I mentioned, this time in the context of commendable open science practices in the field of education. One paper I especially liked was [zotpressInText item=”{DQQSVRBD}”].

Like many, my eyes were opened to the problems plaguing psychology with the publication of the Reproducibility Project [zotpressInText item=”{LFX5G4UJ}”][note]There has been a debate about what the terms ‘replication’ and ‘reproducibility’ mean. Patil et al. (2016) provide an overview of the differences between the terms and why this is important. These definitions are generally accepted, therefore the project should have been called the ‘Replication Project’.[/note]. Another key paper was ‘False Positive Psychology’ [zotpressInText item=”{5421944:ZAGY4CXI}”], which I think everyone should read.

I don’t believe that removing statistical significance will be particularly helpful in changing our approach to science/inference[note]For a discussion around this topic, see here.[/note]. Whatever replaces arbitrary thresholds will likely be used in the same way unless there is a genuine change in how we approach inferences; embracing and understanding the uncertainty around any inference. I therefore both agree and disagree with a recent article arguing for the retirement of statistical significance and appreciation of uncertainty [zotpressInText item=”{5421944:BDUMVXBV}”].

Sharing data ethically is vitally important and Michelle Meyer’s paper on the topic is an excellent resource on it [zotpressInText item=”{5421944:CFDG8RUU}”].

References

[zotpressInTextBib style=”apa” sort=”ASC”]