Improving Scientific Reproducibility with Applied Philosophy of Science

Concerns about the replicability of research have continued to interest researchers across a number of scientific disciplines, particularly during the last decade. Within psychology, numerous explanations for low replicability have been proposed. Most notable are the problems of current publication standards, which disincentivise the publication of null findings, as well as incentivising the use of questionable research practices (QRPs). Estimates of the prevalence of QRPs have suggested that they may be alarmingly common within psychology. However, researchers themselves, as well as independent judges, often rate the use of most QRPs as defensible in certain situations. One pertinent and hitherto unanswered question is how these QRP justifications fit into the wider epistemological frameworks which underpin the practices of many researchers.

The question of epistemic justification for QRPs comprises one part of a wider topic that is particularly deserving of examination: to what extent do psychology researchers actually take note of, and account for, the difficulties of causal inference in their field? It has been suggested that insufficient adherence to an appropriate and well-defined epistemology is an important contributory factor to the poor replicability of findings in social psychology. While both interesting and plausible, this is no more than speculation at present. Consequently, I am proposing a project which will explore the extent to which a lack of epistemological consideration may be hindering the replicability of research, and how it can be improved. This project is both novel and ambitious insofar as, while authors have theoretically discussed how research could be improved with philosophy of science, the topic has yet to be addressed experimentally. I am proposing to collect qualitative and quantitative data on researchers' beliefs and opinions, collaboratively analyse logical consistencies with philosophers of science, and apply the ideas generated in randomised control trials aimed at testing if research practices can be improved.