Reproducibility and Open Science

This is a past event

58 people went

Location image of event venue

Details

Join us for three short talks on open science and reproducible research, given by Fiona Fidler, Hannah Fraser and Mathew Ling.

5:45pm – Refreshments – Staff Tea Room, Peter Hall Building, University of Melbourne (https://maps.unimelb.edu.au/parkville/building/160)
6:15pm – Talks – Theatre 1, Old Geology, University of Melbourne (https://maps.unimelb.edu.au/parkville/building/155/g04)
7:30pm – Dinner – University Cafe, Carlton

Talk 1: Hannah Fraser – Reproducibility and Questionable Research Practices

Psychology, economics and some areas of medicine are in crisis due to the low reproducibility of their research findings. In psychology, researchers find contradictory results around half of the time when they redo studies. In ecology, it is more difficult and expensive to redo studies and we expect far more variation in results due to environmental stochasticity, so we are unsure of how reproducible results are. However, in other fields low reproducibility has been attributed to a number of practices that increase the chances of finding a significant result where no relationship exists. Hannah will discuss her work on Questionable Research Practices and describe a new large-scale project investigating the extent to which different (but still correct) analysis techniques are associated with differences in results and conclusions.

Hannah Fraser is a research fellow at the University of Melbourne working in the Interdisciplinary Meta-Research Group (IMeRG). She is a quantitative ecologist by training but in addition to her love of conservation, has a strong interest in improving research methods. Hannah is particularly passionate about raising awareness about Questionable Research Practices, increasing the uptake of Registered Reports and investigating how replicable research is.

Talk 2: Fiona Fidler – What are solutions to the replication crisis fixing?

Pre-registration and Registered Reports are related proposals to help mitigate publication bias, and the problems (inflated false positive rates and effect sizes) that stem from it. There is mounting evidence that they are successful in important ways. There are also important questions about their scope, and what, if anything, might be more appropriate for research outside the Hypothetico-Deductive model that they presuppose. In this talk, I will briefly outline current issues in this area.

Fiona Fidler is an Associate Professor and ARC Future Fellow at the University of Melbourne with a joint appointment in the School of BioSciences and the School of Historical and Philosophical Studies. She leads the Interdisciplinary MetaResearch Group (IMeRG.info) and is lead PI of the repliCATS project (https://replicats.research.unimelb.edu.au)

Talk 3: Mathew Ling – Open and Reproducible research as Health Behaviours.

Open and reproducible research practices are non-default behaviours, but engagement in these behaviours can lead to improved outcomes for both the individuals that engage in them, and the broader community. In this way, engagement in reproducible research practices is comparable to behaviours like vaccination, and the models implemented there might provide insights into improving the robustness of modern research as practiced. This presentation will examine this comparison and discuss how reproducible research practices can be normalised.

Mathew Ling is a Lecturer in Psychology at Deakin University, where he teaches and conducts research in the Misinformation lab. He is also an advocate for open and reproducible research through the Melbourne Open Research Network and as an Ambassador to the Center for Open Science

You can find more information about the Statistical Society of Australia and join the Society at http://statsoc.org.au/