Abstracts for workshops, hackathons and unconferences
NB: We are still accepting proposals! Submit your proposal here.
Estimation: Why and How, now with R (Geoff Cumming & Robert Calin-Jageman)
The New Statistics (estimation and meta-analysis) is an easily-understood and easily-accessible strategy for moving beyond NHST and meeting the needs of Open Science. I will summarise the case for adopting this strategy and outline how it can be used for planning experiments, analysing and reporting results, and integrating results over studies. I will describe a range of R-based tools that can help. Routine use of estimation should result in the quiet withering of p values.
Getting your Work Read by 1M People: Open Science via Wikipedia and its Sister Projects (Thomas Shaffee)
Wikipedia and its sister projects are the world’s largest open access project and often the most time-effective outreach platform available. Many Wikipedia articles get more than a million reads per year; Wikidata is fast becoming the web’s unified meta-knowledgebase; Wikiversity hosts diverse online courses; and Wikijournals enables dual-publication of peer-reviewed academic articles as citable versions of record, as well as living versionable documents.
This workshop will detail why and how to edit these platforms, including expected outcomes and impacts. It will also include illustrative examples of successful contributions and collaborations from different research communities
Mapping a Collective Pathway to Collaborative, Open Research (Vanessa Crosby & Fiona Bradley)
A culture shift is needed to overcome organisational barriers to transdisciplinary and open research. Theory of Change approaches are used in social impact settings to describe processes of planned social change. This workshop will ask participants to reflect on what needs to be in place at institutions, in disciplines, and individually for researchers to move towards interdisciplinary research and open science. Using outcome mapping tools we will identify blockers, and the enabling behaviours and supports needed for an "ideal state." Participants will map a Theory of Change and identify their sphere of influence, potential partners, and skills that they can build on after the conference.
Rmarkdown for Scientists (Nicholas Tierney)
A scientific report must be reproducibible to be completely credible. Data and code used for statistical analysis should be available for others to reproduce. R Markdown is a tool that allows you integrate your code, text and figures in a single file in order to make high quality, reproducible reports. A paper published with an included R Markdown file and data sets can be reproduced by anyone with a computer. This workshop will teach you how to:
Create R Markdown documents
Create, caption, and reference figures & tables that update with your data
Export R Markdown to PDF, HTML, & Word
Cite research articles and generate a bibliography
Improve workflow with keyboard shortcuts
Taking your R skills to the next level: four great strategies for reproducible research (Saras Windecker)
Growing concern about the irreproducibility of many published results means the credibility of research and the reliability of decisions they inform may be questioned. The aim of this workshop is to teach researchers using the language R best practices for making their research computationally reproducible – meaning the ability to rerun an analysis and reproduce the same results. This workshop will introduce our top strategies for increasing the reproducibility of your work in R.
Developing Resources in Contemporary Philosophy of Scientific Practices for Scientists (Eden Smith)
Many scientists re-examining the foundations of their practices are looking for collaborations with philosophers of science. Yet it can be difficult to find contemporary introductions to philosophy of scientific practice for scientists. Most attempts to integrate some philosophy of science fall back on Popper, Kuhn, Lakatos, and not much beyond. To overcome this, we are developing a set of resources introducing contemporary philosophy of scientific practices for scientists (not philosophers). This will be the second workshop in this series: we will present earlier feedback and facilitate small-group discussions on how to develop these resources to better support collaborations between philosophers of science and scientists themselves.
Maximising the diversity and inclusivity of AIMOS (Hannah Fraser)
We have made an effort to make AIMOS diverse and inclusive but recognise that there is much room to grow in this area. We'd love your help to strategise about the best ways to make AIMOS a welcoming and accessible society that runs diverse, friendly and productive conferences.
Quality in preclinical science (David Howells, Emily Sena & Glenn Begley)
Translational neuroscience faces many difficulties just because of the complexity of the brain. However, it appears that choosing the easiest preclinical experiments, introduction of bias, and poor understanding of statistical power requirements, make the search for therapies for human disease even more difficult. Evidence supporting these contentions will be provided and the audience encouraged to probe this data and help formulate a plan for the future.
Adapting metascientific research and reform to improve the legal system (Jason Chin & Rachel Searston)
We propose an unconference to discuss and promote interdisciplinary collaboration in applying metascience to law. This is important because openness and transparency are central to the legal system and opaquely conducted science has contributed to wrongful convictions. Some questions include:
Conducting forensic science more transparently (e.g., improving forensic science journals, balancing transparency with interpretability to give enough information to lay jurors without confusing them).
Retroactive disclosure statements for expert witnesses to uncover dangerous convictions.
Improving expert witness codes of conduct (e.g., to be more in line with practices like CONSORT).
Challenges in Open Science - Statisticians’ Perspectives (Ian Gordon & Sue Finch)
Day-to-day statistical consulting provides the opportunity to see close-up the breadth of academic applications of quantitative research, and the range of skills and approaches that researchers bring to the research and the analytic process. In this session, we would like to discuss the need for balancing prescribed rigour with contextually sensitive analytics, and the challenges many researchers may face in attempts to make their own science open.
Science Wiki: A viable alternative to journal publishing? (Andrew Vonasch)
We will discuss the feasibility of an alternative form of publishing. Science Wiki would spread information more quickly, thoroughly, and easily than journals. Access to new information would be instant and free. All scientific knowledge would be organized and linked by topic, rather than siloed into individual articles. Publishing would be faster and easier because desiloing removes the need for lengthy formal introductions and discussions. All contributions would be visible, rewarding scholars for currently-unrewarded work like unpublished studies, and reviewing. Challenges include upending the status quo, incentivizing good practices, and sharing control over and organization of topics.
Supporting the modernisation of research practices: Efficiency, Scalability, Openness (Matthew Ling, David Groenewegen, Linda Kalejs, Tyne Sumner & Pablo Ulloa)
There are many opportunities for individual researchers and teams to adopt more efficient, reproducible, accessible, and impactful research processes, tools and workflows. Facilitating and encouraging engagement with such professional development remains a challenge where institutions may be unwilling to invest, and academics are under time pressure, and reluctant to change well-practiced and proven effective strategies. Monash’s Data Fluency, UniMelb’s Research Platform Services, and ANZORN will share how they address these challenges, and facilitate a discussion about what can be learned from their models, what they can learn from you, & how we can collectively uplift research practices.
Why Aren’t There More Replications? (Bob Reed)
Let X = the optimal rate of replication in the social sciences. Let Y = the actual rate of replication. The difference represents the “replication gap”. While data are scarce, it is widely believed that the replication gap is positive. That is, that there are too few replications. This session will do two things. First, we will come up with a list of hypotheses about why there aren’t more replications. Second, we will brainstorm research designs that would allow us to test these hypotheses.