Post provided by David Warton
A few leading reasons for going to a conference are: to present your work and get feedback on it, to find out what others are doing, to meet collaborators and to network. But a conference can also be a great setting for generating completely new ideas. I find that conferences are one of my most likely places for a “eureka moment”.
Surrounded by researchers working on a range of different problems in interesting and often original ways, I’m encouraged to think about things from a different angle. Idea generation is perhaps one of the main benefits of going to a conference – but is the typical conference format is the best way to facilitate that? Or does it focus too much on giving researchers a platform to report on previous research ideas?
Structuring Conferences to Generate Ideas
So what can you do differently, to get the most out of a conference scientifically? As an attendee, I would firstly recommend trying not to burn out on too many talks – no shame in skipping a session to have a chat to someone interesting. And I would make a point of meeting people, especially those who gave talks that piqued your interest, to find out more in an informal setting about their work and how they think about things.
What could conference organisers do differently? The first thing I’d suggest is to make sure you have decent-sized breaks between sessions! (To better enable the above.) If you are game you could experiment with some other ideas too, explicitly aimed at generating new science.
At UNSW Sydney we have hosted a couple of Eco-Stats conferences, the main purpose being to encourage more collaboration between ecologists and statisticians. The format was to pair up invited speakers across disciplines to speak on complementary topics, then give speakers and interested participants some time to think about potential collaborative opportunities, and propose a couple of paper ideas.
For example we asked Otso Ovaskainen (University of Helsinki), who has expertise in model-based approaches to the analysis of multivariate data, to speak alongside Doug Yu (University of East Anglia), who has expertise in metagenomics and related modern techniques for rapid biodiversity assessment. We also put Rachel Fewster (University of Auckland), with expertise in capture-recapture methods and the incorporation of auxiliary data (e.g. genetic markers, incorporating uncertainty identifying individuals) together with Paul Sunnucks (Monash University), who uses genetics in ecology and conservation, for example in estimating the number of northern hairy-nosed wombats in the wild using data from various sources. These researchers would not normally attend the same conference let alone be paired up at one, and most had not met before (although we later found out Otso and Doug had).
On the final afternoon of the conference we had 90 minute break-out sessions to brainstorm potential collaborative opportunities. In our most recent conference, speakers stuck around the following day to generate draft abstracts for ten different papers, which came out of six conference sessions.
Technological Advances at the Interface between Ecology and Statistics
The April 2017 issue of Methods in Ecology and Evolution has a Special Feature highlighting some of the outcomes from this conference, “Eco-Stats ’15: Technological Advances at the Interface between Ecology and Statistics”. The ten proposed papers didn’t all make it over the finish line – one was dropped, a few missed the submission deadline, one chose to submit their paper to a different journal that better suited the idea they came up with. The papers are a mixture of partially formed ideas people came into the conference with and ideas generated in discussions at or shortly after the conference. See what you think!
Next time you go to a conference, will one of your goals be to come away with new paper or project ideas?
The full ‘Technological Advances at the Interface between Ecology and Statistics’ Special Feature will be freely available for a limited time.