In the run up to that I heard about this interesting project lead by James Davenport, tracking the gender of astronomers who asked questions at the American Astronomical Society meeting in Jan 2014.
I ended up discussing this via a Twitter conversation with Jonathan Pritchard (@jr_pritchard) and we decided running the same experiment at NAM would be really interesting. After getting permission from the LOC, RAS and some help from James Davenport setting up the data collection web form we were ready to go.
The first ever NAM Hack Day gave the project a kick start and collected some other co-authors. I'm delighted to be able to say that the results it was published in A&G this month (
As you'll see we got a great response to the call for data collection at NAM - we were able to collect data on questions asked in about 70% of the talks at the conference which was absolutely fantastic.
We did find a clear gender difference in the rate of asking questions, and below I reproduce my favourite plot from the paper, showing how particularly in the first question, men were much more likely to ask, but by the fourth question the questions came from men and women in proportion to their representation at the conference.
|Plot from Pritchard et al. 2014, A&G|
When you read the report you'll see we found similarities, but also subtle differences with the results from the AAS (allowing us to quote Winston Churchill "we are two nations divided by a common language").
I also really enjoyed the research we did placing out result in the context of wider social science research into the psychology of asking questions, and also that we decided to end with a set of concrete suggestions to help improve the gender balance of those asking questions at astronomy conferences.
Our suggestions were:
- Younger scientists should be explicitly encouraged to ask questions (i.e. this should be stated in introductory remarks by the chair), and favoured if there is a choice of questioners.
- If there is a choice between male and female questioners for the first question, a question from a woman should be given priority.
- Questioners should be asked to identify themselves by name.
- If possible, Q&A sessions should not be cut short before at least four questions have been asked (if they need to be ended early). To enable this, session organizers should schedule enough time for questions and speakers should not be allowed to run over time.