ABC’s Q&A and panellist gender

Following up of my examination of the political affiliation of Q&A guests, I have done some very simple analysis of gender on Q&A, based on episodes since February 2009. Two results stand out: (1) the average Q&A panel had 2.8 men and  2.0 women and (2) even allowing for that, a man on a panel had around 50% more opportunities to speak.

The overall effect was that around two-thirds of words spoken by panellists on Q&A are spoken by a man (excluding the host, Tony Jones, of course).

A few definitions to understand the tables that follow: an appearance is a person appearing on a panel; a speech is an uninterrupted contribution from a single person in the transcript.

All episodes (N=199)
  Appear-ances Speeches Words Appearances per episode Speeches per appearance Words per speech Words per appearance
Male 548  19,970  938,356  2.8  36 47  1,712
Female 392  9,738  497,743  2.0  25 51  1,270

The results are not very different for ‘complete’ panels (those with exactly 5 panellists, the standard format), so this result is not explained by special formats (e.g. one-on-one with Kevin Rudd, or the Labor leadership debate between two men).

Complete panels (N=159)
  Appear-ances Speeches Words Appearances per episode Speeches per appearance Words per speech Words per appearance
Male 458  16,352  745,585  2.9  36 46  1,628
Female 337  8,379  410,628  2.1  25 49  1,218

This suggests two issues for Q&A: (1) the panel gender balance is not right, and (2) the moderator does not do a very good job of ensuring that women get equal opportunities to participate in the discussion. The first of these would seem fairly easy to fix; the second might be more difficult.

One suggestion might be to use a female host. We can test this, to some extent, since on five occasions in my dataset Q&A has been hosted by Virginia Trioli in place of Tony Jones (28/06/10, 18/07/11, 26/09/11, 09/07/12, 29/07/13). This suggestion fails: the results were not much different (compare speeches per appearance), and probably not statistically different (although I haven’t tested this.)

Episodes hosted by Virginia Trioli (N=5)
Appear-ances Speeches Words Appearances per episode Speeches per appearance Words per speech Words per appearance
Male 14  523  23,875  2.8  37 46  1,705
Female 11  292  14,280  2.2  27 49  1,298

What else could ABC do to improve this?

5 comments

  1. Fascinating stuff, Andrew. How did you count words?

    A small-N quantitative analysis of representative shows could be useful.

    And why use words as the relevant measure? Surely what we care about is impact. A bit mischievous, I grant you, but what if women put their points across in a more concise manner?

    1. At the moment I just split up each speech into words based on whitespace. There are more sophisticated word parsers out there, but I don’t think it would make much difference here. Also, for gender, the key difference is in number of speeches not words per speech. The implication is that women enter the conversation less often, but when they do they hold the floor for just as long as men. So in that sense the actual number of words (per speech) isn’t really a factor. Perhaps this aligns better with what you think of as impact.

      Alternatively, perhaps one can be very impactful by remaining silent for most of the show, then presenting a compelling argument in one single speech at the end. Of course, quantitative methods such as mine will never pick that up.

  2. I’m actually surprised these results aren’t worse. Leaving aside the words spoken, a 2.8 – 2.0 male/female ratio seems quite close, given that most of the Q&A panels have both a government and an opposition politician in it. Parliament has its own representation issues, and appearances are often dictated or invitations declined by the “party machine” rather than the Q&A editors.

    Perhaps Q&A has more capacity to be proactive with its non-politicians panel members. And of course, politicians do seem to do a lot of talking, so I would be interested to see how similar the results are once the sitting politicians are pulled out.

    When they select former politicians, where I presume they have more flexibility, I remember seeing Amanda Vanstone, Kristina Keneally or Helen Coonan as often as seeing Mark Latham or John Hewson…but maybe I’m having my own confirmation bias.

    Another interesting line to look into would be questions from the floor – that is the unscripted questions that Tony selects – does any gender bias come up in that? I stumbled upon this blog after watching a Ted Talk by Facebook COO Sheryl Sandberg (https://www.youtube.com/watch?v=18uDutylDa4) who suggests women are less likely to put their hand up, and hold it up as long (and hence are less likely to be selected in the Q&A), and I wonder if these views are reflected Q&As data.

    Real thought-provoking work!

    1. Thanks Lucas. I actually have pushed this result a bit further, trying to classify the politics of non-politician guests, but it proved fairly unsuccessful and it’s on hold for now. I agree it would be interesting to look at the gender of questioners, although I’m not sure if you can distinguish scripted vs unscripted questions from the transcript. Another thing I wanted to do was look at the selection of tweets that get displayed, although there are some challenges to that too.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s