Gender Breakdown for Followers of Digital Humanities on Twitter

[Sterling Fluharty]

Is anyone interested in seeing a more complete analysis? I wouldn't
mind putting something together if it would be useful. I would also
appreciate some feedback, since there was some skepticism about the
percentages I presented earlier today on twitter.

[Ben Brumfield]

I'd love to see that. I had a lot of trouble reconstructing today's
discussion of follower ratios and the tool's methodology, since nobody
(including me) was using hashtags. I feel like I missed a lot -- like
whether sterflu, melissaterras, and cliotropic would appear as 1F/0M
or 1F/2M.

[Sterling Fluharty]

Hello All,

I apologize that my methodology was hard for many to follow in the
tweet stream. You will find it explained below in more detail. I
also report on my further analysis and provide some recommendations.

I used this twitter tool to help me generate a list of individuals who
were most likely to be affiliated with the digital humanities
community:

http://twiangulate.com/search/

For my analysis, I first decided to included tweeters who had at least
200 followers. That way the dataset was manageable and the results
would have more statistical validity.

My second step was to include individuals in the dataset if they were
following at least 20 individuals that were also being followed by
@dhnow. I used the follower list of Digital Humanities Now as a
common reference point because I thought it was most likely to reflect
the range of individuals in the DH community. This tool helped me to
do the relevance filtering:

http://twtrfrnd.com/

Lastly, I used this tool to determine the gender breakout for the
followers of DH tweeters:

http://www.twitteranalyzer.com

The final version of the dataset I produced, which contains 164 DH
tweeters, is available here:

http://spreadsheets.google.com/pub?key=tm3QNs91-nQDdj-R0IQ5RhQ&single=true&gid=0&output=html

I should make a few caveats about this analysis. Several of the
individuals who were excluded from this analysis were actually
noteworthy figures in DH, with several serving in administrative
positions at academic units or centers for DH. I noticed that
occasionally the twitter analyzer tool underreported the number of
followers. This happened in probably less than five percent of
cases. I doubt this skewed the overall proportions of male and female
followers in the dataset very much, since both male and female
followers were equally likely to be undercounted in these few cases.
And certainly some followers were not counted because they did not
provide Twitter with an actual given name that could be correlated
with being either male or female.

I determined the gender of all but five DH tweeters in this dataset.
Only 65 out of 164 (40 percent) were women. This is surprising for a
community that draws its practitioners primarily from the academic
humanities, public history, archives, and libraries. But it is worth
pointing out that this gender ratio is fairly consistent with a study
from a little less than a year ago found that 57 percent of twitter
users are male:

http://www.prweb.com/releases/2009/04/prweb2367764.htm

The apparent gender disparity among DH tweeters continues when we
examine who is following whom. If we ignore the two outliers with
more than four thousand followers, we find that the average female DH
tweeter has 564 followers while the typical DH tweeter has 779
followers. We should be asking ourselves why these numbers aren't
more equal.

The following statistics indicate that women are the majority in
almost all of the fields that typically connect with DH:

Government and professional surveys have shown that a little over 80
percent of librarians, and around two thirds of archivists, are women:

http://www.libraryjournal.com/article/CA423789.html
http://caliber.ucpress.net/doi/abs/10.1525/tph.2009.31.1.155

This article explains that women earn nearly 60 percent of all
doctorates in the humanities but only about 40 percent of those in
history:

http://blog.historians.org/articles/543/finding-women-in-history

This recent survey shows that nearly 66 percent of all public
historians are women:

http://www.historians.org/Perspectives/issues/2009/0909/0909pub1.cfm

As you and Shawn have already discussed, there are obviously some
pipeline issues at work that have flip-flopped the proportion of women
one would expect to find in a field like digital humanities. This
report offers some thoughtful analysis and recommendations for women
in the historical profession that may have some relevance for the DH
community:

http://www.historians.org/governance/cwh/CWH-Report_5.20.05.pdf

One of my hopes is that the list of DH tweeters I put together can
help us to discover DH tweeters that we may have overlooked when
deciding who to follow. By expanding the pool of people we follow, we
could likely increase our sense of community and help women feel more
welcome. For instance, the average female DH tweeter has a following-
to-followers ratio of 0.69 while among male DH tweeter this ratio
averages 0.57. It looks like some of the men have been a little too
selective in who have they followed and could probably afford to catch
up with the women.

Best wishes,
Sterling Fluharty

[Andrea Odiorne]

I used the twitteranalyzer and ran through some individual followers
to see their gender. There were many people who were assigned a
gender that was different than the one I, or probably they, would have
assigned them. If they I using names to determine gender then might I
understand why Jo (yes without the e) would show up male, but why
would Tracey, Kristen or Loren? Also, it seems that few men,
(actually none that I found) had the wrong gender label This is a
very unscientific look at twitteranalyzer, but I wonder why the tool
seems more likely to misgender people as male.

[Dorothea Salo]

On Fri, Jan 22, 2010 at 8:44 AM, Andrea Odiorne <andrea...@gmail.com> wrote:
> I used the twitteranalyzer and ran through some individual followers
> to see their gender.  There were many people who were assigned a
> gender that was different than the one I, or probably they, would have
> assigned them.  If they I using names to determine gender then might I
> understand why Jo (yes without the e) would show up male, but why
> would Tracey, Kristen or Loren?  Also, it seems that few men,
> (actually none that I found) had the wrong gender label  This is a
> very unscientific look at twitteranalyzer, but I wonder why the tool
> seems more likely to misgender people as male.

I've run any number of these gender analyzers on my own work,
particularly my erstwhile blog, and the most feminine I've ever
managed is androgynous -- right in the middle. My suspicion is that
the supposed textual markers of femininity are highly stereotyped; or
alternately, that male writing is taken as the "norm" and the
analyzers look for specific (and probably again stereotyped)
deviations therefrom.

Dorothea

[Heather Prescott]

I missed the beginning of this discussion so please excuse any redundancy.  I'm a relative newcomer to Twitter (been using it for about a year).  Most of my colleagues in history, male and female, have not heard of it, and the few  who have think "what's the point"?  None of them blog either and still have to be sold on the professional use of social media (although a few are on Facebook but only for fun).

So, before considering whether or not Twitter is welcoming to women, first of all think about why would any historian, male or female, would want to use it?  In other words, how will it benefit them professionally, especially if they are untenured and trying to fulfill multiple professional and personal obligations?  Mills Kelly made a good comment on my blog about why he doesn't tweet -- he has too much to do and something had to go.  [my blog is hmprescott.wordpress.com]

[Sterling Fluharty]

Thanks for all of the replies. I appreciate all of you double
checking and critiquing the accuracy of twitter analyzer's gender
disambiguation. Maybe one way to verify my analysis of the proportion
of women within the DH community would be to tabulate who @dhnow is
following. It is probably important to remember that twitter analyzer
tells us the gender of a tweeter's followers and nothing directly
about who they are following.

I think it would surprise many in the DH community just how many
historians are on twitter. I am building some lists that, probably by
this weekend, will contain nearly two thousand tweeting historians
that I have found with tools like tweepsearch. Of course, the
majority of these are non-academic historians. So maybe the question
we should be asking is why academic historians are so behind the curve
when it comes to adopting this platform for distributing their work.
I think you have to put Mills Kelly in context. He became an academic
dean around the time that twitter caught on in the DH community. I
noticed that his blogging also declined precipitously, so he probably
was really busy. And even though he mocks twitter as stalking, he
just recently got back on facebook. So maybe there is still hope for
him. ;-)

Best wishes,
Sterling Fluharty

On Jan 22, 7:53 am, Heather Prescott <hmpresc...@gmail.com> wrote:
> I missed the beginning of this discussion so please excuse any redundancy.
> I'm a relative newcomer to Twitter (been using it for about a year).  Most
> of my colleagues in history, male and female, have not heard of it, and the
> few  who have think "what's the point"?  None of them blog either and still
> have to be sold on the professional use of social media (although a few are
> on Facebook but only for fun).
>
> So, before considering whether or not Twitter is welcoming to women, first
> of all think about why would any historian, male or female, would want to
> use it?  In other words, how will it benefit them professionally, especially
> if they are untenured and trying to fulfill multiple professional and
> personal obligations?  Mills Kelly made a good comment on my blog about why
> he doesn't tweet -- he has too much to do and something had to go.  [my blog
> is hmprescott.wordpress.com]
>

> On Fri, Jan 22, 2010 at 9:44 AM, Andrea Odiorne <andreaodio...@gmail.com>wrote:
>
>
>
> > I used the twitteranalyzer and ran through some individual followers
> > to see their gender.  There were many people who were assigned a
> > gender that was different than the one I, or probably they, would have
> > assigned them.  If they I using names to determine gender then might I
> > understand why Jo (yes without the e) would show up male, but why
> > would Tracey, Kristen or Loren?  Also, it seems that few men,
> > (actually none that I found) had the wrong gender label  This is a
> > very unscientific look at twitteranalyzer, but I wonder why the tool
> > seems more likely to misgender people as male.
>

> > On Fri, Jan 22, 2010 at 1:32 AM, Sterling Fluharty <ster...@gmail.com>

> > wrote:
> > > Hello All,
>
> > > I apologize that my methodology was hard for many to follow in the
> > > tweet stream.  You will find it explained below in more detail.  I
> > > also report on my further analysis and provide some recommendations.
>
> > > I used this twitter tool to help me generate a list of individuals who
> > > were most likely to be affiliated with the digital humanities
> > > community:
>
> > >http://twiangulate.com/search/
>
> > > For my analysis, I first decided to included tweeters who had at least
> > > 200 followers.  That way the dataset was manageable and the results
> > > would have more statistical validity.
>
> > > My second step was to include individuals in the dataset if they were
> > > following at least 20 individuals that were also being followed by
> > > @dhnow.  I used the follower list of Digital Humanities Now as a
> > > common reference point because I thought it was most likely to reflect
> > > the range of individuals in the DH community.  This tool helped me to
> > > do the relevance filtering:
>
> > >http://twtrfrnd.com/
>
> > > Lastly, I used this tool to determine the gender breakout for the
> > > followers of DH tweeters:
>
> > >http://www.twitteranalyzer.com
>
> > > The final version of the dataset I produced, which contains 164 DH
> > > tweeters, is available here:
>

> >http://spreadsheets.google.com/pub?key=tm3QNs91-nQDdj-R0IQ5RhQ&single...

[Matthew Kirschenbaum]

A back-of-the-envelope scan of the registration list at DH09 at
Maryland this past June suggests roughly a 2:1 male female ratio. Matt

--
Matthew Kirschenbaum
Associate Professor of English
Associate Director, Maryland Institute for Technology in the Humanities (MITH)
Director, Digital Cultures and Creativity (DCC, a new Living/Learning
Program in the Honors College)
University of Maryland
301-405-8505 or 301-314-7111 (fax)
http://mkirschenbaum.net