The Worldwide Skeptical MovementTM,1 has found itself in the unenviable position of one of Van Wilder‘s2 clients, namely asking, “How do we get ladies to come to our events?”3. Fortunately, The Worldwide Skeptical MovementTM seems to be asking for more serious reasons than the Lambda Omega Omegas. Unless you are the Augusta Country Club, you want your group demographics to mirror those of the society within which your group is embedded. Among many other benefits, this shows that your message is successfully reaching the entire society, not just a specific niche.
This is what is known in the business as “hard”.
The Worldwide Skeptical MovementTM does not look like the larger society. It does look a lot like many IT departments – i.e., white and male. For the purposes of this brain effluvia to which I am subjecting you, we are going to focus on the under-representation of women4. Although the Skeptical Oligarchs have begun to realize that this may be an issue for their plans to be “Worldwide” and a “Movement” of lasting import, understanding this problem has not moved beyond the anecdotal.
I like numbers, and not just because I am uncomfortable with my own emotions. In this case, I like numbers for four reasons. First, measuring the problem demonstrates clearly that you are aware of the issue and signals that you want to correct it. Second, numbers tell you how big the problem is. Third, numbers give you a way to determine whether attempted remedies are making things better, worse, or doing nothing. Fourth, they let you evaluate the approaches of different groups through a single, shared criteria.
Without doing The Worldwide Skeptical MovementTM‘s job for them, I though I might take a cursory look. But, where to get data, if no one is collecting data on Skeptical demographics?
I figured, hey, if the big corporations can buy your privacy from Facebook, I might as well get in on the action too. A number of Skeptics in the Pub (SITP) groups5 organize through Facebook. In order to evaluate whether and/or how much women are under-represented SITP groups, we need to establish the background level of representation in the larger society. In this case, the larger society is Facebook. What is the representation of women of drinking age (these events are occurring in pubs, after all) on Facebook? In the US, women represent 56% of Facebook members above the age of 187 (Figure 1).
For my Skeptical data set, I looked at 17 SITP groups (out of 44 listed)
from the UK and US across a range from membership sizes (n=20-283) with
the requirement that the SITP chapter maintain a Facebook group page8. In total, I reviewed 1096 SITP members. While there were individuals who held memberships in multiple groups, I did not adjust for this minority group, mainly because if you want that kind of rigor, tell the leasing Skeptical organizations to get off their keisters and do it themselves. To the left, you can see a visually stunning pie-ish chart showing the representation of females in each SITP group (Figure 2). The relative size of each group is indicated by its position in the set of concentric circles, with the largest group on the outside (n=283) and the smallest group (n=20) on the inside (circumferences not to scale). The females are in orange. Obviously, that is not even close to 56%9.
Females represent 31% of SITP members. This relationship holds across all SITP group sizes (Figure 3). Female representation in these Skeptical organizations is just barely over half of what we would expect by chance (i.e., Facebook users just randomly choosing to join groups).
That is pretty bad. Extremely bad. But, it gets worse. If we look at the group administrators10, a paltry 12% are female (Figure 4).
SITP have only half the female members they should and, among those current members, only about a third of the female leaders. That’s how serious the problem is. I know misogynistic religious groups that have better numbers than this.
Maybe someone should do something about this? As i said above (before we launched into “lovely” Excel graphs), an excellent first step would be to quantify the problem. Is my quantification great? No. Does it have problems? Yes. But the real question is why is there nothing better than a postdoc spending a few hours browsing Facebook?
Hey, The Worldwide Skeptical MovementTM, show us you really care.
1: I think the trademark rights to this particular Skeptical Movement moniker belongs to the James Randi Educational Foundation. Cough, branding, cough, cough.
2: For those of you with a less prurient sense of humor (a condition known as basculus in ano), allow me to hesitantly recommend Will Smith romantic comedy vehicle Hitch. The movies share the same general plot, but Hitch omits penis enlarger jokes and comes free of the general stain of being a post-Christmas Vacation movie by National Lampoon.
3: I know. I know. You were expecting me to drop some line about solving the problem by having a sexy scientist (or engineer) like the rugbyologist attend, but that would be ridiculous. I can’t be everywhere, people. Geez.
4: I focused entirely on the demographic group of gender not because of its importance relative to other demographic categories like age and race for several reasons. First, while
gender identity is binary for most individuals and is less dependent on group identification. Second, gender can often be inferred rapidly from photo or name, speeding data
collection relative to the calculations necessary for age. Third, gender
identity is usually provided as public information.
5: I restricted my search to Facebook “groups” as the “Page/Likes” dynamic that is becoming more common may not represent actual participation rather than general support, and is much harder to tally.
6: I’m operating under the assumption that male/female/none choice on Facebook
represents the individual’s self-reported gender identity, not their
complement of chromosomes, which would be their sex.
7: I could not find a breakdown that demarcated at 21.
8: All data collection was based on group pages status on 5 May 2010.
9: Unless, perhaps, you are trying to develop a rule for the number of fingers that should be on the human hand.
10: Admittedly, Facebook group admins/officers may not reflect the true leadership structure of a SITP group. In addition, two groups were excluded from the leadership analysis (Orange County and Glasgow), because their admin policy clearly differed dramatically from other groups, leading to disproportionately large numbers of admins for each group and to these groups dominating the data set.