2

There was a recent SE blog post about promoting diversity and niceness at SE, and an associated SE Meta question.

Food for thought for us at Academia SE!

Could someone provide some recent stats regarding gender distribution of Academia SE participants? Academia has been male dominated for a long time (in some fields more than others; and although women are becoming more and more present as time goes on), and it's my impression that that imbalance may carry over to our participation levels at Academia SE. What do the stats say? I realize there are lots of participants who have not indicated their gender. But it would be interesting to see what is known about gender distribution here.

The blog post contains a brief survey questionnaire. I hope Academia SE participants will take part.

3

tl;dr- This post's a data dump for:

  • demographic data and other diversity facts;

  • links to SE content about diversity.

This post is a community wiki, so please feel free to contribute edits!


Demographic info, informal polls

Several SE sites have performed informal polls, though it should be understood that informal polls like these tend to suffer from strong sampling biases.

  1. SE.Academia poll and corresponding discussion.

  2. SE.TeX poll and corresponding discussion.

  3. SE.ReverseEngineering, no corresponding discussion.

  4. SE.Mathematica, no corresponding discussion.

Demographic info, for StackOverflow

Most SE sites don't collect demographic info, but the big exception is the annual developer survey for StackOverflow:

Figures below are from the 2018 developer survey.

Geographic location

Gender:

  • Male: 92.9%;

  • Female: 6.9%;

  • Non-binary, genderqueer, or gender non-conforming: 0.9%.

Race/ethnicity:

  • White or of European descent: 74.2%

  • South Asian: 11.5%

  • Hispanic or Latino/Latina: 6.7%

  • East Asian: 5.1%

  • Middle Eastern: 4.1%

  • Black or of African descent: 2.8%

  • Native American, Pacific Islander, or Indigenous Australian: 0.8%

Sexual orientation:

  • Straight or heterosexual: 93.2%

  • Bisexual or Queer: 4.3%

  • Gay or Lesbian: 2.4%

  • Asexual: 1.9%


Demographic info, by field in American academia

Bachelor's degrees in the US, though probably reasonably reflective of graduate degree distributions as well as the student body academics at universities see everyday:

      

Key observations:

  • Computer Science and Engineering are both heavily male-dominated.

  • Health Professions, Public Administration, Education, and Psychology are heavily female-dominated.


StackExchange posts on diversity issues

Q&A about general diversity issues:

  1. From SE.Academia:

  2. From SE.ComputerScienceEducators:

  3. From SE.Skeptics:

Q&A about StackExchange-specific diversity issues:

  1. Related to the recent blog post, "Stack Overflow Isn’t Very Welcoming. It’s Time for That to Change." (2018-04-26):

  2. General discussion unrelated to the recent blog:

  3. Specific cases of reported issues:

2

I think a particularly valuable part of this question is the encouragement for everyone on Academia SE to take part in the SE survey, which allows a space for some short comments, as well as an opportunity to volunteer for follow-up research. The more that active users engage with that, the better a picture we'll get of what's working and what opportunities there are for change.

The more that this is qualitative research, rather than just counting people up in categories, the more useful I think this effort will be. That sort of user experience work can surface complex issues and potential solutions. On the other hand, it's hard to tell what benchmark Academia SE demographics should be compared to.

  • For instance, it's a worldwide site, so the meaning of certain race/ethnicity categories is hard to interpret.
    • People of South Asian and East Asian descent are 16.5% of the SE developers (according to Nat's great answer), but 35% of all people in the world live in China and India.
    • Should it be benchmarked against the world's English-speaking population? Maybe not, since most academics around the world are incentivized to work in English.
    • More deeply, asking about race/ethnicity is not based on some idea that all people who are white or of European descent are interchangeable or share a deep set of characteristics. Rather, the category provides useful information in the context of a given country; a particular race/ethnicity means they are likely to have been treated in a certain way or that we are more likely to be able to predict other correlated traits.
    • So far in the academic literature, I haven't seen treatment of how to ask meaningful world-wide questions about race/ethnicity, other than tailoring the questions to different countries/cultures to capture the distinctions that matter in that society. (I am very aware that many people find the idea of "race" itself offensive, particularly Europeans.)
  • SE users are self-selecting in a lot of ways.
    • I speculate that Academia SE users are more likely to be early career rather than late career academics, which might make the pool more demographically diverse. (Not sure the extent to which undergraduates and grad students are involved.)
    • Because SO and SE are built on programming questions, people who do programming and computational work are more likely to discover and sign up on Academia SE. I believe that even within academic disciplines in the U.S., more quantitative and computational work tends to be done by men. (Definitely my own experience; I believe I've seen documentation of this, and would edit in a reference if anyone has one off-hand.)
    • Are there elements of SE culture that are further causing self-selection? In bad ways?

A lot of measurements seem like their meaning should be self-evident, but that's rarely the case. It's also very easy for people to list statistics and for others to infer blame from those statements. Further, when people know stats but aren't used to social science or stats about people, it's easy to come up with calculations that are technically correct but misleading or misinterpreted. It may be most productive if we can agree on some basic descriptive facts and withhold normative judgment from those numbers alone.


When I saw this question, I became really worried it would explode into the conflict we've seen on other parts of the site in response to this issue. My initial thoughts were that the most productive way forward might be to not press the issue but to keep doing useful things, like Atlanta's rebranding itself as "The city too busy to hate."

I don't think avoiding the real problems people experience is the right thing to do, but in this case I think that those problems are better explored by people discussing their experiences, rather than reading into demographic numbers.


Edit: I took a look at the questions in the poll Massimo Ortolano commented about, and those seem like they may be helpful for Academia SE to understand its users.

Edit: Why am I hesitant about this conversation? Because it's being addressed in different ways across SE/SO. This answer on SE Meta covers a lot of it. Then there are various ones (responses to the blog post and others on the topic) that demonstrate how volatile these discussions can be: SE Meta 1, SE Meta 2, SE Meta 3, Interpersonal Skills (IPS) Meta 1, IPS Meta 2, IPS Meta 3, SO Meta 1, SO Meta 2, SO Meta 3, SO Meta 4, SO Meta 5, SO Meta 6, SO Meta 7

  • 2
    Thanks for your thoughtful post. // "I am very aware that many people find the idea of "race" itself offensive, particularly Europeans" -- what's that about? If you don't want to take the time to explain that, could you suggest something to read (something not too long) -- I didn't know about this. Is this idea most commonly found among white Europeans, or is it ascribed to by a representative slice of the population in certain European countries? // "I became really worried it would explode into the conflict we've seen on other parts of the site in response to this issue." I didn't know... – aparente001 May 3 '18 at 20:59
  • 1
    ... about that. If my question creates a mess at Academia SE, I apologize, that wasn't my intention. I found out about the blog post by accident, with a delay, and I thought that others might not be aware of it. But could you clarify, when you say "other parts of the site" -- do you mean other stacks? I didn't know about that either. – aparente001 May 3 '18 at 21:01
  • 2
    "I speculate that Academia SE users are more likely to be early career rather than late career academics": oh, well, there are a few old timers around too ;-) – Massimo Ortolano May 3 '18 at 22:15
  • @aparente001 I'm having trouble finding a good source about European hesitation to use the term "race," but this Wikipedia entry may help clarify; I am guessing that the norms are probably rooted in the aftermath of the Holocaust for Europe, vs. the aftermath of slavery for the U.S. Years ago I reviewed the comments on a demographic survey given by a large international website, with many people expressing the sentiment, "I am European and humans are all one race. It is heinous to imply otherwise." – cactus_pardner May 4 '18 at 1:05
  • 2
    @aparente001 and cactus_pardner: Some twenty years ago I was on my second trip to the US and one day I visited an Italian friend who was going to move in the US. He had to apply for the SSN and I went with him to the office to help with the application. The application form had a "race" field. We were both shocked and our reaction was: "Why on earth do they want to know your race?!" Write "human", I said. We eventually asked at the desk how to fill that field... So, yes, at least in my ears, the need to specify or speak about a race evokes disturbing ideas. – Massimo Ortolano May 4 '18 at 11:58
0

It’s not possible to report gender data, because it’s not something that Stack Exchange asks for or collects.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .