5
$\begingroup$

The research output from UK and Germany are disproportionately large compared to other European countries (which you can see based on the number of Marie Curie or ERC grants awarded/capita to people in those countries versus others) and in general. They are research powerhouses in general, and also specifically in matter modeling research (consider the countries of origin for software like MOLPRO, CFOUR, NECI, HANDE, etc.).

There's currently only 1 user from UK in our list of top 24 users by reputation, and 0 from Germany (the following list is not too different if consider "most active" users by number of total posts instead):

Rank Username Country
1 Nike Dattani Canada
2 Camps Brazil
3 Jack China
4 Susi Lehtola Finland
5 Tyberius USA
6 ProfM UK
7 Thomas India
8 Tristan Maxson USA
9 Cody Aldaz USA
10 Andrew Rosen USA
11 taciteloquence USA / Taiwan
12 Geoff Hutchison USA
13 Hitanshu Sachania India
14 SR Maiti India
15 Etienne Palos Mexico / USA
16 Anoop Nair India
17 Chi Kou Algeria
18 Anyon undisclosed but possibly USA
19 B. Kelly Canada
20 ksousa Brazil
21 Pauli Bao Canada
22 Kevin J. May USA
23 Anibal Bezerra Brazil
24 Alone Programmer USA

I stopped there because I don't know where megamance is located, but I'll say that:

  • The next highest-rep user who is certainly in UK is Phil Hasnip (#31) and the top user in Germany is wzkchem5 (#26).

Why might this be? Our first instinct might be to think it's a cultural thing, but:

What might be the reason for this? The good news is that this data indicates that this site has a huge potential to grow from where it is right now!

$\endgroup$
5
  • $\begingroup$ The short answer: Randomness. With the proportion of users from StackOverflow (linked in Tyberius's answer) we would expect about 1 German and 1 British user among the top 24. Finding none is not so unlikely as to conclude that the distribution is different here. If you want the exact p-value, that would make a good stats.SE question, but the problem is that n=24 is a tiny sample. $\endgroup$
    – Pere
    Sep 3, 2021 at 15:48
  • $\begingroup$ @Pere Nice to meet you, and welcome to MMSE! Why is it that you say we would expect about 1 German and 1 British? Some of what you're saying has truth to it, but we could also look at the top n users on several SE sites, and base our estimations off of that, and then we'd expect more people from UK our top n users list. Also, matter modeling is more popular in UK and Germany than in almost every country present in our top 24, whereas "programming" (the subject of SO) is more universal. $\endgroup$
    – Nike Dattani
    Sep 3, 2021 at 15:59
  • $\begingroup$ According to the linked data, 6.75% of StackOverflow users are from Germany. In a sample of 24 people from the same population the expected number of German users is 6.75% of 24, which is between 1 and 2. The same for the expected number of British users (5.37% of 24). Furthermore, random deviations from the expected value (en.wikipedia.org/wiki/Expected_value) are usual, and therefore it doesn't seem strange not to find any British or German user among a 24 user sample - even without doing the maths. In fact, finding just one UK user is what happened. $\endgroup$
    – Pere
    Sep 3, 2021 at 16:09
  • $\begingroup$ (0.0675 x 24) is closer to 2, not 1. A lot of what you're saying is correct, but it's not taking into account the disproportional number of UK and Germany matter modeling researchers compared to many of the countries in this top users list. Furthermore, China has over 1 billion people so I wouldn't be surprised if there were a lot of them on SO, but UK will still likely have far more people in the top users list because UK will have more people fluent in writing professional-level (or native-level) English. India is #2 in the survey but nowhere near there in "top users". Not randomness. $\endgroup$
    – Nike Dattani
    Sep 3, 2021 at 16:23
  • $\begingroup$ Well, it's even worse - I work in Germany but I'm Chinese. If I wrote China in my Country panel instead, Germany would seem to be even more underrepresented... $\endgroup$
    – wzkchem5
    Oct 8, 2021 at 7:56

1 Answer 1

Reset to default
2
$\begingroup$

One of the tricky parts with this is that most users don't have their location as part of their profile, leaving the demographics of most of the site's ~320 avid users or ~3000 total user unknown. So hypothetically, the demographics of the rest of our user base might be closer to those of Stack Overflow where the number of users by country is ordered

  1. USA
  2. India
  3. UK
  4. Germany
  5. Canada
  6. France
  7. Brazil
  8. Poland
  9. Netherlands
  10. Italy

As to why the highest rep users don't match this ordering, I don't have a full answer, but I expect its related to your other recent Meta post: We hit rock bottom last week, in voting. What can we do now to encourage people to vote more?. Most of our highest rep users joined during the first few months the site existed, when site activity (questions/answers/voting) was at its highest.

So the demographics of our highest rep users are roughly the demographics of the users recruited during and shortly after the private beta. At that stage, we were mostly spreading the site to colleagues, which I would expect would lead mostly to people recruiting from within their own country.

Now, our traffic/growth is presumably driven more by Hot Network questions for other SE user or external search engines for entirely new users. I expect the users that find the site those ways roughly fit the demographics of the Stack Overflow survey, but due to lesser amounts of voting, it may take longer for these users to work their way up to the top of the rep leader board.

$\endgroup$
6
  • $\begingroup$ I disagree about users lacking location as part of their profile, because the title of my post says "active" users, by which I mean the users in my above list who do have their location as part of their profile. Furthermore, the part about the highest rep users being likely to be colleagues or people from the home countries of the early recruiters sounds on the surface like it makes sense, but it's not quite true. I heavily recruited people from UK and Germany (two countries where I've worked extensively and have the most matter modeling colleagues) but the highest rep users quickly $\endgroup$
    – Nike Dattani
    Aug 27, 2021 at 19:55
  • $\begingroup$ turned out to be a spread of people that I didn't know before starting the site. This might seem like a surprise but in the list of users in my post, I only knew 3 of them before launching this site proposal. $\endgroup$
    – Nike Dattani
    Aug 27, 2021 at 19:58
  • $\begingroup$ @NikeDattani But did these UK/Germany recruits join/stay? I guess my main point is that the high rep users are basically just the earliest users that stuck around, with Anoop and Jack being the closest to exceptions having only joined roughly a year ago. Tying this into the low voting post, more recent users like mykd or roma ichenko (who don't have their locations listed) have been reasonably active, but they probably have half the reputation they would have received if they had made the same posts during the Beta. So it moves the question to "why were the demographics of the Beta this way?". $\endgroup$
    – Tyberius Mod
    Aug 27, 2021 at 20:25
  • $\begingroup$ In your last sentence I think you mean "private" beta, but ProfM, Tristan, S R Maiti, Anyon, Kevin May, wzkchem5, megamance, JensenPang, livars98, etc. (in addition to Jack, Anoop) became active too late to get the private beta badge. I'm still surprised that we have so few users from UK and Germany (and my original post said that if you switch from "total rep" to "most active" the results look almost the same). The "most active" users correlate (very) far less with joining early than the highest-rep users do. $\endgroup$
    – Nike Dattani
    Aug 27, 2021 at 21:24
  • $\begingroup$ I suppose I mean the Beta and the month or two after, where we were still very actively trying to sure they didn't pull the plug on the site and so daily activity was artificially high. For the question at hand, unfortunately the people that would best be able to answer it are exactly the people we seem to be missing (UK/Germany matter modelers). $\endgroup$
    – Tyberius Mod
    Aug 27, 2021 at 21:44
  • $\begingroup$ I think we all agree that the highest-rep users are the ones who joined early, but it doesn't explain why the UK and Germany matter modelers aren't here. You're right that the best people to answer it might be the people who are not here, but I do wonder if people who are here have any theories as to why we don't have a lot of representation from those countries. For example maybe ProfM (UK) and wzkchem5 (Germany) can weigh in on it, or others who know why their UK/Germany friends are saying no to joining, or know about some other forum that UK/Germany people use (China has one!). $\endgroup$
    – Nike Dattani
    Aug 27, 2021 at 21:51

You must log in to answer this question.

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