Short summary (what we know from the 2023 census). Feel free to add an additional observations:
Total responses: 531 (about 8% of ~6,500 registered users at the time)
Age: the distribution peaks in the 30–39 bracket (author: “average age … a bit older than other social platforms”).
Gender: strong male skew — 441 identify as “Man”, 32 as “Woman”, 25 as “Non-binary” (plus a few others)
Location: ~87% of respondents are from Canada
Education & employment: high education levels (e.g., 223 reported a Bachelor’s, 55 Master’s, 19 PhD) and ~70.8% employed full-time.
So your typical answer is coming from a fairly well educated mid 30’s Canadian guy. Huh.
As others mentioned, this would be an interesting data point to validate. I’m not familiar with the server side of Lemmy, but does the server provide any logs which could be used with GeoIP to get a sense of the relative number of connections from different countries? While there is likely to be some misreporting due to VPN usage and the like, it’s likely to be a low enough number of connections to be ignored as “noise” in the data. Depending on the VPNs in question, it may also be possible to run down many of the IP addresses which are VPNs in the connections logs and report “VPN user” as a distinct category. This would also be interesting to see broken out by instance (e.g. what countries are hitting lemmy.world versus lemmy.ml versus lemmy.ca etc.).
All that said, thank you for sharing. These sorts of exercises can be interesting to understand what a population looks like.
The reason the survey shows that many canadian people is probably much simpler. The survey was hosted on lemmy.ca
OP didn’t knew a lemmy server was federated with other lemmy servers. They thought lemmy was centralized and canadian