@Mehrad The SO survey always has some really bizarre categorization stuff. I find it very nice because it has a lot of people answering it, and you can certainly identify certain trends, but more so than other similar projects, you need to invest a lot of time in reading it correctly- because it's not easy!
@yo
The issue is that they are not providing the raw (or cleaned) data to be analyzed "correctly". I think there are a lot of statistical assumptions that are not clarified in the methodology. Also lack of reporting the variance makes it pretty unreliable about trends.
For example the report says 0.6% of participants have primary and elementary school education!
Perhaps the only "reliable" part is the age-range of users:
https://survey.stackoverflow.co/2024/developer-profile#6-age
@Mehrad Oh, they don't? I think they used to do this in previous editions.
I've always thought there's a couple of interesting data analysis things to do there. (E.g. show stuff which is different when you fix one variable; for example, how devs from country X differ from the average.)
Shame.
@yo
It might be there, but at least I didn't find it. I would be happy if someone can point me to the direction of finding the data for proper analysis.