Comment by wainstead | Hacker News Reader

Top New Best Ask Show Jobs

Comment by wainstead | Hacker News Reader

Beware: the Stack Overflow Developer Survey is not a scientific survey. Participants are self-selected, not randomly sampled.

For true random sampling of the software developers in the United States, use the Bureau of Labor Statistics, who employ real statisticians:

https://www.bls.gov/oes/current/oes_nat.htm#15-0000

You can also break them down by U.S. state:

https://www.bls.gov/oes/current/oessrcst.htm

And by MSA (Metropolitan Statistical Area):

https://www.bls.gov/oes/current/oessrcma.htm

As an interaction/ux/product designer, I always find it odd that government forms never have a relevant job role in any of its data. They lump us in with graphic design. :/

Why do you find it odd? A "a interaction/ux/product designer" did not really exist 10 years ago or wasn't very popular. Do you think the government moves at the pace of a Silicon Valley startup updating job titles every year?

I don't think the Stack Overflow data is fantastic, but I'm not clear that the BLS stats are any better for the fancy credentials of their collectors.

In either case, the problem is coming up with a reasonably accurate reference class. At least in my case, I'm not well represented by the BLS breakdowns (which underestimate what I can expect to get by a factor of about four.)

The Stack Overflow approach, using industry-specific features (developer-friendly keywords/job titles/languages) is definitely smarter, though still wildly insufficient. (It's only off by a factor of about three.)

Reply on news.ycombinator.com