[wesleytodd]

I noticed that as well. I have always thought there might be a correlation between self-taught and self selection for success in a topic.

Anecdotally, I was self-taught, and many of the highest performing engineers I know where also. Most of the average performing engineers I know were traditionally educated in CS.

[ddoolin]

I'm self-taught and that salary is within a couple $K of my own. Regarding your own experience though, I have noticed the inverse wherever I go. I moved from Charlotte (NC) two years ago, and there nearly everyone was self-taught and the skill level, generally, was much lower. The ones with degree educations were still stronger. When I got to SoCal, way more engineers were college-educated and still stronger than the self-taught ones, myself included.

[suresk]

I think it is hard to infer much from some of these data points.

There may be some correlation between "high performance" and "self-taught", since it takes a decent amount of motivation and talent to even get your first few jobs as someone without the educational background, but I don't know how long that is significant for.

I'm self-taught, and while it has worked out well for me so far, I do wish I'd gotten a CS degree. I've spent a ton of time studying stuff you'd learn in a CS degree, but I do feel like - especially for the first 5 years of my career - I made a lot of mistakes I probably wouldn't have, spent more time "inventing" things that had already been invented, and designed things in less optimal ways than I could have.

Plus, it took me about 10 years to understand just how fascinating and fun some of this stuff is. Although, who knows if I would have appreciated it in college?

[ddoolin]

Agreed, it's kinda hard to qualify, particularly as a side-effect rather than the central focus. Most of this boils down to personality.

I also sometimes wish I had gotten my degree for similar reasons, and still do plan to, though perhaps not in CS. Definitely math, and maybe something else. That isn't to say I wasn't boosted in my own ways, unencumbered by not trying to design things to fit a preconceived model, and learning what was practically useful. "Inventing" those things yourself can give you more insight and appreciation for the way they are or should be built. Unfortunately it seems to come full circle, and eventually you need those models and that less-practical knowledge.

[wesleytodd]

Yep, anecdotal evidence :)

I will say, where I am now (bay area FANG company) it is WAY more common to have traditional CS background and also be a high performer.

I attribute that to a biased selection in SV, and not an actual correlation of CS education to higher skill level.

[ddoolin]

Regarding the bias, I think it's the same way here, to be honest. My sample size for California (employer) is n = 1, and we do work that is both technically challenging and fun, so we attract and can only realistically take on those kinds of people.