A master's degree, assuming no experience, is more like a warning signal for someone who's stayed in academics too long and has no practical skill. It's like you're hiring a really expensive junior developer if they've never built real world software under business constraints that needed to like be bug free, and scalable. It's like the code written in graduate college gave a shit about memory use, file or network i/o performance, a good ui design, or like actually doing the thing it's supposed to. Sorry, ranting due to bad experience in the past.
The thing is, more education will enable a good developer to be even better. It's doesn't seem to help a struggling developer in any shape or form. Unfortunately sometimes I get the impression that people stay in school to address a problem that can't be addressed with more classroom time and end up doubling down on a field that maybe isn't right for them.
At the very least try to intern early on and or successfully help a popular open source project before graduate school.
> The thing is, more education will enable a good developer to be even better. It's doesn't seem to help a struggling developer in any shape or form.
In my (limited) experience, the people who struggled in undergrad run screaming away from any possibility of doing grad school (my experience here is consistent across econ and CS). The people who were good at the field consider grad school because the psychic costs of learning more are low.
If you find a person with an MS who can't pass a simple programming test, then (and this is very general, I'm sure there are exceptions) take a look at the school her or she attended and stop interviewing people from that school, problem solved.
> take a look at the school her or she attended and stop interviewing people from that school, problem solved.
This is really bad advice, because much about learning programming is about an individual effort. It has little if anything to do with the school.
> If you find a person with an MS who can't pass a simple programming test
You actually need more than a simple programming test that solves a single problem. You need to see enough code to have some architecture, designed by the candidate. Interviews are too short for this so experience whether at some other company or open source, or just if skilled people can vouch for you goes a long way.
I also agree in doing a black mark against the school however it helps to have a look at (in terms of Masters, PhD) the quality of work within that department / within the graduating year. It's a way to go beyond the marketing fluff in non top-tier schools.
Replying to leobelle here since the thread got too deep.
I agree that learning to program is individual, but you don't learn to program in grad school. If a school gave you an MS and you can't do Fizzbuzz, that school isn't serving its signalling purpose (that I mentioned in my first comment). So while that school may be fine, if lax, if the employer is using education as a signal, a school that doesn't fail students doesn't provide a strong signal.
Relating to your comment about needing to see more code, education is actually perfect for that. Most companies don't let you take your source code with you, schools do. So provided you took some courses that required significant engineering projects, a student has a big pile of (hopefully) reasonably-designed code to show off to an interviewer. A person who worked for a less-than-progressive company (since some maintain open source libraries and such) does not. They both have their open source contributions.
The thing is, more education will enable a good developer to be even better. It's doesn't seem to help a struggling developer in any shape or form. Unfortunately sometimes I get the impression that people stay in school to address a problem that can't be addressed with more classroom time and end up doubling down on a field that maybe isn't right for them.
At the very least try to intern early on and or successfully help a popular open source project before graduate school.