...it shows that you're dedicated enough to spend 4 years of your life to get a piece of paper, which shows that you're less likely to be the type of person who will quickly jump ship.

My experience belies that assertion quite handily. I don't mean to imply that I think that those with degrees are inherently unreliable, or anything even remotely similar. What I mean to say is this: It's relatively easy to obtain a degree during four or more years spent living on Daddy's generosity, as contrasted with gaining substantially the same knowledge of one's chosen field, within or without academia, while coping with the struggles of the real world.

Several years ago, I had an enlightening experience. I'd landed a fairly large and technically complex project, and subcontracted some work out to two programmers with whom I had no previous experience. One was self-taught, the other sporting a Ph.D. in CS, with a Masters in mathematics for a kicker. I had to ride the latter like a rented mule to finally get some incredibly sloppy code out of him, which I then delivered to the self-taught guy and asked him for his opinion. Within minutes, he gave not only the title and page numbers of the text from which the code was mostly lifted, but pointed out the logic flaws in the more or less original contributions. The project continued without the Piledhigher.Deeper, and the self-taught individual not only persisted, but he rose to an even higher group of challenges and performed admirably.

My experience belies that assertion quite handily. I don't mean to imply that I think that those with degrees are inherently unreliable, or anything even remotely similar.

I won't get into my thoughts on grade inflation, diploma factories, etc, but your example showed my point to be true -- the guy didn't leave. Now, he didn't produce anything of value, but he didn't cut and run.

What he did do, was over promise and under deliver, which I already commented on. Personality is just as important as skill when it comes to a job -- is the person going to work towards the goals of the project? How do they handle setbacks? Do they communicate when they're running into problems and/or ask for help?

This is why it's better to get someone who admits to their failings, and is willing to ask for help. The person who thinks they know everything will assume that there's nothing more to learn, and will make no attempt at improving.

There was a question on Slashdot yesterday, with someone complaining about Behavioural interviews during job applications. They're important, be it interviews, or checking references, etc. (and for references, there are certain questions you can't ask, so start with the easiest one -- would you hire this person back?)

...your example showed my point to be true -- the guy didn't leave.

Respectfully, I assert that my example does not prove your point to be true, and I apologize for not providing adequate details to avoid the appearance that it might. The point isn't proven by my example because "The project continued without..." doesn't address the manner of his leaving, and unless the implied assumption that I abruptly terminated his contract holds true there's no basis for proof. The missing details that might have avoided the appearance of proof: After three weeks the (only) code he (ever) delivered was less than 30 lines, most inaccurately transcribed from a common text, with a light frobbing. It contained three logical errors, all within the frobnication, and five lexical errors, three within the inaccurate transcription and two within the frobbed portion. He presumably expended no more than a half hour's effort -- and copied code samples from a $20 book are not something a reasonably intelligent person pays Ph.D. wages to acquire, anyway. After conducting the contractually mandated code review in which he refused to participate we respectfully asked him to rework the code to substantially but not necessarily perfectly meet the clearly written specification, and to ensure that it would not invoke die() when run with the strict pragma used. He refused, stating that "his" work was "... a simple, elegant, robust solution". He might not have left in the most strict meaning of the word, but his presence was only to the extent of displacing air. You might, as I do, consider that jumping ship without getting wet. Yes, I terminated his contract because he willingly breached it, but not until offers of compromise and other diplomatic efforts failed.

I came away from the experience thinking that the guy was suffering from some kind of a mental or emotional disorder, to so thoroughly self-sabotage while refusing numerous and respectful offers of assistance including the reassignment of some of his responsibility. He interviewed brilliantly. He impressed me in the thinking aloud portion of the interview, and particularly in the realm of the problem domain that kicked him in the head later. I was tickled to have hired such a brilliant guy and eager to integrate him into the group. It remains a mystery to me that he didn't even attempt to implement the solution at which he'd arrived during the interview, even after I explicitly asked him to try his hand at it well before he delivered the borrowed then busted code. Beware the genius incompetent.

Of course, one example does not constitute a trend, and I wholeheartedly agree with you that it's the person and not the degree that matters. It just happens that I stand by my assertion that "It's relatively easy to obtain a degree during four or more years spent living on Daddy's generosity, as contrasted with gaining substantially the same knowledge of one's chosen field, within or without academia, while coping with the struggles of the real world."

Thanks for an interesting bit of discussion. It seems that we don't disagree by very much.

Or that you've learned to be cuthroat enough to know how to game the system. I know a lot of people who copied their way though university. :-(

a week later, I had gotten 3 job leads, even though I wasn't actually looking for work

It's frustratingly easy to find work when you don't want it. It's often much harder to find it when you do. I call it "The World Hates Me" theorem. ;-)

My advice to anyone looking for a career change -- don't oversell yourself. If you promise, and can't deliver, you're screwed.

*shrug* Honest people always tell the truth, and nobody wants liars to prosper. Either way, reminding people to tell the truth so they don't get caught just helps the jerks. Let the lying twits find out the hard way! I hate people who say they can do a job when they can't, especially when they poach jobs at honest people's expense. They deserve to be fired (or fired upon!). (Who me, bitter? Yup, still bitter. Yup, still cleaning up messes in production by overzealous people who couldn't pull off what they promised. :-( )

My advice to the OP is simpler: "programming" doesn't have to be hard. "Programming" can be exactly hard as you want it to be. In some sense, "programming" is just another word for "configuration". I am, in a very fragile (and unconventional) sense of the word, "programming" this message into the PerlMonks system right now; that is, I'm setting it up to be displayed when I come back to it, and to let other people view it. It's so simple that no one will pay me to do it; but in general, if you know how to to configure something that someone else is unable or unwilling to do, you can usually get paid to do it. Heck, people will pay you to program their VCRs for them!

Programming can be as simple as taking a big, long command that you type in and creating a short cut for it; or as hard as building a control system for a space shuttle.

Most of the time, programming just involves automating something that you already know how to do. Ocasionally, it involves thinking out the steps to do something non-obvious (and then, typically, publishing the results so that no one *else* wastes their time thinking as hard as you did); in very rare cases, it involves extremely hard research into the world's great unsolved problems.

Depending on the job you want, you can get paid to re-install someone's Window's drivers on a dozen computers across a network (perhaps by writing a quick Perl script to help), or you can work on control systems that make automated subway trains run. It's a big continuum; you don't have to know it all right away.

Good Luck!

--
Ytrew