I don't know... how often will one be taking exams as an official PhD holder? Exams are sometimes better at measuring how one takes exams than what one knows.
Take job interviews for example. I hold a CS degree from a major university and my business is a website that so far gets 500,000 pageviews a month and is profitable. It was coded and designed by me from scratch; its server is administered and monitored by me; its community and customer support is managed by me. Given these accomplishments I would say I have a reasonable grasp of the art of web programming. But put me in a job interview situation where an interviewer asks me to write an implementation of something basic like a linked list and I'll probably manage to screw it up somehow. Then when I get sent home and I sit down in front of my monitor, I'll have it done in 5 minutes. Does my being unable to cope with programming interview situations make me a bad programmer? Does being unable to cope with a math exam make one a bad math researcher?
Edit: though on the whole I do agree that degrees nowadays are being handed out like candy and far to often to people completely undeserving of them.
And you may well be a very competent programmer, and deserve that recognition.
But the recognition of a PhD is something altogether different. It's not something that any person has a right to. And it seems to me that the only way to get one ought to be to prove that one has earned it, and deserves it. To have a shortcut that says "he could have earned it but for this disability" changes the degree from one of recognizing achievement to a (subjective) judgment of potential.
The way universities are pumping out PhDs it's not surprising that a whole bunch of requirement are simply being swept under the rug. Universities figured out a long time ago that graduate students are basically slave labor and they accept way too many students that fumble their way through the program without really accomplishing anything. So get off your high horse and put the blame where it should be placed.
My CS courses were basically all dependent on exam scores - I would estimate that on average 75% of the course grades came from exams.
In some of the courses, even as late as my 3rd and 4th years, people weren't able to complete projects that had even moderate amounts of coding, even in Java. I would much rather have seen a more 50/50 grading scale (at least in courses that it made sense for - I don't see much of a way to get around exams in theory courses) so that a good grade would require strength in both application and theory.
At my university, the CS department requires students to pass both the homework/project portions of the class and the exam portion. If you get a 95 on the final and don't do any of the projects, you fail. It's not exactly a 50/50 grading scale, but the idea is similar.
It's worth noting that the CS department is the only department at this university that applies this grading requirement- I'm a little surprised the other engineering programs haven't picked it up.
They haven't, because putting a large amount of weight on assignments leads to endemic cheating. Exams are highly weighted because they take place in a controlled environment where students can't easily copy answers or crib off the Internet.
I suspect if they ran source-level comparisons on the programming projects at your university, they would find a huge amount of copying between students.
Not huge... I was with a friend who was a marker of assignments in my university... He says that there are always cheaters when he marks the assignments, but its only 1 or two out of a hundred. He marks only one part of the assignment... so if we suppose there are 5 parts to each assignment and each assignment is copied similarly... the rate of copying is only ~5%, which isn't a "huge amount" IMO.
The most worrying is not copying but students outsourcing their assignments. I've googled coding questions before just to find them on rentacoder.com.
Just last week a student offered me money to do their assignment for them. (but of course I flatly declined.) I've heard from some first years the market price for completing that assignment was $75.
I can certainly see that concern, but actually we do source-level comparisons, and we make it very clear up front that this takes place. It happens, but not at the rate you're suggesting.
I'm the same way. Anxiety makes me draw blanks and stutter. No amount of mental effort or preparedness seems to make it go away. Funny how age changes you; I used to be pretty good at exams and tests in school.
Personally I'd hate to go through the university process and make it all the way to PhD only to choke on the exam and be denied the qualification. That's a lot of time, money, and effort lost on such an archaic practice.
I know this may not sound very helpful (I too resisted at first), but have you tried to seek specialized help? A good therapist should be able to help you deal with whatever issues you have that could be causing the condition. It's like hiring an experienced bug fixer.
Take job interviews for example. I hold a CS degree from a major university and my business is a website that so far gets 500,000 pageviews a month and is profitable. It was coded and designed by me from scratch; its server is administered and monitored by me; its community and customer support is managed by me. Given these accomplishments I would say I have a reasonable grasp of the art of web programming. But put me in a job interview situation where an interviewer asks me to write an implementation of something basic like a linked list and I'll probably manage to screw it up somehow. Then when I get sent home and I sit down in front of my monitor, I'll have it done in 5 minutes. Does my being unable to cope with programming interview situations make me a bad programmer? Does being unable to cope with a math exam make one a bad math researcher?
Edit: though on the whole I do agree that degrees nowadays are being handed out like candy and far to often to people completely undeserving of them.