Would I fail if, for example, during the discussion a mistake in my comparison and evaluation approach is discovered which makes my results meaningless? Although I discussed my approach and everything I did with my advisor (postdoc not the prof), yet I still fear that I might have done something wrong and then everything collaps.
This should not happen unless:
- The mistake is blatantly obvious and something you should have really known better. And with blatantly obvious, I mean something that somebody who has not even studied computer science or a related field could easily spot. Even then you may have chances, if the rest of your work is appropriate – almost everybody brainfarts now and then.
- You made not only one mistake but a lot of big mistakes.
- It becomes obvious that you deliberately ignored that mistake, to avoid being stuck or to get “nicer” results.
(In the first two cases, your advisor is in big trouble, too.) The point of the master thesis is that you should demonstrate that you can investigate a scientific question under supervision (or something similar – check the regulations, if you wish to know). One mistake does not change this.
Moreover, in some examination regulations I am aware of, there is a procedure for the case that some important but localised flaw is detected in your thesis. For example, you could be given a month to amend your thesis.
How would one fail a master thesis defense?
I have not experienced or heard of such a case but from what I have gathered, you pass if you:
- Give a talk about your thesis.
- Be able to answer questions about it.
- Are not detected to have been cheating.
Even if you suffer from a mental breakdown due to nervousness or similar, you probably can repeat the defense – at least with any reasonable examination regulation and examination committee. I am aware of one case where somebody had a nervous breakdown for understandable reasons¹ and the examination board let him repeat the exam. If you look into your examination regulations, there are probably some clauses that allow the examination board to do some things at their own discretion in exceptional cases.
Be aware though that being very difficult to fail does not make the defense unimportant. A bad defense may seriously (and in particular more than nominally) affect your thesis degree, which in turn has a huge impact on your total degree. In particular having passed a thesis with the lowest possible degree (or something close to it) is something you definetely do not want to have in your vita².
¹ It’s very complicated, but you might compare it to the following: In the middle of the defense, somebody who was the defender’s girlfriend until six months ago enters the room being obviously pregnant in her ninth month.
² Unless you are in one of those few disciplines where this is the norm.