MSc defence tips?
November 11, 2004 6:46 PM   Subscribe

I'm giving my MSc defence on Monday -- any tips on how to structure the presentation and answer an hour's worth of questions from the committee? [more inside]

Presentation is 20 minutes long, using Powerpoint (blech). I mainly want to know how much of my written dissertation I should rehash, and what to do if I get caught on a tough question (apparently saying "I don't know" is the best bet). The field is Human-Computer Interaction. Cheers!
posted by krunk to Education (11 answers total)
 
For nasty questions: make sure that you understand what they ask. Don't be afraid to repeat/ask what the want. Depending on the question - do they want facts or do they want you to speculate or elaborate, etc. - start by laying down some basic facts, which should buy you some time to think the question over. At least, that got me through my 3-hour M.Sc. and my Ph.D. defense. Bear in mind that they may ask question to which they themselves don't know the answers, just to see how you manage. Mind you, that's how they do it in Holland.

For the presentation: who will be the main audience ? Is it just the committee (who hopefully have read your thesis) or is it for parents, friends, etc. ? 20 minutes is not a lot, so choose the target audience: elaborate "light" introduction with the main conclusions or in-depth analysis, skipping the trivia. How did your fellow students approach their presentation ?

Anyway, Good luck!
posted by swordfishtrombones at 7:08 PM on November 11, 2004


I'd rehash the most interesting things to the audience of the talk, as swordfishtrombones mentions. So, if the main audience is a technical one, go into the technical details. You're in charge of the talk portion, so you get to pick the bits that are intersting. I would not dwell on the process too much ("First I thought that people were like this, but then I did x and discovered y, so I went in a different direction. But that wasn't right either...") and instead show your results / design / gadget / whatever in the context of their use. The best way that I've found to do this is to set up a scenario and quickly get to a problem that needs solving. And then show your work as a (partial) solution to that problem.

For the questions, definitely say you don't konw if you don't know something. I'd also recommend a "Here's what I might do to answer that question... approach for some hard hypothetical questions. (For example: "I'd design a latin square experiement to determine which design is best for x, and then I'd try to implement the system in such a ways as to y...").
posted by zpousman at 8:02 PM on November 11, 2004


It's an open defence, meaning that anyone who is interested can watch, but can't ask questions. Only the defence committee (5 professors, including one external to my department) gets to torture me.

I'm just worried that the presentation will be too similar to the written dissertation, which I'm sure they will all read for the first time in the hallway on the way to the defence. Actually, maybe it's a good idea to repeat it all!
posted by krunk at 8:29 PM on November 11, 2004


I'm not sure how relevant this is to your experience, but in my master's exam (in math) I answered one particularly tricky question with, essentially, "I'm afraid there's not enough time for me to go into that," which was met with both approval and a smile. (And I suspect it was correctly interpreted as "I don't know, and I don't want to waste your time speculating.") But I had a friendly committee, and only an hour for both my presentation (which was lengthy) and questions, so, you know, grain of salt and all.
posted by evinrude at 8:38 PM on November 11, 2004


If you do get a question you don't know the answer to, don't fake it. Someone on your committee may smell blood and escalate the issue. It's best to just admit that you don't know, and point out that the question is indeed a truly excellent one.
posted by shoos at 8:41 PM on November 11, 2004


Try substituting 'that's outside of the scope of the study' for some of the 'I don't knows', its most appropriate to do this in relation to conclusions and project parameters rather than underlying theory.

Include any deadends you went down in your research as part of your presentation, ensuring you mention why you concluded/decided they were deadends and you ended up choosing the route you did eventually take.

Naturally you will have practiced the presentation a number of times, including under times conditions and with someone else present (at least to keep you honest, at best to give informed feedback).

It is entirely possible that not everyone on the panel will have read the dissertation fully. It is even possible that no-one on the panel will have read the dissertation fully.

I'd disagree with zpousman to some extent, don't try and make things too simple to cover the whole audience. Aim everything at your 5 professors.

Actually, maybe it's a good idea to repeat it all!

That's what you're there for. One (you might even say the) key element of a defence is to show that it's your work and that you understand how it was done, how the conclusions developed and that you understand the work.
posted by biffa at 4:33 AM on November 12, 2004


Also include lots of powerpoint animations and sound effects so that your committee will get so annoyed that they will beg you to quickly wrap up your defense.
posted by shoos at 4:43 AM on November 12, 2004


note to self: remember this for next May.
posted by ruelle at 4:58 AM on November 12, 2004


Have you talked to your committee about this sort of thing? You really really should ask someone on it.

I mean, unless things are either *very* different in Canada, or very different in your field, you shouldn't expect an oral defense to be a particularly big deal. You're going to walk in with a pass, and you're going to have to actually work to fuck up badly enough to be failed.

If there were any nontrivial problems with your thesis, it's your committee's job to tell you ahead of time. You should expect them to be asking hard questions, but from one of two places. First, they might be asking to honestly clarify something, or because there's some small issue that you're going to have to take care of with a normal revision and they're leading you there. Second, because they want to see what you do with it.

Unless your committee has told you otherwise, ignore any non-committee audience, and do not plan anything for them. Anyone who's not part of the decision about your degree is irrelevant.

Assuming things are vaguely similar to US poli-sci:

Have you been to any job-talks? That's what you're aiming for.

Your presentation is going to be very similar to your thesis. You're presenting your thesis research.

A normal presentation might look like:

(1) Here's a problem in the lit, or similar
(2) Here's what I'm doing
(3) Here's why you should give a shit
(now you're 5 minutes in)
(4) Here's my question again
(5) Methods
(now you're 10 minutes in, unless you're using especially advanced methods)
(6) Results -- pick the most salient points to highlight, have slides ready for *ALL* of your results in another powerpoint file or as overheads to deal with questions, discuss what in the hell it all means, and again why someone who's not you would give a shit.
posted by ROU_Xenophobe at 7:21 AM on November 12, 2004


you may be surprised how little some of the people on the committee know. there's probably going to be one person who asks the most stupid introductory questions. on the other extreme, you'll probably also have someone who knows more than you. so it's not easy :o)

remember that, unless you're really borderline, this makes no difference at all. they've already maybe up their minds.

20 mins is very little. how about 1/3 general intro, 1/3 focussing on one special application/topic in more detail, and 1/3 on conclusions? stress that the middle third is just an example of what you've done, to fit in the time limits.

also, while what i said above about things being ok is generally true, do think about any politics. is there anyone on the committee you've argued with or annoyed? if so, have some good slides on hand to argue with (don't show them unless needed). a relation of mine got absolutely screwed by her own supervisor a while back. it was awful, but she could have seen it coming, and perhaps helped herself a little.

anyway, ignore that and good luck ;o)
[oops, found this in a tab written ages ago - apologies if now repeating things others have said]
posted by andrew cooke at 7:27 AM on November 12, 2004


Include any deadends you went down in your research as part of your presentation, ensuring you mention why you concluded/decided they were deadends and you ended up choosing the route you did eventually take.

I don't agree with this particular piece of advice for a 20 minute talk. 20 minutes is quick; you'll barely have time to summarize your work. I agree with andrew cooke that the best way to go is to focus on one topic in great detail, but to make it clear that it's only a portion of the work. This also gives you an advantage in the questioning: you might be able to redirect some of the questions to other aspects of your work, which you can then talk about knowledgeably and in detail.
posted by mr_roboto at 11:54 AM on November 12, 2004


« Older How do I stop this new noise with my radiators?   |   Gifts for Girls Newer »
This thread is closed to new comments.