kirideth: (Computer Science)
[personal profile] kirideth
What do you do if you spend an extended period of time working on your thesis, and right near the end come to the conclusion that it can't be done? I mean, the work would still be there and still be significant, but the overall conclusion would be that it's a failed idea, at least with this specific approach.

Because right now, I'm absolutely terrified that'll happen to me.

Date: 2009-05-09 01:42 pm (UTC)
From: [identity profile] yashakizu.livejournal.com
This happened to me. I agonized over whether I had developed my research plan well enough to get the data I needed and oh, I could have done this better and should have done this better... But when I actually started to put it all together in my paper I realized it was good data and I'd done well. Could it be anxiety about the paper? Are you sure it's a failed idea with your approach? *hugs* Take a bit of time away from the paper to see if your perspective might change on it.

What is your thesis on?

Date: 2009-05-09 02:08 pm (UTC)
From: [identity profile] kirideth.livejournal.com
I haven't even proposed it yet, so at this point it's fear and paranoia. But the current idea is developing an architecture unit for a microprocessor that is optimized for making genetic algorithms run faster.

Date: 2009-05-09 03:32 pm (UTC)
From: [identity profile] yashakizu.livejournal.com
Would it help to run that by another engineer? I could always run it by my dad...your topic is right up his alley. I don't want that to come across as patronizing my asking another engineer (because you are one intelligent engineer) but I found that when developing my thesis topic that discussing my options with an experienced teacher helped to make sure I wasn't waaaaaay off basis.

I'm using my new Spock icon for you since I know you'll appreciate it. :)

Date: 2009-05-09 10:45 pm (UTC)
From: [identity profile] kirideth.livejournal.com
It might. I ran it by Dr. Hudson, but he still hasn't replied about it. The suspense! It is getting to me! I've asked a few friends to read it over, but none of them are engineers, so the consensus has been "it sounds like you think you can do it" which is important, but not the only deciding factor in getting it accepted.

Mmm.... Spock makes everything better.

Date: 2009-05-13 07:24 pm (UTC)
From: [identity profile] yashakizu.livejournal.com
You might even want to run your idea past a mathematician because of the algorithms. At any rate, I hope it works out! :)

Date: 2009-05-09 07:17 pm (UTC)
luinied: I have only seen an aurora once in real life. (worried)
From: [personal profile] luinied
I don't know, but I'm pretty scared of this possibility myself.

I think your chances of getting by with a negative result thesis are much better than getting negative results into a journal. But, really, it would be something that, if it happens, you just have to bring up to your committee, and they'll let you know what sort of thing you should do.

Date: 2009-05-09 10:48 pm (UTC)
From: [identity profile] kirideth.livejournal.com
You're right, it will be ultimately up to the committee. A committee which doesn't exist yet, unfortunately. I ran the initial idea by one of the professors I wanted on my committee because he's The FPGA Guy, but he tried to steer me towards doing something else because he thought it sounded like too much. Maybe the formal proposal will go over better. Also, I'd like your opinion on the idea from a software perspective, if you'd be willing to help. The project itself is, of course, hardware, but closely tied to software.

Date: 2009-05-09 07:39 pm (UTC)
From: [identity profile] the-red-baron.livejournal.com
I have not actually started grad school yet, but I'm accepted and everything and will start in the fall, though it's in a drastically different field, obviously.

But...I'm not sure failed results, in themselves, need to be a problem? I mean I understand the anxiety & disappointment, but of course negative results are still advancing knowledge. Would it be feasible to rewrite the paper to reflect that? Like, here's this theory I have tested, the approach failed for the following reasons, here's what we learned from that, here's where I suggest research should go in the future based on this? Basically as long as you can draw really significant conclusions from the failure, and emphasize that and make them sound important, it sounds to me like you've still got a perfectly workable thesis, though of course maybe I don't know enough about how your corner of academia works.

What does your advisor say? Or is your advisor useful for questions like this?

Date: 2009-05-09 10:56 pm (UTC)
From: [identity profile] kirideth.livejournal.com
Karl's right. It's up to my committee, ultimately. And while it sounds plausible to present a failed idea in a positive way, I'm worried that due to the results-driven nature of engineering that it might not be enough. Though perhaps writing the paper detailing what it *can* do, even if it doesn't do what I thought it would, could possibly be used as a stepping stone for someone else to use the ideas in a way I hadn't thought of. I mean... bipolar junction transistors were discovered through a failed attempt to invent MOSFETs.

As for my advisor, he's been on emergency leave for the past year, and no one knows when or if he's coming back. (So I'm currently working with the department head, who's very busy, and it's a bit of a frustrating and suspenseful experience.)

Profile

kirideth: (Default)
kirideth

December 2011

S M T W T F S
    123
45678910
11121314151617
18 19 2021222324
25262728293031

Expand Cut Tags

No cut tags