11

I am writing my dissertation, and things came out somewhat different to what I had thought three years ago. Back then, I had three hypotheses. Two turned out to be true. The third, last and flimsiest, looks false. The question is: how should I describe the third hypothesis in the short list of hypotheses?

  • I could phrase it the way I did three years ago. In the results and conclusions section, I could try refuting it.

  • I could phrase it the other way around, as if I always thought it was untrue. There a billion problems to that, the least being: how does one then describe the experimental plan? I can't really say I wanted the null hypothesis to come true? Surely different kinds of experiments would have been more useful.

  • I could leave that section out. That is madness, it's a lot of work down the drain.

  • 8
    I don't understand the question. You have proven a (reasonable) hypothesis to be false? Congratulations! That's a valuable result. Write it down so others don't have to test that hypothesis again. Don't forget to write down how you derived that hypothesis. – Roland May 30 '15 at 13:09
  • 1
    Voting to close. This is completely unanswerable except by somebody familiar with the work in question. You should be asking your advisor, not random strangers on the internet who don't even know what you're working on. – David Richerby May 30 '15 at 17:55
  • 1
    I'm voting to close this question as off-topic because questions on how to refute hypotheses are not strictly within the domain of academia.se and might be best located in a discipline specific conversation. – RoboKaren May 30 '15 at 18:28
  • 2
    I disagree: I think it's drawn some nice general answers. – jakebeal May 30 '15 at 18:54
  • As a slight tangent, it can sometimes help fill out the greater context and place of a work if you at some point (usually in an introduction) detail how the entire project got started. Frequently the best way of writing out a paper is wildly different from how it proceeded in reality--I recently had a paper which started with a clear and specific idea for the starting point, and this ended up being the last thing talked about, as a simple application, in the actual paper--, and this can make it unclear how the problem was originally approached or even found. Those things are valuable to know – zibadawa timmy May 30 '15 at 23:28
22

It is hard to give good advice on this without knowing the concrete dissertation and the expectations of your community. Hence, this is one of those famous questions where the right answer really is ask your advisor.

However, in the dark I am not sure why you would not just use Option 1 from your list - you had a (hopefully reasonable) hypothesis, you set up experiments and tested it, and found no evidence to support your hypothesis. Assuming that the hypothesis wasn't bad to begin with and the experiments were sound, why can't you just write it down like this?

| improve this answer | |
14

I agree with @xLeitix's advice (+1): use the first option, that is, leave your hypotheses untouched. There is nothing wrong with not confirming certain hypotheses - negative results are also valuable (for example, see this paper, this journal and this workshop). I would suggest to complement textual reporting with a summary table, with a minimum of three columns: hypothesis, result of hypothesis testing, statistical significance of the test. In addition, I would recommend (and this is expected in a dissertation report) to include your interpretation of all results (including the negative ones) in the Discussion section of the Results chapter.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy