Has anyone here proven the null hypothesis in their thesis?

[Grashupfer]

Hey all,

 

I'm wondering if anyone here has worked on a research project and has gotten data that does not prove their hypothesis.  Im my case I am studying sulfur isotope ratios in pyrite (that's right, fool's gold!) in a lake environment and seeing if it has any link to climate cycles supposedly seen in lake sediments.  I have found that there is no link between these two.  This is for my master's thesis.  My advisor tells me not to worry and that it is defendable and perhaps publishable material.  I am a little worried about my defense, which will be in maybe 6-8 weeks.  So has anybody had success defending or presenting this type of outcome for their research?  Should I be worried?

[GeoDUDE!]

You should be fine.

 

Masters work doesn't even have to be publishable. 

 

I think its an interesting work if it was common thought that there was a relationship between two things and it turns out there is not. It probably isn't the sexiest of papers, but its work that needs to be done, and thats a really good masters work imo. 

[Grashupfer]

Geodude, you are always there to help a fellow geologist! (You've posted in another topic that I made).  Nobody really knows. There have been 0 studies on this subject that I have come across.  I thought that there would be a relationship, but there is in fact none.  I'm just trying to think of possible explanations as to why there is no relationship.

[fuzzylogician]

If there was no work on this in the past, then any finding should be interesting, even if it's the null hypothesis. The more interesting question, as you say, is the 'why.' But either way, it sounds like you did enough for a Masters thesis. Depending on what you have to say about your results, I'd say there might also a publication there (though for this part I am judging by my own field, I don't know anything about yours, but people tend to be interested in the first results about something noone has looked at before, no matter what the outcome).

[happy little pill]

I'm in the humanities, so the science-related stuff is lost on me but as a general presentation suggestion, I would play up the reasons why you thought that there would be a correlation between these two things. Obviously your idea didn't come from thin air and I assume that your POI didn't think that the idea was completely beyond the scope of reason either or I assume they would have said something. If you have a good basis for talking about how you came to even think that this was a possibility and how that informed and affected your research process, then I think that that will serve as a good base for going into your defence, especially if your research method was solid. 

[TakeruK]

My Masters thesis was a null result. We ran simulations of one theory of the planet formation process and used as much physics as computationally possible. Our result was that our simulations was not able to create the solar system we see today. It's a null result but still an interesting one -- since we can't really go out and observe planetary systems being created, simulations is the best we can do, and if we can't simulate it, then the theory is either missing something, or our simulations are still missing some key physics. But if we didn't try and fail, then no one would know that this particular set did not work.

 

In fact, I would argue that most of science research is finding the null result. We find the truth by first eliminating all the things that do not work. Science progresses incrementally, one null result at a time!

[fuzzylogician]

http://www.jasnh.com/

[geographyrocks]

Like others have said, a null result is still a result. 

 

As for an explanation as to why there is no relationship, from what I understand, pyrite seems to flourish when around organisms.  Is it possible that the lake you are studying has not had a steady supply of organisms?  I don't know what kind of lake you're looking at or what time frame so I really just have guesses coming from out of the dark.  I did a quick google search and it seems that there is quite a bit of research on the biological aspect of sulfur isotopes in pyrite.  Maybe the biology of that particular lake doesn't reflect climatic shifts? 

[Grashupfer]

Like others have said, a null result is still a result. 

 

As for an explanation as to why there is no relationship, from what I understand, pyrite seems to flourish when around organisms.  Is it possible that the lake you are studying has not had a steady supply of organisms?  I don't know what kind of lake you're looking at or what time frame so I really just have guesses coming from out of the dark.  I did a quick google search and it seems that there is quite a bit of research on the biological aspect of sulfur isotopes in pyrite.  Maybe the biology of that particular lake doesn't reflect climatic shifts? 

After doing a little more work, I think I actually found a link!  I just needed to plot more lake level data.  It is a closed basin lake from the Triassic.  There is quite a bit of research, but a lot of it is lab based, and what isn't is mostly about pyrite formation in marine settings.  I think my relationship has to do with sulfur recycling occurring at low lake levels, which can have a drastic effect on isotope values of pyrites.

[juilletmercredi]

I have, indeed, failed to reject the null hypothesis. In fact, I would say most of my proposed hypotheses in my dissertation didn't work out. Actually, that was one of the first things my advisor mentioned to me - not to worry if I failed to reject the null; I could still graduate, lol. He was right; I found some statistically significant differences and some that were not, and still defended successfully and graduated.

 

The thesis isn't really about the findings; it's about the process of completing a major research project independently.

[scarvesandcardigans]

I have, indeed, failed to reject the null hypothesis. In fact, I would say most of my proposed hypotheses in my dissertation didn't work out. Actually, that was one of the first things my advisor mentioned to me - not to worry if I failed to reject the null; I could still graduate, lol. He was right; I found some statistically significant differences and some that were not, and still defended successfully and graduated.

 

The thesis isn't really about the findings; it's about the process of completing a major research project independently.

 

True. This is pretty much what my professor told me when I completed my own research. I think it even opens the doors for more research questions for you to research in the future.

[Grashupfer]

I have, indeed, failed to reject the null hypothesis. In fact, I would say most of my proposed hypotheses in my dissertation didn't work out. Actually, that was one of the first things my advisor mentioned to me - not to worry if I failed to reject the null; I could still graduate, lol. He was right; I found some statistically significant differences and some that were not, and still defended successfully and graduated.

 

The thesis isn't really about the findings; it's about the process of completing a major research project independently.

This makes me feel good.  I guess it's really about how you frame your results(?).

[Powerup McMisterpants]

I have. Chapter in my thesis and published in a decent journal. The key is to have an interesting null and make it into a good story when submitting to a journal. Just as good science as finding support for the alternative.

[Igotnothin]

http://www.jasnh.com/

 

I think I've heard of this journal. Wouldn't recommend submitting to it though. Better to submit null results to a journal in your field than a generic "negative results journal."

 

I also wanted to point out a few statistical issues that are relevant. First, note that you really mean "fail to reject the null" rather than "prove the null." There is a class of study designs known as equivalence testing or non-inferiority testing which allow you to prove that two or more groups have equal means, but I doubt this is what we're dealing with here.

 

Second, whenever you do a study and find no association you need to talk about statistical power. Ideally the study was powered to detect some pre-specified effect with 80% or 90%  probability. But if it wasn't designed carefully or if it was a secondary data analysis then it could be drastically underpowered. That would make your result less interesting and harder to publish because you just didn't have the sample size to answer the question adequately. Definitely a good idea to look into this so you can defend your findings when reviewers or committee members bring it up.