A friend of Eastern European origin once shared this phrase with me, and I love it.
I'm currently trying to turn shit into ice cream in the form of my next chapter/paper. It's the least interesting part of my of my Ph.D. work. The data were collected mainly to provide baseline information for another study, but it took so much effort that I wanted to get something more from it (remember the counter I had going on my sidebar a while back? it's that project). Unfortunately, some of the findings contradict one of the main conclusions from the first chapter/paper, and I think it's because the design of this second study kind of sucks. However, I can't just say that, nor can I present explicitly contradictory data. For one thing, I don't want to completely undermine the ongoing study before I even collect the data from it.
So, what to do? The design has specific weaknesses that are quite common in my field. But honestly, I think that the natural phenomenon I am trying to describe interacts with those weaknesses in a way that enhances them. When others have tried to address the same question, they sometimes find result A, occasionally find result B, and sometimes find no difference between A and B. I think I need to take a look at the experimental design that generated each result to see if there are trends. Perhaps result A is only possible to detect if you have the proper design? Maybe all the studies that find B or no difference use this common, yet crappy design? I suppose that would be sort of interesting.
9 comments:
Yes, it would - and probably rack up the citations too, which is no bad thing. Go for it!
I was just going to point out what JaneB said. That it would be pretty interesting and explain why your field keeps getting different results. Pretty nifty in its own right (at least I'd think so).
Thanks for the positive reinforcement! Maybe today I'll be able to pull finger and make it happen (I've been procrastinating on this chapter/paper because I didn't know how to handle the discussion).
Ugh, I have the same problem with my dissertation -my results go against Big Theory and I think it's because the design of my project (designed by my DCMs who rejected my original design) sucks. In fact, it doesn't address Major Methodological Weakness, even though my original design did. So now I have to explain these results and "you designed a shit study" probably won't fly!
I agree that it sounds like a good idea. However, depending on how much politics is involved in the design of those studies, you might have a hard time with referees (or at least that's what happens in my field). Good luck!
I've never heard that saying before. I love it! Unfortunately it also reminds me of the current manuscript I am working on from the second chapter of my dissertation. It doesn't contradict previous data. it's just BORING as hell and 50% negative data. Good luck teasing out your problem! Does your PI have any thoughts or helpful suggestions?
Did you develop a new technique out of any of this work? Perhaps just turn it into a technical paper.
TJ - it's not a new technique. Both designs are completely observational, but one is much better than the other. The problem is that there are very few studies capable of the good design. My first chapter has it, but my second does not.
MXX - I haven't talked to my PIs about dealing with the paper this way. We've discussed the contradictory data, but so far not any solutions to the problem. I need to finish reviewing the literature to determine if my approach will work. Also, I think negative data is better than contradictory data!
Amelie - I don't think that will be an issue, since with these studies you either have the good design or you don't. There's not a lot for people to take personally, and the bad design is all we have in most cases, so it's not going anywhere.
PG - Suck that you got stuck with a bad design that was not even your idea! Mine was my own doing, at least.
I love the saying!
Good luck with sorting this.
Post a Comment