The first chapter of my dissertation involves this very cool dataset that is comprised of data from archived samples going back twenty years. I have been working with these data for the last four years and still have not finalized the story. There is always one more thing to check, one more group of samples that turn up and need to be analyzed, or more way to look at the results. It’s so frustrating. I thought I finally had it all sorted out, but there was a glaring ambiguity that was somewhat disquieting. However, I thought there was nothing I could do about it so I was ready to start working on the manuscript. Then another set of samples were discovered, so I analyzed them and got the data back today. Guess what. The f’ing ambiguity still exists. The new numbers didn’t help much at all. I wouldn’t care so much (the data are the data), but there is one really important conclusion that I can either make or not make. I need to be convinced that the data support one story over the other before I can persuade anyone else. This dataset eventually will be included in a meta-analysis or synthesis of some kind (hopefully) which will then influence policy (hopefully). I really want to feel confident in my conclusion because I think this is an important study.
I will talk to my advisor about it on Monday. That will help me feel secure about however we decide to discuss the results, but she’s not going to offer some new idea to solve the problem. We’re out of options.
Phooey.
1 comment:
I totally know how you feel. It's so frustrating when the data are too contradictory or ambiguous to draw conclusions. If I wanted ambiguous, I could just play with a Magic 8-Ball instead of analyzing data!
Post a Comment