![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Data Analysis Doldrums
Jun. 11th, 2013 05:59 pmI've been stuck in the Data Analysis Doldrums for weeks, now.
I've completed three major cricket experiments here, but I've collected some of the same information across all three experiments, and so I've been trying to assess these things across the three experiments (e.g. how much food the crickets ate across the 13 different diets and how that affected mass gain).
The trouble is, I'm like that person with two watches, who never knows what time it is, because for all of the times when the results are consistent across the three experiments, there are many times where the results aren't, and I don't have a decent way of explaining why not, other than because they were completed at different times (meaning that a bunch of different little things that were out of my control could have happened).
This is only a serious problem inasmuch as I am trying to say something about cricket feeding patterns, "in general." If I wish to make the declaration that there is no "in general," I need to provide fairly exhaustive (exhausting) evidence as to why not (basically, playing whack-a-mole against a bunch of little "whatif" goblins).
As I think things over and try to summarize what's going on, I need to toggle back and forth between twenty-two (and counting) spreadsheet pages.
It's hard to concentrate on this while simultaneously starting up another experiment and managing undergraduates. Time to wedge myself back in the "cave," my third-floor office-hallway (why do I keep winding up in "offices" that are actually entryways to other spaces?). I need to also remind myself that it's okay to struggle with this because it's new, and I *am* occasionally finding some interesting gems amongst the rubbish.
I've completed three major cricket experiments here, but I've collected some of the same information across all three experiments, and so I've been trying to assess these things across the three experiments (e.g. how much food the crickets ate across the 13 different diets and how that affected mass gain).
The trouble is, I'm like that person with two watches, who never knows what time it is, because for all of the times when the results are consistent across the three experiments, there are many times where the results aren't, and I don't have a decent way of explaining why not, other than because they were completed at different times (meaning that a bunch of different little things that were out of my control could have happened).
This is only a serious problem inasmuch as I am trying to say something about cricket feeding patterns, "in general." If I wish to make the declaration that there is no "in general," I need to provide fairly exhaustive (exhausting) evidence as to why not (basically, playing whack-a-mole against a bunch of little "whatif" goblins).
As I think things over and try to summarize what's going on, I need to toggle back and forth between twenty-two (and counting) spreadsheet pages.
It's hard to concentrate on this while simultaneously starting up another experiment and managing undergraduates. Time to wedge myself back in the "cave," my third-floor office-hallway (why do I keep winding up in "offices" that are actually entryways to other spaces?). I need to also remind myself that it's okay to struggle with this because it's new, and I *am* occasionally finding some interesting gems amongst the rubbish.
