I've previously described the process of scientific data analysis like making a pot. I've equated it with walking two steps forward, one step back, and one step sideways. As you know, I'm currently in the middle of a data analysis on larval behavior. But this time around, the data analysis doesn't feel like pottery or a crazy walk. It feels like carving a tree trunk with a chainsaw.
|These graphs are about 1/3 of what I started with. Each of |
them show valuable information about our larvae. For the
record, though, this is far from the largest number of graphs
I've ever analyzed at once.
Then the carving began.
Sometimes, it was a deep, major cut with a giant chainsaw. Other times, it was just woodchips flying away. We eliminated an entire experiment (one of four) because its results were not trustworthy. We eliminated potentially erroneous data and metrics that had meaningless results. We have shaved and sliced and sanded away at the data until finally, a carving began to emerge.
I once heard sculpture described as freeing an artwork from the prison of its material, and I really like that image. I'm finally at a point with the larvae data where I can see the end product, the scientific work of art, being freed from the superfluous surrounding mess. I'm excited to continue shaping the analysis, carving away as I go.
With a chainsaw.