So I got to thinking…
It doesn’t seem right to me that the number of trees I planted each day was unrelated to the difficulty of the land (using price/tree as a proxy for land difficulty). I think I was missing a factor in the last post. I think the factor was time. Two things changed over the course of the season. First, tree prices went up:
But, my daily income increased as well:
As these things were happening, the number of trees I was actually planting each day was pretty random:
(there is a minor hump trend here, but it’s not much, quadratic fit yields p=0.01 but R?=0.22.)
So is it possible that tree prices really were based on the difficulty of the land? Yes, if: the land got increasingly difficult for some reason, and if: my skill as a planter was going up at a matched rate because of practice through the course of the season. If that’s true, then my original hypothesis that the number of trees planted per day would be related to tree price might not be exactly wrong, it’s just untestable with this data because of the obscuring factor of variability in my planting skill.
You’d think I’d be able to remember wether or not the land got harder through the season. Frankly though, it’s all mostly a blur. Land always looks hard at the start of the year, even if it’s cream. Clear cuts are scary until you get used to them. And I suppose they do naturally get greener as the summer wears on.
I think there might be some additional insight to be had from the numbers using covariate analysis (theta coefficients? zeta coefficients?) but that might be too much effort.