NFT, that's a tour de force.
Looks like a 6mm PPC ?
I don't like the idea of fitting a single curvy line that runs exactly through all the points on your graph. A regression curve that allows for some uncertainty of measurement in each groups size would be more convincing. You can get some idea of that from the way your 3 group average varies by up to 0.2" from the single measurement on the left. Looks like around CBTO 1.915 should be your sweet spot. Nevertheless, very good work and I wish I had data like that.
Bookmarks