search

forc-db / Global_Productivity

Creative Commons Attribution 4.0 International
2 stars 0 forks source link

refine use of term "linear" #90

Closed teixeirak closed 3 years ago

teixeirak commented 4 years ago

from BBL: careful about terminology here. In statistics a "linear model" includes polynomial models, which are linear in their parameters.

bpbond commented 4 years ago

https://blog.minitab.com/blog/adventures-in-statistics-2/what-is-the-difference-between-linear-and-nonlinear-equations-in-regression-analysis

beckybanbury commented 3 years ago

@teixeirak what do we want to do about this? Can we still use linear and polynomial to refer to the various models in the tables etc, and just clarify in the methods section that we are referring to a linear model with a polynomial term?

teixeirak commented 3 years ago

Good question. The different uses/definitions of "linear" vs "nonlinear" in ecology/statistics/math leave something to be desired... "First-order linear" is not a concise term, so it would be annoying to use that throughout. I've edited the text in the methods to read:

We tested models with latitude as a first-order linear term (corresponding null: model without latitude) and as a second-order polynomial (corresponding null: model with latitude as a first-order linear term). For brevity, we henceforth refer to first-order linear models as "linear" and second-order polynomial models as "polynomial.

@bpbond , please let us know if this doesn't adequately address your concern.

bpbond commented 3 years ago

Perfect I think. Thanks!