Premium
When are two pieces better than one: fitting and testing OLS and RMA regressions
Author(s) -
Friedman Jane,
Bohonak Andrew J.,
Levine Richard A.
Publication year - 2013
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/env.2213
Subject(s) - ordinary least squares , linear regression , mathematics , simple linear regression , phase (matter) , regression , least squares function approximation , statistics , econometrics , estimator , physics , quantum mechanics
This paper describes a new method of choosing between a simple linear and a two‐phase linear reduced major axis regression model. For continuous two‐phase linear models, differential evolution is used to estimate the parameters. The method can be used with either ordinary least squares or reduced major axis regression. There are no other methods in the literature for fitting two‐phase reduced major axis regressions. Our new method is applied to problems in zebrafish kinetics bedload transport and a classic linear model of cricket chirping. Copyright © 2013 John Wiley & Sons, Ltd.