Earth and Planetary Sciences
8 years ago

New study reveals best method for fitting lines through data

Journal of Evolutionary Biology
Ordinary least squares regression is indicated for studies of allometry
Joseph T. Kilmer, Rafael L. Rodrı́guez

Paper Summary

The best method for studying allometry is using ordinary least squares (OLS) regression on data with low measurement error. While reduced major axis (RMA) regression is often suggested, it has drawbacks. OLS is more reliable and easy to correct for any errors that may arise. It's better to account for any measurement error in the data rather than using RMA regression.

Related papers

Theme background image
Computer Science
a year ago

Revolutionizing Time Series Regression: Better Predictions, Smarter Decisions!

Neural Networks
Theme background image
Mathematics
5 years ago

New robust regression techniques protect against outlier influence in data analysis

Multicollinearity
Regularization Methods
Statistical Control
Theme background image
Mathematics
8 years ago

Robust estimators in linear regression provide more reliable weight estimates.

Multicollinearity
Statistical Control
Multivariate Analysis