Abstract
The objective of this paper is to discuss the eta coefficient and to point out some limitations and misconceptions about the coefficient.
Specifically, we will discuss the fact that; 1) the eta coefficient is a global measure of curvilinearity; 2) the eta coefficient has limited
interpretability; 3) there are a number of other curvilinear relationships that might be of more significance and of more interpretability; 4) these other curvilinear relationships do not suggest nor encourage grouping of data as does the eta coefficient; and 5) these other curvilinear relationships may tend to be more amenable to replication than is the eta coefficient.

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Copyright (c) 1970 Keith A. McNeil (Author)