Is it possible to obtain an effect size (such as correlation coefficient or some measure of goodness of fit) from a locally weighted scatterplot smoothing plot?

The reason for obtaining an effect size is to use the data in a meta analysis.

More Sylvia Seulbe Lee's questions See All
Similar questions and discussions