Solved – Correlation between fitted and observed values in linear model

Question

This is about linear model. How do I show/calculate the Correlation between the fitted values and observed values?

Answer 1

It depends on the response variable. If the response variable is continuous, try the paired T-test; if the response variable is categorical, try the chi square test in the contingency table.

However, I don't understand your purpose to calculate the correlation between fitted values and observed values. The correlation must be strong if the model you applied is kind of appropriate for the data. If you simply want to check if the model fitting is good, there are goodness-of-fit tests, like R square and lack-of-fit sum square.