Question

Hi,

In the multiple regression output I got R-squared 0.52387 or 52.4% and adjusted R-squared 0.2858 (28.6%).

Is the regression model good or bad, can it be used to forecast the dependent variable?

Answer 1

Solution:

The given values of R-squared derived for the multiple regression are too low. A relatively strong or good R-squared value is considered to be at least 70% or 0.70, for academic regressions (for research purpose, any value above 50% is also considered to be good enough). So, the regression model is not good.

With such low value for adjusted R-squared, there is a possibility that your regression model includes some of the explanatory variables which are an extra burden, meaning they do not explain the dependent variable well and thus, must be eliminated (which would not decrease R-squared value by much, but increase adjusted R-squared). No, it's not good enough, or better, strong enough model to build forecasts about the dependent variable.