Is there a simple way to assess if there is enough variation in your dependent variable, or is it best to just run the regression and asses the R-squared value?