How do you know whether your estimated model is a good one or not? Explain.
To confirm whether my estimated model is good or not, i will approve whether the model meet satisfy assumptions.
By composing a plot with standardized outstanding against expected value and confirm if there is or no extreme point it will help in proving the validity of estimated model.
To confirm whether my estimated model is good or not, i will inspect possible influential point(s).
I will check data such as DF Beta to discover whether a specific point is considerably altering the regression results.
To confirm whether my estimated model is good or not, i will check essential correlations.
In case there is main independent variable, I will confirm whether it is relating to other independent before making independent variables’ interpretation. Leaving interactions unadjusted can bias estimate.
Comments
Leave a comment