GOF measures

#1
When we validate a statistical model, should we only consider the traditional Goodness-Of-Fit measures?

Are there other ways in which the model should be validated?

Should the validation be application specific or purely statistical?

Do we have any examples that we can share with the class?