Coming across various data ranging from social to scientific fields, I have never understood what matters in deciding whether to use a given or come up with definitely new model for the given data, for instance when using Cobb-Douglas models and not just a regression model.

Similar questions and discussions