Refer to the exhibit:
SAS output from the RSQUARE selection method, within the REG procedure, is shown. The top two models in each subset are given.
Based on the exhibit, which statement is true?
A00-240 Exam QuestionsBrowse all questions from this exam
A00-240 Exam - Question 18
Show AnswerHide Answer
Correct Answer: B
In model selection, the SBC (Schwarz Bayesian Criterion, also known as BIC) is typically more concerned with parsimony than the AIC (Akaike Information Criterion). Parsimony refers to the simplicity of the model, in this context meaning fewer predictor variables. While the AIC is mainly focused on the goodness of fit, it does not penalize additional parameters as heavily as the SBC does. Therefore, the SBC champion model is generally more parsimonious than the AIC champion model because it tends to favor models with fewer parameters.
Discussion
2 commentsjevanthonyOption: B
Aug 5, 2020Why is it SBC instead of AIC?
MrMister
Sep 7, 2020I believe because AIC is not concerned with the number of parameters, or predictor variables (parsimony) whereas SBC *is* concerned about parsimony. It's pretty common for a champion model to have both the lowest SBC value AND the lowest AIC value. But, regardless, SBC is concerned with having the least number of parameters, or predictor variables, as needed to explain a model.
UnorichOption: B
Oct 8, 2020Also, SBC champ has 6 models and AIC has 8.