Question

Please indicate which of the following statements are TRUE or FALSE. Would also appreciate a sentence explaining the choice for each.

1> Penalization in linear regression models means penalizing for complex models, that is, models with a large number of predictors.

2> For large sample sizes, Mallow’s Cp penalizes model complexity more than leave-one-out cross validation but less than the Bayesian Information Criterion.

3> At the same step in their respective algorithms, backwards stepwise selection and forwards stepwise selection will compute a criterion value for the same number of models.

4>Two highly skilled statisticians will always choose the same model after performing variable selection on the same set of variables.

Answer 1

1) Penalization in linear regression models means penalizing for complex models, that is, models with a large number of predictors.

TRUE

2) For large sample sizes, Mallow’s Cp penalizes model complexity more than leave-one-out cross validation but less than the Bayesian Information Criterion.

TRUE

3) At the same step in their respective algorithms, backwards stepwise selection and forwards stepwise selection will compute a criterion value for the same number of models

TRUE

4) Two highly skilled statisticians will always choose the same model after performing variable selection on the same set of variables.

FALSE