Restricting attention to the subset of models with AIC within 6 of the ‘‘best’’ model (i.e., applying the idea of Occam’s window to the AIC), find the model-averaged regression coefficients and their...

Restricting attention to the subset of models with AIC within 6 of the ‘‘best’’ model (i.e., applying the idea of Occam’s window to the AIC), find the model-averaged regression coefficients and their estimated variances for the baseball salary regression. Burnham and Anderson indicate that model averaging based on the AIC and BIC tends to produce results that are more similar than model selection based on the two criteria. Does that hold true here? Note: Following Burnham and Anderson, include 0 coefficients (with variances of 0) in the averages.