A SAS data scientist uses PROC HPREG with a SELECTION statement to perform variable selection. After the selection, the analyst wants to examine diagnostics for influential observations. Which PROC step should be run after PROC HPREG to examine Cook's Distance for the final selected model?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — you re-fit the final selected model in PROC REG and ask for Cook's Distance with the INFLUENCE option in the MODEL statement. PROC HPREG does the selection, but PROC REG gives you the regression diagnostics you need.
Full Explanation
PROC REG is correct because it provides extensive regression diagnostics, including Cook's Distance via the INFLUENCE option in the MODEL statement or OUTPUT COOKD=. After PROC HPREG selects variables, you re-fit the final selected model in PROC REG for diagnostics. The claim that PROC HPREG automatically outputs Cook's Distance when OUTMODEL= is specified is wrong because OUTMODEL= stores model information, not Cook's Distance. PROC GLMSELECT with DIAGNOSTICS is wrong because it handles mixed models; although it can produce diagnostic output, Cook's Distance for OLS regression is most readily available in PROC REG. PROC LOGISTIC with INFLUENCE is wrong because it applies only if the target is binary.