What is the difference between the datasets OUTFILE1 and OUTFILE_2?

Question

Options

• AOUTFILE_1 contains the final parameter estimates while OUTFILE_2 contains the newly scored probabilities.
• BOUTFILE_1 contains the model goodness of fit statistics while OUTFILE_2 contains the newly scored probabilities
• COUTFILE_1 contains the model goodness of fit statistics while OUTFILE_2 contains the newly scored logits.
• DOUTFILE_1 contains the final parameter estimates and Wald Chi-Square values while OUTFILE_2 contains the newly scored probabilities.

• A
  87% (41)
• B
  2% (1)
• C
  6% (3)
• D
  4% (2)

Explanation

Option A is correct because in statistical modeling procedures (such as SAS PROC LOGISTIC), OUTFILE_1 is generated by the OUTEST= option, which stores the final parameter estimates (regression coefficients) from the fitted model, while OUTFILE_2 is produced by the SCORE statement's OUT= option, which applies those estimates to data and outputs predicted (scored) probabilities.

Options B and C are wrong because goodness-of-fit statistics (e.g., AIC, -2 Log L) are printed in the procedure output log/tables, not stored in OUTFILE_1 - that dataset is specifically dedicated to parameter estimates. Option C adds a second error by claiming OUTFILE_2 contains logits; the scored output dataset stores probabilities (P), not raw logits.

Option D is a near-miss trap: while Wald Chi-Square values do appear in the printed results alongside parameter estimates, they are not stored in the OUTEST= dataset - only the estimates themselves are. Mixing in Wald values makes D inaccurate.

Memory tip: Think "1 = the model itself (estimates), 2 = what the model does (scores new data)." OUTFILE_1 captures what you built; OUTFILE_2 captures what it predicts.

No community discussion yet for this question.