options nodate nonumber ps=2000; proc glmselect data=ncaa; model y=x1-x19/selection=backward select=sl sls=1E-40 stats=all ; run; Note that BIC on this SAS output is not the usual BIC. SBC is a version of the usual BIC. The GLMSELECT Procedure Data Set WORK.NCAA Dependent Variable y Selection Method Backward Select Criterion Significance Level Stop Criterion Significance Level Stay Significance Level (SLS) 1E-40 Effect Hierarchy Enforced None Number of Observations Read 97 Number of Observations Used 94 Dimensions Number of Effects 20 Number of Parameters 20 The SAS System The GLMSELECT Procedure Backward Selection Summary Effect Number Model Adjusted Step Removed Effects In R-Square R-Square AIC AICC BIC 0 20 0.8563 0.8194 379.2090 5.1919 391.8737 ---------------------------------------------------------------------------------------------- 1 x16 19 0.8563 0.8218 377.2279 5.1568 389.3426 2 x19 18 0.8562 0.8241 375.2669 5.1227 386.8232 3 x14 17 0.8559 0.8260 373.4445 5.0911 384.3829 4 x1 16 0.8555 0.8277 371.7252 5.0615 382.0126 5 x11 15 0.8546 0.8289* 370.2891 5.0357 379.8310 6 x15 14 0.8524 0.8284 369.7381 5.0201 378.2496 7 x10 13 0.8500 0.8278 369.2505 5.0060 376.7936 8 x12 12 0.8472 0.8267 368.9709 4.9949 375.5689 9 x13 11 0.8454 0.8267 368.1121 4.9783 374.0143 10 x18 10 0.8426 0.8257 367.7764* 4.9680 372.9178 11 x17 9 0.8389 0.8237 367.9855 4.9642* 372.3386* 12 x6 8 0.8328 0.8192 369.4457 4.9743 372.8968 13 x8 7 0.8166 0.8040 376.1334 5.0407 378.0684 14 x9 6 0.8110 0.8002 376.9975 5.0457 378.5811 15 x7 5 0.7901 0.7807 384.8295 5.1255 385.4916 16 x4 4 0.7708 0.7632 391.0866 5.1890 391.3756 17 x5 3 0.7539 0.7485 395.7801 5.2365 396.0971 18 x3 2 0.7069 0.7037 410.2146 5.3881 410.4210 19 x2 1 -.0000 -.0000 523.5776 6.5927 522.6062 * Optimal Value Of Criterion Backward Selection Summary Effect Step Removed CP SBC PRESS ASE F Value Pr > F 0 20.0000 430.0749 5797.4541 36.9146 ---------------------------------------------------------------------------------------------- 1 x16 18.0149 425.5505 5539.8313 36.9220 0.01 0.9032 2 x19 16.0456 421.0462 5314.5682 36.9373 0.03 0.8605 3 x14 14.1856 416.6805 5232.3751 37.0072 0.14 0.7056 4 x1 12.4075 412.4179 5102.1400 37.1179 0.23 0.6326 5 x11 10.8552 408.4385 4970.9485 37.3412 0.47 0.4953 6 x15 10.0181 405.3443 5003.4292 37.9213 1.23 0.2713 7 x10 9.2510 402.3133 4969.7148 38.5363 1.30 0.2581 8 x12 8.6779 399.4904 4940.0967 39.2481 1.50 0.2248 9 x13 7.6389 396.0883 4925.2443 39.7275 1.00 0.3199 10 x18 7.0615 393.2094 4892.7222 40.4372 1.48 0.2268 11 x17 6.9891* 390.8752 4878.9538* 41.3987 2.00 0.1613 12 x6 8.1009 389.7921* 4890.1510 42.9510 3.19 0.0778 13 x8 14.4378 393.9364 5177.6485 47.1099 8.33 0.0049 14 x9 15.3596 392.2573 5160.3685 48.5674 2.69 0.1045 15 x7 24.0946 397.5460 5608.3748 53.9225 9.70 0.0025 16 x4 32.0193 401.2598 5953.5209 58.8734 8.17 0.0053 17 x5 38.7296 403.4100 6304.7106 63.2185 6.64 0.0116 18 x3 60.9413 415.3012 7356.7828 75.2964 17.39 <.0001 19 x2 422.9949 526.1209 24670.9955 256.9029 221.89 <.0001 * Optimal Value Of Criterion