Example use of the SAS VAMS/FSR macro. %FSR(ind = NCAA, KT = 19, fsr = 0.05, amax = 0.3, B= 500) The REG Procedure Model: MODEL1 Dependent Variable: y avg_grad Number of Observations Read 97 Number of Observations Used 94 Number of Observations with Missing Values 3 Summary of Forward Selection Variable Number Partial Model Step Entered Label Vars In R-Square R-Square C(p) F Value Pr > F 1 x2 act25 1 0.7069 0.7069 60.9413 221.89 <.0001 2 x3 oncampus 2 0.0470 0.7539 38.7296 17.39 <.0001 3 x5 size 3 0.0169 0.7708 32.0193 6.64 0.0116 4 x4 ft_grad 4 0.0193 0.7901 24.0946 8.17 0.0053 5 x7 bbindex 5 0.0208 0.8110 15.3596 9.70 0.0025 6 x17 accept 6 0.0087 0.8197 12.8709 4.20 0.0433 7 x15 pop 7 0.0077 0.8274 10.8834 3.86 0.0527 8 x6 tateach 8 0.0053 0.8327 10.1715 2.68 0.1056 9 x9 board 9 0.0059 0.8386 9.1181 3.09 0.0826 10 x8 tuition 10 0.0071 0.8457 7.4498 3.83 0.0536 11 x12 sf_ratio 11 0.0026 0.8484 8.0866 1.43 0.2350 12 x10 attend 12 0.0021 0.8505 8.9921 1.15 0.2864 13 x13 white 13 0.0019 0.8524 10.0258 1.02 0.3163 14 x18 l_pct 14 0.0023 0.8546 10.8552 1.24 0.2697 15 x11 full_sal 15 0.0009 0.8555 12.4075 0.47 0.4953 16 x1 top10 16 0.0004 0.8559 14.1856 0.23 0.6326 17 x14 ast_sal 17 0.0003 0.8562 16.0456 0.14 0.7056 18 x19 outstate 18 0.0001 0.8563 18.0149 0.03 0.8605 19 x16 phd 19 0.0000 0.8563 20.0000 0.01 0.9032 The Selection Results By the FSR Method Method Based on Estimating rRE ALPHA_RE KI_RE VARIABLE_SELECT 0.044 6 x2 x3 x5 x4 x7 x17 Method Based on Estimating rER ALPHA_ER KI_ER VARIABLE_SELECT 0.036 5 x2 x3 x5 x4 x7 Now a second run with B=5000: %FSR(ind = NCAA, KT = 19, fsr = 0.05, amax = 0.3, B= 5000) The Selection Results By the FSR Method Method Based on Estimating rRE ALPHA_RE KI_RE VARIABLE_SELECT 0.038 5 x2 x3 x5 x4 x7 Method Based on Estimating rER ALPHA_ER KI_ER VARIABLE_SELECT 0.034 5 x2 x3 x5 x4 x7