use http://www.philender.com/courses/data/gpa, clear sw regress gpa greq grev mat ar, pe(.1) pr(.15) forward begin with empty model p = 0.0003 < 0.1000 adding ar p = 0.0120 < 0.1000 adding grev p = 0.0756 < 0.1000 adding mat p = 0.0385 < 0.1000 adding greq p = 0.2135 >= 0.1500 removing ar Source | SS df MS Number of obs = 30 ---------+------------------------------ F( 3, 26) = 13.96 Model | 6.43865846 3 2.14621949 Prob > F = 0.0000 Residual | 3.99600819 26 .153692623 R-squared = 0.6170 ---------+------------------------------ Adj R-squared = 0.5729 Total | 10.4346666 29 .359816091 Root MSE = .39204 ------------------------------------------------------------------------------ gpa | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------- greq | .0049255 .0017006 2.896 0.008 .00143 .0084211 grev | .0016122 .0010605 1.520 0.141 -.0005676 .003792 mat | .0261191 .0087314 2.991 0.006 .0081716 .0440667 _cons | -2.14877 .9054056 -2.373 0.025 -4.009858 -.2876821 ------------------------------------------------------------------------------ sw regress gpa greq grev mat ar, pe(.05) pr(.1) forward begin with empty model p = 0.0003 < 0.0500 adding ar p = 0.0120 < 0.0500 adding grev Source | SS df MS Number of obs = 30 ---------+------------------------------ F( 2, 27) = 14.36 Model | 5.3789825 2 2.68949125 Prob > F = 0.0001 Residual | 5.05568415 27 .187247561 R-squared = 0.5155 ---------+------------------------------ Adj R-squared = 0.4796 Total | 10.4346666 29 .359816091 Root MSE = .43272 ------------------------------------------------------------------------------ gpa | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------- ar | .329625 .104835 3.144 0.004 .1145215 .5447286 grev | .0028514 .0010586 2.694 0.012 .0006794 .0050234 _cons | .497178 .5765156 0.862 0.396 -.6857342 1.68009 ------------------------------------------------------------------------------ sw regress gpa greq grev mat ar, pe(.05) pr(.1) /* note: there is no forward option */ begin with full model p = 0.2135 >= 0.1000 removing ar p = 0.1405 >= 0.1000 removing grev Source | SS df MS Number of obs = 30 -------------+------------------------------ F( 2, 27) = 18.87 Model | 6.08344211 2 3.04172105 Prob > F = 0.0000 Residual | 4.35122454 27 .161156464 R-squared = 0.5830 -------------+------------------------------ Adj R-squared = 0.5521 Total | 10.4346666 29 .359816091 Root MSE = .40144 ------------------------------------------------------------------------------ gpa | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- greq | .0059763 .001591 3.76 0.001 .0027118 .0092408 mat | .0308074 .0083646 3.68 0.001 .0136446 .0479702 _cons | -2.129377 .9270377 -2.30 0.030 -4.031501 -.2272527 ------------------------------------------------------------------------------