Forward Stepwise Regression of Divorce Rate (per 1000 married females) (1920-1996) --------------------------------------------------------------------------------------------- log: Z:\mydocs\s209\m9divorcestep.log log type: text opened on: 24 Mar 2006, 11:42:12 . use "Z:\mydocs\s209\divorce.dta", clear . sw regress divmf unemp flfprt marumf trend brthrt brtw1544 milperk, pr(.10) pe(.05) forward > begin with empty model p = 0.0000 < 0.0500 adding flfprt p = 0.0000 < 0.0500 adding brtw1544 p = 0.0000 < 0.0500 adding marumf p = 0.0005 < 0.0500 adding brthrt Source | SS df MS Number of obs = 77 -------------+------------------------------ F( 4, 72) = 230.29 Model | 2265.455 4 566.363751 Prob > F = 0.0000 Residual | 177.070203 72 2.45930838 R-squared = 0.9275 -------------+------------------------------ Adj R-squared = 0.9235 Total | 2442.52521 76 32.1384896 Root MSE = 1.5682 ------------------------------------------------------------------------------ divmf | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- flfprt | .4627424 .0250104 18.50 0.000 .4128851 .5125996 brtw1544 | -.2222111 .0318244 -6.98 0.000 -.2856519 -.1587703 marumf | .1151048 .0205593 5.60 0.000 .0741206 .156089 brthrt | .6544097 .1792721 3.65 0.000 .2970373 1.011782 _cons | -6.346091 2.432583 -2.61 0.011 -11.19536 -1.496825 ------------------------------------------------------------------------------ . * note variables that were NOT entered by the automatic procedure . * also, this model is purely cross-sectional; dynamics of process needs to be investigated . vif Variable | VIF 1/VIF -------------+---------------------- brthrt | 16.30 0.061361 brtw1544 | 11.93 0.083855 flfprt | 2.68 0.373721 marumf | 2.25 0.444417 -------------+---------------------- Mean VIF | 8.29 . sw regress divmf unemp flfprt marumf trend brtw1544 milperk, pr(.10) pe(.05) forward begin with empty model p = 0.0000 < 0.0500 adding flfprt p = 0.0000 < 0.0500 adding brtw1544 p = 0.0000 < 0.0500 adding marumf p = 0.0018 < 0.0500 adding trend p = 0.0034 < 0.0500 adding milperk Source | SS df MS Number of obs = 77 -------------+------------------------------ F( 5, 71) = 199.74 Model | 2280.4023 5 456.080461 Prob > F = 0.0000 Residual | 162.122904 71 2.28342119 R-squared = 0.9336 -------------+------------------------------ Adj R-squared = 0.9290 Total | 2442.52521 76 32.1384896 Root MSE = 1.5111 ------------------------------------------------------------------------------ divmf | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- flfprt | .8548008 .1027604 8.32 0.000 .6499024 1.059699 brtw1544 | -.1101161 .0126572 -8.70 0.000 -.1353539 -.0848782 marumf | .1593358 .0213952 7.45 0.000 .1166751 .2019966 trend | -.217904 .0507758 -4.29 0.000 -.3191481 -.1166599 milperk | -.041197 .0135986 -3.03 0.003 -.0683118 -.0140822 _cons | -12.54132 3.117228 -4.02 0.000 -18.75689 -6.325741 ------------------------------------------------------------------------------ . * note different model obtained by just removing one variable from the list . log close log: Z:\mydocs\s209\m9divorcestep.log log type: text closed on: 24 Mar 2006, 11:48:08 ---------------------------------------------------------------------------------------------