Specification Searches (sometimes called "data mining.") * * * * M U L T I P L E R E G R E S S I O N * * * * 1. Omitted Variable Bias Listwise Deletion of Missing Data Equation Number 1 Dependent Variable.. EXPEND2 Vacation expenditures Block Number 1. Method: Enter AGE COLLEGE Variable(s) Entered on Step Number 1.. COLLEGE College Degree=1, no degree=0 2.. AGE Age Multiple R .21145 R Square .04471 Adjusted R Square .04384 Standard Error 1124.57713 Analysis of Variance DF Sum of Squares Mean Square Regression 2 129573637.84757 64786818.92378 Residual 2189 2768370763.41109 1264673.71558 F = 51.22809 Signif F = .0000 ------------------ Variables in the Equation ------------------ Variable B SE B Beta T Sig T AGE -8.385996 1.435329 -.122058 -5.843 .0000 COLLEGE 439.304004 53.519100 .171483 8.208 .0000 (Constant) 909.378923 72.081438 12.616 .0000 End Block Number 1 All requested variables entered.        * * * * M U L T I P L E R E G R E S S I O N * * * * Listwise Deletion of Missing Data Equation Number 1 Dependent Variable.. EXPEND2 Vacation expenditures Block Number 1. Method: Enter AGE COLLEGE INCOME Variable(s) Entered on Step Number 1.. INCOME Income (in thousands) 2.. AGE Age 3.. COLLEGE College Degree=1, no degree=0 Multiple R .32431 R Square .10517 Adjusted R Square .10385 Standard Error 1110.94553 Analysis of Variance DF Sum of Squares Mean Square Regression 3 293318729.81919 97772909.93973 Residual 2022 2495552352.57420 1234199.97655 F = 79.21967 Signif F = .0000 ------------------ Variables in the Equation ------------------ Variable B SE B Beta T Sig T AGE -6.439265 1.508539 -.090337 -4.269 .0000 COLLEGE 193.264522 59.216745 .073910 3.264 .0011 INCOME 13.874906 1.177372 .268278 11.785 .0000 (Constant) 415.131811 85.914202 4.832 .0000 End Block Number 1 All requested variables entered. 2. Irrelevant Variable Bias Drop your insignificant variables if they are not required by theory. If they are required by theory, leave them in the model or your model might be improperly specified resulting in bias in some of the other coefficients. Also, be careful with irrelevant variables. When trying to construct an example, my irrelevant variable (zip code/household size) was significant! 3. Functional Form Equation Number 1 Dependent Variable.. EXPEND2 Vacation expenditures Block Number 1. Method: Enter AGE INCOME Variable(s) Entered on Step Number 1.. INCOME Income (in thousands) 2.. AGE Age Multiple R .31737 R Square .10072 Adjusted R Square .09984 Standard Error 1112.74412 Analysis of Variance DF Sum of Squares Mean Square Regression 2 281112911.82728 140556455.91364 Residual 2027 2509830361.17223 1238199.48750 F = 113.51681 Signif F = .0000 ------------------ Variables in the Equation ------------------ Variable B SE B Beta T Sig T AGE -6.323956 1.505841 -.088923 -4.200 .0000 INCOME 15.296814 1.095356 .295698 13.965 .0000 (Constant) 413.060737 85.910876 4.808 .0000 End Block Number 1 All requested variables entered. Listwise Deletion of Missing Data Equation Number 1 Dependent Variable.. EXPEND2 Vacation expenditures Block Number 1. Method: Enter LNAGE LNINC Variable(s) Entered on Step Number 1.. LNINC 2.. LNAGE Multiple R .31231 R Square .09754 Adjusted R Square .09665 Standard Error 1114.71317 Analysis of Variance DF Sum of Squares Mean Square Regression 2 272222555.57246 136111277.78623 Residual 2027 2518720717.42704 1242585.45507 F = 109.53877 Signif F = .0000 ------------------ Variables in the Equation ------------------ Variable B SE B Beta T Sig T LNAGE -224.746247 64.169877 -.074327 -3.502 .0005 LNINC 463.893975 33.315943 .295495 13.924 .0000 (Constant) -33.934013 276.590815 -.123 .9024