Students just starting out in regression analysis are often confused by the different computer output generated by different statistics programs. The truth of the matter is that the computer printouts are really more similar than they are different. Information is organized a little bit differently from one program to the next and sometimes different programs use different names for the same terms. In this unit, we will present the same regression analysis run in Stata, SAS, and SPSS. We will use Agresti and Finlay's crime dataset (with Washington, D.C. removed) from Chapter 9.
Source | SS df MS Number of obs = 50 ---------+------------------------------ F( 5, 44) = 23.58 Model | 3123844.66 5 624768.933 Prob > F = 0.0000 Residual | 1165780.56 44 26495.0127 R-squared = 0.7282 ---------+------------------------------ Adj R-squared = 0.6973 Total | 4289625.22 49 87543.3718 Root MSE = 162.77 ------------------------------------------------------------------------------ crime | Coef. Std. Err. t P>|t| Beta ---------+-------------------------------------------------------------------- pctmetro | 7.541909 1.170283 6.445 0.000 .5525032 pctwhite | -2.489705 2.579502 -0.965 0.340 -.093095 pcths | 3.300208 7.251138 0.455 0.651 .0628064 poverty | 21.28312 10.12292 2.102 0.041 .3083509 single | 80.88412 20.28576 3.987 0.000 .4032581 _cons | -1173.33 632.2908 -1.856 0.070 . ------------------------------------------------------------------------------
Model: MODEL1
Dependent Variable: CRIME
Analysis of Variance
Sum of Mean
Source DF Squares Square F Value Prob>F
Model 5 3123844.6346 624768.92693 23.581 0.0001
Error 44 1165780.5854 26495.013304
C Total 49 4289625.22
Root MSE 162.77289 R-square 0.7282
Dep Mean 566.66000 Adj R-sq 0.6973
C.V. 28.72497
Parameter Estimates
Parameter Standard T for H0:
Variable DF Estimate Error Parameter=0 Prob > |T|
INTERCEP 1 -1173.329937 632.29078909 -1.856 0.0702
PCTMETRO 1 7.541909 1.17028250 6.445 0.0001
PCTWHITE 1 -2.489704 2.57950225 -0.965 0.3397
PCTHS 1 3.300204 7.25113773 0.455 0.6513
POVERTY 1 21.283118 10.12291688 2.102 0.0413
SINGLE 1 80.884132 20.28576012 3.987 0.0002
Standardized
Variable DF Estimate
INTERCEP 1 0.00000000
PCTMETRO 1 0.55250322
PCTWHITE 1 -0.09309493
PCTHS 1 0.06280634
POVERTY 1 0.30835078
SINGLE 1 0.40325817
* * * * 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.. CRIME
Block Number 1. Method: Enter
Variable(s) Entered on Step Number 1.. SINGLE
2.. PCTMETRO
3.. PCTHS
4.. PCTWHITE
5.. POVERTY
Multiple R .85337 Analysis of Variance
R Square .72823 DF Sum of Squares Mean Square
Adjusted R Square .69735 Regression 5 3123844.63463 624768.92693
Standard Error 162.77289 Residual 44 1165780.58537 26495.01330
F = 23.58062 Signif F = .0000
------------------ Variables in the Equation ------------------
Variable B SE B Beta T Sig T
PCTMETRO 7.541909 1.170282 .552503 6.445 .0000
PCTWHITE -2.489704 2.579502 -.093095 -.965 .3397
PCTHS 3.300204 7.251138 .062806 .455 .6513
POVERTY 21.283118 10.122917 .308351 2.102 .0413
SINGLE 80.884132 20.285760 .403258 3.987 .0002
(Constant) -1173.329937 632.290789 -1.856 .0702
End Block Number 1 All requested variables entered.
Call:
lm(formula = crime ~ pctmetro + pctwhite + pcths + poverty + single)
Coefficients:
(Intercept) pctmetro pctwhite pcths poverty single
-1796.322 7.609 -4.486 8.658 26.250 109.452
summary(m1)
Call:
lm(formula = crime ~ pctmetro + pctwhite + pcths + poverty + single)
Residuals:
Min 1Q Median 3Q Max
-533.262 -87.243 -9.825 110.443 397.775
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -1796.322 668.614 -2.687 0.0101 *
pctmetro 7.609 1.295 5.875 4.79e-07 ***
pctwhite -4.486 2.777 -1.615 0.1132
pcths 8.658 7.826 1.106 0.2745
poverty 26.250 11.082 2.369 0.0222 *
single 109.452 20.354 5.377 2.59e-06 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 180.2 on 45 degrees of freedom
Multiple R-squared: 0.8499, Adjusted R-squared: 0.8332
F-statistic: 50.94 on 5 and 45 DF, p-value: < 2.2e-16