Statistical Methods for Psychology

Coefficientsa

Unstandardized Standardized

Coefficients Coefficients

Model B Std. Error Beta t Sig.

1 (Constant) 1089.294 44.390 24.539 .000

Expend 2 20.892 7.328 2 .381 2 2.851 .006

2 (Constant) 1147.113 16.700 68.688 .000

Expend 11.130 3.264 .203 3.410 .001

LogPctSAT 2 78.205 4.471 2 1.040 2 17.491 .000

aDependent Variable: SAT

522 Chapter 15 Multiple Regression

Mean

965.92

5.90526

3.1573

SAT

Expend

LogPctSAT

Std.

Deviation

74.821

1.362807

.99495

N

50

50

50

Descriptive Statistics

R

.381a

.941b

Model

1

2

R Square

.145

.886

Adjusted

R Square

.127

.881

Std. Error

of the

Estimate

69.909

25.781

Model Summary

aPredictors: (Constant), Expend

bPredictors: (Constant), Expend, LogPctSAT

Sum of

Squares

39722.059

234585.6

274307.7

243069.3

31238.381

274307.7

Model

1 Regression

Residual

Total

2 Regression

Residual

Total

df

1

48

49

2

47

49

Mean

Square

39722.059

4887.220

121534.649

664.646

F

8.128

182.856

Sig.

.006a

.000b

ANOVA

aPredictors: (Constant), Expend

bPredictors: (constant), Expend, LogPctSAT

cDependent Variable: SAT

Exhibit 15.1 Multiple regression predicting SAT from Expend and LogPctSAT

asked SPSS to first produce the regression using just Expend as the predictor and then to

add LogPctSAT and run the regression again with both variables. I normally would not do

that (I would just run the one regression with both predictors) but it makes it easier for us

to see what is happening. I have left out some of the printout to save space.

Notice that each table has two parts—one where Expend is the sole predictor and

another where both Expend and LogPctSAT are the predictors. The first table that I want to