Statistical Methods for Psychology

270 Chapter 9 Correlation and Regression

SAT Verbal Score

Test Score

SAT Verbal Score

Test Score

SAT Verbal Score

Test Score

Pearson Correlation

Sig. (1-tailed)

N

SAT. Verbal

Score

1.000

.532

.

.002

28

28

Test Score

.532

1.000

.002

.

28

28

Correlations

R

.532a

Model

1

R Square

.283

Adjusted

R Square

.255

Std. Error

of the

Estimate

53.13

Model Summary

Sum of

Squares

28940.123

73402.734

102342.9

Model

1 Regression

Residual

Total

df

1

26

27

Mean

Square

28940.123

2823.182

F

10.251

Sig.

.004a

ANOVAb

aPredictors: (Constant), Test score

bDependent Variable: SAT Verbal Score

aPredictors: (Constant), Test score

Coefficientsa

Model

1 (Constant)

Test score

B

373.736

4.865

Std. Error

70.938

1.520

Beta

.532

t

5.269

3.202

Sig.

.000

.004

Unstandardized

Coefficients

Standardized

Coefficients

aDependent Variable: SAT Verbal Score

Exhibit 9.1 (continued)

as the dependent variable, even though it was actually taken prior to the experiment. The

first two panels of Exhibit 9.1 illustrate the menu selections required for SPSS. The means

and standard deviations are found in the middle of the output, and you can see that we are

dealing with a group that has high achievement scores (the mean is almost 600, with a stan-

dard deviation of about 60. This puts them about 100 points above the average for the SAT.

They also do quite well on Katz’s test, getting nearly 50% of the items correct. Below these

statistics you see the correlation between Score and SATV, which is .532. We will test this

correlation for significance in a moment.

In the section labeled Model Summary you see both Rand R^2. The “R” here is capital-

ized because if there were multiple predictors it would be a multiple correlation, and we