Statistical Methods for Psychology

Exercises 287

These data do not really reveal the pattern that we would expect. What do they show?

(In Chapter 15 we will see that the expected pattern actually is there if we control for other

variables.)

9.23 In the study by Katz, Lautenschlager, Blackburn, and Harris (1990) used in this chapter and
in Exercises 7.13 and 7.29, we saw that students who were answering reading comprehen-
sion questions on the SAT without first reading the passages performed at better-than-
chance levels. This does not necessarily mean that the SAT is not a useful test. Katz et al.
went on to calculate the correlation between the actual SAT Verbal scores on their partici-
pants’ admissions applications and performance on the 100-item test. For those participants
who had read the passage, the correlation was .68 (N 5 17). For those who had not read the
passage, the correlation was .53 (N 5 28), as we have seen.
a. Were these correlations significantly different?
b. What would you conclude from these data?

9.24 Katz et al. replicated their experiment using subjects whose SAT Verbal scores showed con-
siderably more within-group variance than those in the first study. In this case the correla-
tion for the group that read the passage was .88 (N 5 52), whereas for the nonreading group
it was .72 (N 5 74). Were these correlations significantly different?

9.25 What conclusions can you draw from the difference between the correlations in Exercises
9.23 and 9.24?

R

.453a

Model

1

R Square

.205

Adjusted

R Square

.188

Std. Error

of the

Estimate

65.49

Model Summaryb

aPredictors: (Constant), Current expenditure per pupil—1994–95

bDependent Variable: Average combined SAT 1994–95

SPSS

Sum of

Squares

50920.767

197303.0

248223.8

Model

1 Regression

Residual

Total

df

1

46

47

Mean

Square

50920.767

4289.197

F

11.872

Sig.

.001a

ANOVAb

aPredictors: (Constant), Current expenditure per pupil—1994–95

bDependent Variable: Average combined SAT 1994–95

Coefficientsa

Model

1 (Constant)

Current expenditure

per pupil—1994–95

B

1112.769

2 23.918

Std. Error

42.341

6.942

Beta

2 .453

t

26.281

2 3.446

Sig.

.000

.001

Unstandardized

Coefficients

Standardized

Coefficients

aDependent Variable: Average combined SAT 1994–1995