Statistical Methods for Psychology

Value

95% Confidence Interval

Lower Upper

Odds Ratio for Fault

(Little / Much)

For cohort Guilt 5 Guilty

For cohort Guilt 5

NotGuilty

N of Valid Cases

4.614

1.490

.323

358

2.738

1.299

214

7.776

1.709

.486

Value Approx. Sig.

Nominal by Phi

Nominal Cramer’s V

Contingency Coefficient

N of Valid Cases

2 .097

097

096

825

.005

.005

.005

Value df

Asymp. Sig.

(2-Sided)

Exact Sig.

(2-sided)

Exact Sig.

(1-sided)

Pearson Chi-Square

Continuity Correctionb

Likelihood Ratio

Fisher’s Exact Test

Linear-by-Linear Association

N of Valid Cases

7.710a

6.978

7.358

7.701

825

1

1

1

1

.005

.008

.007

.006

.007 .005

150 Chapter 6 Categorical Data and Chi-Square

Exhibit 6.1b SPSS output on death sentence data

Exhibit 6.1c Measures of association for Unah and Boger’s data

Exhibit 6.1d Risk estimates on death sentence data

will contain a W or an NW, depending on the race of the defendant. The second column

(labeled Sentence) will contain “Yes” or “No”, depending on whether or not a death sen-

tence was assigned. Finally, there will be a third column giving the frequency associated

with each cell. (We could use numerical codes for the first two columns if we preferred, so

Chi-Square Tests

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 22.72.

b Computed only for a 2 3 2 table.

Symmetric Measures

Risk Estimate

Race of Defendant*Sentence Crosstabulation

Count

Sentence

No Ye s Total

Race of Defendant Nonwhite

White

Total

251

508

759

33

33

66

284

541

825