# Suggested answers to exercise: Cough during the day or at night and history of bronchitis, 3

Question 3: Now make cough the row variable and history of bronchitis the column variable. (This was how the table was presented in the lecture.) Carry out the chi-squared test of association and calculate the odds ratio and the relative risk. How do the results change?

We choose Analyze, Descriptive Statistics, Crosstabs. Put the variable 'Bronchitis by age 5' into Row(s): and 'Day or night cough at age 12' into Column(s):. Click Statistics, Chisquare, Continue. Click OK.

You should get the case processing summary and then the following:

Day or night cough at age 12 * Bronchitis by age 5 Crosstabulation
Count
Bronchitis by age 5 Total
Yes No
Day or night cough
at age 12
Yes 26 44 70
No 247 1002 1249
Total 273 1046 1319

Chi-Square Tests
Value   df   Asymp. Sig.
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 12.180a 1 .000
Continuity Correctionb 11.145   1 .001
Likelihood Ratio 10.609   1 .001
Fisher's Exact Test .001 .001
Linear-by-Linear Association 12.171   1 .000
N of Valid Cases 1319
a 0 cells (.0%) have expected count less than 5. The minimum expected count is 14.49
b Computed only for a 2x2 table

Risk Estimate
Value   95% Confidence Interval
Lower Upper
Odds Ratio for Day or
night cough at age
12 (Yes / No)
2.397 1.448 3.970
For cohort Bronchitis
by age 5 = Yes
1.878 1.358 2.598
For cohort Bronchitis
by age 5 = No
.784 .653 .940
N of Valid Cases 1319

There are the following differences:

1. In the cross-tabulation, the rows have become the columns and the columns have become the rows, as we requested.
2. The odds ratio row in the Risk Estimate table has a different label, but the numbers are the same as before.
3. The second and third rows in the Risk Estimate table have different labels and different numbers in them.
The chi-squared and other tests of significance are unchanged. The odds ratio and its confidence interval are unchanged.