Hypothesis: A higher proportion of women trust in the police than men.

FREQUENCIES trstplc.
RECODE trstplc (0=0) (1 thru 10=1) into trstplc_cat2.
VARIABLE LABELS trstplc_cat2 ‘Trust in the police_cat2’.
VALUE LABELS trstplc_cat2 0 ‘no trust’ 1’somehow trust’.
FREQUENCIES trstplc trstplc_cat2. *Note: It is better to recode it in the following way: 0-5 into 0, 6-10 into 1.

FREQUENCIES gndr.
RECODE gndr (1=0) (2=1) into gndr_cat2.
VARIABLE LABELS gndr_cat2 ‘gender_cat2’.
VALUE LABELS gndr_cat2 0 ‘Male’ 1 ‘Female’.
FREQUENCIES gndr gndr_cat2.

LOGISTIC REGRESSION trstplc_cat2 with gndr_cat2.

*If you want to create a graph then this is how you create the contingency table showing the percentages in the direction of the independent variable:.
CROSSTABS trstplc_cat2 by gndr_cat2 /CELLS=COLUMN.

1-somehow trust
0-no trust
0-male
1-female

The logarithm of the odds of trust in the police among women
are 0,054 times as low as among men.
The logarithm of the odds of trust in the police among men is 3.098.
The logarithm of the odds of trust in the police among women is (3.098-0.054).
The odds of trust in the police among women are 0.948 times as low as among men.
The odds of trust in the police among men are 22.161.
The odds of trust in the police among women are (22.161-0.948).

Conclusion: The results refute our hypothesis since the p-value is higher than 0,05. (If the p-value were lower than 0,05 then we should also check what does the value of the odds ratio shows us. (Exp(B) for b1)).