Odds Ratios are useful in summarising the effect size or magnitude of a difference in a 2x 2 table.
For example
|| '''Group''' ||<38%>'''Sober'''||<45%>'''Drunk'''||
||<17%>Complied||<38%>106||<45%>9||
||<17%>Refused||<38%>59||<45%>56||
Odds Ratio = (106/9) / (59/56) = (106x56) / (9x59) = 11.18
The compliers were 11 times more likely to be sober than the refusers.
You can obtain a confidence interval for the odds ratio [[FAQ/oddsr|using some simple calculations]] or by clicking on risk under statistics in the crosstabs procedure in SPSS.