Theoretical examples

Example 1 Suppose we want to examine the association between gender (x) and educational level (y) by means of a simple ordinal regression analysis. Gender has the values 0=Man and 1=Woman, whereas educational level has the values 1=Low, 2=Medium, and 3=High. Now, we get an OR of 1.62. This would mean that women have higher educational attainment compared to men.

Example 2 Here we want to examine the association between having young children (x) and number of pets (y). Having young children is measured as either 0=No young children and 1=Young children. Number of pets has the values 1=No pet, 2=1-2 pets, and 3=3 or more pets. Let us say that we get an OR that is 1.29. We can hereby conclude that families with young children own more pets than families without young children.

Practical example

Dataset

StataData1.dta

Variable name	educ
Variable label	Educational level (Age 40, Year 2010)
Value labels	1=Compulsory 2=Upper secondary 3=University

Variable name	bullied
Variable label	Exposure to bullying (Age 15, Year 1985)
Value labels	0=No 1=Yes

sum educ bullied if pop_ordinal==1

ologit educ bullied if pop_ordinal==1, or

When we look at the results for bullied, we see that the odds ratio (OR) is 0.71. Put differently, a unit increase in bullied is associated with lower educational level. This means that those who were exposed to bullying are less likely to reach a higher level of educational attainment. 

The association between bullied and educ is statistically significant, as reflected in the p-value (0.000) and the 95% confidence intervals (0.63-0.82). 

Summary Those who were exposed to bullying at age 15 are less likely to reach higher levels of educational attainment at age 40 (OR=0.71; 95% CI=0.62-0.82).