prev next front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |19 |20 |review

    next example

    100 lung cancer cases and 100 healthy controls were asked if they drink alcohol.

lung cancer
healthy
alcohol intake
yes/ no
71/29
52/48
yes/no ratio
2.45
1.08

Suppose you feel that patients of lung cancer tend to drink alcohol and you actuarially asked 100 patients of lung cancer if they drink or not. You also asked healthy 100 subject of same age, sex and race, if they drink or not.

The results are indicated in the table
if you divide 2.45 by 1.08, you get 2.27.
This is also statistically significant.

This means lung cancer patients is 2.27 times more likely to drink than healthy people.