Confounding, Chance and Power Questions

Confounding, Chance, and Power Researchers want to study the association between children’s television watching habits and obesity.

They recruit a sample of 500 1st grade students and ask the parents to report the number of hours each day the child watches television.

At the same visit, they measure the child’s weight and height to determine if s/he is obese or not.

They are concerned that the child’s diet might be a confounder of the association between television watching and obesity, so they also ask the parents a number of questions about the child’s diet, including consumption of soda.

They obtain the following results: Overall Sample ≥2 hours/day of TV <2 hours/day of TV Obese 115 63 Not obese 120 202 500 Kids Who Drink Soda ≥2 hours/day of TV <2 hours/day of TV Obese 60 33 Not obese 60 102 255 Kids Who Do Not Drink Soda ≥2 hours/day of TV <2 hours/day of TV Obese 55 30 Not obese 60 100 245

1. What type of study design is this?

2. What is the odds ratio for television and obesity among all children? Interpret this result.

3. What is the odds ratio for television and obesity among children who consume soda? Interpret this result.

4. What is the odds ratio for television and obesity among children who do not consume soda? Interpret this result.

5. Is soda consumption a confounder of the association between television watching and obesity? Why or why not?