#15: Suppose that we conduct a survey and ask a random sample of students to report their daily amount of exercise and report whether or not they have a dog. Suppose that we find the result to be significant: that there is a significant difference in the average minutes of daily exercise between the two groups. p-value = 0.003. There are three possible explanations:

1. Having a dog causes students to get more exercise.
2. Those who have a dog and those who do not have a dog have different amounts of daily exercise to begin with (groups differed at baseline)
3. Random chance.

Can we rule out explanation #2?