Question

You want to study the effect of exercise on appetite. You have a subject pool of 10,000 students, 5500 of which choose to be in the treatment group where they exercise for one hour each day. The other 4500 subjects choose to be in the control group which never exercises. You find that over a month, the people in the treatment group consume an average of 1900 calories per day. Those in the control group consume 2035 calories per day. You find that the difference is statistically significant at the 1% level.

	a.	We cannot conclude anything about the impact of exercise on appetite from this study because there are more people in the treatment group.
	b.	We cannot conclude anything about the impact of exercise on appetite from this study because the treatment was not randomly assigned.
	c.	You conclude that the exercise is an appetite suppressant on average and ceteris paribus.
	d.	You conclude that the exercise is an appetite suppressant.
	e.	We cannot conclude anything about the impact of exercise on appetite from this study because the subject pool is students.

Answer 1

Option b is correct because the treatment is not randomly assigned, there can be selection bias in this case that is the students who chose to excercise might be significantly different from the students who chose not to excercise, the possible factors which are systematically different between these two pool of students can be health status, height, nutrition, weight and therefore the results that is the significant difference which we got might be because of these factors and not because of excercise. Therefore the results can not be solely attributed to excercise.

b.	We cannot conclude anything about the impact of exercise on appetite from this study because the treatment was not randomly assigned.