Question: An experimenter wants to determine the relationship between weight change Y over a specified period and the use of a specially designed diet. The predictor variable X is a categorical variable indicating whether or not a person is on the diet. A total of 200 volunteers signed on for the study; a random selection of 100 of these were given the diet and the remaining 100 continued their usual diet.
(a) Record the experimental design.
(b) If the results of the study are to be applied to the population of all humans, what concerns do you have about how the study was conducted.
(c) It is felt that the amount of weight lost or gained also is dependent on the initial weight W of a participant. How would you propose that the experiment be altered to take this into account?