Question

An educational researcher has designed two different methods for teaching a statistics course and would like to investigate which is more effective. He designs an experiment as follows. He will obtain a group of volunteer statistics students. Volunteers will be randomly assigned to one of two groups. Group 1 will be taught the statistics course by Anna Clark using method A and group 2 will be taught by Bob Beauman using method B. At the end of the semester, students will all take the same test. Then the scores for the two groups will be compared. Identify the flaw(s) in this experiment and suggest a way of improving the experiment.

Answer 1

There is a flaw in this experiment. The two groups are being taught by two separate teachers. So if the scores for the two groups are different, they will be different not only due to different teaching methods but also due to the different teachers, which is not accounted for.

The experiment can be improved if the variation in teachers is removed by appointing the same teacher to both the groups. Another way to improve the experiment can be repeating the experiment after alternating the teachers for both groups and then comparing the vector of 2 scores for the two groups.