In: Statistics and Probability
Experimental Study Design
Suppose that you have been hired by the university administration to design a randomized experimental study to assess the following hypothesis among undergraduate students: Attending weekly yoga classes leads to better performance on exams.
Be sure that your study design addresses the following elements:
1. Find some volunteers and divide them into two group. Each group should have almost same mean marks, s.d. in terms of marks and almost same range.
2. If there are any students who are unable to attain yoga classes for physical reasons, exclude them. Exclude the outlier (excellent academic result or very poor academic results)
3. To avoid any placebo effect, don't describe the actual goal of the study. One can say they ar testing whether yoga hepls with muscle strains. With that in mind, give two group same task.
4. One group will attain yoga and the other will not do any physical activities in that time. Make sure each group does similar thing except the yoga practice.
5. If you are collecting data over a year, collect data after each quizzes, med term and semester exams.
6. You have to assign similar diet, sleep plan to both group. Each group should be allowed to study as their own way as different study methods works differently for each group. To maximize our compliance, we should let them function at their best.
7. By banning one group from doing limited physical activities , we may diminish their physical and mental health slightly during the study.
8. Their study methods, any other factors that they think have helped/obstructed their study.
9. Take the avg increase/decrease in their marks. Let group A didi yoga and group B didn't. Find out the mean values for each group and their s.d.. Let the null hypothesis be that the mean value of group A is same as that of group B. Compare the mean value using t test for different group. If the difference turns out to be significant, we can conclude that yoga has positive effects on performance on exam.