Obrain a list the students in your class in order from highest test average to lowest. From the list of students, using a stratified random sample, select every second student on the list and randomly assign half of them to drink 32oz. of Diet Coke before the math exam. The other half will drink nothing but water before the exam. Have each student take the exam and compare the scores between the students who drank Diet Coke and the students who drank water.
Sampling method: Stratified random sample
Independent variable: Consumption of Diet Coke
Dependent variable: Score on the math exam
Control: Group drinking nothing but water
Experimental: Group drinking 32 oz of Diet Coke
You would evaluate the results by comparing the mean score of the water drinkers' test to that of the Diet Coke drinkers. If the Diet Coke drinkers perform poorly, then we can generalize the results to the class.