1. Chance or Real Effect?

The main question to be answered is, “In the population, is there a preference for Coke or for Pepsi?”

As a budding statistician, design an experiment that will answer this question and at the same time eliminate glass-preference as a possible explanation.

Use the repeated measures design, and an N of 20 with each subject tasting both Coke and Pepsi once in separate tastings. For this question, make up the sample data such that the results are signifi cant favoring Pepsi. Evaluate the data using the sign test with   0.052 tail. To make the analysis interesting, use sample scores such that the obtained probability exceeds alpha by the smallest amount possible with an N of 20.