Question: Suppose the president of an advertising agency carried out an experiment to determine the perception of a set of ads for a grocery product. Each participant in the experiment was randomly drawn from the population and each was randomly assigned to one of three groups. Group 1 was a control group that just watched a TV show segment, then filled out a questionnaire about their perception of the product. Group 2 saw the TV show segment, including a very inexpensive – to-produce commercial, then filled out the same questionnaire. Group 3 saw the TV show segment, including a very expensive commercial, then filled out the same questionnaire as the other groups.

a. Assuming interval data, how would you instruct the researcher to analyze the data for the agency president? (What is the hypothesis and how would you test it?)

b. Suppose the main response was nominally scaled as 1 = like the commercial; 2 = not sure; 3 = can’t stand it. What would you suggest for analysis of the data collected in this type of experiment?