At this point, Coca-Cola announced that Pepsi’s taste test was biased and unfair. Coca-Cola
pointed out that variables other than taste were affecting volunteers’ choices. One extraneous
variable is that people have a natural preference for the letter M over the letter Q. As a result, the
preference for product M could be based on taste or could be a subconscious preference for the
letter.
In extensive testing, when people were asked to pick either Q or M, 78 percent chose M and 22
percent preferred Q. When people were asked to choose a number from 1 to 4, 70 percent chose
2 or 3, and only 30 percent chose 1 or 4. How do your class results compare?
Part 2
Before Coke introduced its reformulated “New Coke” in 1985, it conducted almost 200,000
blind taste tests with consumers. The following are the results.
New Coke (55 percent) chosen over original Coke (45 percent)
New Coke (52 percent) chosen over Pepsi (48 percent)
However, after New Coke was introduced, it failed miserably in the market. The original formula
was reintroduced a few months later as “Coca-Cola Classic.”
You can replicate the taste test comparing Coke Zero, Coca-Cola Classic, and Pepsi in the
following manner:
1. Get 40 small paper cups, and label 10 with the letter R, 10 with S, 10 with T, and 10 with
W.
2. Outside the room, have a student volunteer randomly assign Coke Zero, Coca-Cola
Classic, and Pepsi to the letters R, S, and T. Write down which soft drink goes with which
letter.
3. At the start of the class, select 10 students as taste testers. The subjects should be regular
consumers of non-diet cola (at least six 12-ounce bottles in the last month). Place the
students at the front of the classroom.
4. Outside the room, the student volunteer should be filling each cup with the appropriate
soda. Fill the W cups with water.
5. Put an R, S, T, and W cup in front of each student, and hand each student a copy of the
Cola Taste Test Form provided.
6. To eliminate order bias, have three of the students begin the taste test with cup R, three
with cup S, and four with cup T. Have them take a sip of water between colas and continue
to sample and test in any order they wish. They can resample as needed to fill out the
questionnaire.
7. Have a student tabulate the answers during class, and share the results at the end of class.
The form could even lend itself to cross-tabulations (between preferences and answers to