The Cola/Pepsi Challenge

Cover Sheet

 

Author  Adapted by Tanya Sharon from Solomon, P.R. (1979). Science and television commercials: Adding relevance to the research methodology course. Teaching of Psychology, 6, 26-30.

 

Case Purpose and Relation to Course Objectives This case gives students practice in designing a solid, valid experiment in a context that many may not have thought to apply science, namely people’s preferences for one product over another.  In this way it also shows the diversity of scientific inquiry- science applied to human psychology for practical business purposes.

 

Case Materials

Paper cups and 1 large bottle each of Pepsi and Coke, labels hidden or removed (if removed, make sure to mark the bottle so you can tell them apart).

 

 

Case Notes

This case gives students practice in designing a solid, valid experiment in a context that many may not have thought to apply science, namely people’s preferences for one product over another.  In this way it also shows the diversity of scientific inquiry- science applied to human psychology for practical business purposes.

 

Explain to students that you have a burning question today: which is better, Coke or Pepsi? If they jump right in with their opinions, point out that ‘better’ is ill-defined.  ‘Better’ could mean better for removing rust from nuts and bolts that are rusted together (a well-known trick among mechanics.) Clarify you mean ‘better tasting’. You want to find out which is really the best-tasting cola and need their help.  If they suggest simply taking a poll, remind them that you are a scientist.  People may only think they like a particular cola. Once you like one cola, you tend not to retry the other! You’d like empirical evidence.  The students in class today will be your sample. Pull out with 2 identical cups and fill each with a small amount of cola from your 2 different bottles.  Each student will try both- but first, you need to design your test. Challenge them to design a clean, robust taste test, one that rules out confounds and extraneous variables. They will probably assume this is easy. Show them it’s not so simple- here are some issues you can raise:

 

• Which should be sampled first? Which second? Should one always be first or should the order change?
• Does it matter if one is the diet version and the other is not?
• How can you assure that the 2 colas are presented at the same temperature? Does it matter what temperature you use? (e.g., cold versus room temp)
• How can you assure that features of the colas other than taste (eg, color) don’t influence ratings? Should tasters be blindfolded? (a true ‘single-blind’ experiment)
• Should tasters rinse their mouths between tastings? If so, with what?
• How can you assure that all participants are starting the taste test under the same conditions? Anyone who has just had a meal or another drink may have residual tastes in their mouths.
• Are the ratings to be made right after each cola has been sampled or after both have been?
• How will the ratings be made? A simple preference? Ratings on a scale from 1-5?
• Should you track participants’ age? (taste buds become less sensitive with age)
• Should participants’ pre-existing preferences be publicly recorded?

 

Once the major flaws have been addressed, conduct the test according to the agreed upon protocol. Tally the data for Mystery Cola 1 versus Mystery Cola 2, then reveal their true names. It could also be interesting to tally the number of students whose preference in the blind test was different from their previously-identified preference.  In fact, market research by both Pepsi and Coke apparently shows (see the first web link below) that in blind tests people generally prefer Pepsi.  Depending on what your data show, and students’ initially reported preference, this can lead into a discussion of the influence of marketing and the accuracy of our perceptions.  

 

Links:

http://www.snopes.com/cokelore/newcoke.asp

 

http://www.nationmaster.com/encyclopedia/Pepsi-Cola