Discussion

 

Weaknesses:

 

No matter how well-designed a study is, it is bound to have flaws and weaknesses.  The most prevalent weakness in my study was non-response bias.  Of the 120 people that I had selected from the sample, 3 were excluded because they went to Polaris full-time.  That left the amount of surveys that actually went out to 117.  Of the 117 surveys that went out, 95 responded.  Counting the three Polaris kids that were excluded from the survey, less than 80% (actually 79.1666%) of those sampled responded to the survey.  Those 25 responses, even if they all had said they preferred Pepsi, would not have changed the outcome of the study since 63 out of the 95 people preferred Coca-Cola.  The results of the significance test might have been different, as we may have ended up failing to reject the null hypothesis.  None of the surveys that I had were voided because of writing in a response that was not on the sheet.  There was some selection bias in the study, especially because there were 23 junior girls in the original sample of the classes, but as far as the other classes (freshmen, sophomores and seniors), there was almost no bias toward one particular gender.  In a simple study such as Coke vs. Pepsi, there is almost no way to have any bias.  In addition, there are no real confounding variables in the study that are apparent.

 

Extrapolation:

 

I could extrapolate the results to about all teenagers in the United States.  I got a pretty diverse sample of teenagers in most areas.  There is no known demographic appeal for whether or not people of certain backgrounds prefer one or the other, so as such, I think that my findings could be extrapolated to all teenagers in the U.S.

 

Further Research in the Field:

 

As I had discussed earlier, Dr. Read Montague's studies in the field with his fMRI study in whether or not why people choose Coke or Pepsi is truly the most definitive study available in the field right now.  As a result, there is not too much work that needs to be done in the field, but I would like to see if one group of people has a particular appeal by income level (whether or not rich or poor people tend to prefer one type or another), gender (whether or not males or females tend to prefer one or another since my results were semi inconclusive), and by race and ethnic background.  That would determine whether or not Pepsi or Coke is more popular in different parts of the world.  Any of these studies would be valuable marketing tools for Coke and Pepsi.