Running head: COOKIE PREFERENCE BASED ON TASTE-TESTS

Abstract

Due to the number of brands and various fat contents available when purchasing cookies, an experiment was run to see if these two variables affected actual cookie preference. Thirty Lewis and Clark College students and faculty members were involved in the experiment. A double- blind, counterbalanced study was conducted involving three cookies: Nilla Wafers, Reduced Fat Nilla Wafers, and Safeway Brand Vanilla Wafers. A non-parametric Friedman Test was used to interpret the data. Though the mean ranking for each type of cookie differed, no statistical significance was found, p=.5308. This finding shows that the participants did not prefer any one cookie over the other two.

Effects of Brand and Fat Content on Cookie Preference

With the numerous brands of cookies available and the various fat contents of each, a question is raised as to whether or not these variables matter when looking at cookie preference. Previous research done in the area of cola tasting has shown that specific brands of colas can be identified based on taste when the participants are given the names of colas that they will be testing (Thumin, 1962). However this study focused on identification of different colas, rather than on cola preference. The present study introduced the participants with three cookies, and asked for a ranking based on preference. Thus, without knowing which brands were being tested, nor the fat content of each, the experimenters and the participants could see if preference deviated from the regular, name-brand cookie. Our hypothesis was that there would be no preference for the name-brand cookie over the generic brand cookie, both of regular fat content; there would however be a preference for these two brands of cookies over the reduced fat type of the name-brand cookie. This paper was written in accordance with the Publication Manual of the American Psychological Association (1994).

Methods

Thirty Lewis and Clark College students and faculty members participated individually in the experiment. Participants entering or leaving the Watzek Library were asked if they would like to participate in a taste-test experiment.

A typed sheet of instructions was given to each participant, along with a small cup of water and a cardboard tray containing three cookies. The three cookie types were "Nilla Wafers," "Reduced Fat Nilla Wafers," and "Safeway Brand Vanilla Wafers." A black arrow was drawn on the left (pointing right) to indicate the order which the cookies were to be eaten. The bottom of each tray was marked with one of six possible combinations of the numbers one, two and three (counterbalanced.) The order of the numbers on the bottom of the trays indicated the order in which that participants were to have their cookies arranged in the tray. A black box was drawn above each cookie (three in total) providing space for the participant to rank his/her cookie preference.

Each participant signed a consent form, and was given a typed instruction sheet , a cardboard tray containing three cookies, and a cup of water. The instructions were as follows:

Please taste each cookie from left to right. Between each trial please rinse your mouth with water. After eating all three cookies please mark the box above your favorite cookie as "best," the box above your least favorite cookie as "worst," and the box above your middle preference as "ok." This means you need to remember where each cookie was. Thank you for your participation.

One experimenter decided which cookie type would represent each number (one, two, and three,) and she remained the only person who knew this information until all tests were run (double-blind.) A second experimenter gave the participants the water cups, while the other experimenters handed trays to the taste-testers. The participants ate their cookies in order from left to right according to the numbers on the bottom of their tray, and rinsed their mouths out with water in between each cookie. In order to reduce errors due to the sequence in which each participant ate each cookie, counterbalancing was used, making six groups of different combinations, with five participants in each. After all three cookies were eaten, the participants ranked their preference of the three cookies as follows: their favorite was to be labeled "best," their second choice as "okay," and their least favorite as "worst." The experimenters then took the labeled trays and another experimenter recorded the order of preference with regards to the order indicated on the bottom of the tray. The results were then entered in a "Statview" program by participants that did not know which cookie represented which number. The actual experiment lasted approximately one hour.

Results

The results obtained through the use of a non-parametric Friedman Test indicated that no preference for any one cookie type was found. The participants responses of "best," "ok," and "worst" were converted into the numbers "1," "2," and "3" for ease of interpretation. The rank for the Nilla Wafers (M= 2.167), the rank for the reduced fat Nilla Wafers (M=1.933), and the rank for the Safeway Brand Vanilla Wafers (M=1.900) was not statistically significant, as seen by x2(2, N=30) =1.267, p = .5308.

Discussion

Participants did not show any preference for any one cookie more than the others. This would imply that they could not tell the difference between the name-brand cookie, the reduced fat name-brand cookie, or the generic brand cookie. These findings go against our original hypothesis, as no preference over the reduced fat brand was found.

After the experiment was finished, the experimenters discussed the fact that there were some confounding variables that need to be remedied if another similar study were to be conducted. For example, since all of the participants were tested in the same room, there individual comments and responses could have been heard by other participants. These comments could have some effect on the responses of others. Also, many participants disliked this particular type of cookie altogether, which could have a remarkable effect on their rankings. In addition, many participants mentioned that all three cookies tasted the same, or that two of the cookies must have been the same brand. Despite the experimenters reassurance that no deception was involved, some participants may have ranked the cookies randomly, as they thought that they were all the same nonetheless.

In order to address these confounding variables in a future replication of the study, participants could be tested in individual rooms or booths. Also, a direct statement to the participant before the cookies are given could be included, stating clearly that no deception will be used, and that all three cookies are of different types.

The experimenters found it interesting that the mean ranks for the Safeway Brand cookies was higher than for the other two, which poses a number of interesting questions. First, why then is there such a price difference between the name-brand cookies and the generic cookies if participants didn’t prefer one over the other? Second, what would the results be if the participants knew that they would be tasting one generic brand and one name-brand? Third, what sort of results would be obtained if the experimenters asked for the identification of the three types instead of preference? Would the participants be able to identify the name-brand, the reduced fat, and the generic brand? These are questions that cannot be answered by the present study, but may have interesting implications if studied.

References

Publication Manual of the American Psychological Association (4th ed.). (1994). Washington, D.C.: American Psychological Association.

Thumin, F. J. (1962). Identification of Cola Beverages. Journal of Applied Psychology, 46(5), 358-360.