Running Head: THE WAFER COOKIE TASTE TEST

Abstract

Three types of Vanilla Wafers are tested to determine if there is a preference for regular fat and calorie content wafers, of varying brands, over reduced fat and calorie wafers. The results of a within subject, Non-Parametric Friedman test show that there is not a significant difference in the preference of the three types of wafers tested, Nabisco Nilla Wafers, Reduced Fat Nabisco Nilla Wafers, and Safeway Vanilla Wafers. The data reports that there is not a noticeable difference in preference for any of the cookies.

The Vanilla Wafer Cookie Taste Test

This study was performed in an attempt to identify a noticeable difference in taste between three different types of vanilla wafer cookies, Nabisco Nilla Wafers, Reduced Fat Nabisco Nilla Wafers, and Safeway Brand Vanilla Wafers. This study is important because although there have been taste tests performed on other cookies, it is important consumer knowledge to know if the money being spent on low-fat cookies or generic versus national brand cookies makes a difference.

Inspiration and information for this study was drawn from two sources. The study was based on a cola taste test, Identification of Cola Beverages, conducted by Thumin (1962). Although the cola experiment investigated the ability of the participants to distinguish between brands and the cookie test investigated the preference of wafers of varying types, the need to make a distinction between brands was a factor in both experiments. The APA Publication Manual (1994) has been used as a guide in the format of this paper.

The hypothesis of the experiment is that there would not be a significant difference in preference between the two different brands of cookies with the same fat and calorie content; that is, the Nabisco Nilla Wafers and the Safeway Vanilla Wafers. There would though be a significant difference in the preference of these cookies over the Reduced Fat Nilla Wafers.

Method

Thirty people were opportunistically selected from the Lewis & Clark Campus between the time of 10:00am and 11:00am on a Tuesday morning. The participants were students and employees of the college as well as any other person who happened to be on campus. There was no distinction made for age or gender of participants.

The materials for this experiment include one box each of the following three cookies: Nabisco Nilla Wafers, Reduced Fat Nabisco Nilla Wafers, and Safeway Vanilla Wafers. Brown paper and tape were used to completely cover the cookie boxes and a black felt tip pen was used to label the boxes with its appropriate number. Thirty small paper plates were used to hold the cookies for the participants and thirty small cups and bottled water were used for mouth rinsing.

In designing this experiment, we used a within subject, double blind test on 30 subjects. This ensured that each subject would give us a rank of the cookie samples so that for each subject, there would be a preference. The subjects were asked to order the cookie preference as ‘Best’, ‘Okay’, and ‘Worst’.

Before the experiment was conducted, we had to accurately set it up. One of the experimenters secretly labeled the boxes as A, B, and C. This was to ensure that both the experimenter interacting with the subjects and the subjects themselves would not know the order in which the subjects were eating the cookies.

Counterbalancing and appropriately labeling each plate with the order in which the cookies were to be eaten was also performed. This was to ensure that the order in which the cookies were eaten did not significantly effect the overall preferences. Since there were three different cookies, there were six possible combinations of cookie eating order. Therefore, we placed five bowls in each of the six categories and labeled them accordingly.

We placed the appropriate letter on the underneath side of the bowls with the corresponding cookie on the bottom of the bowl. For example, five bowls were placed in the category A, B, C; five bowls were placed in the category C, B, A and so forth. An arrow was placed on the side of the bowl where the subject was supposed to start. Once this was completed, the bowls were randomly mixed. On the edge of the bowls, above each cookie, was a box for the subject to mark which cookie they preferred Best, Worst, and in the Middle.

The person who secretly labeled each box of cookies then placed one-third of each cookie onto the appropriate places and only experimenters who were blind to the cookie order interacted with the participants. Subjects approached the table on their own freewill. Before participating in the experiment, they read and signed a consent form, and then were given the following directions to read:

"Rank Your Cookie Preference

Thank you for your participation."

Once the subject completed the experiment, the marked bowl was given to another experimenter who also did not know which cookie corresponded to which letter. This person also did not interact with the subject. Their job was to mark down in three columns labeled "A", "B", and "C", where the preferences lay.

Results

The data was analyzed using a Non-Parametric Friedmen Test, as the cookies were ordinally ranked. The results for Friedman Chi Square are (2, N=30) = 0.5308, p<.05, therefore it is not significant. In a count of thirty participants, the Sum Ranks and Mean Ranks, for the three cookies are as follows: The Nabisco Nilla Wafers Sum Rank is 65.00 and Mean Rank is 2.167; The Reduced Fat Nabisco Nilla Wafers Sum Rank is 58.00 and Mean Rank is 1.933; The Safeway Vanilla Wafers Sum Rank is 57.00 and Mean Rank is 1.900.

Discussion

Upon statistical analysis, the data collected fails to support half of the original hypothesis. The results show that there is not a significant difference in the preference of the three different cookies. Neither the reduction of fat nor the difference in brand made a difference in the preference.

During the experiment participants often mentioned that they could not easily tell the difference in the cookies, or that they preferred the cookies equally. Using the ordinal system of measurement may not have been the most appropriate testing strategy. A test using a point scale might have been more appropriate and given more statistically significant results.

There were some confounding variables in the experiment. Often times there were a number of participants taste testing at one time. The participants were commenting out loud about their preferences. This may have influenced other participants in their decision making. Secondly, although there was counterbalancing, some participants did not follow the directions and ate the cookies in the wrong order. Other participants either did not drink water in between each cookie, or they did not drink water before the taste test to clear their palate. A couple of participants also expressed their general dislike for Vanilla Wafer cookies, which may have influenced their decisions. Finally, because the participants were forced to choose a favorite, middle, and least favorite cookie, guessing may have been involved since it appeared to be difficult for many participants to taste a difference in the cookies. These confounding variables should be accounted for in any further studies of this nature.

References

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

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