Discriminating Taste 10.12

This taste discrimination experiment was the second of two designed to explore the nature of science generally, and the scientific method specifically. Here is the final report:

INTRODUCTION AND PREDICTIONS

The purpose of this experiment was to see if 6^th grade students were able to distinguish by taste the difference between three brands of locally available cola drinks here in Dakar, SENEGAL: Coca-cola, Magic Time Cola and Afri-Cola.

The class was mixed in their predictions. Most were unsure of the result.

PROCEDURE

The subjects for this experiment were 38 6^th grade science students attending the International School of Dakar. They are 11 and 12 years old, 19 girls and 20 boys, all good looking and above average.

Subjects were asked to compare the taste of two small samples of cola, served in small, plastic bathroom-style cups, to determine whether the samples tasted the SAME or DIFFERENT. Each student received three trials in succession. Between each sample, and between trials, subjects were instructed to sip fresh water, with the intent of cleansing their palettes.

In addition to determining SAME or DIFFERENT, subjects were also asked to provide a confidence rating for each trial: on a scale from 1 to 5, how confident are you of your response, where 5 means that you’re certain, and 1 means that you guessed.

Students designated as POURERS prepared all samples in isolation in the rear of the classroom, using a master list of subjects x trial x cola. EXPERIMENTERS worked directly with the SUBJECTS, guided and monitored their sampling, and recorded their responses.

The POURERS did not prepare their own samples. Instead, volunteer students prepared their samples from a master list kept hidden from the POURERS.

This experiment was double blind: neither the EXPERIMENTER nor the SUBJECT knew which brands of cola were being served. Actually, the sneaky teacher mixed up the colas bottles, and regrouped them only by number – cola 1, cola 2, and cola 3 – so that even the POURERS didn’t know what brand they were serving.

Three factors were randomized: the assignment of students to subject numbers; the assignment of the SAME/DIFFERENT conditions to trials; and the assignment of cola brands to cups. All this was done using a random number generator available on the internet at http://random.org.

Note the master data sheet, used by the POURERS, attached. Note also a sample subject response form, used by the EXPERIMENTERS, also attached.

RESULTS AND CONCLUSIONS

The results of our experiment are summarized in the top image above.

Based upon these results, we conclude the following:

· = Students are able to distinguish between the three brands of cola, but the effect is small, only about 60%.

· = The differences between boys -girls (about 1%) and homerooms (about 3%) are not significant.

· = Students who rated their response with a confidence of 5 did in fact perform more accurately, but the difference is very small, only about 5%.

· = Our results suggest that subjects were more accurate in judging differences than similarities, with the highest accuracy in the SAME condition 9-20% lower than the DIFFERENT conditions.

DISCUSSION

The purpose of this experiment was to investigate whether children are able to accurately discriminate among brands of cola drinks. In this case, we compared Cola-Cola with two less expensive generic brands.

Our results suggest that our subjects were able to distinguish amongst the three brands, but the effect was weak. There were no significant gender or homeroom class differences. It does seem that differences are more easily detected than similarities. The reasons for this are particularly interesting and worth exploring.