This will help to eliminate the possibility of psychological effects on the results. Group a will receive the caffeinated coffee and Group B. will receive the Decaf. Group a will serve as the test group. Group B. will serve as the control group. The independent variable will be caffeine and the dependent variable will be short-term memory. This study will measure the effect of caffeine on short-term memory.
Both groups would be instructed to refrain from consuming any food containing caffeine for one week before the test. The test would be administered first thing in the morning. Both groups would be instructed to fast after midnight on the night before the test. They would be instructed to consume nothing prior to taking the test. This procedure was designed to eliminate as many confounding variables as possible. For instance, the consumption of protein or sugars might affect the test and skew the results. Subjects were instructed to get at least 8 hours of sleep the night before the test to eliminate the possibility of fatigue and sleep on short-term memory. Subjects were asked not to smoke cigarettes for one hour before the test.
Prior to taking the memory test, the subjects were given a questionnaire to make certain that they had followed instructions prior to the test procedure. On the day of the test, one group would drink two 8 oz. cups of decaf coffee 30 minutes prior to taking a test using Sternberg's Short-Term Memory Test (SMT). The other group would consume two 8 oz. cups of caffeinated coffer before administration of the test. Differences in scores between the two groups was compared to determine the effect of caffeine on short-term memory scores.
Sternberg Short-Term Memory Test asks subjects to judge whether a test symbol is contained in a short memorized sequence of symbols. The mean reaction time increases linearly with the length of the sequence. In this test, improved memory results in greater accuracy and decreased reaction time in the responses. (Sternberg, 1966). The SMT involves the subjects viewing a random series of numbers between 0 and 9. The length of the series varies between groups of 4, 6, and 8. Each number is displayed for 750 milliseconds. Upon completion of a series, a warning signal is given, followed immediately by the test number. When the test number was presented the subject was asked whether or not the test number was included in the series (Sternberg, 1966). Both the accuracy and speed of the response was recorded.
Measures of Central Tendency
The following chart summarizes the accuracy and speed of the trials in both groups.
Accuracy Length 4 Series Length 6 Series Length 8 Speed Series Length 4 Series Length 6 Series Length 8 Group a 1 100 96.5-79.5-1-0.68 0.68 0.96 2 100 94.2-83.2-2-0.69 0.7-0.94 3-97.9-87.5-81.5-3-0.74 0.71 0.92 4-98.4-91-83 4-0.71 0.72 0.92 5-99-92.5-80.8-5-0.69 0.71 0.97 Mean 99.06-92.34-81.6 Mean 0.702 0.704 0.942 Median 99.03-92.42-81.55 Median 0.696 0.707 0.941 Mode 100 #N/a #N/a Mode 0.69 0.71 0.92 Group B. 1-96.4-96.5-68 1-0.75 0.75 0.95 2-97.8-95-75 2-0.76 0.76 0.96 3 100 89-81.3-3-0.64 0.78 0.97 4-99.5-91.2-85.2-4-0.78 0.83 0.96 5-97.2-93-83.4-5-0.74 0.87 0.96 Mean 98.18-92.94-78.58 Mean 0.734 0.798 0.96 Median 97.99-92.97-79.94 Median 0.745 0.789 0.96 Mode #N/a #N/a #N/a Mode #N/a #N/a 0.96
Measures of Variance
Accuracy Length 4 Series Length 6 Series Length 8 Speed Series Length 4 Series Length 6 Series Length 8 Group a 1 100.00-96.50-79.50 1.00 0.68 0.68 0.96 2 100.00-94.20-83.20 2.00 0.69 0.70 0.94 3-97.90-87.50-81.50 3.00 0.74 0.71 0.92 4-98.40-91.00-83.00 4.00 0.71 0.72 0.92 5-99.00-92.50-80.80 5.00 0.69 0.71 0.97 Variance 0.89-11.50 2.40 Variance 0.00 0.00 0.00 STD DEV 0.94 3.39 1.55 STD DEV 0.02 0.02 0.02 Range 2.10-11.0-3.7.04.04.05 Group B. 1-96.40-96.50-68.00 1.00 0.75 0.75 0.95 2-97.80-95.00-75.00 2.00 0.76 0.76 0.96 3 100.00-89.00-81.30 3.00 0.64 0.78 0.97 4-99.50-91.20-85.20 4.00 0.78 0.83 0.96 5-97.20-93.00-83.40 5.00 0.74 0.87 0.96 Variance 2.33 8.87-49.80 Variance 0.00 0.00 0.00 STD DEV 1.53 2.98 7.06 STD DEV 0.05 0.05 0.01 Range 3.6-3.5-13.12.12.01 Conclusion
When one first examines the results of this data set, it would appear that a significant difference is apparent between Group a (Caffeine Group) and Group B (Decaf Group) exists. However, when one begins to examine the variance between groups, it becomes apparent the results demonstrate inconsistency, without pattern. Standard deviations show a wide variability as compared to the mean. There was little consistency as far as the mode was concerned in some cases.
Although one could conclude that there is a significant differences between the control and test group upon first appearance, it would be difficult to determine that the differences between groups was significant due to the high degree of variance within the groups. This study cannot conclude that the hypothesis holds true and that caffeine improves short-term memory. The results are inconclusive at best.
This study obtained similar results to that of Terry and Phifer (2006) that was discovered in the literature review. One of the key difficulties in determining that caffeine did not significantly improve short-term memory is that they too found a high degree of variance among their results. They used a different measure of short-term memory, but still found a high degree of variance, as was found in this study.
The literature review revealed that there are several external factors that can affect short-term memory. It is not known what role these factors played in this research, but it is likely that they might be responsible for the variance observed in the data set. The methodology attempted to eliminate as many of the confounding variables as possible. However, there is no guarantee that others might have affected the results.
This study compared two groups of individuals that might have a high degree of variability in their abilities. In the final study this variability will have to be accounted for in the research design. One way that the researcher might use to reduce the possibility of external factors that could affect the results is to conduct a baseline study of the participants in the study population. One could take a baseline reading using the memory test on a group of subjects that has not had caffeine. The researcher could then give this same group caffeine and measure the results of the short-term memory test again.
This method has design problems, as much as this research study. For instance, there might be some improvement on the second test simply due to the fact that the subjects have seen it before. This research method could be used to enhance the present research design and help to eliminate some of the variability within groups. It is expected that some variance will be eliminated statistically in a larger sample population.
However, as we mentioned earlier, this is a pilot study that is intended to demonstrate areas that need to be improved in the larger study. It is expected that a larger sample population will reduce the variance within the groups and that more conclusive trends will emerge. There are many factors that could account for the variance between individuals within the same groups, including individual's differences in short-term abilities.
However, it one takes a qualitative approach to the data, it appears that there is a definitive trend in short-term memory between caffeine and non-caffeine users. A larger study group would help to provide more definitive results. From the results obtained in this pilot study, one cannot determine that caffeine improves short-term memory.
Anderson, K., Revelle, W., & Lynch, M. (2004). Caffeine, impulsivity, and memory scanning: A comparison of two explanations for the Yerkes-Dodson Effect. Neuroscience Letters.
367 (3), 327-331.
Bichler, a.; Swenson, a.; & Harris, M. (2006). A combination of caffeine and taurine has no effect on short-term memory but induces changes in heart rate and mean arterial blood pressure. Amino Acids. 31-940, 471-476.