Date of Award


Document Type


Degree Name

Master of Arts (MA)




Caffeine, a central nervous system stimulant, is said to produce an elevation in level of arousal. Humphreys and Revelle (1984) have proposed a model to account for changes in information processing efficiency resulting from differences in personality, situational moderators such as caffeine, and type of information processing task. According to the model, a high level of state arousal will result in poor performance on shortterm memory tasks, while improved performance will occur at a low level of arousal. In this experiment, subjects' trait level of arousal, as measured by the impulsivity subscale of the Eysenck Personality Inventory, was manipulated by administering caffeine, to produce a state level of arousal. Subjects' performance on a free recall task was examined to see if caffeine affected short-term memory performance on a theoretically based task.

Seventy-four male subjects were divided into groups of high and low impulsives. These two groups were further divided according to amount of caffeine administered, 0 mg/kg, 2 mg/kg or 4 mg/kg of caffeine, resulting in six groups. Subjects were administered caffeine according to dose level and body weight.

Following a 30 minute absorption period, subjects listened to twelve lists of words presented at four rates which varied from one word every five seconds to two words per second. Following the presentation of each list, subjects wrote their recall of the list. Subsequently, subjects wrote a final free recall of words from all lists.

The proportion of words recalled from the primacy, middle and recency portion of each list was computed. A 2(Impulsivity) x 3(Dose) x 4(Rate) x 3(Serial Position) x 3(Practice) ANOVA was computed with the proportion of words recalled as the dependent variable.

The primary result was a dose x rate interaction with caffeine impairing recall under the 2 mg/kg dose for words presented at the slow rate, and caffeine facilitating recall under the 4 mg/kg dose for words presented at the fast rate. Differences in recall between the slow and fast rate declined as the dose of caffeine increased. Thus, it appeared that with caffeine, subjects were failing to benefit from the slow rate of presentation.

The results of this study indicate that caffeine affects the efficiency of processing in short-term memory tasks differentially according to dose and rate.