The interactive effects of caffeine and personality on autonomic responding and vigilance performance

Eysenck (1967, 1982) has long proposed that individual differences between introverts and extraverts are based on differences in chronic levels of arousal. Alternatively, Stelmack (1990) has proposed that I-E differences are due to differences in reactivity or sensitivity to sensory input. Recent work utilizing caffeine as an arousal manipulator has demonstrated differential effects of decreased arousal for introverts under conditions of high arousal. Other work in the area of vigilance has demonstrated greater attentiveness for introverts as compared to extraverts. It was hypothesized that caffeine would differentially affect I's in that they would evidence decreased electrodermal responding and an accelerative heart rate response in an auditory OR paradigm as arousal conditions increased. It was also hypothesized that I's would be differentially negatively affected by arousal increases in their performance on a visual, sensory type vigilance task. The present study utilized two levels of caffeine and a placebo as arousal manipulations in three 5 tone OR series at 65, 85 and 105 dB. Heart rate and EDA measures were recorded throughout part one in extreme group personality subjects. Part two recorded EDA only in the same subjects as they participated in a 42 min. visual vigilance task. Task order and other variables were completely counter-balanced. Results for the OR task failed to support the arousal or sensory reactivity hypotheses. Heart rate results were supportive in terms of greater "stimulus rejecting", accelerative responses for I's as stimulus intensity increased. Results offered stronger support for the sensory reactivity hypothesis. Vigilance task results did not support introvert superiority over extraverts in vigilance task performance. Examination of EDA related to missed targets, a new measure, demonstrated larger SCR to missed than to hit targets. Results suggest that I's perceived or identified more targets but did not respond behaviorally. Results while partially supportive of Eysenck's theory appear to offer stronger support for Stelmack's theory.