As discussed earlier, the combination of behavioral and cognitive measures, physiological data, and questionnaires is beneficial in studying pharmacological interventions to enhance performance (as well as a whole range of other topics). Without this wide variety of measures, some important effects may have been missed altogether, and it would not have been possible to know whether observed performance and mood changes were attributable to physiological shifts in arousal, changes in motivation, or possibly even some form of experimenter or subject bias. The ability to collect EEG data in order to assess physiological arousal, questionnaire data to monitor subjective impressions of mood and alertness, computerized test data to evaluate basic cognitive skills, and simulator and/or aircraft flight data to assess actual operational performance all contribute to answering the research questions. For example, analysis of the SYNWORK data from the simulator dextroamphetamine studies indicated a 10-minute test was too short to show effects of sleep deprivation, however by having physiological data as well as an actual flight-performance measure, the deleterious effects of sleep deprivation were captured. Had a full array of tests not been implemented, it may have been erroneously concluded that sleep loss would not cause operationally significant problems or that dextroamphetamine would not be an effective countermeasure. Another example is the SSS used in the triazolam study. For aviators who are reluctant to express that they are less than fully capable when on flight duty, this test was not sensitive to sleepiness encountered throughout the day; however, the EEG measures that were collected in conjunction with the SSS were not susceptible to subjective factors. These data revealed that although subjects were not expressing alertness problems, decrements in CNS arousal were in fact present.
Another advantage of utilizing electrophysiological measures in conjunction with more standard performance tests is that physiological measures can provide the explanation for observed performance decrements or suggest that performance may begin to deteriorate at some future time that has yet to be established. An example of the explanatory utility of electrophysiological techniques can be seen in the zolpidem-napping study in which changes in both mood and performance occurred concurrently with objectively measured reductions in physiological arousal. In addition, the fact that zolpidem-induced naps attenuated the negative aspects of sleep loss better than placebo naps was at least partially attributable to the better sleep quality (measured electrophysiologically) after zolpidem. An example of the utility of EEG measures for suggesting a future performance problem could be seen in the dextroamphetamine studies. Note that although dextroamphetamine was effective for attenuating the impact of sleep deprivation, the medication did create modest reductions in the quality of recovery sleep. Although these re-