This study focused on athlete’s sleep behavior, dietary intake, and exertion of energy.  Through self-report, the subjects’ food intake and level of exercise was recorded. Sleepiness was assessed by subjective measures, specifically the Epworth Sleepiness Scale, the Stanford Sleepiness Scale, and a question inquiring how sleepy one feels after completion of the questionnaire (Question 71). Objective measures, particularly the Sleep Latency Test, were also applied to determine somnolence in the athletic participants.
Alcohol and Caffeine Consumption
        There was no significant difference between those who experienced sleep onset and those who did not involving the consumption of alcohol or caffeine. Therefore, self-report of alcohol ingestion did not influence sleep onset.  Caffeine, which also revealed no significant difference between those who slept and didn’t sleep, was not a factor swaying sleep onset latency.
Exercise
        The findings did not support the hypothesis stating exercise intensity and SOL are negatively correlated. The results also rejected the hypothesis predicting Epworth scores and level of exertion to be positively correlated.
Food
        The current findings rejected the hypothesis that carbohydrate and fat intake have a negative relationship with sleep onset latency. Also, carbohydrate and fat consumption did not act as a predictor for somnolence.  However, there was a significant correlation between fat intake and the Epworth. Caution is advised to those who solely administer the Epworth in studies involving food.  The reason being that while the Epworth reported a significant correlation, the objective measure of sleepiness reported no such thing.
Sleep
        The hypothesis regarding sleep onset, particularly that sleep onset latency would be less than 10 minutes, was supported in the findings.  3.8 percent of the subjects fell asleep prior to 5.5 minutes. This places these specific subjects in the pathological category where impairment and somnolence is present.  31.0 percent of the participants had a sleep onset latency of less than 10 minutes.  These subjects fall under the pathological grey area.  This simply states that they are sleepier than the average adult and possibly, but not necessarily, suffering from a sleep disorder. Variables that could influence sleep onset, particularly in the 55.2 percent that did not fall asleep or had a sleep onset latency of greater than ten minutes, are the artificial environment of the sleep lab or the anxiety of being observed.
        It is imperative to note that 44.8 percent of the participants experienced sleep onset prior to ten minutes.  It appears that they are more tired than the general college population.  It should also be noted that the sleep evaluation occurred equivalent to the time in which the participants would have been studying. Therefore the participants were not only sleepy, but more importantly, were incredibly sleepy during the period in which they were to be completing tasks for school.  Holmgren and her colleagues (2000) have reported findings of a decreased alertness and memory in athletes (2000).  It is quite possible that these deficits in attention and memory are related to somnolence.
Sleep Latency Test
        The Sleep Onset Latency Test (SLT) was a manipulation of the Multiple Sleep Latency Test (MSLT) measuring the somnolence of an individual at a critical point in time.  This test focused specifically on the period of time when the participants would have been studying for college classes. Caution must be taken when generalizing the results of a SLT to overall somnolence as measured by the MSLT. For example, it is possible that the circadian rhythm of the athlete induces sleep during this period of time, similar to the “lunch time dip,” but not throughout the entire day.
Subjective Versus Objective Tests
        An interesting finding of the current study revealed that the standardized subjective measures of sleepiness did not correlate with each other.  The two self-report tests assessing sleepiness, the Epworth Sleepiness Scale and Stanford Sleepiness Scale, had no correlation. Also, the findings revealed that neither of the two standardized subjective sleepiness measurements correlated with the objective measurement, sleep onset latency.
        On the otherhand, there was a significant correlation when asking the participant to report the likeliness of falling asleep after completing the questionnaire (Question 71) and the Epworth scores. A significant correlation also existed between Question 71 and the Stanford Sleepiness Scale. And surprisingly enough, Question 71 was significantly correlated with SOL.  It appears that this question was more accurate at assessing somnolence than other forms of self-report regarding sleepiness.  Kibbe and colleagues (1999) had similar findings demonstrating how Question 71 was a better assessor of sleepiness than the Epworth Sleepiness Scale.
        One possible explanation for the discrepancy among the different self-report sleepiness tests is that they are measuring different constructs.  Another possibility is that individuals are unable to perceive their actual biological somnolence.  This is supported by the findings reporting 7 out of the 18 subjects who fell asleep did not even realize that they had actually slept.
Further Research
        More research is needed in the area of subjective and objective measurements of sleep and somnolence. If, indeed, subjective measures of sleepiness are inaccurate in assessing somnolence when compared to objective measures, the use of these measures can cease in the clinical setting.  The use of subjective tests would be meaningless in screening and diagnosing patients if it lacked the sensitivity to assess sleepiness.  Perhaps the construction of a new self-report test that is highly correlated with objective measures, such as Question 71, is in need in the area of sleep.
        Given the controversy of self-report tests’ validity and the clinical demand of the MSLT, perhaps we can utilize the SLT as an assessment tool for somnolence.  While the SLT is inferior to the MSLT, it appears to be more valid than self-report tests of sleepiness. It is possible that the SLT may replace the administration of self-report tests while increasing the accuracy of the reported degree of somnolence. More research is needed in assessing somnolence, specifically involving subjective and objective measures.