Experiments involving dosed automobile drivers have occasionally been used over the years to investigate the effects of alcohol on operation of automobiles. These studies have generally involved fairly low speed and low BACs. There have been very few examples of experiments to determine the effects of alcohol on motorcycle operation using subjects operating motorcycles on a closed course. Some examples were found as articles in the motorcycling magazines. These experiments were run by staff of the magazines involved and efforts were made to adhere to protocols necessary for a valid controlled research study. However there were sufficient problems with the methodologies that these studies can not be considered entirely valid. If future studies are performed by professional researchers, they will be subject to approval of an IRB.
Because of the increased vulnerability of motorcyclists, the tasks involved in such studies need to be fairly simple and need to be performed at lower speeds and lower BACs than studies with car drivers. It would be important that the closed course be free of fixed objects (e.g., poles, guardrails) that would pose a threat to a fallen rider even at low speeds. It would also be important that riders not be required to do anything that they are uncomfortable doing when unimpaired; obviously, riders should be capable of performing the task safely at baseline before attempting to perform it with a BAC.
There are known motorcycle skill tests that involve operating at as low a speed as possible which are extremely challenging and present little risk of injury and minimal damage to motorcycles in case of a crash. Some police motorcyclists perform these tests on a regular basis. They may be willing to participate in a dosed-rider study. Other types of expert riders may be good candidates for such a study because they may already be insured for risky riding, may be able to operate at relatively high BACs, and may show impairment while having reduced risk of injury. These would include motorcycle racers and or stunt riders. Use of such professional riders as subjects, and operation at lower speeds could make it easier to obtain IRB clearance for a closed-course study, however the relevance of findings to non-professional riders at higher speeds, would be questionable.
Another possibility would be to conduct an off-street study using motorcycles equipped with outriggers to prevent the motorcycle from falling over. These have been used in the past for research. This would potentially reduce the chances of a complete capsize and fall-down. However the outriggers will likely effect handling and maximum lean angles, such that findings may not reflect normal riding circumstances. Their presence may alter rider risk-taking behavior. Further, the outriggers may not completely prevent any possibility of injury. For example, a rider could still fall off an outrigger-equipped motorcycle, or, depending upon the outrigger configuration, have a foot trapped between the outrigger and the pavement.
As with the simulator study, it will be extremely important to provide enough practice on the measure to guarantee that there is no confounding between alcohol effects and practice effects. The study design must also acknowledge that there is a potential difference between effects of a given BAC while drinking, as BAC is going upward, versus the same BAC after drinking when the BAC is coming down.
Many of the advantages for the simulator study apply to the closed-course study as well: face-validity due to measurement of motorcycle operation, ability to use same riders in both impaired and unimpaired conditions, ability to compare motorcycle and automobile operation (though obviously two different courses would be necessary for the two different vehicles). Although it would not be possible to measure as many aspects of operation as can be measured with a simulator, it would still be possible to measure aspects of operation (e.g., stopping distances, low-speed dropping of the motorcycle, lane excursions) that would not be measurable using archival data, interviews, surveys, and so on. BAC data would be of the highest quality. As with the simulator study, riders could be videotaped performing the test and footage could be used as a teaching tool.
Closed-course studies share many of the disadvantages of simulator studies. Getting IRB approval for such a study would likely be more difficult than for a simulator study. The extent to which this becomes an obstacle will vary depending on the funding agency and local rules and regulations. Like simulator studies, closed-course studies gain some validity from the fact that they involve actual operation of motorcycles. However, depending on the tasks involved, closed-course operations may bear little resemblance to those involved in on-street operations. As an example, extreme low-speed (i.e., 5 mph or less) handling of a motorcycle is a very different skill than operation at normal speeds. It would be difficult to say much about the connection between study results using this approach and actual crash likelihood. Tests of skill that require riders to go as slow as possible will not reflect tendency for drinking riders to use poor judgment by riding faster than appropriate, though they may still allow riders to exhibit poor judgment in some other form. Although it may be possible to run a closed-course study in such a way as to minimize crash likelihood and severity, there is still the possibility of injury. Liability issues will be important to whoever does such a study. It may be easier to get clearance to run such a study by using professional riders (e.g., motor officers, racers) but these are not representative of the skill level of the general rider population. Findings from a closed-course study are not completely applicable to on-street operation.
To create a low-speed course will not require a great deal of real estate. Existing training facilities could be used. Motorcycle police in some jurisdictions are currently conducting this type of exercise regularly as a part of training and retraining. Often these exercises are conducted in parking lots. It may be possible to obtain riders on a volunteer basis or it may be necessary to pay subjects. Subject participation fees should not be terribly costly, depending on the number of subjects. It was suggested that this study would be most effectively performed in conjunction with a simulator study. In that case it must be acknowledged that available funds would have to pay for both studies. Additional costs to be considered are the possible need for additional insurance and the cost of repair and maintenance.