This approach would be used to obtain data comparable to that collected from a roadside survey (e.g., BAC, demographic data). These data would then be paired with archival crash data, such as FARS. Fuel stations within the sampling region would either be randomly selected or selected based upon close proximity to a crash site. Motorcyclists would be approached at these stations while refueling and asked to participate in the study. Willingness to participate will be an issue, as with any survey. If response rates aren’t reasonably high, results may be of questionable validity. If the survey is kept short, it may be possible to administer it during the time it takes to refuel. Because this would be a survey, issues of OMB clearance will be in effect. Ethical issues when the survey team identifies a rider who is clearly impaired will also be a concern. It may not be necessary to have the cooperation of local police for this study, but it will be necessary, instead, to get the cooperation of fuel station owners and managers. The OECD uniform motorcycle crash data collection protocol contains procedures for conducting such a survey (Organisation for Economic Co-operation and Development, 2001). While this method has been used in Europe and Asia, it has not yet been used to collect alcohol data. An attempt to survey riders at fuel stations in Southern California met with several obstacles. Managers at fuel stations were unwilling to give permission to interviewers to approach people in their station. Senior officials at fuel companies also refused permission. A researcher counted the number of motorcycles at a fuel station in a period of several hours. There were very few motorcyclists. The potential difficulty in obtaining permission to survey, and the limited cost-effectiveness if too few riders can be surveyed, raise some doubts about the use of this method to collect comparison data in the United States (Prof. Hugh H. Hurt Jr., personal communication, 2003).
Riders would already be stopped and, therefore, may be more willing to participate. Avoiding the need for police cooperation may greatly simplify the project, and may reduce the cost if police agencies require that officers be paid out of project funds for their participation.
Because different motorcycles have different gas tank capacities and burn fuel at different rates, some motorcycles need to stop more frequently. This increases the likelihood of these riders being surveyed, which would cause them to be overrepresented in the survey results. Also, riders may be less likely to refuel after they’ve been drinking, either because of decreased willingness to stop or increased likelihood of deciding to refuel at the beginning of an outing, eliminating the need to refuel later. Because there are relatively few motorcycles on the road, interviewers may have large amounts of downtime between surveys. It may not be possible to obtain permission from owners/managers to survey riders on fuel station property. Conducting surveys during nighttime hours (e.g., 11 p.m. to 2 a.m.) may also present some security issues for the survey team. Conducting surveys in an area with some vehicular traffic raises safety concerns for researchers. This study would likely involve IRB clearance and OMB clearance would almost certainly be necessary.
$250,000 – $500,000
This study would likely be less expensive to perform than a roadside survey involving traffic stops. Some savings may come from the fact that it would probably not be necessary to use law enforcement personnel. The time necessary to set up a safe area into which to direct survey traffic would be saved as well. On the other hand, not working with law enforcement may make it more necessary to pay subjects to participate.