Evaluating DUI task force effectiveness through random sobriety checkpoints

DUI

experiment design test for DUI

Experiment to test DUI Task Force

Police departments today must be very careful about how they spend their scarce resources. This experiment is designed to assess the efficacy of the DUI task force, specifically to see if random DUI checks reduce the number of alcohol-related accidents. "Unlike a regular traffic stop, police do not need reasonable suspicion to stop or question suspects at a sobriety checkpoint. While the driver is still in the car, the police officer will look for signs that the person may be intoxicated. For example, the officer may take note of slurred speech, smells of alcohol, or uncoordinated physical movements" (DUI sobriety checkpoints, 2013, Legal Match). This is assuming that the task force is located in one of the 38 states in which such stops are legal (Sobriety checkpoint laws, 2014, GHSA).

Pretest

The number of citations issued for DUI will be catalogued via a statistical analysis of six months of previous tickets.

Treatment

Random DUI stops will be created throughout the town. Individuals will be randomly stopped and if they show signs of intoxication they will be given a field sobriety test. This intervention will take place over the course of six months.

Random assignment to a control group and treatment group

The number of citations issued for DUI will be compared with the number of citations issued before the intervention. The pre-intervention period will act as a control.

Internal and external validity

Threats to external validity include the fact that there could be a random spike in DUI (or an increase in drinking and driving due to seasonal factors such as holidays or the weather) when the checkpoints are instituted and that the increase in citations would have occurred without the stops. Threats to internal validity might include the fact that by virtue of the creation of the program, the police are more vigilant in prosecuting DUI cases than they would be if their methods were not under scrutiny. Causality (i.e., the stops generating more needed arrests and the spike in tickets) might be difficult to determine (Threats to validity, 2014, Research methods: Philosophy of Science and Research Design).