Study 1
Sahin and colleagues (2016) conducted a randomized controlled experiment in the city of Adana, Turkey, in which drivers exceeding the established speed limits were waved off the road and randomly assigned, by a senior officer on the scene, to be approached by either treatment or control officers. If multiple cars were pulled off the road at the same time, treatment and control cars were separated by 30 feet to ensure that drivers were unable to tell that they were being treated differently; control officers were instructed not to interact with treatment drivers and vice versa. Randomization was predetermined on pads of report forms provided by researchers to the officers. Researchers were on the scene to observe all speed-control operations, ensure that control officers did not take on any treatment behaviors, and review all paperwork after the traffic stops to ensure that randomization occurred correctly.
Treatment officers provided drivers with information about local speed-related injuries and deaths, gave drivers an opportunity to explain their driving behaviors and ask questions, and issued a speeding ticket in a polite and respectful manner. Control officers behaved as they normally would in traffic stop situations, which typically involved stopping the car, approaching and informing the driver that the speed limit had been exceeded, asking for license and registration, writing a ticket, and handing it to the driver.
The initial sample of 500 police–citizen traffic encounters included 254 respondents in the treatment group and 246 in the control group. The full working sample included 458 police–citizen traffic encounters after 42 respondents were dropped from the initial sample due to interview non-response. The sample of drivers was on average 37.2 years of age, 88.4 percent male, and 41.9 percent college-educated. Because the treatment group members were significantly more likely to be college- educated and have a family member or acquaintance who was a member of the Turkish National Police, those variables were controlled statistically in the analyses.
All participating drivers in the study were interviewed by the researcher immediately after the speeding ticket was issued by the officers, using an instrument that measured agreement or disagreement with nine statements related to police legitimacy and procedural justice concepts, as well as demographic information on age, gender, education, district of residency, occupation, and income. All interviews were conducted in Turkish and took between 2 and 4 minutes to complete. Data were analyzed using estimate bivariate-normal regression models with control variables. No subgroup analyses were conducted.