Exploring Racial Bias in Drivers' Behavior at Pedestrian Crossings

Kimberly Barsamian Kahn, Portland State University


The proposed project is designed to answer the research question, “Are drivers’ behaviors altered when they interact with pedestrians of different races?” We hypothesize that drivers’ implicit racial attitudes may subtly influence drivers’ behavior when interacting with African American versus White pedestrians. We will conduct a field experiment to test this hypothesis by observing drivers’ behavior with African American and White research team members crossing the street as pedestrians. We chose an experimental approach in a field setting order to control for all other variables aside from race and determine causality. By doing this research in a naturalistic setting, we are also able to capitalize on drivers’ real life behavior, and thus maximize external validity as well. The field experiment will be conducted at a marked crosswalk at an unsignalized intersection in Portland that has been extensively researched by the research team. The trained research team confederates (3 White, and 3 African American) will simulate an individual pedestrian crossing situation under normal traffic conditions. Each trial will consist of one of the confederates stepping up to the marked crosswalk, waiting for a car to stop, and then crossing to the other side of the street. All other confederates will be out of view. There will be a 1 minute break between trials before the next confederate crosses the street. The race of the confederates will alternate with each trial. Each confederate will cross the street 25 times (150 total trials). The experiment will be repeated across two days. To assess the drivers’ behavior, we will take a multi-method approach. First, trained observers will catalog the number of cars that do not stop for the pedestrian, how long it takes for a car to stop after the confederate has stepped up to the crosswalk, and estimates of the distance between the stopped car and the crossing pedestrian. Next, video recordings of the crossings will be used to confirm and further analyze observations made in the field. Finally, the confederates will have small sensors that they carry as they cross the street that catalog the precise distance between themselves and the driver that stops. Using all 3 metrics, we will be able to accurately assess any variations in driver behavior by pedestrian race. We will analyze the results of this project using between-groups ANOVAs on the dependent variables described above. We predict that more cars will pass waiting minority pedestrians, it will take longer for minority pedestrians to get the opportunity to cross the street, and there will be reduced stopping distance between the car and the minority pedestrians. The results of this study would illuminate how subtle psychological biases may threaten the safety of minority pedestrians and ultimately discourage minorities from choosing to walk. Implications for improving safety through policy and design will also be suggested. We expect to produce at least one journal article from this pilot work, to be submitted to psychology and travel behavior journals. We anticipate presenting this work at both a social psychology as well as a transportation conference, in order to expose this research to both disciplines. As a pilot study as part of a larger cross-disciplinary program of research on the effect of psychological attitudes on intermodal conflict, this project provides an initial examination of potentially discriminatory behavior directed differentially toward pedestrians. We plan to use this initial study to determine whether this differential behavior toward minority pedestrians is evident in these interactions. As part of the next steps in this line of research, we plan to focus on the causal role of implicit racial attitudes in producing these results. We will also use this pilot study to ground-truth the experimental methodology and usefulness of the sensors and the camera observations for coding driver behavior. We also expect to extend this line of research to replicate the study at different types of crosswalks, to test whether crosswalk design – like curb extensions or special signage – might negate the impact of racial bias, and to test at different times of day and under different traffic conditions. We will replicate this study in neighborhoods with different socio-demographic characteristics. Continuing, we will also employ the methodology and sensor equipment developed through this study to explore drivers’ implicit racial or gender biases in their interactions with bicyclists, as part of the larger program of research. Finally, we hope that this project and larger line of research will positively contribute to the burgeoning relationship between social psychology and transportation researchers at Portland State and in the two broader fields.

Project Details

Project Type:
Project Status:
End Date:
March 31,2014
UTC Grant Cycle:
NITC Small Starts Round 1
UTC Funding: