1. Attitudes Towards Police Use of Force
Brady C. Palecek & Aaron M. Tryhus
Faculty Mentor: Jeffrey A. Goodman
University of Wisconsin – Eau Claire
Introduction
Methods
Results
The relationship between the public and law enforcement is complex. While most contact between police and citizens seems smooth, there is a mutual undercurrent of fear and distrust.
In 2015, Americans’ confidence in the police hit a 22-year low, with Democrats and People of Color showing the sharpest declines in confidence (Gallup, 2015). This changing perception is believed to be the result of recent highly publicized instances of deadly use of force against unarmed African Americans, and the belief that US police forces have become increasingly militarized (Cook, 2015). Surprisingly, there is a paucity of research on Americans’ attitudes toward police officers’ use of force.
In the present study, we aim to shed light on this important social issue and see if there is a difference on how individuals view police use of force when we manipulate variables such as race/ethnicity and sex.
We hypothesize that participants’ perceptions of police use of force will vary as a function of the driver’s race/ethnicity and sex. Our findings will contribute to research into factors affecting attitudes toward and support of the police, and will also shed light on a timely and relevant social issue.
Primary Dependent Measure
A 9-item measure of Perceptions of Police-Driver Interaction (α = .96)
e.g., The police officer’s use of force was appropriate given the driver’s level of resistance
e.g., The officer provided sufficient opportunities for the driver to comply
Supplemental Measures (Gau, 2013; Tyler, Fagan & Geller, 2014; Tyler, Jackson & Mentovich, 2015)
A 4-item Trust in Police Scale (α = .91)
e.g., The police are generally honest
A 3-item Obligation to Obey Scale (α = .82)
e.g., It would be hard to justify disobeying the police
A 3-item Police Respect People Scale (α = .87)
e.g., The police treat people fairly
A 2-item Quality of Decision Making Scale (α = .79)
e.g., The police make decisions based on fact and law, not their personal opinions
A 8-item Positive Police Item Scale (α = .92)
e.g., The police usually treat everyone equally regardless of their race
e.g., It is your duty to do what the police tell you to do, even if you do not understand or agree with their reasons
A 9-item Negative Police Item Scale (α = .89)
e.g., Sometimes you have to bend the law to get things to come out right
e.g., People in power try to use the law to try to control people like me
A 3-item Personal Suspicion of Police Scale (α = .90)
e.g., The police are generally suspicious of people like you
A 3-item Police Normative Alignment Scale (α = .90)
e.g., The law represents the moral values of people like you
A 3-item Trust in Police Motives Scale (α = .93)
e.g., The police make decisions that are good for everyone in the community
3 attention check questions were asked to assess how well the participant was paying attention to the scenario
Gender, race/ethnicity and force used by officer
We also asked participants to describe the police officer they pictured mentally while reading the scenario
Most participants responded: White, male, middle-aged
To finish, participants completed a number of demographic items and then were debriefed.
e.g., Sex/Gender, Age, Socioeconomic Status, Race/Ethnicity
A series of 3 (Use of Force: High vs. Medium vs. Low) ×2 (Race of Driver: Caucasian vs. African American) ×2 (Sex: Male vs. Female) ANOVAs revealed only a main effect of scenario type for perceptions of the appropriateness of the police-driver interaction, and a trend toward the 2-way interaction between use of force and race of driver.
A 1-way ANOVA revealed that Participants of Color (N = 21), compared to Caucasian participants (N = 174), perceived the police-driver interaction to be less appropriate.
More generally, Participants of Color, compared to Caucasian participants, reported lower confidence in US police officer’s decision making, less positive impressions of police, a lower belief than police values are aligned with their own, and higher levels of police suspicion and negative impressions of police.
Significant group differences can be observed on the graphs by non-overlapping standard error bars.
Methods
Participants
N = 224 (71% Female; 89.5% Caucasian; mean age = 27.42)
23 participants were excluded for failing attention checks
Procedure
Participants accessed the survey through SONA or Qualtrics online.
Participants were randomly assigned 1 of 12 scenarios in which we manipulated the race/ethnicity and sex of the driver and the amount of force used by the police officer.
“An officer has pulled over a car that they observed crossing the center traffic… the officer noticed that the occupant was a 30 year old Caucasian/African American male/female… the officer could smell alcohol on his/her breath… The driver refused to exit the vehicle and submit to a sobriety test for the officer.”
Participants then read a paragraph that described the officer’s warning, the driver’s response to the warning, and the officer’s use of force given the response of the driver. In each of the 3 use of force conditions, the officer described two options.
Low use of force – given option to exit vehicle or officer would be forced to remove the driver (verbal threat only); driver complied.
Medium use of force – given same option as above; driver noncompliant; officer opens door and extracts driver.
High use of force – given same options as above; driver initially noncompliant then locks door, closes window, buckles belt; officer responds by presenting options of exiting vehicle or breaking the window and forcibly remove driver; officer follows through and pulls driver through the broken window.
Conclusion
Only one of our primary hypotheses was supported – as the level of police force went up, people were less approving of the use of force and the police officer’s behavior. Race and sex of driver were not found to influence perceptions of the officer.
Our lack of support for our hypotheses is likely due to us currently having too few participants. This resulted in our study being underpowered. An a-priori power analysis indicated a required sample size of 25 participants per scenario (300 total) to achieve 80% power to detect a medium effect size. Hopefully as we continue to collect data, we will be able to surpass our required sample size to determine if current trends become meaningful.
Although older than a typical university sample, participants were relatively homogenous. Most of our participants were female, white and young. We hope to obtain a more diverse group of participants.
Although we had few Participants of Color, our findings that they hold more critical views of the police than Caucasian participants are consistent with recent national polls and are relevant to contemporary events.
We thank UWEC’s Learning & Technology Services for printing this poster.