Impact of Cell Phone Videos on Public Perceptions of Police
School
Franklin University**We aren't endorsed by this school
Course
CJAD 710
Subject
Sociology
Date
Dec 12, 2024
Pages
19
Uploaded by SuperGuanacoMaster1190
summarize the article. Within your summary, identify the analytic method, sample size, theoretical framework, findings and limitations.Anecdotal evidence suggests that recent video-recorded police–citizen encounters have undermined police legitimacy and fueled civil unrest across the United States. Drawing from the process-based model of policing, social cognitive theory, and past research on media effects, we assess the influence of viewing cell phone videos of police–citizen encounters on perceptions of law enforcement. Using quasi-experimental methods and video footage of an actual police–citizen encounter captured on cell phones, the effects of viewing these videos are assessed using a series of repeated measure ANOVAs. Results indicate that viewing cell phone videosof police–citizen encounters significantly impacts perceptions of law enforcement, though little evidence of differing effects based on point-of-view, number of video exposures, or ordering of video exposures was found. The process-based model of policing should consider further incorporating the contributions of technology to provide a more holistic account of the factors influencing perceptions of police.Keywords: police-citizen encounter; police legitimacy; camera footage; procedural justice; camera perspective biasAnybody who's got a phone in their hand becomes a video journalist, and we've all seen it, right? With the emergence of YouTube and Facebook and everything ... I tell people that all the time now—"Always assume you're on camera."—Police sergeant (Campeau, [17], p. 674)Technology, media, and institutional legitimacy are closely intertwined, particularly for law enforcement (Neyroud & Disley, [58]; President's Task Force on 21st Century Policing's, [63]). From televised broadcasts of Bull Connor aiming fire hoses at civil rights protestors to the beating of Rodney King, visual accounts of police–citizen interactions have shaped public perceptions of the appropriateness, fairness, and acceptability of police behavior (Lawrence, [48]; Skolnick & Fyfe, [76]; Weitzer & Tuch, [94]). Issues of police legitimacy have gained renewed attention in the past several years due to a number of recorded police–citizen encounters from across the US. Some videos have come directly from departments via officer body-worn cameras, such as those showing the death of James Boyd, while others such as those showing the deaths of Eric Garner and Walter Scott, were generated by bystanders using cellular phones. Many of these recordings were later uploaded to social media sites likeFacebookandYouTubewhere they could be viewed by individuals around the world. Suffice to say, these and other videos, as well as the technology used to generate and disseminate them, have serious implications for public perceptions of police (Campeau, [17]; Goldsmith, [32]; Walker, [90]).Although it appears that some citizen-recorded videos have influenced civil unrest and negative perceptions of law enforcement, little is known regarding the effects ofexposure to videos of police–citizen encounters on police legitimacy (Pyrooz, Moule,
& Decker, [65]). We situate the potential influence of these videos within Tyler's ([87]) process-based model of policing, which focuses on the social sources of perceptions of police (e.g. vicarious and direct experiences; Brunson, [14]; Rosenbaum, Schuck, Costello, Hawkins, & Ring, [71]; Tyler, [87]). A parallel body of research has examined the influence of media on perceptions of police, but has concentrated largely on perceptions of racial bias and misconduct (e.g. Donovan & Klahm, [22]; Dowler & Zawilski, [23]; Sigelman, Welch, Bledsoe, & Combs, [74]; Weitzer & Tuch, [95]), to the neglect of legitimacy and the process-based model.Due to the prevalence of cell phones and social media (Pew, [60], [61], [62]), and the ease with which they allow recordings to be viewed and disseminated (Moule, Decker, & Pyrooz, [56]), these technologies provide the next step for understanding perceptions of police legitimacy. We characterize viewing video recordings of police–citizen encounters as technology-mediated exposures, as they provide intimate or experiential first-person accounts of police–citizen encounters (i.e. sights and sounds) without directly involving the viewer. These videos thus reflect a "middle ground" between the direct and indirect experiences common to the process-based model of policing (Tyler, [87]). Further, videos also present unique points-of-view, providing viewers with distinct pieces of information that can differentially influence subsequent perceptions (Lassiter, Geers, Handley, Weiland, & Munhall, [46]). Exposure to encounters through videos may then offer an addendum to the process-based model. Given the apparent role of these recordings in generating civil unrest, understanding whether these videos impact perceptions of police legitimacy is a timely and important topic to address (Klinger, Rosenfeld, Isom, & Deckard, [42]; Pyrooz, Decker, Wolfe, & Shjarback, [64]; Pyrooz, Moule, et al., [65]).The current study examines whether technology-mediated exposure to a real police–citizen encounter influences perceptions of police legitimacy. Using a quasi-experiment with a large sample of students, and theories of procedural justice, social cognition, and media effects, we ask four research questions: (1) Does viewing videos of police–citizen encounters influence perceptions of law enforcement? (2) Does this influence vary based on the point-of-view presented in a video? (3) Does this influence vary by the ordering of exposure to these videos?, and (4) Does this influence vary by whether individuals see multiple videos of an encounter? Answering these questions will explicate the contribution of a new generation of technology to perceptions of law enforcement. We begin by discussingthe process-based model of policing.The Process-Based Model of Policing and Police LegitimacyLegitimacy is "a property of an authority that leads people to feel that the authority or institution is entitled to be deferred to and obeyed" (Sunshine & Tyler, [79], p. 514), which generates voluntary compliance from individuals interacting with institutional actors (Hinds & Murphy, [35]; Mazerolle, Antrobus, Bennett, & Tyler, [53]). Acquiring and maintaining legitimacy is of particular importance for police, as they require the voluntary assistance and cooperation from citizens in the day-to-day execution of their duties (i.e. ensuring public safety, fighting crime and order maintenance) (Huang & Vaughn, [29]; Reiss, [70]; Tyler, [85], [87]). Tyler's ([87])
process-based model provides an accounting of the development and persistence ofpolice legitimacy.Central to legitimacy is the concept of procedural justice (Lind & Tyler, [49]; Tyler, [84]). Procedural justice reflects how people feel they are treated in interactions with institutional actors and the decision-making processes used by those actors (Hinds & Murphy, [35]; Tyler, [87]). In the context of policing, procedural justice rests on several elements, including citizen participation in an interaction, the quality and fairness of the treatment by police, as well as the quality of decision-making on the part of the officer (Mastrofski, Snipes, & Supina, [52]; Mazerolle, Bennett, Davis, Sargeant, & Manning, [54]; Sunshine & Tyler, [79]; Tyler & Huo, [89]). For an encounter to be considered procedurally just, citizens must feel they are engaging in a dialog with the officer, that they are being heard, and that they have been treated with respect (Reisig, Bratton, & Gertz, [68]). Finally, police are considered procedurally just when they show themselves to be objective in their decisions and fair in their treatment of citizens (Tyler, [87]).In the process-based model, legitimacy is gained or lost during police–citizen interactions (Tyler, [87]). When citizens view officers as acting in a procedurally just manner, they are more likely to see the police as legitimate (Mazerolle, Antrobus, etal., [53]; Tyler, [85]). Procedural justice thus engenders legitimacy, and is a natural preceding factor in this process (Gau & Brunson, [31]; Sunshine & Tyler, [79]). Positive treatment of citizens enhances perceptions of legitimacy, while negative treatment reduces legitimacy (Skogan, [75]; Sunshine & Tyler, [79]; Wolfe, Nix, Kaminski, & Rojeck, [101]). Stronger perceptions of police legitimacy, in turn, produce compliance with the law and cooperation with police (Donner, Maskaly, Fridell, & Jennings, [21]). The positive outcomes associated with procedural justice have led subsequent research to expand on Tyler's ([87]) framework. In doing so, research has begun to provide a more holistic accounting of the social sources influencing perceptions of police legitimacy.Social Sources of Police LegitimacyTyler's ([87], pp. 5, 94–114) process-based model emphasizes the role of direct experiences for assessments of procedural justice and police legitimacy. Direct experiences involve in-person contact with police, whether voluntary, such as calling for assistance or reporting a crime, or involuntary, as in cases of traffic stops or stop-and-frisk searches (Carr, Napolitano, & Keating, [18]). Perceptions of police also develop through indirect, or vicarious, experiences. These experiences involve information learned from outside sources and are not personally experienced; common examples include hearing stories of others' encounters with police (Brunson, [14]; Horowitz, [37]).As most individuals do not come into contact with the police (Eith & Durose, [25]), indirect experiences are salient in shaping perceptions of law enforcement (Horowitz, [37]). Indeed, individuals draw on vicariously obtained information to evaluate the police generally, as well as use this information to interpret their experiences if they do come into contact with an officer (Rosenbaum et al., [71]; Wolfe et al., [101]). For example, Warren ([91]) found that individuals who heard
about negative experiences with police were four times more likely to report being disrespected by an officer. Weitzer and Tuch ([95]) similarly found that residing with someone who had a negative police experience resulted in lower levels of trust in the police.In addition to the vicarious experiences conveyed by friends or family, media portrayals of police also impact public perceptions (Eschholz, Blackwell, Gertz, & Chiricos, [26]; Sigelman et al., [74]). These portrayals vary from fictional accounts, found in popular films and television, to non-fictional broadcasts on the evening news. For instance, media coverage of police misconduct has been found to reduce favorable views of the police, and lead citizens to believe that misconduct and racialprofiling are commonplace (Weitzer & Tuch, [94], [95]). Although not explicitly examining police legitimacy, Weitzer ([93]) found lower levels of public support for police following media coverage of police misconduct.Taken together, this research demonstrates that experiences with police, either directly or vicariously, alter individual perceptions (Augustyn, [4]). That such diverse experiences impact perceptions of police raises provocative questions aboutother sources of these perceptions. New technologies, such as cell phones and the Internet, are one such source, as they allow for the recording and dissemination of encounters (Pyrooz, Moule, et al., [65]). Videos recordings contain elements of direct experiences (sights and sounds), while not directly involving the viewer. The mix of direct/vicarious experiences offered by new technologies suggests a possible addendum to Tyler's ([87]) model. To better situate the potential contribution of these technologies to perceptions of police, we draw on social cognitive theory (SCT).Social Cognitive Theory and the Process-Based Model of PolicingThe ability of direct and vicarious experiences to shape an individual's perceptions is consistent with Bandura's ([5], [6]) SCT. SCT holds that people learn about the world from a variety of sources, and that these sources create and reinforce definitions which guide future perceptions, interactions, and behavior (Bandura, [6],[8]). Learning is a transactional event, whereby perceptions are a consequence of internal factors such as personality, lived experiences, and the interaction between the two (Bandura, [7]).SCT specifies that individuals learn new information through three processes: live, instructional, and symbolic modeling (Bandura, [6]). Live modeling occurs through direct experiences. Instructional modeling occurs when individuals are given verbal or written information, vicariously or in-person. Symbolic modeling occurs via vicarious means, where individuals learn by observing others' actions (Bandura, [8]). These processes parallel those found in the process-based model (see Figure 1). Live modeling is analogous to the direct encounters in Tyler's ([87]) theory. Vicarious experiences with police, found in discussions with friends and family (Feagin, [27]; Feagin & Sikes, [28]), are analogous to the instructional aspects of SCT. The remaining process, symbolic modeling, is useful for understanding why viewing videos of police–citizen encounters would influence perceptions of police legitimacy.
Graph: Figure 1 Overlap between process-based model of policing and social cognitive theory.Symbolic modeling constitutes a major part of individuals' daily lives, as "people gain understanding of causal relationships and expand their knowledge by operating symbolically on the wealth of information derived from personal and vicarious experiences" (Bandura, [7], p. 267). The sources of symbolic modeling aresimilar to those found in media effects research on perceptions of police (e.g. news/entertainment media), and exposure to these sources can alter individual perceptions (Adkins & Castle, [3]; Bushman, [15]; Donovan & Klahm, [22]). Just as seeing "bad" information (e.g. violence, racial profiling) can result in negative knowledge and perceptions, so too can seeing "good" information (e.g. courteousness) improve perceptions (Maguire, Lowrey, & Johnson, [50]). Thus, SCT recognizes the capability of media exposure to influence perceptions, insofar as individuals must process new information as they are exposed to it. This acknowledgment comes with one caveat: each encounter involves multiple points-of-view, which may convey different or inconsistent information about an event, resulting in different changes to subsequent perceptions.Point-of-View and Perceptions of InteractionsCamera perspectives influence viewer perceptions of situations (Kraft, [43]; Mandell& Shaw, [51]), as point-of-view affects the information extracted from observations (Lassiter et al., [46]; McArthur, [55]). For example, perspective-related biases have been found for taped confessions (Lassiter & Irvine, [47]; Ratcliff, Lassiter, Schmidt, & Snyder, [66]) and observer testimony (Hennessy & Jakubowski, [33]; Landström &Granhag, [44]). These biases result from theperspectiveitself, as individuals receive different images or information about a situation (Jones & Nisbett, [39]; Snyder, Lassiter, Lindberg, & Pinegar, [78]); related to perspective issalience, where viewers overestimate the importance of the actor featuring most prominentlyin a recording (Lassiter et al., [45]; Taylor & Fiske, [82]); and, finally,temporal proximityreflects when a recent perspective is most influential for recall and perceptions (Hirose, Kennedy, & Tatler, [36]). These biases, in essence, reflect anticipated differences between points-of-view, with each featuring in recordings of police–citizen encounters.Boivin, Gendron, Faubert, and Poulin ([10]) examined the possibility of a body-worn camera perspective bias. Using a sample of undergraduate students and police candidates, participants were randomly assigned to view a recording of a police–citizen encounter from the perspective of either a body-worn or surveillance camera. No perspective bias was found among the undergraduates. Among the police candidates, however, assessments of a lethal use of force differed based on whether the candidate saw footage from an officer-worn camera or a security camera. This finding suggests the multiple points-of-view present in an encounter be accounted for, as they may uniquely influence viewer perceptions (Boivin et al., [10], p. 13). Further evidence for this consideration is found in the work of Snyder etal. ([78]), who examined how dual-camera perspectives—those capturing the words and facial expressions of suspects and interrogators simultaneously–contribute to
viewer perceptions. Using side-by-side recordings, Snyder and colleagues ([78]) found that this approach did not bias viewer perceptions, compared to videos focusing primarily on the suspect or the interrogator.The side-by-side dual-camera approach used by Snyder et al. ([78]) is unlikely to occur in videos of police–citizen encounters. Rather, recordings of encounters likely involve seeing only one point-of-view, such as that of a bystander. Alternatively, some may see multiple videos of an encounter, also encompassing diverse points-of-view. Hirose et al. ([36]) demonstrated that viewer memories are generally oriented toward more recently acquired information (e.g. those occurring after a shift in point-of-view), suggesting that the most recently viewed video seen may be more influential for perceptions. As a result, exposure to videos in different orders would be expected to produce differences across viewers. These considerations inform the current study.Current StudyTechnology and legitimacy are closely intertwined, and the ubiquity of cell phones, social media, and the Internet provide new challenges to the legitimacy of police. Over the past several years, a number of recordings of police–citizen encounters, filmed by citizens and disseminated online, gained prominence in the media and have allegedly contributed to the civil unrest in many American cities. A sizable literature has examined perceptions of police legitimacy (Donner et al., [21]; White, Mulvey, & Dario, [98]), but has overlooked the contributions of media to these perceptions, even as the influence of media exposure on perceptions of police misconduct is well documented (Chermak, McGarrell, & Gruenewald, [19]; Weitzer, [93]). These considerations are especially prominent given the prevalence of technologies allowing for the immediate recording and dissemination of police–citizen encounters.It is necessary to understand whether viewing videos of police–citizen encounters impacts perceptions of the police. Drawing on theories of procedural justice, cognition, and media effects, the current study seeks to answer four questions:(1) Does viewing cell phone videos of police–citizen encounters influence perceptions of law enforcement?(2) Does this influence vary based on the point-of-view presented in a video?(3) Does this influence vary by the ordering of exposure to these videos?(4) Does this influence vary by whether individuals see multiple videos of an encounter?By examining these questions, we assess the relevance of emerging social phenomena (cell phones, social media) for the process-based model (Tyler, [87]). Given the benign nature of the encounter, we anticipate exposure to this type of video will improve perceptions of police. Further, we anticipate that, because groupsreceive different information about the encounter, based on the point(s) of view they are exposed to, the ordering of video exposures, and the number of videos
they are exposed to, that effects will be significantly differ across groups. Specifically, we hypothesize that:Hypothesis 1: Viewing videos of a benign police–citizen encounter will produce positive changes in perceptions of police.Hypothesis 2: The effects of viewing videos of an encounter will vary depending onthe point-of-view a viewer is exposed to.Hypothesis 3: Seeing different numbers of videos of an encounter will result in significantly different perceptions across groups.Hypothesis 4: The ordering of video exposures will produce significant differences in perceptions across groups.In the next section, we detail the data and methods used to test these hypotheses and answer our research questions.Data and SampleData for the current study are derived from self-administered, online surveys distributed to adults enrolled in a large university located in the American southwest. Surveys were administered during the spring 2015 semester as part of abroader project examining technology and perceptions of law enforcement. To recruit participants, all individuals listed as teaching at least one class (in-person or online) in the criminology and criminal justice department were contacted via email,provided an overview of the project, and asked to provide the information to their students. Instructors willing to participate provided a digital survey links to their students.Consistent with IRB requirements, students were assured that responses were anonymous and their participation in the study was voluntary prior to completing the survey. Participants were asked about their perceptions of police, shown a randomly assigned video of the police–citizen encounter, and again questioned regarding their perceptions. The survey took roughly 20 minutes to complete. Although the exact number of students invited to participate is unknown, 45 instructors teaching a total of 48 undergraduate and 21 graduate level classes werecontacted, resulting in 573 individuals participating in the study. Owing to missing data through item non-response (< 1%), the total sample for this study was 567 individuals.The sample is split fairly evenly between women (56%) and men (44%). With respect to the racial and ethnic breakdown of the sample, 64% were White, 6% were African-American, 3% were Asian, 3% were Native American, and 24% reported being of another race. One-third of respondents (33%) reported being Hispanic.[1] Participants ranged in age from 18 to 66 years old, though the majorityof the sample (55%) was 23 years old or younger. The majority of the sample (60%) were undergraduate upperclassmen, 18% underclassmen, and approximately one-fifth (22%) of participants were graduate students; most participants (88%) were
majoring in criminology and criminal justice. Summary statistics are provided in Table 1.Table 1 Summary statisticsGroup 1—Citizen POV (n = 141)Group 2—Officer POV (n = 144)Group 3—Cit./Off. POV (n = 138)Group 4 Off./Cit. POV (n = 144)Total sample (N = 567)RaceAfrican American3.5% (5)5.6% (8)5.0% (7)10.5% (15)6.2% (35)Asian2.8% (4)3.5% (5)2.9% (4)2.1% (3)2.8% (16)Native American3.5% (5)2.8% (4)3.6% (5)3.5% (5)3.4% (19)White66.4% (95)65.3% (94)59.4% (82)62.9% (90)63.6% (361)Other23.8% (34)22.9% (33)29.0% (40)21.0% (30)Hispanic24.1% (137)Yes31.0% (44)33.0% (47)34.3% (46)35.4% (51)33.4% (188)No69.0% (98)67.1% (96)65.7% (88)64.6% (93)66.6% (375)SexMale45.5% (65)38.1% (55)52.9% (73)39.6% (57)44.0% (250)Female54.6% (78)61.8% (89)47.1% (65)60.4% (87)56.1% (319)Class statusFreshman10.5% (15)7.6% (11)9.4% (13)7.7% (11)8.8% (50)Sophomore6.3% (9)11.8% (17)8.7% (12)9.1% (13)9.0% (51)Junior21.7% (31)28.5% (41)31.2% (43)31.5% (45)28.2% (160)Senior35.7% (51)36.1% (52)26.1% (36)30.1% (43)32.0% (182)
Group 1—Citizen POV (n = 141)Group 2—Officer POV (n = 144)Group 3—Cit./Off. POV (n = 138)Group 4 Off./Cit. POV (n = 144)Total sample (N = 567)Graduate student25.9% (37)16.0% (23)24.6% (34)21.7% (31)22.0% (125)Criminal justice majorYes92.3% (132)91.7% (132)83.5% (116)86.0% (123)88.4% (503)No7.7% (11)8.3% (12)16.6% (23)14.0% (20)11.6% (66)Video ExposuresThe two videos used in the current study come from a police–citizen encounter that occurred in Pontiac, Michigan on 27 November 2014.[2] The citizen, Brandon McKean, an African-American male, was stopped by a white male Oakland County sheriff's deputy in response to a call from a business owner who felt Mr. McKean may have been "casing" his establishment. Mr. McKean began filming the officer as he was being interviewed. Shortly after Mr. McKean began filming on his cell phone, the officer also began filming the encounter on his cell phone. This brief encounter involved only a verbal exchange, and no use of force by the officer. Following a "high-five" from the officer, Mr. McKean was allowed to continue on his way. He lateruploaded his video toFacebook, where it went viral, receiving over three million views within the first week (Abbey-Lambertz, [1]). A week following McKean'sFacebookposting, the Oakland County Sheriff's Office released the deputy's video.Using videos of this police–citizen encounter is advantageous for three reasons. First, when officers are filmed or are filming on duty, only one point-of-view is typically captured: that of the officer, the citizen, or a nearby third party. The videos in this study are, to the best of our knowledge, unique in that they capture the near totality of the encounter, from the first-person perspectives of the citizen and the officer.[3] Second, these videos capture a racial dynamic (white officer, African-American citizen) that has featured prominently in the media. Third, this encounter does not involve any use of force, and while a departure from typical studies of police behavior (the majority of videos receiving media attention), more closely reflects the modal category of police–citizen encounters (Adams, [2]; Terrill, [83]; Westmarland, [96]). In other words, we assess whether even benign encounters caninfluence perceptions of police.From the two videos, four possible groups were created to which participants were randomly assigned via Qualtrics. Group 1 saw the video of the citizen's perspective (looking at the officer). Group 2 viewed the video of the officer's perspective (looking at the citizen). Group 3 was shown both videos, first the citizen's perspective, then the officer's. Group 4 also viewed both videos, first seeing the officer's perspective then the citizen's. Appendix A contains sample images from
each video. Overall, the data are well-suited for assessing whether exposure to cell phone footage of police–citizen encounters influences individual perceptions of law enforcement.VariablesAll measures of the process-based policing model used in the current study were drawn from Reisig et al. ([68]). Items were measured on 4-point Likert scales indicating whether respondents (1) strongly disagree, (2) disagree, (3) agree, or (4) strongly agree with each statement. Higher scores are indicative of more positive views of police. Principal component analysis with promax rotation found the items loaded into four distinct outcome variables: obligation to obey the law, willingness to cooperate with police, procedural justice, and distributive fairness. Appendix B contains a list of all items.Consistent with prior research (Reisig et al., [68]), we use four global measures to assess antecedents, dimensions, and consequences of police legitimacy.[4]Obligation to Obeyis a two-item summed scale designed to gage participants' belief in obeying the directives of police officers. The scale showed adequate psychometric properties (r= .510).Willingness to Cooperateis a four-item summed scale measuring individual willingness to share information with or contact police (α= .862, avg. inter-itemr= .620).Procedural Justiceis an eight-item summed scale gaging individuals' perceptions of how procedurally just officers are (α= .920, avg. inter-itemr= .597).Distributive Fairnessis a two-item summed scale measuring the degree to which individuals feel police treat all citizens fairly (r= .693). To reduce the potential for learning effects between tests, the ordering ofboth blocks of survey items as well as the questions within each block was randomized for each respondent.Analytic StrategyTo examine if viewing videos of police–citizen encounters influences individual perceptions of police, and whether points-of-view distinctly impact these perceptions, the current study uses a randomized four-group pretest-posttest design(Campbell & Stanley, [16]; Shadish, Cook, & Campbell, [73]). Respondents were randomly assigned to one of four treatment conditions: citizen video, officer video, citizen-officer videos, or officer-citizen videos. Respondents were surveyed on their perceptions of police, exposed to a treatment condition, and then re-surveyed. The analysis proceeds in three steps. First, comparisons of means within and across each treatment condition are examined. Second, we compare means within and across groups who viewed only one video (Groups 1 and 2) or both videos (Groups 3and 4). Third, we compare means within and across groups who saw both videos, albeit in different orders (Groups 3 and 4). To assess differences in pre- and post-test outcomes, we use a series of repeated measure ANOVAs. Analyses were conducted using SPSS 22 (IBM Corp., Armonk, NY).Results
A series of repeated measures ANOVAs were conducted to compare the effect of viewing a specific police–citizen encounter on the outcome variables, before and after viewing the videos. Table 2 displays the results of the repeated measures ANOVAs examining the effect of seeing either the officer's or the citizen's point-of-view.Table 2 Repeated measures ANOVAs—point of viewPre-testPost-testF (treatment effect)F (treatment effect by video group)VariableVideo groupMeanSDMeanSDObligation to obeyOff. POV 5.1681.465.2481.4101.208 .037Cit. POV5.1601.465.2601.473Distributive fairnessOff. POV 5.2271.7315.3171.755.2692.929Cit. POV4.4651.6555.2991.802Procedural justiceOff. POV 22.6544.60523.2184.4446.251**1.183Cit. POV23.2014.42623.4244.845CooperationOff. POV 13.3982.79813.4562.7292.4641.004Cit. POV13.9722.67614.2362.4981 *p< .05, **p< .01.There was a significant effect of viewing the videos on individuals' procedural justicescores (Wilk's Lambda = .975,F(1, 243) = 6.251,p= .013). There was no significant effect for the video group and procedural justice interaction, indicating that the specific point of view was not as impactful as just seeing the event.[6] Thisfinding provides only partial support for Hypothesis 1 and 2, with significant pre–post changes occurring across groups.[7]Table 3 displays the results of the repeated measures ANOVAs examining the outcomes based on the number of videos seen. There was a significant effect for the number of videos seen on an individuals' obligation to obey scores (Wilk's Lambda = .965,F(1, 565) = 20.412,p= .000). There was a significant effect for the number of videos seen on an individuals' procedural justice scores (Wilk's Lambda = .976,F(1, 565) = 14.156,p= .000). Additionally, there was a significant effect for the number of videos seen on individuals' cooperation with the law scores (Wilk's Lambda = .975,F(1, 567) = 14.473,p= .000). There was only one significant interaction between the number of videos: obligation to obey the law
scores (Wilk's Lambda = .990,F(1, 565) = 5.903,p= .015). This finding provides limited support for Hypothesis 3, as only changes in obligation to obey significantly differed based on whether respondents saw one or both videos.[8]Table 3 Repeated measures ANOVAs—number of videosPre-testPost-testF (treatment effect)F (treatment effect by video group)VariableNumber of videosMean SDMeanSDObligation to obey15.181.475.2631.4520.412**5.903**24.971.375.2521.30Distributive fairness15.411.685.3611.78.3672.47025.141.715.2451.69Procedural justice123.06 4.4323.424.5714.156**.131222.524.1422.964.44Cooperation113.63 2.8413.882.6814.473**.034213.622.6313.892.372 *p< .05, **p< .01.Finally, the ordering of the videos was considered with ANOVAs examining the outcome variables with which order the videos were seen (either officer's perspective first, followed by the citizen's, or the reverse). As Table 4 shows, ordering of the videos had a significant effect on individuals' obligation to obey (Wilk's Lambda = .9.27,F(1, 280) = 22.058,p= .000), procedural justice (Wilk's Lambda = .975,F(1, 280) = 7.131,p= .008), and cooperation scores (Wilk's Lambda = .971,F(1, 280) = 8.288,p= .004). There were no significant interaction effects for the video ordering and outcome variables.[9] This indicates that while seeing two videos in total had an effect for several of the outcome measures, the order in which they were viewed was did not significantly impact individuals' scores.Overall, these results provide negative evidence regarding Hypothesis 4; it does notappear that the ordering of video exposures results in significantly different perceptions.Table 4 Repeated measures ANOVAs—ordering of videos
Pre-testPost-testF (treatment effect)F (treatment effect byvideo group)VariableOrder of videosMean SDMeanSDObligation to obeyCit./Off.5.061.435.411.3922.058**1.222Off./Cit.4.891.305.101.20Distributive fairnessCit./Off.5.421.615.491.582.304.195Off./Cit.4.881.765.001.76Procedural justiceCit./Off.23.12 4.1623.324.437.131**2.085Off./Cit.21.944.0522.634.44CooperationCit./Off.13.52 2.8713.912.528.288**1.304Off./Cit.13.712.3713.882.223 *p< .05, **p< .01.There was no significant effect of viewing the videos on individuals' distributive fairness scores, in any of the ANOVAs examined. Further, confirmatoryt-tests indicated no statistically significant differences in the distributive fairness post-test scores compared to the pre-test scores regardless of treatment condition (see Appendix A). We reserve discussion of these findings for the following section.DiscussionOver two decades ago, the videotaped beating of Rodney King spurred a national conversation about the treatment of minorities and the appropriate uses of force by police (Skolnick & Fyfe, [76]). In recent years, as new names have entered the national lexicon—Michael Brown, Eric Garner, and Walter Scott, among them—the American public, the law enforcement community, policymakers, and criminologists have continued to grapple with these pernicious issues. These issues persist, in partas a consequence of technologies like cell phones, social media, and the Internet. These technologies allow citizens to share their experiences not only with friends and family, but potentially the entire nation. Anecdotal evidence suggests that videos of some encounters negatively affected police legitimacy and fueled civil unrest in American cities (Pyrooz, Decker, et al., [64]). Using a large sample of students, and drawing on theories of procedural justice, cognition, and media effects, we examined whether technology-mediated exposure to a real police–citizen encounter, and variations in these exposure (points-of-view, number of videos, ordering of videos) within that encounter, influenced perceptions of police. Our findings warrant three points of discussion.
First, we found that exposure to these videos influenced perceptions of police. Specifically, viewing video recordings of this specific encounter produced positive changes in perceptions of procedural justice, obligation to obey, and willingness to cooperate with law enforcement. This finding suggests expanding the scope of the process-based model to account for the influence of new technologies. These technologies appear to act as additional sources of legal socialization. Additionally, this finding creates new avenues for research beyond the direct/ vicarious dichotomy within the process-based model. Individuals can now learn of or experience encounters far removed from the actual event. As time passes between events, how does this impact the influence of each experience? Are these experiences naturally reinforcing? We cannot answer these questions, but raise them as a natural consequence of our findings, and future researchers would do well to consider them moving forward.Additionally, future research should work to more finely parse the relative impact of direct, vicarious, and technology-mediated experiences on perceptions of police. This should be coupled with assessments of the long-term influence of these experiences, how the influence of these experiences changes over time (Augustyn, [4]), and how repeated viewing of recordings influences perceptions. Indeed, repeated exposure to negative encounters with law enforcement should result in consistently less favorable views of police (Wolfe, [99]). For example, minorities are consistently more distrustful of police than Whites (Garofolo, [30]; Hindelang, [34]; Schuman, Steeh, Bobo, & Krysan, [72]; Tyler, [86]), and this should be attributable to disparate treatment by criminal justice actors. The current study has focused on general changes in perceptions of police, rather than racial differences in receptivityto these exposures, or heterogeneous treatment effects. Further exploration of these topics is warranted (Kappeler, Sluder, & Alpert, [41]; Tyler, [84]).Second, despite evidence of the impact of these videos on public perceptions of police, we also found little evidence supporting point-of-view, number of videos, or ordering effects. The only treatment effect found was that individuals' obligation to obey the law scores were influenced by the number of videos they saw. No other outcome was impacted significantly by these various treatment conditions. This may be attributable to the fact that the majority of both the officer and citizen videos were capturing the same encounter, except for some minor differences in length. Given that the content of the videos were not substantially different from each other it is relatively unsurprising that there were no effects found based on differing conditions. Additionally, this finding (or lack thereof) may be driven in part by the benign nature of the encounter. Future research would do well to vary and compare encounters containing escalating levels of conflict (e.g. uses of force) and more diverse content to better understand how videos influence future perceptions. As Maguire and colleagues ([50]) recently demonstrated, video simulations of encounters may be one means of doing so.Third, the influence of videos and the technologies used to produce and disseminatethem need not be limited to cell phones or social media; police technologies such asbody-worn or on-officer cameras will likely play a role in the legitimacy of law enforcement as they become more widely adopted. Indeed, on-officer cameras are
already providing video evidence from the officer's point-of-view (Jennings, Fridell, &Lynch, [38]; Ready & Young, [67]; White, [97]). These videos can offer insights into the interactions between citizens and officers, as well as the moments preceding each encounter. Given comparable recording capacities (e.g. camera resolution, audio capturing), we would expect footage from body cameras or other personal recording devices to similarly impact perceptions of police legitimacy. To that end, our findings have relevance for policy. Generally speaking, results here suggest that law enforcement agencies should encourage and train officers to assume and behave as though they are always being filmed. The ubiquity of cell phones and the use of officer-worn cameras reinforces such an assumption (Campeau, [17]). Although there is some skepticism and concern among command staff toward body cameras (Smykla, Crow, Crichlow, & Snyder, [77]), our results also suggest that such footage can, at least in some circumstances, offer opportunities to improve legitimacy and trust with the communities that officers serve (President's Task Forceon 21st Century Policing's, [63]).This study is not without limitations. First, we used recordings of a single encounter, one entailing no substantive use of force. This is consistent with the modal categoryof police–citizen encounters, though the encounter is fairly unique, given the method of filming used by both parties and its conclusion. Thus, we cannot say that the reactions to this video would generalize to other encounters. We focused only on this type of encounter as a preliminary assessment of the consequences of videoexposure for perceptions of police. Although we do not have other encounters to compare to, we anticipate negative encounters (e.g. uses of force or problematic behavior) would likely result in less favorable views of police (Maguire et al., [50]; Skogan, [75]; Taylor, [81]). Future research should consider how exposure to variousencounters and uses of force impact perceptions of police legitimacy. Likewise, the videos used here capture a particular racial dynamic that has featured prominently in the media; parsing how other race and gender combinations influence viewer perceptions is necessary.The videos used here did receive media attention around the time of the incident. Many videos see substantial drops in popularity shortly after "going viral" (Broxton, Interian, Vaver, & Wattenhofer, [13]), suggesting that respondents would, at best, be passingly familiar with these videos months later. Only a few weeks after these videos emerged, media coverage focused on the assassination of two New York City police officers (Mueller & Baker, [57]) and other lethal police–citizen encounters. Respondents had little knowledge of this encounter, as only 7% reported viewing the videos prior to the study, and no more than 10% in a given group reported previous viewership. We do not anticipate that this biased our results, though futureresearch should elaborate on how the framing of videos in the media, and selection into particular forms of media, contribute to individual perceptions of police (Zillmann & Bryant, [102]).Certain limitations are also inherent to the research design and the sample. Although the internal validity of pre–post designs is robust (Campbell & Stanley, [16]), external validity can be a concern. External validity refers to whether observed results may be generalized across persons, places, and times. Certainly,
much of the country is grappling with issues of race, uses of force, and police–community relations (President's Task Force on 21st Century Policing's, [63]; Wolfe &Nix, [100]), and these issues are increasingly caught on film and disseminated via social media (Bonilla & Rosa, [11]). As we reiterate below, we have no reason to believe college students would be uniquely susceptible to cell phone footage of police–citizen encounters. Concerns about learning effects must also be noted. Blocks of questions, and questions within each block in the pre- and post-test portions of the survey, were randomized to reduce the possibility of learning effects.The caveats associated with student samples (e.g. homogeneity and generalizability) must also be addressed. With regard to the former, the university from which the sample was drawn is culturally, racially/ethnically, and intellectually diverse (Reisig, Wolfe, & Pratt, [69]). This diversity is reflected in the sample demographics (see Table 1). With respect to generalizability, recent research suggests that concerns about student samples may be overstated (Druckman & Kam, [24]), and the contributions of student samples to criminological research are well documented (Payne & Chappell, [59]). We can think of no reason why college students would be especially susceptible to the influence of cell phone videos. If anything, we anticipate that students would be less sensitive to the kinds of videos used here, making findings more conservative than may otherwise be anticipated. Nonetheless, given these limitations, until replicated with alternative samples and methods, our findings should be considered with care.In the end, new technologies will continue to be features of everyday life and influence the legitimacy of prominent social institutions. There can be little doubt that such sentiment is applicable to law enforcement, and as the opening quote from a police sergeant demonstrates, officers are cognizant of these technologies and their consequences for perceptions of the police. This study raises questions about the extent to which the process-based model sufficiently accounts for the potential effects of technology-mediated exposure to police–citizen encounters. Results from the present study lends support to this possibility, though much remains to be explored. By integrating insights from the process-based model of policing, media effects, and SCT, we have provided a foundation for future research,and we encourage our colleagues to embrace the practical and theoretical implications of new technologies for criminology and criminal justice.Disclosure StatementNo potential conflict of interest was reported by the authors.AcknowledgementsWe thank Cody Telep, Scott Decker, Mike White, Scott Wolfe, Danielle Wallace, Bryanna Fox, and Rick Trinkner for their comments on earlier drafts of the manuscript. Previous versions of this paper were represented at the 2015 meeting of the American Society of Criminology and the 2016 Michigan State University Interdisciplinary Conference on Cybercrime.Appendix A. Post-hoc T-tests
Table 5 Outcome variable means and two-tailedt-tests by groupPre-test (SD)Post-test (SD)DifferencetObligation to obeyGroup 1—Citizen POV5.198582 (1.48)5.312057 (1.42).11347521.6256Group 2—Officer POV5.159722 (1.46)5.215278 (1.47).0555556.6748Group 3—Cit./Off. POV5.057971 (1.43)5.405797 (1.39).34782614.0521***Group 4—Off./Cit. POV4.888889 (1.30)5.104167 (1.20).21527782.5714**Procedural justiceGroup 1—Citizen POV22.9078 (4.44)23.42553 (4.30).51773052.9185**Group 2—Officer POV23.20139 (4.43)23.42361 (4.85).22222221.0309Group 3—Cit./Off. POV23.11594 (4.16)23.31884 (4.43).20289861.1662Group 4—Off./Cit. POV21.94444 (4.05)22.62500 (4.44).68055562.4512*Distributive fairnessGroup 1—Citizen POV5.347518 (1.71)5.425532 (1.77).0780142.8828Group 2—Officer POV5.465278 (1.70)5.298611 (1.80)−.16666671.6997Group 3—Cit./Off. POV5.420290 (1.61)5.492754 (1.58).0724638.8443Group 4—Off./Cit. POV4.875000 (1.76)5.006944 (1.76).13194441.2809CooperationGroup 1—Citizen POV13.28671 (3.00)13.52448 (2.81).23776221.6417
Pre-test (SD)Post-test (SD)DifferencetGroup 2—Officer POV13.97222 (2.70)14.23611 (2.50).26388891.9325Group 3—Cit./Off. POV13.51799 (2.90)13.90647 (2.51).38848922.5059**Group 4—Off./Cit. POV13.71329 (2.40)13.88112 (2.22).16783221.4396*p<.05, **p<.01, ***p<.001.Table 6 Outcome variable means and two-tailedt-tests by number of videosPre-test (SD)Post-test (SD)DifferencetObligation to obeyOne video only5.178947 (1.47)5.263158 (1.45).08421051.5596Two videos4.971631 (1.37)5.251773 (1.30).28014184.6723***Procedural justiceOne video only23.05614 (4.43)23.42456 (4.57).36842112.6333**Two videos22.51773 (4.14)22.96454 (4.44).44680852.6965**Distributive fairnessOne video only5.407018 (1.68)5.361404 (1.78)−.045614.6874Two videos5.141844 (1.71)5.244681 (1.69).10283691.5299CooperationOne video only13.63066 (2.84)13.88153 (2.68).25087112.5255*Two videos13.61702 (2.63)13.89362 (2.37).27659572.8614***p<.05, **p<.01, ***p<.001.
Appendix B. Sample Image Stills from Citizen and Officer Cell Phone VideosAppendix C. Instrument-Outcome MeasuresObligation to obey the law(1) You should accept police decisions even if you think they are wrong.(2) You should do what the police tell you to do even if you disagree.Procedural justice(1) Police treat citizens with respect(2) Police take time to listen to people(3) Police treat people fairly.(4) Police respect citizens' rights.(5) Police are courteous to citizens they come into contact with.(6) Police make decisions based upon the facts.(7) Police explain their decisions to the people they deal with.(8) Police make decisions to handle problems fairly.Distributive fairness*(1) Police give minorities less help because of their race.(2) Police provide better services to wealthier citizens.Cooperation(1) Call the police to report a crime.(2) Call the police to report an accident(3) Report suspicious activity near your house to the police.(4) Provide information to the police to help find a suspected criminal.*Reverse coded