THESI'S Date llllllllllllllllllllllllllllll'lllllllllllllllllllllllllll 293 01568 4727 This is to certify that the thesis entitled THE EFFECT OF "COPS" ON UNDERGRADUATE STUDENTS ' PERCEPTIONS OF LAW ENFORCEMENT presented by Sean Baker has been accepted towards fulfillment of the requirements for H. S . degree in Wstice (m Maj rofessor Hay 1/97 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Mlchlgan State Unlverslty PLACE N RETURN BOX to roman thb checkout from your rooord. TO AVOID FINES return on or bdoro duo duo. ll DATE DUE DATE DUE DATE DUE MSU Is An Milan-tho ActIon/Equol OppommIty InotItutIon m ulna-9.1 THE EFFECT OF “COPS” ON UNDERGRADUATE STUDENTS’ PERCEPTIONS OF LAW ENFORCEMENT BY Sean Baker A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE School of Criminal Justice 1997 ABSTRACT THE EFFECT OF “COPS" ON UNDERGRADUATE STUDENTS’ PERCEPTIONS OF LAW ENFORCEMENT BY Sean Baker This thesis examines the relationship between college students’ perceptions of police officers and the reality- based television program “COPS." An experiment was conducted using students enrolled in a undergraduate Introduction to Criminal Justice course (n=150). Generally, students hold positive opinions of the police.~ Evidence was discovered supporting cultivation theory. It was found that for the television program “COPS” media effects are not direct. A one-half hour exposure to “COPS" is not sufficient enough to produce a change in student attitudes about police officers. Also, respondents with heavy previous exposure to “COPS" tend to hold a more positive opinion of law enforcement than those with minimal prior exposure. ACKNOWS I would like to thank everybody who made this accomplishment possible including: Vincent Hoffman, Christina DeJong, Peter Manning, Folu Ogundimu, Ken Jackson, JCJS, my parents, and especially Jennie for all of her support. iii For Grampa Swenson iv TABLE OF CONTENTS List of Tables ........................................... vi INTRODUCTION .............................................. 1 CHAPTER 1 Literature Review ......................................... 4 Media Effects Research ................................. 4 Public Opinion and the Police .......................... 9 Reality-Based Television .............................. 12 CHAPTER 2 Methodology .............................................. l4 Instrument ............................................ 15 Experiment Design ..................................... 17 CHAPTER 3 Data Analysis ............................................ 19 Data Description ...................................... 19 Factor Analysis Results ............................... 30 Bivariate Comparisons ................................. 33 Multivariate Comparisons .............................. 36 CHAPTER 4 Discussion and Conclusion ................................ 41 Recommendations ....................................... 45 Exploratory Considerations ............................ 45 Summary ............................................... 48 APPENDICES Appendix A—Survey Instrument .......................... 50 Appendix B—Variable Receding .......................... 56 Appendix C-Statement Means by Group ................... 57 Appendix D-Statement Means by Time .................... 58 Appendix E—Factor and Additive Score Mean Comparisons..59 LIST OF REFERENCES ....................................... 63 Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table LIST OF TABLES l-Respondent Characteristics by Group .............. 2-Significant Demographic Group Differences ........ 3-Statement Means for Groups for All Cases ......... 4-Confirmatory Factor Loadings by Group ............ 5-Factor Score Means by Group Pre-test vs. Post—test .................................... 6-Factor Score Means by Time Control vs. Experimental ................................. 7-Post-test Significant Statement Means by Group 8-Post-test Significant Factor Means by Group ...... 9-Post-test Significant Additive Means by Group 10—Mean Additive Scores by Reality—Based Exposure ll-Mean Additive Scores by CJ TV .................... 12-Mean Additive Scores by Perceived Realism ........ 13-Factor Score Means by Group Pre-test vs. Post—test .................................... 14-Factor Score Means by Time Control vs. Experimental ................................. 15-Additive Score Means by Group .................... 16-Additive Score Means by Time ..................... vi 20 22 24 31 35 36 .38 4O .40 .47 47 48 59 6O 61 61 INTRODUCTION The mass media, especially television, are one of the most pervasive social structures within our society. Daily, television sets are turned on about seven hours, and individuals watch about three hours (Surette 1992; Morgan and Signorielli 1990). A substantial amount of this viewing includes images of the criminal justice system, where 20 percent of local news, 12 percent of national news, and 25 percent of prime time programming consists of some aspect of criminal justice (Surette 1992). Since many people in the United States rarely are directly exposed to criminal justice agencies, they receive much of their knowledge and impressions of it through the mass media. Criminal justice agencies supervise and receive power from the general public. Therefore, policy, procedures, and standards of operations are defined by society suggesting that social researchers need to explore perceptions in addition to actual functions of the system. The process by which people obtain a sense of the world is often labeled social construction of reality (Berger and Luckman 1967). It refers to the manner in which people configure their own reality through the interplay of individual knowledge and social interactions. As a result, a person acquires agreement and attaches meaning to situations and experiences (Macionis 1991). The mass media, notably television, is a major contributor to an individuals reality construction (Surette 1992). Other sociological theories describe the relationship between media, crime, and justice. Reflection theory, at its most basic model, asserts that “culture is the mirror of social reality" (Griswold 1994, p.22). Television programs and images are conceived as salient cultural indicators. The relationship between images and popular culture is more complex. Mediated messages by nature produce a warped and distorted description of the world which is inherently biased by the constraints of the presentator (Shoemaker and Reese 1996). Societal values, ideologies, justice philosophies, and humanitarian beliefs are projected to the masses by the media. As a result normative structures are internalized, and equated with common sense knowledge (Cantor and Cantor 1992). More specifically, the criminal justice system is legitimized by the mass media. Creechan (1992) concludes that an increase in newsprint statements about unlawful violence results in a less critical View of police officers, and one outcome is increased acceptance of social authority which formalizes dominate control structures. By examining the association between media content and the effect it has on the public's perception of social reality, one obtains an improved understanding of society in general. This study's purpose is to analyze the dynamic relationship between mediated messages and public perceptions of authority figures. RESEARCH QUESTIONS To investigate potential media effects, I ask if reality—based television programs about criminal justice agencies affect public perception of the criminal justice system. Specifically, do positive portrayals of police in the television program “COPS” lead to positive opinions of law enforcement? This analysis focuses on interpersonal aspects of policing, however, other dimensions exist an will be discussed in the literature on public opinion and police section. Seven aspects of police functions will be examined. They are job difficulty, effectiveness, public protection, social skills, power, degree of performing social intervention functions, and planning. Regarding higher education as a mediated message, I ask to what extent does taking an introduction to criminal justice course affect public perceptions of law enforcement? If a change exists, do students' perceptions become more congruent with the content of the lectures and reading materials? Chaptor 1 LITERATURE REVIEW MEDIA EFFECTS RESEARCH The possible influence of television on public opinion has long been debated. Previous research has found evidence for and against a television effect. However, the paradigm is shifting towards the position that television effects exist. Many authors determined that television does influence public perception. One such author is George Gerbner, a pioneer of television analysis. He and his colleagues worked extensively on their “Cultural Indicators" research during the 1960's and 1970's (Gerbner and Gross 1976). This team observed the content of television dramas, and determined the influences the messages had on viewers. They labeled this process “cultivation analysis.” Looking at public perceptions of crime, they found that heavy viewers of television violence and disorder came to believe that incidents of violence were greater than actual rates (Gerbner and Gross 1976). They called those responses “television answers.” Their principle conclusion is that heavy viewing of television brutality cultivates people into believing the world is a mean and violent place. This hypothesis also holds for newspapers (Creechan 1992). Gerbner and his colleagues extended cultivation theory into two new mechanisms: mainstreaming and resonance (Gerbner, Gross, Morgan, and Signorielli 1980). Mainstreaming is the idea that heavy viewers’ perceptions of social reality merge into similar attitudes and beliefs, regardless of their different backgrounds and environments. Television produces a consensus reality in the minds of heavy viewers. Resonance is defined as: When what people see on television is most congruent with everyday reality (or even perceived reality), the combination may result in a coherent and powerful “double dose" of the television message and significantly boost cultivation (Gerbner, et al. 1980, p.412). Currently, the cultural indicators project continues to investigate the ongoing long term shifts in audience beliefs, attitudes, and behaviors. These researchers examine and document television content and survey respondents after air- time (Signorielli and Morgan 1990). Criticisms of cultivation theory exist. First, the causal process is unclear. By testing for gradual variations in public opinion, they fail to address the issue of who changes what. Do public preferences alter television content, is the opposite true, or is it a reciprocal effect (Slater and Elliott 1982)? Second, there is a lack of evidence supporting cognitive processes relating the connection between first and second order beliefs (Hawkins and Pingree 1990). A first order belief is regarded as the matching of mediated reality to social reality (e.g. feelings that violent crime is rising). Second order beliefs are only implied by media content (e.g. fearful of walking alone at night because violent crime is on the rise). Second order cognitions dominate individual behaviors and attitudes (Hawkins and Pingree 1990). Research has found little evidence connecting the two. In fact, the two types of beliefs are independently affected from television exposure (Hawkins and Pingree 1990). Third, cultivation theorists assume that viewing is non-selective. However, this may be inaccurate. The amount of attention one invests in television viewing fluctuates across time. People often “tune-out" television which makes it difficult to determine which messages affect perceptions (Condry 1989). A final criticism asserts that individual demographic differences and direct experiences may impact television effects on public perceptions (Slater and Elliott 1982). A person's existing world view is ignored by cultivation theorists. Slater and Elliott (1982) suggest that perceived program realism, not amount of exposure are Gerbner proposes, is the best predictor of television influence. Television has a greater impact on respondents who believe that television programs represent the real world. Some people have no direct experience with the criminal justice system. They lack the personal knowledge to compare with mediated messages. In this situation, Those who lack firsthand experience with the legal system will probably construct their mental images of it from the media’s disproportionate coverage of violent and sensational crimes and its focus on law enforcement (Hans and Dee 1991, p.136). Much of the research on mass media effects has examined the relationship between exposure and voting patterns. The agenda-setting hypothesis is a foundation of media effects research during presidential elections (McCombs and Shaw, 1972; Weaver, Grabner, McCombs and Eyal 1981). Agenda— setting states that problems which secure salience in national news become the most important in the public’s minds (Iyengar and Kinder 1987). Findings suggest that agenda- setting is confirmed during presidential elections (McCombs and Shaw 1972; Weaver, Grabner, McCombs, and Eyal 1981). Also, research on economic issues (Iyengar and Kinder 1987), public fear of false crime problems (Brodt 1987), views of sentencing (Roberts and Doob 1990), and influences on prosecutorial discretion (Prichard 1986) supports the agenda- setting hypothesis. A major facet of agenda—setting research involves issue obtrusiveness. Opinions regarding issues with minimal direct personal experience (unobtrusive) are more likely than ones with extended personal experience (obtrusive) to change form media exposure (Weaver, Grabner, McCombs, and Eyal 1981). A third aspect of media effects research involves the distinct influence across media sources. News commentaries and experts produce the most notable impact on public perceptions (Page, Shapiro, and Dempsey 1987; Jordan 1993). Further, the stylistic form of an information source may alter media effects (Leff, Protess, and Brooks 1986). “Like a gifted comic it is not the subject matter that counts but the way he (information source) tells it" (Lewis 1991, p.35). Most communication researchers found evidence supporting the existence of mass media influence on perceptions. Recent studies show that police television programming influence attitudes toward law enforcement (Carlson 1985). The current analysis specifies the cultivation perspective. A major criticism of cultivation theory lies in the inability to address the process where content acquisition evolves into long term memory (Potter 1993). As discussed earlier, the link between first and second order effects is unclear. To address this problem I employ Tapper's (1995) re—conceptualization of cultivation theory. He suggests that theorists need to further examine the choice of programming individuals watch and perceptions of program realism. Borrowing form agenda-setting research this examination addresses the obtrusive/unObtrusive issue (Weaver, Grabner, McCombs, and Eyal 1981). Utilization of this concept assists to determine the specific nature of media effects. Methodologically, Oliver (1995) recommends that experimental formats will clarify the effects of reality—based programming on attitude formation and re- formation. Before considering methodological issues, discussions of public opinion and the police and reality- based programming are necessary. PUBLIC OPINION AND THE POLICE The existing findings on public opinion and the police are extensive (see Walker 1992; Radelet 1995). Generally, a majority of Americans have positive attitudes toward police officers. Policing tends to receive higher ratings than many other professions. However, opinions vary by demographic characteristics. Racial minorities consistently hold less favorable attitudes toward police. Young people also have less favorable attitudes. Poor, less educated, and crime victims rate police services lower than the general public. Large city residents tend to possess poorer opinions of policing. The only gender differences appear to be that females tend to View police work as more service orientated than males. These generalizations are consistent over time (Maguire and Pastore 1994). These relationships are more complex than aggregate statistics demonstrate. Contextual variables need consideration when examining public perceptions of law enforcement. The nature of prior experiences with police officers is an important factor when considering citizen attitude toward police. Individuals who perceive that they receive unfair or unsatisfactory services are more likely to hold negative opinions (Correia, Reisig, and Lovrich 1996). Perceived treatment may interact with race and age variables. 10 Some theorists suggest that minorities are socialized to fear officers resulting in less favorable sentiments during police encounters (Dunham and Albert 1988; Sullivan, Dunham, and Albert 1987). Thus, race effects may be explained by a neighborhood culture which is anti-police (Decker 1981). Youths tend to value individual liberties and commit a disproportionate amount of crime, therefore, they may experience an increased amount of negative contacts with police (Gaines, Kappeler, and Vaughn 1994). Frequency of contacts is inversely related to positive opinions (Walker 1992). However, this relationship needs clarification. The most frequent police/citizen encounters are traffic related (Radelet and Carter 1994). Receiving a traffic citation is one of the largest producers of negative public opinion of policing (Correia, Reisig, and Lovrich 1996; Cox and White 1988; Radelet and Carter 1994). Thus, it is apparent that contact intensity may actually indicate that high traffic citations cause negative opinions. Individual and contextual differences may influence the results of this study. Thus, these factors will be considered during data analysis. It is suspected that the current subjects' attitudes will vary accordingly to previous research on public opinion and the police. As stated earlier, these dimensions of policing are purely interpersonal. Bitner (1990) is a leading theorists on functions of the police. He states that most perceptions of law enforcement are derived from the broad aspect of ll upholding the law. Bitner (1990) further states that policing is a tainted occupation. For most people the police are both feared and admired. While attempting to control the undesirables in our society, the police have the power to withdraw rights of almost every citizen (Bitner 1990). This dichotomy produces the fear/admiration perception of the police. Deployment and other investigative activities are seeded within societal prejudices. The police have been described as an instrument of political actors within the dominate social structure (Bitner 1990; Manning 1977). A third structural trait of police work discussed by Bitner (1990) involves the legitimization of force. Three themes emerge from this discussion. The public legitimizes use of force as a form of self-defense officers must employ when confronting violent suspects. Second, some force is necessary for custodial treatment of incarcerated or deranged people. Finally, force is legitimized through institution of an authoritative agent like a law enforcement agency (Bitner 1990). The final argument is most relevant to this discussion. A society deems it necessary to give up personal freedoms for the protection of all members. In this sense the police are conceived as administrators of justice with a heightened freedom to Violate personal freedoms for the sake of social well being (Bitner 1990). Bayley (1979) provides a good summary of aspects of police work beyond the maintenance of public order. The following is a attenuated list of police functions: 12 “protection of life and liberty, enforcing criminal laws, investigating criminal offenses, patrolling public places, advising about crime prevention, guarding persons and facilities, regulating traffic, issuing ordinances, supervising jails, impounding animals or lost property, and promoting community crime prevention activities" (p.111-112). Even though the current analysis focuses on limited aspects of police, the interpersonal dimensions are valid and important components which need examination. REALITY- BASED TELEVI S ION Reality television programs like “COPS” are depicted as true representations of police work. However, these programs are actually socially constructed masquerades (Andersen 1994; Manning 1996). Most employ similar editing strategies which result in an artificial action—packed dramatizing of policing. Andersen (1994) states that producers coach officers to perform in a “televisually appealing tone” (p.9). The television verit technique reduces the social distance between viewers and program events (Manning 1996). As a result, exposure becomes an interactive affair. Networks encourage this process to enhance internalization of meaning and boost ratings (Caldwell 1995). The “COPS" format address the third aspect of media effects research. It heavily utilizes expert police commentaries during each episode. The innovative and stylish technique is consistent with the conclusion that extravagant television segments are more l3 likely to impact public opinion (Leff, Protess, and Brooks 1986). The reasons for including “COPS" in this analysis are fourfold. First, “COPS" is the pioneer of the cinema verit technique, which consists of the use of hand held cameras to reproduce live audio and visual representations. Second, it's popularity soared since the premiere in 1989. Third, media critiques have honored “COPS" with four Emmy nominations and bestowed it Best Reality Show at the 1993 American Television Awards. Finally, it has inspired numerous “imitator” programs across diversified occupations and situations (Fox-29 WFTC 1995). In sum, “COPS" is the quintessential reality—based television program. Chapter 2 METHODOLOGY An experimental format is well suited to examine immediate and direct impacts of television on viewers' perceptions (Maxfield and Babbie 1995). Using a laboratory setting enables this analysis to confidently determine that the change in subjects’ opinions is exclusively a product from viewing the television clips. Previous cultivation research, in general, analyzed media content and subsequently ' surveyed respondents' attitudes about the preceding messages. During an experiment the researcher is in complete control of the stimuli, thus, the causality problem of prior theorists does not apply. Therefore, the present analysis uniquely contributes to the understanding of the interaction between mass media and audiences. An experiment was conducted using registered students in an introduction to criminal justice course offered Fall term, 1996 at Michigan State University. Two hundred students were enrolled for this class. The stimulus l4 15 consisted of 27 minutes of pre-recorded video clips of the television program “COPS". It was created from a random sample of “COPS" during the summer of 1996. The clips are consistent with previous content analyses of reality—based programming (Oliver 1994). Oliver (1994) states that a majority of episodes result in successful police resolutions. Police officers on “COPS" and similar reality-based police programs are portrayed as effective criminal investigators and as performing social intervention functions (Oliver 1994). Also, episodes are often orientated toward social work functions of police work (Hallett 1995). Since higher education instruction is a form of mediated messages, I attended class lectures and examined the course textbook to determine the substantive content of the subjects' class experience. INSTRUMENT The questionnaire (See Appendix A) contains 35 statements asking respondents to rate their opinions on a five point Likert type scale ranging from “strongly agree" to “strongly disagree." It addresses the same issues presented in the content of the television clips. These concepts are: difficulty of police work, police effectiveness, police officers' concern for the general public, degree of police functioning as social workers, police social skills, amount of power needed by law enforcement, and police officer planning. These constructs were devised directly from the 16 themes within the stimulus itself. I have personally viewed many “COPS” episodes, and I believe that the stimulus is a valid sample. Also, previous content analyses of “COPS” have found similar themes (Hallett 1995; Oliver 1994). The questionnaire taps each variable with four to seven statements. Seven of the 35 items were borrowed from existing surveys of attitudes about policing (Klyman and Kruckenberg 1974; Rundrquist and Sletto 1936; Tuohy and Wrennall 1995). The instrument measures subjects' overall attitudes about law enforcement. Surveys limit the number of questions one can ask and still obtain reliable data, therefore, I limited the concepts to seven. To assist in Tapper's (1995) re-conceptualization of cultivation theory the questionnaire requests the subjects to list which and how much crime related television programs they watch. Issue obtrusiveness (Weaver, Grabner, McCombs, and Eyal 1981) is measured by surveying personal contacts with police and desire to become a criminal justice practitioners. Numerous demographic questions are also included. The instrument was pre-tested to assess its reliability. The pre—test group consisted of students enrolled in an undergraduate criminal justice research methods course. Originally, eight statements were written to measure each concept. A confirmatory factor analysis was conducted on the pre-test responses. Based on the results, items with poor factor loadings were eliminated from the instrument. Items were dropped with loadings below .50. 17 The edited instrument has an acceptable reliability score (alpha = .80), thus, a strong correlation exists between all items on the scale. I am confident that the edited questionnaire accurately measurers each concept. EXPERIMENT DESIGN In the classroom, the subjects were pre-tested at the end of the fifth week of classes. Two weeks later the stimulus was presented and the subjects responded to the post—test questionnaire. Prior to that event, a stratified approach was used to appoint subjects to control and experimental groups. Stratification was based on respondents’ major, television viewing habits (choice of programming), and desire to work within the criminal justice system (obtrusive/unobtrusive). Specifically, one-half of the students who are criminal justice majors, watch reality- based police programs, and plan to become criminal justice practitioners were assigned to each group. The remaining respondents were randomly assigned to control or experimental groups. At the end of the ninth week of classes the subjects were separated during the class period. The control group was moved to another room, completed the questionnaire, and were debriefed. Subjects in the experimental group received the stimulus, then responded to the questionnaire, and finally were debriefed. Before analysis the response categories were coded ordinally from 1 to 5, representing most negative to most positive opinions, respectively. 18 Confirmatory factor scale reductions were utilized to combine index items into corresponding variables. Seven variables should result. This process was repeated for each of the three groups. Comparisons of control, experimental, and pre-test group scores will allow this investigator to ascertain to what extent mediated messages affects subjects' perceptions of law enforcement officials. The new variables were categorized by standard deviation units, resulting in discrete values. Analyses of variances should sufficiently compare group attitudes. Initially, pre-test scores were compared to determine of pre-existing group differences were present. Next pre-test and post—test differences for both groups were compared. Finally, examinations of the impact of demographic differences and television viewing habits were conducted. To accomplish this group means were compared across demographic and viewing habit categories. For example, the means of heavy and light viewers were contrasted to determine if differences exist. Chapter 3 DATA ANALYSIS DATA DESCRIPTION Scores for demographic, prior police contact, and television viewing habit variables are listed in Table 1. Limited cases in some variable responses were recoded into dichotomous response categories (See Appendix B for recoding protocols). At face value all demographic characteristics appear similar across groups. Respondents in all groups tend to be about 19 years old and Caucasian. Roughly one half are freshman and they are equally likely to be male or female. About 55% are employed, however, very few currently hold positions within the criminal justice system. Just over one half of the respondents plan on becoming criminal justice practitioners after graduation, and of those who stated a preference a majority desire to work as law enforcement officials. Across all three groups, respondents’ previous contact with law enforcement officials shows similar results, except for the average number of police contacts. As displayed in Table 1 the pretest and control groups have analogous means, however the experimental mean is considerably lower. A vast majority of respondents have had at least one experience 19 20 Table 1 Respondent Characteristics by Group Question Variable Protest Post-test Post-test Number Scores Control Experimental (n=150) Scores Scores (n=69) (n=60) Means 7 Age. 19.12 19.13 18.97 11a Number of police contacts. 631* 5.44* 3.63“ 11c Nature of last contact. 4.52 4.86 4.27 13a Number of police known. 2.55 2.63 2.53 14 Number of police who are relatives. .30 .28 .28 8 Hours of TV viewing a week. 10.87 11.13 10.27 9 Hours of TV viewing about CJ a week. 1.93 2.33 1.77 1 Perceived program realism.l 3.32 W2 5 Percent male. 56.7 52.2 50.0 5 Percent female. 43.3 46.4 50.0 2 Criminal justice major. 35.3 40.6 35.0 3 Freshmen in college. 45.3 43.5 46.7 15 Currently employed. 53.3 58.0 58.3 15a Currently employed in C] field. 7.3 5.8 3.3 16 Planning CJ employment. 553* 60.9“ 50.0 11 Experienced contact with police. 88.6 97.1 88.3 13 Personally know any police officers. 70.0 80.9 73.3 14 Relatives who are police officers. 22.7 21.7 20.0 10 Regularly view C] TV programs. 32.7 43.5’ 21.7* 9a Exposure to C] reality based 55.3 52.2““ 48.3* Prom 1 Experimental group only. 2 These questions are dichotomous and coded as 0=no and l=yes. Therefore, the mean represents the percent who responded ’yes’. * T-test Significant at <.05. 21 with an on—duty police officer, and around 75 percent personally know an officer. Only one-fifth of the respondents have relatives who are police officers, and for all respondents the average number is well below one. Considering only those who have relatives who are police officers, the average number is slightly over one. At face level television viewing habits are also consistent across groups. On average, respondents watch between ten to eleven hours of television a week and about two hours are spent viewing programs which involve the criminal justice system. Results from the regular viewing of topical criminal justice programs vary across groups. On average, the control group has the highest percentage (43.5%) of respondents who regularly watch criminal justice type programming. Next is the pretest group with 32.7 percent, and 21.7 percent of the experimental group stated that they consistently watch television programs regarding criminal justice. As an initial analysis these figures describe the typical respondents in the experiment. It is necessary to determine if any statistical differences exist between pretest, control, and experimental groups. To accomplish this, t-tests were performed to compare group means. Means for all demographic, police contact, and television viewing variables were compared across the three groups. Three separate t-tests were conducted to determine if the means statistically vary between groups. For example, 22 t-tests on average age of pretest and control, pretest and experimental, and control and experimental groups were carried out. results from these tests, significant differences exist between groups. may affect the outcome of this analysis. Since Table 2 lists the significant and indicates that some These results Table 2 Significant Demographic Group Differences Variables Comparison T-test Groups Si ' canoe Experienced contact Pretest .03 with police. Control Number of police Pretest .05 contacts. Control Number of police Pretest .01 contacts. Experimental Number of police Control .05 contacts. Experimental Exposure to Cl reality based Control .01 programm. Experimental Regularly view C] Control , .01 ‘ TV programs. Experimental pre-existing differences exist between groups, those differences may effect the final results of the experiment and produce erroneous conclusions about the relationship between prior police contact, perceptions of law enforcement. respondents have, on average, contacts with police officers. television viewing habits and For example, post—test experienced less personal Previous research has found that frequency of contacts is inversely related to positive opinions (Walker 1992). show that experimental subjects’ have an elevated opinion of law enforcement, Therefore, if the final results it may be an artifact of the pre-existing -111 .1 I... ‘1__ I . | 23 group differences. Respondents in the control group report that 52.2 percent have been exposed to reality based criminal justice television programming. This is only four percent more than the experimental group’s exposure. Hopefully, the minimal difference will not affect the results of this analysis. The difference between control and experimental groups’ regular viewing of criminal justice based programming is substantial. Respondents in the control group are twice as likely to consistently watch television programs regarding some aspect of our criminal justice system. Since the number of experimental respondents in this category is low, any meaningful analysis is not practical. Generally, it seems that the combination of stratified and random sampling strategies produced uniform groups. The existing differences may cause some problems and will be considered during the final analysis. Respondents hold above neutral to positive opinions of policing and law enforcement officials. Table 3 displays the means for all public perception of the police statements. The mean for all statements is about 3.4 on a scale from 1 to 5 where 1 represents most negative opinion and 5 represents most positive. The difficulty of work construct has the highest mean followed by the public protection statements. Next is the degree of officer planning with officer social skills, amount of necessary power, and social intervention functions clustered just below the protection variables. Mean attitudes regarding 24 Table 3 Statement Means by Group for All Cases (n=150) Variable Pretest Control Experimental DHTI 426 404 4J8 DHTZ 396 4J7 408 DIFF3 3.31 3.30 3.30 DET4 3%) 406 4a) DIFFS 4.12 4.33 4.30 DIFF ALL 3.91 3.98 3.97 EFFl 2.02 2.01 2.43 1§T2 285 304 3J3 EFF3 2.75 2.78 2.85 13T4 296 304 328 EFF ALL 2.65 2.72 2.92 POWER] 3.86 4.00 3.75 POWER2 2.91 2.91 2.85 POWER3 3.01 3.01 2.93 POWER4 3.35 3.35 3.45 POWTRAQI. 328 332 325 PREPI 3.01 3.19 3.30 PREP2 4.03 3.75 3.65 PREEB 364 364 365 PREP4 3.42 3.41 3.50 PREPS 3.48 3.50 3.50 PREP ALL 3.52 3.50 3.50 PROTECT 1 3.35 3.59 3.38 PROTECI‘Z 3.92 3.90 3.75 PROTECI‘3 3.89 3.94 4. 13 PROTECT 4 3.32 3.51 3.47 PROTECT ALL 3.62 3.74 3.68 SOCIAL] 3.18 3.29 3.10 SOCIAL2 4.42 4.51 4.45 SOCIAL3 2.39 2.58 2.57 SOCIAL4 3.19 3.33 3.12 SOCIALS 3.85 3.88 3.98 SOCUU6 351 304 327 SOCIAL7 3.15 3.29 3.20 SOCIAL ALL 3.38 3.42 3.38 SOCW 1 3.78 4.01 3.77 SOCW2 3.17 3.28 3.20 SOCW 3 2.21 2.51 2.30 SOCW4 3.30 3.59 3.42 SOCWG 344 371 363 SOCW6 3.65 3.88 3.78 SOCW ALL 3.26 3.50 3.35 ALL MEAN 3.37 3.45 3.43 25 police effectiveness are considerable lower than the other statements. Before discussing statistical findings it is necessary to describe the conceptual relationship between the stimulus and the questionnaire. The stimulus consisted of eight prerecorded television clips of the program “COPS." The clips ranged from one minute to seven minutes long, and the total time was 27 minutes. Specific aspects of the stimulus relate to the perceptions of policing constructs used in this experiment. The first started with a panoramic shot of a conference room filled with twelve police officers. The caption at the bottom of the images read “Roll Call.” For about one minute the camera showed interested but relaxed officers discussing the latest criminal activity and successful techniques to control crime. This clip addressed officer preparedness. The roll call and discussions of current criminal activity represents police officer job planning. The second clip is the longest with a running time of seven minutes. It started with two officers in a patrol car explaining that a nearby apartment complex is often the location for the sale and consumption of illicit substances. Shortly after they pulled over a woman who failed to drive her car at a green traffic light. Upon checking her license and vehicle registration they discovered that she has a warrant for her arrest. Subsequently, they searched her car, and in a few seconds they found paraphernalia, what looked to be illegal drugs, and a fair amount of money. In the 26 meantime, a man (owner of the car) walked to the scene and claimed to be the woman’s friend. The officers talked with the man and woman and ascertained that the drugs actually belonged to the man. Unfortunately, he did not confess so they were forced to continue their arrest of the woman. The officer sympathized with woman’s predicament and attempted to console her. At the end of the clip the officers drove away talking about how tragic it was that the man let his friend “take the fall for him." In the second video clip, the officers efficiently found illegal drugs and paraphernalia in a suspects' vehicle. It is possible that the search was time consuming, but the representation on “COPS" lasted only a few seconds. Also, this clip displayed two officers catching a suspect and quickly bringing the situation under control. The third clip started with two officers on patrol and a domestic disturbance call came over the radio. When the officers arrived they discovered that a thirteen year old boy was assaulting a three year old boy who threw mud on him. The officer explained to the teenager that fighting with such a young child is not appropriate. Also, the officers gave some supportive advice to the teenager about diplomacy skills. Prior to leaving, the officers discussed the situation with the parents and suggested that they should attempt to become better neighbors. As the officers drove away they talked about how many times they are called for non—criminal activities and how they often they “bear the 27 responsibility to raise this city’s kids.” This episode ran for three minutes. Next, the same officers were waved down by a woman in the street. She informed them that an elderly man is lost and confused. They talked with the man and obtained the name of a friend from him. After phoning the friend and discovering the elderly man's address, the offices gave him a ride to his residents. Throughout the three minute clip the police were very helpful in getting the confused man back home. Both the third and fourth parts of the stimulus attempted to address the social intervention and social skills aspects of police work. In each of these representations the officers resolved non-criminal situations in a very helpful and diplomatic fashion. When the officers advised the teenager (third) and helped the elderly man home (fourth) not once did they show any resentment or dissatisfaction with this component of their jobs. In fact, they seemed very happy to assist in both situations. Both of these representations showed that the police care about the general public. By helping the man home and the comment about “raising this city's kids" the officers were portrayed as concerned officials with the duty to assure the safety of the general public. The fifth clip lasted about three minutes and was set during the night. At the onset, the patrol officer reports to the camera operator how dangerous his job is. He was 28 telling a story about a shooting incident that he was involved in when a disturbance call was radioed. The officer arrived on the scene where two men were fighting in the street. He was forced to physically break up the fight, and resolved the situation. Similar to the previous “COPS" clips the next one started inside a police car with the officers talking about their job. A shooting call came over the radio in which a police officer was involved. Hearing the sirens, and seeing the lights the viewers were exposed to a graphic crime scene. As the police car pulls up an ambulance raced away. Next, the cameras showed an image of a person laying in the grass with blood spattered around the scene. For the final minute of this four minute clip the camera focused on the officer who explained the situation to the audience and stated that the injured officer will be fine. Also, he emphasized the dangerousness of the job by recalling some of the violent situations which he had experienced. The fifth and sixth clips address the difficulty of police work. Both started with an officer stressing the violent situations which the police are confronted with while working. In the fifth situation, the officer had to physically separate two fighting adults and constrain them. In the sixth clip, the audience was exposed to a radio call announcing police involvement in a shooting, and graphic displays of the aftermath of this incident. These portrayals emphasize the dangerousness of policing, and 29 hence justify the use of force in these potentially volatile situations. The seventh presentation started with an officer on night patrol. He saw a number of people loitering outside a liquor store. He pulled up and spoke with the obviously, intoxicated people. Instead of arresting them on a minor offense, the officer instructed them to pick up all the garbage in the parking lot for their “punishment.” As the patrol car drove away, the officer explained to the audience that many of those people were the “town drunks" who had been previously arrested many times. He further stated that they do not belong in the criminal justice system and that is was he why “gave them a break.” This clip showed an officer using his discretionary power to exclude some trivial suspects from inclusion into the criminal justice system. The officer showed genuine concern for the loiters and attempted to resolve the situation without official involvement. Finally, the same officer was patrolling near a known illegal gambling establishment. He investigated the building and determined that illegal gambling was most likely occurring at that time. After knocking on the door many people fled from the scene. The officer caught one of the assailants and questioned him. It was discovered that the suspect was a recent immigrant to this country, therefore, the officer informed him that unsanctioned gambling was illegal in the U.S. As the officer walked back to his patrol 30 car he told how distressing it is that a few days before the current incident a person had been murdered over a small argument at this house. The last clip portrayed law enforcement officials as both compassionate and possessing competent investigative skills. The officer correctly determined that illegal gambling was occurring at the suspected house. After breaking up the situation, he advised the newcomer to this country about gambling laws and the recent violence which had occurred at the same residence. This shows that the officer had a concern for the welfare of the suspect. FACTOR ANALYSIS RESULTS Confirmatory factor analyses were performed on each of the concepts measuring public perception of policing. The results are listed in Table 4. These groupings are based on the content of the stimulus and other content analyzes of reality—based police television programs (Hallett 1995; Oliver 1994). Overall, most of the factor loadings are above the accepted cut off point of 0.5. The factor loaded values across groups are consistent. Also, items within each concept are generally, clustered around common values. These results support the notion that each set of statements is actually measuring the associated concept. However, the scale measuring social intervention seems to be an anomaly. The item loadings are poor and reliability is low. In addition, the SOCIAL3 statements in the scale measuring 31 Table 4 Confirmatory Factor Loadings by Group Concept Pro-test Pro-test Post-test Post-test Loadings Alpha Loadings Alpha (n=150) (n=129) Difficulty of police work. .68 .68 DIFF 1 .68 .62 DIFF2 .59 .54 DIFF3 .47 .58 DIFF4 .84 .83 DIFFS .73 .77 Police effectiveness. .58 .48 EFF] .6] .62 EFF2 .86 .57 EFF3 .69 .76 EFF4 .52 .56 Police protection of the public. .43 .68 PROTECT 1 .66 .81 PROTECTZ .60 .73 PROTECT3 .68 .46 PROTECl‘4 .7 1 .8 1 Police officer social skills. .66 .56 SOCIAL] .79 .81 SOCIAL2 .48 .40 SOCIAL3* .15 .09 SOCIAL4 .78 .69 SOCIALS .51 .45 SOCIAL6 .63 .54 SOCIAL7 .67 .65 Police use of power. .63 .65 POWER] .78 .33 POEWR2 .83 .88 POWER3 .84 .84 POWER4 .64 .7] Police as social intervenors. .42 .39 SOCW 1 .56 .36 SOCW 2 .58 .43 SOCW 3 .42 .15 SOCW4 .42 .75 SOCWS .30 .45 SOCW 6 .71 .70 Police officer planning. .69 .77 PREP] .72 .68 PREP2 .57 .71 PREP3 .71 .82 PREP4 .65 .68 PREPS .71 .7] *Since the SOCIAL3 statement has poor loadings across each group it will be excluded from this analysis. 32 police officer social skills have low factor scores across both groups. This statement may measure something other than the intended purpose, and therefore it will be excluded from the factor scores in this analysis. The first statement (POWERl) tapping respondent perceptions of police use of power and force has inconsistent factor loadings from pre—test to post-test. A similar relationship exists between the SOCW3 statements in the police as social intervenors section of the attitudinal index. These two statements have low post-test factor loadings. It is possible that these items are poor measures of the underlying factors. However, the loadings for the pre-test statements are adequate. These inconsistencies may affect the final results and will be considered during the final analysis. Reliability analyses were conducted on the scale for each group. Table 4 lists the results from these procedures. For each concept, nearly all groups have acceptable alpha levels. The post-test alpha for the police functioning as social intervenors is low (.39). The alpha is within accepted levels, and corresponds to a weak relationship. It is possible that this portion of the scale does not produce stable results, or the weak alpha may be a condition of too few items on the scale. Besides this reservation, for both groups a correlation exists between all items on each of the seven scales. 33 B IVARIATE COMPARI SONS During an experiment, respondent withdrawal is often a problem, which lowers the number of cases that can be matched from pre—test to post—test. Because of inconsistent class attendance and failure to list student identification numbers, the total number of students completing the questionnaire was 97. Where possible, the true experimental data will be used. However, since post—test only comparisons produce valid findings, this type of analysis will be explored to increase case numbers (N=129). Before comparing factor scores across groups, it is necessary to examine the individual statement level relationships between groups. An analysis of variance was conducted comparing pre-test and post-test statements (See Appendix C). The means are based on a five point Likert type scale where 5 is most positive opinions and 1 represents most negative. As expected, the group means for each statement are very similar. For all pre-test statements, except SOCIALl, no significant differences exist between control and experimental groups. Only the EFFl statement in the post-test exhibit a statistical difference. Thus, the visible group differences do not introduce biases produced from the allocation of subjects into each group. A similar comparison was conducted comparing control and experimental groups instead of test time contrasts (See Appendix D). It is hypothesized that experimental means 34 should increase from the exposure to the stimulus in relation to the control group. If an increase exists for both groups it is expected that the rate of experimental increase is larger than the control group’s rate. Also, if the means for both groups decrease the hypotheses suggest that experimental differences will be less than control group difference. In general, the means across testing time and groups are consistent. Statements that exhibit differences between control and experimental are minimal. In fact, only one statement (SOCIAL6) has a significant difference across groups. Both means for the SOCIAL6 statement decreased from pre-test to post-test but difference between the two declines is only .02 ( less than .001%). It was predicted that the experimental group’s perceptions would become more positive in relation to those of the control group. At initial analysis, it seems that the stimulus had little or even a negative effect on peoples’ perceptions of law enforcement. To further investigate this relationship the mean factors scores will be examined. Table 5 displays the pre-test and post—test mean factor scores across groups. These scores were created separately for pre-test and post—test groups. Analysis of variance models were run on each mean in Table 5, and the police use of power during the pre-test shows a significant difference. This demonstrates that an attitudinal difference regarding police use of power existed during the 35 Table 5 Factor Score Means by Group Pre-test vs . Post-test (n=97) Cmmqu Imam“ Rmbnu Control Experimental Control Experimental menSamm -Difficu]ty of police work. .034 -.046 .071 -.098 -Police effectiveness. -.064 .086 -. 165 .221 -Police protection of the public. .057 -.076 .051 -.070 -Police officer social skills. .055 -.073 .084 -.]18 -Police use of power. .184* -.247* .079 -. 106 -Po]ice as social intervenors. .047 -.063 .152 -207 -Police officer plannifi. .107 -. 147 .125 -. 167 *Significant at <05 pre-test. Specifically, pre-test subjects in the control group are more likely to support police officers’ use of power and force than the experimental pre-test respondents. It was expected that significant changes would be found in this the experimental group’s factor score means. However, is not the case. These findings suggest that the experimental group’s exposure to the “COPS" video clips did not significantly change their opinions about law enforcement. In addition to the above comparisons, a similar table to Table 5 was constructed for control and experimental group factor scores across time. The difference is that the factor scores are first broken into groups and then pre-test and post-test means are compared. Table 6 displays these findings. As in Table 5 it was hypothesized that experimental group means should increase relative to control this condition was not found. No group means. Again, significant differences were discovered in Table 6. The lack of significant findings indicates that the stimulus did 36 Table 6 Factor Score Means by Time Control vs. Experimental (n=97) Concept Control Experimental Pro-teat Post-test Pit-test Post-test Factor Scores -Difficulty of police work. .034 .071 -.046 -.098 -Police effectiveness. -.064 -.165 .086 .221 -Po]ice protection of the public. .057 .051 -.076 -.070 -Police officer social skills. .055 .084 -.073 -.1 18 -Police use of power. .184 .079 -.247 -.106 -Police as social intervenors. .047 .152 -.063 -.207 -Police officer planning .107 .125 -. 147 -.167 I"Significant at <05 not alter experimental groups’ perceptions of policing. Factor reductions were conducted using a data file where control and experimental group scores separately. Also, an additive scale was constructed. Similar results were found, and are listed in Appendix E. MULT IVARIATE COMPARISONS Considering the fact that few significant results were found from each of the previous tests it is necessary to determine if differences exist across demographic, prior police contact, and television exposure variables. Because of limited cases in the true experimental data set, the post-test only data were utilized in these comparisons (n=129). Also, some variables have little or no variability, such as race and year in school. Therefore, the number of controls is reduced (See Appendix B). However, it must be noted that even for post—test only multivariate analyses the cell frequencies for many statements are at or below thirty cases. Also, post—test 37 only results can not consider pre-existing differences between control and experimental groups. Table 7 lists the significant results from the statement level elaborated analysis of variance. For each of the 54 response categories listed in Table 7 the highest means are divided equally between control and experimental groups. Some consistencies emerge from the results displayed in Table 7. For statements DIFF3, DIFF5, PREP3, and SOCW6, control group respondents with two or fewer prior police contacts significantly posses the most positive opinion of law enforcement officials. The two elaborations by age (POWER2 and SOCIAL4) show that control group respondents who are nineteen years or older hold the highest regard for the police. Experimental subjects that are eighteen years old or less follow second behind the older control respondents. Both of the significant time since last police contact statements (POWER2 and PROTECTZ) show that experimental subjects with less than three months since their last police encounter have the most positive opinion of the police. The few apparent patterns do not support a general theme throughout the attitudinal index. An examination of the factor and additive means may provide such evidence. The significant factor means for the elaborated analyses of variances are listed in Table 8. Subjects in the experimental group with three or more police contacts have a more positive opinion regarding the difficulty of 38 Table 7 Post-test Significant* Statement Means by Group (n=129) Concept and Controlling Statement Response Control Experimental Difficulty of police work. -DIFF2 -Planning CJ employment. Yes 4.00 4.03 No 4.44 4.13 -DIFF3 -Number of police contacts. 2 or less 3.45 3.11 3 or more 3.20 3.61 -DIFF4 -CJ major. Yes 3.96 4.29 No 4.12 3.85 -Number of police contacts. 2 or less 4.23 3.89 3 or more 3.98 4.17 -DIFF5 -CJ major. Yes 4.32 4.67 No 4.34 4.10 Police effectiveness. -EFF2 -Nature of last police contact. Positive 3.15 2.94 Negative 2.92 3.36 Police use of power. -POWER1 -Exposed to reality based CJ Yes 4.33 3.55 TV programs. No 3.9] 3.87 -POWER2 ~Age. 18 or less 2.62 2.90 19 or more 3.20 2.81 -Date of last police contact 3 or less 2.82 3.09 (months). More than 3 3.00 2.59 -POWER3 -Nature of last police contact. Positive 3.37 2.84 Negative 2.79 3.04 -Hours of TV viewing a week. 8 or less 3.06 2.66 9 or more 2.97 3.25 Police officer planning. -PREP1 -Nature of last police contact. Positive 3.30 3.13 Negative 3. 10 3.50 -PREP3 -Number of police contacts. 2 or more 3.86 3.43 3 or less 3.57 3.74 -PREP5 -Freshman in college. Yes 3.6] 3.34 No 3.67 3.68 39 Table 7 (cont’d) Concept and Statement Response Control Experimental Police protection of the public. -PROTECI‘2 ~Date of last police contact 3 or less 3.82 3.97 (months). More than 3 3.94 3.48 -PROTECI‘3 -Sex. Male 4.08 4.00 Female 3.81 4.27 -Hours of CI TV viewing a week. 1 or less 4.09 4.03 More than 1 3.81 4.29 Police officer social skills. -SOCIAL] -Nature of last police contact. Positive 3.70 3.13 Negative 3.05 3.07 -Planning CJ employment. Yes 3.29 3.33 No 2.30 2.87 -SOCIAL3 -Sex. Male 2.75 2.43 Female 2.3] 2.67 -Exposed to reality-based CJ Yes 2.07 2.64 TV programs. No 2.72 2.50 -SOCIAL4 -Age. 18 or less 3. 12 3.24 19 or more 3.54 3.00 -CJ major. Yes 3.07 3.00 No 3.5] 3.33 Police as social intervenors. -SOCW1 -Planning CJ employment. Yes 3.83 3.93 No 4.30 3.60 -SOCW2 -C.l major Yes 3.46 2.90 No 3.15 3.36 -SOCW4 -Exposed to C1 reality-based Yes 3.80 3.09 TV programs. No 3.54 3.61 -SOCW6 -Number of police contacts. 2 or less 4.27 3.73 3 or more 3.75 3.87 *Significant at <05 40 Table 8 Post-test Significant Factor Means by Group (n=129) Concept and Conuomng§Mtement Response Control Experimental Difficulty of police work. -CJ major. Yes -.O95 .358 No .093 -.222 -Number of police contacts. 2 or less .203 -.224 3 or more -.089 .310 Police officer social skills. -Nature of last police contact. Positive .58] -.077 Negative -.271 -.058 policing construct, while the opposite is true for control group respondents. Also, those with a positive last police contact in the control group hold the highest opinions of police officer social skills. Both of these findings are true for the additive scale means. Table 9 below shows these results. Finally, experimental subjects who are exposed to more than nine hours of television per week tend to support law enforcement agencies use of power and force. Overall, the results of this experiment produced few significant findings. Next, these will be discussed in addition to some possible rationale for the poor results. Table 9 Post-test Significant Additive Means by Group ( n: l 2 9 ) Concept and Controllingflatement Response Control Experimental Difficulty of police work. -Number of police contacts. 2 or less 4.07 3.86 3 or more 3.93 4.15 Police use of power. -Hours of TV viewing a week. 8 or less 3.37 3.08 9 or more 3.28 3.43 Police officer social skills. -Nature of last police contact. Positive 3.6] 3.38 Negtive 3.29 3.39 Chapter 4 DISCUSSION AND CONCLUSIONS This analysis has used an experimental format to determine if exposure to the television program “COPS" has an effect on people's perceptions of law enforcement. It must be noted that any findings cannot be generalized past the subjects in this study. These finding represent public perceptions of undergraduate students who are inclined to enroll in an introduction to criminal justice course. Results from this experiment are inconclusive. In fact, very few significant control and experimental group differences were found at any level of analysis. Of the 35 attitudinal statements only one (EFF4) significantly changed in the hypothesized direction. Opinions changed in the opposite direction for the other significant mean difference. Essentially, these initial results support cultivation theorist’s perspective that media effects are not direct (Gerbner, et al. 1980). Findings from the factor score means similar. No apparent differences exist between pre-test and post-test for the experimental group. The control group did show that for the police functioning as social intervenors construct opinions became more positive from pre—test to post-test. 41 42 This difference can be attributed to the exposure to classroom lectures and reading materials. Similar results were found for the additive score mean comparisons. No significant differences exist between pre- test and post-test means for the experimental group. As with the factor scores, additive score control group means for the social intervention function construct increased from pre-test to post-test. These results indicate that exposure to the “COPS" video clips did not influence overall perceptions of law enforcement. Because of high subject drop out rates, the number of cases in this experiment is limited. Even when using post—test only data many of the cells in the analysis of variance tables have low frequencies. Any conclusions drawn from the post—test only results should be taken with caution. Findings from the elaborated analysis of variances support some of the previous literature on public perceptions of law enforcement officials. Prior theorists suggest that people with positive experiences with police officers are more likely to have positive opinions of the police (Correia, Reisig, and Lovrich 1996). The present study supports this position. Control group subjects with a positive last experience with an officer hold significantly higher regard for police officer social skills. Unfortunately, this result does not address media effects questions. 43 The experimental means for hours of television viewing per week contribute to this paradigm. Cultivation theorists suggest that heavy television viewers often perceive social reality as similar to the reality presented on television (Signorielli and Morgan 1990). Experimental subjects with nine or more hours of television exposure a week have the most positive opinions of police use of power and force. One attribute of “COPS” is that officers tend to be confronted with violent or fleeing suspects, who must be physically caught and subdued to make an arrest. On “COPS” the police almost always accomplish this goal (Andersen 1994). This experiment supports the notion that heavy viewers of television internalize the justified use of force on “COPS" and accept that as a normative aspect of policing. Explanations for the lack of significant findings are multifaceted. The most obvious reason lies in the limited number of cases. It is possible that these results are tainted by the statistical artifact of low case numbers. Another reason, the questionnaire may be deficient. It is very difficult to measure or substantiate evident shifts of public opinion. As small as a three percent change can be considered a substantial variation (Morgan and Signorielli 1990). The stimulus could have been too short. The content of the video clips may not of had the necessary stylistic integrity in invoke a change (Lewis 1991). Television exposure in an university classroom is drastically different 44 that at home or dorm room exposure. Any attitudinal changes induced from university classroom experiences may have introduced problems into this experiment. Many students’ prior knowledge of the functions of our criminal justice system may have changed between pre- test and post—test. Education may cause subjects to differently interpret pre—test and post—test statements, causing inappropriate comparisons. Similarly, college students may experience more benign police contacts. This may cause them to have an elevated opinion of law enforcement compared to the general public. With higher baseline opinions, the stimulus may not have been adequate to produce significant changes. It is possible that the relationship between mass media and public perceptions of reality is excessively complex to test with an experiment. Other processes could govern how an individual internalizes mediated messages. For example, peer relations, personal experiences, inter-group communication, or other information sources could influence attitude formation. Two weeks between administration of the tests may not have been long enough. Subjects may have attempted to remember their previous responses and attempted to “help" the researcher by answering with similar responses. Finally, even though the evidence suggests otherwise, the assignment of subjects into groups may have introduced bias into the experiment. 45 RECOMMENDATIONS Future experimental research on the effects of reality- based television programming can benefit a great deal from this analysis. First, case numbers should be increased. This allows for more sophisticated statistical procedures, and may reduce the significance problem evident in the present experiment. The stimulus and presentation of it could be improved. An increase in exposure time may help future researchers. Also, the setting should be more intimate. Experimenters should attempt to create an environment which resembles the everyday television viewing experience. In addition to survey responses, future researchers may direct discussions during and after the stimulus presentation to account for any qualitative components of attitudinal change. Multiple samples and exposures over time to the same groups may aid in any future research. Other variables should be considered. Prior knowledge of subject matter, contact with other mediated messages, interest in the television program, personal television viewing characteristics (e.g. watch while eating or doing other tasks vs. complete attention), or other personal beliefs would contribute to the complex mechanisms at work during television exposure. EXPLORATORY CONSIDERATIONS 46 Even though this experiment produced few significant findings, it does contribute to the media effects research paradigm. Very little research has been conducted on the relationship between reality-based television programming and public perceptions of reality. Thus, a discussion of the overall themes discovered in this analysis is appropriate. For the statements listed in Table 3, all four job effectiveness statements are larger for the experimental group. Three of the effectiveness statements (numbers 4, 15, and 25 in Appendix A) are concerned with effective and efficient crime controlling police functions. The final effectiveness statement (number 23) asks about officers wasting time and energy on unimportant activities. Thus, the respondents who were exposed to the “COPS" clips have a more positive opinion of crime control functions of the police. One theme throughout “COPS" is that criminal always are caught by the police (Oliver 1994). The current stimulus also showed two clips where successful investigations produced arrests. The concerned police officer representations helping the older man home directly addresses the wasting time on unimportant activities statement. As a result, the two clips may have influenced experimental group's opinions regarding the effectiveness construct. However, without significance these differences may be a product of existing variation between the groups. 47 While controlling for television viewing habits, other patterns emerged from the mean additive score comparisons. First, subjects who stated that they had been previously exposed to reality-based criminal justice television programming have more positive opinions of the police than those with no exposure. This is true for all constructs except for difficult of work. Table 10 displays these means . Table 10 Mean Additive Scores by Reality-based Exposure ( n= 9 4 ) Concept Not Exposed Exposed Difficulty of police work. 3.93 3.91 Police effectiveness. 2.71 2.79 Police protection of the public. 3.63 3.75 Police officer social skills. 3.43 3.56 Police use of power. 3.38 3.44 Police social intervention functions. 3.34 3.36 Police officer planning. 3.26 3.35 Second, the amount of criminal justice programming exposure seems to affect one’s opinions about law enforcement. Table 11 lists the additive mean scores by amount of exposure to criminal justice programming. For all Table 11 Mean Additive Scores by Amount of CJ TV (n=94) Concept Less Than 1 More Than 1 Hour Hour Difficulty of police work. 3.84 4.0] Police effectiveness. 2.67 2.78 Police protection of the public. 3.56 3.82 Police officer social skills. 3.40 3.56 Police use of power. 3.3] 3.51 Police social intervention functions. 3.28 3.31 Police officer planning. 3.54 3.66 48 constructs, respondents who are exposed to more than one hour of television about our criminal justice have more positive opinions than those with less than one hour. A final note concerns respondents' perceptions in the experimental group the accurately the “COPS” stimulus represented reality. Table 12 lists the experimental group mean scores by degree of perceived program realism. Table 12 Mean Additive Scores by Perceived Realism (N=59) Concept Very-Real Real Difficulty of police work. 3.96 3.98 Police effectiveness. 3.03 2.85 Police protection of the public. 3.80 3.59 Police officer social skills. 3.45 3.33 Police use of power. 3.38 3.17 Police social intervention functions. 3.34 3.36 Police officeflilannirL 3.52 3.48 Generally, the mean scores are consistent across perceived realism. However, for five of the seven concepts respondents who stated that the stimulus was very real have better opinion of the police. All three of these arithmetic descriptive tests support cultivation theorists' opinion that people with heavy exposure to television tend to View social reality similar to what is presented in television. However, it must be stated that statistical significance was not found for any of these comparisons. Even though, these descriptive statistics do warrant future investigation into the relationship between reality-based television programming and public perceptions of reality. 49 SUMMARY This analysis conducted a experiment to determine if exposure to the television program “COPS" influences peoples' perceptions of law enforcement. The results were minimal. Some evidence was found is support of cultivation theory. It seems that for reality-based television programs like “COPS” media effects are not direct. Generally, students enrolled in an introduction to criminal justice course a half hour video clip is not sufficient to evoke a change in public opinion of law enforcement. APPENDICES APP-.11 A The Questionnaire Directions: Please circle the answer or fill in the blank. PIN Number (student number) 1. On a scale from 1 to 10 where 1=very realistic and 10=not very realistic please rate how realistic these TV clips area real neutral not real 1 2 3 4 5 6 7 8 9 10 What is your major? Are you a: freshman sophomore junior senior or graduate student What is you GPA? 4.0-3.5 3.49—3.0 2.99—2.5 2.49—2.0 1.99-1.5 below 1.5 What is your gender? male female Which of the following best describes you? African American Caucasian Native American Asian American Hispanic Other What is your age? How many of hours of television do you watch a week? How many of these hours are spent watching programs about our criminal justice system? If any, which programs? 50 10. ll. 12. 13. 14. 15. 16. 51 Do you regularly watch any programs about our criminal justice system? yes no If so, which programs? Have you ever been in direct contact with an on duty police officer? yes no If so: How times has this occurred? When was the last encounter? On a scale from 1 to 10 where 1=most positive and 10=most negative please rate this encounter positive neutral negative 1 2 3 4 5 6 7 8 9 10 How frequent are your close friends or close family members in contact with on duty police officers? Regularly Occasionally Sometimes Rarely Never Do you know any police officers? yes no If so, how many? Are any of your relatives police officers? yes no If so, how many? Are you employed? yes no If so, what is your job? Do you plan on getting a job in the criminal justice field? yes no If so, what type of job? Directions: statement. 1=strongly disagree, 3=neutral, 4=agree, 1. Police officers are insensitive to the public. 2. Police work is simple. 3. Police officers face many non—criminal situations while working. 4. Violators of the law are nearly always detected and punishedfi 5. Police officers have good planning skills. 6. Police officers need little power. 7. The police in general are helpfulh 8. Our society praises police officers. 9. Police work is unstructured. 10. The police exist to make society a better place for everyone to live int 11. Police officers do not listen to suspects. 12. Police officers are confronted by complex situations. 13. The primary role of police officers is to catch criminals. 14. Police officers are 52 Please circle the answer that represents most closely how much you agree or disagree with each poor organizers. 2=disagree, and 5=strongly agree. Strongly Disagree 1 Strongly Agree 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 53 Police officers efficiently control crime. Few police officers are injured while working. Police department are well organized. On the whole police officers are honestt Police work is difficult. Police work is stressful. Police officers are well prepared. Police officers are nice people. Police spend much time and energy on too many unimportant thingsfi Police officers are unaware of peoples’ needs. Police officers catch Strongly Disagree 1 criminals in a timely manner. Police offices are needed in our society. Police officers abuse their power. Police work usually involves crime control. It is OK for police officers to perform non-investigative functions. Police officers are hostile while working. All police functions should be crime related. 2 Strongly Agree 5 32. 33. 34. 35. Police officers act as social workers. The police help protect ordinary people like mek Police officers are well educated. The police often use unnecessary forcefi 54 Strongly Disagree 1 2 Strongly Agree 5 55 ENDNOTES 1. This statement was omitted from the control group questionnaire. 2. From Rundquist and Sletto 1936 3. From Tuohy and Wrennall 1995 4. From Tuohy and Wrennall 1995 5. From Rundquist and Sletto 1936 6. From Klyman and Kruckenberg 1974 7. From Tuohy and Wrennall 1995 8. From Tuohy and Wrennall 1995 APPENDIX B variable Recoding For all groups the following protocols were used to recode variables into dichotomous response categories. Variables Recode Response Categories Age. 18 or less; above 18 Major. Non-criminal justice; Criminal justice Year in school. Freshman; Not freshman Type of criminal justice job Non-law enforcement; lanning to get. Law enforcement Length of time since last police contact. 3 months or less; more than 3 months Number of police 2 or less; contacts. More than 2 Nature of last police contact. Positive; Negative Frequency of close associates Rarely; police contact. Sometimes or more Number of police 1 or less; known. More than one Amount of TV 8 hours or less; exposure per week. More than 8 hours Amount of C] TV 1 hour or less; exposure per week. more than 1 hour Type of CI TV Non-reality based; exposure. Reality based Type of CI TV Non-reality based; Reality flgularly exposure. based Perceive program realism Very real; (experimental gr_oup only). Not very real 56 APPHIX C Statement Ileana by Group Ere-teat vs. Post-test (n-97) Concept Pre-test Post-test Control Experimental Control Experimental Difficulty of police work. -DIFF1 4.25 4.37 4.23 4.20 -DIFF2 4.07 4.05 4.23 4.05 -DIFF3 3.25 3.27 3.32 3.22 -DIFF4 3.93 3.83 4.09 4.02 -DIFF5 4.18 4.02 4.36 4.29 Police effectiveness. -EFF] 1.98 2.20 206* 2.46* -EFF2 2.93 3.07 3.13 3.15 -EFF3 2.80 2.90 2.75 2.95 -EFF4 3.18 2.93 3.13 3.37 Police use of power. -POWER1 4.07 3.85 4.09 3.83 ~POWER2 3.7 l 2.93 2.98 2.83 -POWER3 3.27 3.00 3.07 2.95 -POWER4 3.67 3.22 3.45 3.43 Police officer planning. -PREP1 3.16 3.17 3.30 3.34 -PREP2 4.15 4.00 3.91 3.63 -PREP3 3.80 3.66 3.75 3.54 -PREP4 3.51 3.36 3.50 3.39 -PREP5 3.64 3.49 3.67 3.49 Police protection of the public. -PROTECI‘ 1 3.38 3.34 3.61 3.5] -PROTECI‘2 4.04 3.73 3.93 3.85 -PROTECI‘3 3.95 4.02 3.98 4.15 -PROTECT4 3.45 3.44 3.59 3.44 Police officer social skills. -SOCIAL] 3.42* 3.15* 3.34 3.07 -SOCIAL2 4.47 4.49 4.59 4.54 -SOCIAL3 2.39 2.63 2.55 2.66 -SOCIAL4 3.35 3.23 3.34 3.17 -SOCIAL5 3.91 3.98 3.9] 4.02 -SOCIAL6 3.54 3.71 3.09 3.28 -SOCIAL7 3.32 3.20 3.39 3.39 Police as social intervenors. -SOCW1 3.77 3.68 4.00 3.78 -SOCW2 3.25 3.15 3.36 3.12 -SOCW 3 2.16 2.24 2.54 2.24 -SOCW4 3.47 3.27 3.73 3.46 -SOCW5 3.53 3.49 3.71 3.66 -SOCW6 3.7] 3.74 3.88 3.78 *Significant at <05 57 APP-.11 E EACTOR.AND.ADDITIVE SCORE MEAN COMPARISONS It is possible that by creating the factor scores for pre-test and post-test groups separately, biases were introduced into this analysis. Since factor reductions produce scores in terms of standard deviation units, comparing two sets of factor scores from similar populations may result in analogous scores across groups. As a result, the distributions for the two groups may be so alike that significant differences do not exist. To resolve this potential problem, duplicate analyses were conducted on a data file where pre-test and post-test files were merged first and then confirmatory factor reductions were run. Table 13 displays the pre-test and post—test mean factor scores across groups. Results in Table 13 are almost Table 13 Factor Score'Means by Group Pre-test vs. Post-test (n=97) Concept Pro-test Post-test Control Experimental Control Experimental Factor Scores -Difficulty of police work. -.062 -.147 .166 .005 -Police effectiveness. -. 163 -.022 -.048 .306 -Police protection of the public. -.011 -.155 .108 .022 -Police officer social skills. .120 -.352 .026 -.]63 -Police use of power. 235* -. 175* .026 -.]76 -Po]ice as social intervenors. -. 101 -.226 .309 -.060 -Police officer planning. .114 -.135 .1 19 -. 180 1 Factor scores were created with both groups combined. I"Significant at <05 identical to those in Table 5. The pre-test police use of power construct exhibits the only significant differences. 59 60 Once again, this difference may introduce testing bias into this experiment, caused by pre-existing group differences. Table 14 lists the control and experimental groups’ factor means across testing times. Again, the difference is that the factor scores are first broken into groups and then pre-test and post-test means are compared. Table 14 Factor Score'Means by Time Control vs. Experimental (n=97) Concept Control Experimental Pro-test Post-test Pre-test Post-test Factor Scores -Difficulty of police work. -.062 .166 -.147 .005 -Po]ice effectiveness. -. 163 -.048 -022 .306 -Police protection of the public. -.011 .108 -.155 .022 -Police officer social skills. .120 .026 -.035 -.163 -Police use of power. .235 .026 -. 175 -. 176 -Police as social intervenors. -.]01"' .309“ -.226 -060 -Police officer planning. .114 .119 -.135 -.181 1 Factor scores were created with both groups combined. I"Significant at <05 The control group’s attitudes regarding the police functioning as social intervenors have significant differences from pre—test to post-test. No significant changes exist from pre-test to post—test for the experimental group. The final set of bivariate analyses of variance consists of similar models. The distinction is that instead of using factor scores, an additive scale was created from the attitudinal index. Table 15 lists the pre-test and post-test mean additive scores across groups. Similar to Table 5, pre-test means for the police use of 61 Table 15 Additive Score Means by Group Pre—test vs. Post-test (n=97) Concept Pre-test Post-test Control Experimental Control Experimental Additive Scores -Difficulty of police work. 3.94 3.91 4.05 3.96 -Police effectiveness. 2.72 2.77 2.77 2.98 -Police protection of the public. 3.71 3.63 3.78 3.74 -Po]ice officer social skills. 3.66 3.62 3.61 3.55 -Police use of power. 352* 325* 3.40 3.26 -Police as social intervenors. 3.32 3.27 3.54“ 334* -Police officer planning. 3.64 3.53 3.63 3.48 *Significant at <05 power construct are significantly different across groups. Also, post-test results show that control and experimental group attitudes regarding the police functioning as social intervenors differs. Table 16 lists the control and experimental groups’ additive means compared across testing time. Control Table 16 Additive Score Means by Time Control vs. Experimental (n=97) Concept Control Experimental Pre-test Post-test Pre—test Post-test Additive Scores -Difficulty of police work. 3.94 4.05 3.91 3.96 -Police effectiveness. 2.72 2.77 2.77 2.98 -Police protection of the public. 3.71 3.78 3.63 3.74 -Police officer social skills. 3.66 3.61 3.62 3.55 -Police use of power. 3.52 3.40 3.25 3.26 -Police as social intervenors. 332* 3.54‘ 3.27 3.34 -Police officer planning. 3.64 3.63 3.53 3.48 *Significant at <05 group’s attitudes regarding the police functioning as social intervenors have significant differences from pre-test to post—test. Post-test control subjects are more likely than pre—test control subjects to view law enforcement officials 62 pre-test control subjects to view law enforcement officials as performing social intervention functions. No significant change exsists from pre-test to post-test for the experimental group. LIST OF REFERENCES Anderson, Robin. 1994. “'Reality' TV and Criminal Justice." he Humanist. 54:8-13. Bayley, David. 1979. “Police Function, Structure, and Control in Western Europe and North America: Comparative and Historical Studies. In Norval Morris and Michael Tonry (Eds.), im n u ' ' Research. Chicago. University of Chicago Press. Berger, Peter and Thomas Luckman. 1967. The_§eeiel Cegegreegien of Reeliey. Baltimore, MD: Penguin Press. Bitner, Egon. 1990. Aspecge ef Peliee Work. Boston: Northeastern University Press. Brodt, Stephen. 1987: “'Rape Panic': Definition of a Crime Problem and Community Response." Free Inquirx in CIEEEIME Sociology 15:183-187. Caldwell John 1995 Te1oyisualiL21_St¥1ol_Crisis_and Augherigy in Ameriean Televisien. New Brubswick, New Jesery: Rutgers University Press. Cantor, Muriel and Joel Cantor. 1992. Prime—Time Teleyieion Centen; end generol. Newbury Park, California: SAGE. Carlson, James. 1985. Prime Time Lam Enforcemene. New York: Praeger. Condry. John. 1989. Tho_Psycholooy_of_1oleyision. New Jersey: Lawrence Erlbaum. Correia, Mark, Michael Reisig, and Nicholas Lovrich. 1996. “Public Perceptions of State Police: An Analysis of Individual-Level and Contextual Variables." Jee;nel_ef Crimina1_lustioo. 24 17-28. Cox, Terry and Mervin White. 1988. “Traffic Citations and Student Attitudes Toward the Police: An Examination of Selected Interactive Dynamics.” Joernel ef Peliee Seience and_Administration. 16:105-121. 63 64 Creechan, James. 1992. “Media Impact on Criminal Justice Perceptions: The Integration Effect.” Centre for Criminological Research Discussion Paper. University of Alberta. Decker, Scott. 1981. “Citizen Attitudes Toward the Police: A Review of Past Findings and Suggestions for Future Policy.” rnal lice ci n n ' ' i 9:80-87. Dunham, Roger and Geoffrey Albert. 1988. “Neighborhood Differences in Attitudes Toward Policing: Evidence for a Mixed- Strategy Model of Policing in a Multi- Ethnic Setting. h n l rimin l L mi 79: 504- 521. Fox—29 WFTC. 1995. “COPS.” At URL: http://www.fox29.com/shows/best/cops/. Gains, Larry, Victor Kappeler, and Joseph Vaughn. 1994. Polieing in Ameriee. Cincinnati, Ohio: Anderson Publishing. Gerbner, George and Larry Gross. 1976. “Living with Television: The Violence Profile.” Jee;eel_ef Communicagions 26:172-199. Gerbner, George, Larry Gross, Michael Morgan, and Nancy Signorielli. 1980. “Some Additional Comments on Cultivation Analysis." eblie Opinion Querferly 44:408-420. GriSW°1d: WendY- 1994. Cu1tures_and_Sooieties_in_a_Chanoino World. Thousand Oaks, California: Pine Forge. Hallett, Michael. 1995. “Guns and Roses on Cops." Peece Reyiew. 7:233-236. Hans, Valerie and Juliet Dee. 1991. “Media Coverage of Law." Am ri an B h i ra cientis 35:135-149. Hawkins, Robert and Suzanne Pingree. 1990. “Divergent Psychological Processes in Constructing Social Reality from Mass Media Content. " In Nancy Signorielli and Michael Morgan (Eds. ), l i a i n a i 'r i ' _edie_EffeeLe_Reeee_eh. Newbury Park, California: SAGE. Iyengar, Shanto and Donald Kinder. 1987. Newe tha; Meeeere. Chicago: University of Chicago Press. Jordan, Donald. 1993. “Newspaper Effects on Policy Preferences Pub11o_Qoinion_Quorterly 57: 191- 204 Klyman, Fred and Joanna Kruckenberg. 1974. “A Methodology for Assessing Citizen Perceptions of Police." Joernel ef Criminal Jeegiee. 2:219—233. 65 Leff, Donna, Davis Protess, and Stephen Brooks. 1986. “Crusading Journalism: Changing Public Attitudes and Policy Making Agendas." eblie Opinion Quarterly 50:300-315. Lewis, Justin. 1991. The Ideologieel Qetepge. New York: Routlage. Maguire, Kathleen and Ann Pastore Eds. 1994. Soereebeek ef Cr1m1na1_J_stioe_Statistios Washington D-C- Bureau of Justice Statistics. Macionis, John. 1991. Seeiolegy. Englewood Cliffs, New Jersey: Prentice Hall. Manning. Peter 1977 Po11oe_wo_k__Tho_Soo1a1_Qroan1zation ef Pelicing. Cambridge: MIT Press. Manning, Peter. 1996. “Dramaturgy, Politics, and the Axial Media Event " The_Soo1o1oo1o_1_Quarterl¥ 37: 261- 278 Maxfield, Michael and Earl Babbie.1995. Beseereh Megheds fer Cr1_1na1_Just1oe_and_gr1_1no1ooy Belmont. CA: Wadsworth. McCombs, Maxwell and Donald Shaw. 1972. “The Agenda-Setting Function of Mass Media. u 'n o ua . 36:176- 187. Morgan, Michael and Nancy Signirielli. 1990. “Cultivation Analysis: Conceptualization and Methodology." In Nancy Signorielli and Michael Morgan (Eds.), Celeivegien ' . e Di tion 'n e i E f t r . Newbury Park, California: SAGE. Oliver, Mary Beth. 1993. “Portrayals of Crime, Race, and Aggression in 'Reality-Based' Police Shows: A Content Analysis." 0 r a f Br ad s i l r ' M i 38:179192. Page, Benjamin, Robert Shapiro, and Glenn Dempsey. 1987. “What Moves Public Opinion." r' P ' ' ' e Reyiew 81:23-43. Potter, James. 1993. “Cultivation Theory and Research: A Conceptual Critique.” Human Communieatien Reeeareh. 19:564-601. Prichard, David. 1986. “Homicide and Bargained Justice The Agenda-Setting Effect of Crime News on Prosecutors." Public Qpinien Qgerterly 50:143-159. Radelet, Louis and David Carter. 1994. The Police and the Cemmeeiey. New York: Macmillian. 66 Roberts, Julian and Anthony Doob. 1990. “News Media Influences on Public Views of Sentencing.” aw end Human Behevio; 14:451-468. Rundquist, Edward and Raymond Sletto. 1936. Personalipy ip phe Depreseiep. Minneapolis, MN: University of Minnesota Press. Signorielli, Nancy and Nancy Morgan. 1990. “Cultivation Analysis: Conceptualization and Methodology. In Nancy Signorielli and Nancy Morgan (Eds. ) Cplpiyatiop Apelyei1e: New Direepio ne ip Medie Eff eepe Researeh. Newbury Park, CA. SAGE. Slater, Dan and William Elliott. 1982. “Television Effect on Social Reality." Th uart r o rn f 1982:69-79. Surette, Ray. 1992. Media rime n imin Belmont, California: Wasdsworth. Sullivan, Peggy, Roger Dunham, and Geoffrey Alpert. 1987. “Attitude Structure of Different Ethnic and Age Groups Concerning Police." urnal f 'min L Cpiminelogy. 78:177-196. Tapper, John. 1995. “The Ecology of Cultivation: A Conceptual Model for Cultivation Research." gemmunieapion Theepy. 5:36—56. Tuothy, A. P., and M. J. Wrennall. 1995. “Seeing Themselves as Others See Them: Scottish Police Officers' Metaperceptions of Public Opinion.” depppei_ef_gemmppipy end Applied Seeiel Peychology. 5:311-326. Weaver, David, Doris Graber, Maxwell McCombs, and Chiam Eyal. 1981. Media Agenda-Setping in e Presidential Eleetiep; Isepee. Imeges. end Inperesp. New York: Prager. Walker, Samual. 1992. The Eeliee in America: An Intreduetion. New York: McGraw-Hill. "Illllllllllllllllls