<%BANNER%>

Portrayal of school shootings by American newspapers

University of Florida Institutional Repository

PAGE 1

PORTRAYAL OF SCHOOL SHOOTINGS BY AMERICAN NEWSPAPERS By JOHN ROBERT BENNETT A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2003

PAGE 2

ii ACKNOWLEDGMENTS This document is dedicated to my wife, Tosha L. Bennett. Without her support, I would not have been able to complete this study. Special thanks go to Charles F. Gattone, Ph.D. and Joe R. Feagin, Ph.D.; without their guidance, support, and providing a committee for this thesis project none of this would be possible. I thank Terry L. Mills, Ph.D., and Karen F. Parker, Ph.D., for their initial advice on how to go about developing my ideas into a working project which ultimately became my thesis.

PAGE 3

iii TABLE OF CONTENTS Page ACKNOWLEDGMENTS ................................ ................................ ................................ .... ii LIST OF TABLES ................................ ................................ ................................ ............... v ABSTRACT ................................ ................................ ................................ ........................ vi INTRODUCTION ................................ ................................ ................................ ............... 1 1.1 Statement of the Research Problem ................................ ................................ ........ 1 1.2 Background ................................ ................................ ................................ ............. 2 1.3 Key Terminology and Concepts ................................ ................................ ............. 3 1.4 Theoretical Perspectives ................................ ................................ ......................... 5 1.4.1 Conflict-Labeling Criminology ................................ ................................ .... 5 1.4.2 Power-Threat Hypothesis ................................ ................................ ............. 7 1.4.3 Media Criminology ................................ ................................ ....................... 9 1.4.4 Critical-Media Perspective ................................ ................................ ......... 12 1.4.5 Gaps in the Literature ................................ ................................ ................. 16 METHODOLOGY ................................ ................................ ................................ ............. 19 2.1 Summary of Findings ................................ ................................ ............................ 20 2.2 Discussion of Data Characteristics ................................ ................................ ....... 22 FINDINGS AND ANALYSIS ................................ ................................ ........................... 24 3.1 Research Findings ................................ ................................ ................................ 24 3.1.1 Qualitative ................................ ................................ ................................ ... 24 3.1.2 Quantitative ................................ ................................ ................................ 27 3.1.2.1 Non-aggregate Dataset ................................ ................................ ...... 28 3.1.2.2 Aggregate Dataset ................................ ................................ ............. 34 3.2 Analysis of Findings ................................ ................................ ............................. 41 3.2.1 Qualitative Analysis ................................ ................................ .................... 41 3.2.2 Quantitative Analysis ................................ ................................ .................. 48 DETAILED DISCUSSION AND ANALYSIS OF SPECIFIC INCIDENTS ................... 57 4.1 Columbine High School Shooting ................................ ................................ ........ 57 4.2 Thurston High School Shooting ................................ ................................ ............ 65 4.3 John Batram High School Shooting ................................ ................................ ...... 67

PAGE 4

iv 4.4 Lew Wallace High School Shooting ................................ ................................ ..... 67 CONCLUSIONS ................................ ................................ ................................ ................ 69 5.1 Overview of Theories Utilized in this Analysis ................................ .................... 70 5.2 Overview of Important Findings ................................ ................................ ........... 70 5.3 Critique and Suggestions for Future Research of Differential Newspaper Coverage ................................ ................................ ................................ ................. 75 LIST OF REFERENCES ................................ ................................ ................................ ... 78 BIOGRAPHICAL SKETCH ................................ ................................ .............................. 80

PAGE 5

v LIST OF TABLES Table page 2-1. Number of articles written per case. ................................ ................................ ......... 21 3-1. Basic figures for non-aggregate variables. ................................ ............................... 28 3-2. Basic figures for aggregate va riables. ................................ ................................ ....... 35 3-3. Matrix of cases with exceptional qualities. ................................ ............................... 49 3-4. List of cases from lowest to highest affirmation of NTSV status. ............................ 50 3-5. Correlations of aggregate v ariables. ................................ ................................ ......... 53 3-6. Table of significant correlations for aggregate variables.... .....................................54

PAGE 6

vi Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts PORTRAYAL OF SCHOOL SHOOTINGS BY AMERICAN NEWSPAPERS By John Robert Bennett May 2003 Chair: Joe R. Feagin Cochair: Charles F. Gattone Major Department : Sociology The purpose of this analysis is to investigate the way American newspapers portray domestic school shootings. To this date, there have been no investigations of media portrayal of school shootings in America. School shootings make up a small percentage of school violence in America; however, the damage inflicted on the victims of these crimes is devastating. This study is concerned with the coverage and portrayal of these shootings by the news publishers in America. There are no current data sources listing all of the variables of interest to this study. I relied on the Lexis Nexis database of newspapers to find a sample of articles that address this issue. The time frame for the search has been limited to include only articles which were published from 1 st May 1997 through 30 th April 2002. These cases were analyzed using the structural functional and conflict paradigms with a specific interest in the conflict-labeling criminology, power-threat hypothesis, media criminology, and critical-media perspectives.

PAGE 7

vii In conclusion, two key factors seem to explain the relationship between the variables used in this analysis to determine if there is differential coverage of NTSV and TSV. The two key factors are the level of sensational behavior exhibited in a given case and whether or not the behavior exhibited was considered to be expected based upon socially acceptable norms of American society.

PAGE 8

1 CHAPTER 1 INTRODUCTION 1.1 Statement of the Research Problem The purpose of this analysis is to investigate the way American newspapers portray domestic school shootings. To this date, there have been no formal sociological investigations of media portrayal of school shootings in America. School shootings make up a small percentage of school violence in the United State; however, the damage inflicted on the victims of these crimes is devastating. This analysis is not necessarily concerned with the effects of each school shooting. It is more concerned with the coverage and portrayal of these shootings by the news publishers in the U.S. There are no current data sources listing all of the variables of interest to this study. Therefore, I relied on the Lexis Nexis database of newspapers to find a sample of articles that address this issue. The time frame for the search has been limited to include only articles which were published from 1 st May 1997 through 30 th April 2002. I gathered and analyzed articles that span a five year period of time. During this period, 578 articles were published in U.S. newspapers on school shooting. Of these 578 articles, I selected one hundred at random for inclusion in the sample and coded each for the following variables: title, case name, number of words, date of occurrence, date of publication, section of the newspaper in which the article was published, time difference between date of publication and date of occurrence, race of offender and victim, socioeconomic status of offender and victim, number of offenders and victims, number of offenders and victims killed, type of weapons used, number of weapons used, relation of victim to

PAGE 9

2 offender, and level of unique interest. I also developed qualitative descriptions of each article. Of the one hundred articles included in the sample, twenty seven cases emerged. I organized these articles on the basis of case name so that key variables from the individual articles could be represented in a combined form. The variables included in the aggregate dataset are as follows: new trend vs. tradition, case name, number of articles per case, number of articles written within first year after the occurrence, socioeconomic status of offenders and victims, quantity of weapons used, type of weapons used, race of victims and offenders, number of offenders and victims, number of offenders and victims killed, level of unique interest controlling for time, racial make-up of the city where shooting occurred, median family income of the city where shooting occurred, and total population for the city where the shooting occurred. 1 These cases were analyzed using the structural functional and conflict paradigms with a specific interest in the conflict-labeling criminology, power-threat hypothesis, media criminology, and critical-media perspectives. 1.2 Background Government agencies, criminal justice departments, and scholars in the U.S. have not studied school shootings in the way that is proposed here. Among these institutions the tendency has been to focus on data regarding the number of shootings and number of victims, the types of weapons used in offenses, and components related to offenders psychological makeup. The Federal Bureau of Investigation has been involved in compiling profiles for offenders in school shooting cases. Other agencies have developed 1 The 2000 U.S. Census American FactFinder was used to find the racial makeup of the city where the shooting took place and the median family income of the city where the occurred.

PAGE 10

3 programs aimed at trying to prevent school violence and teaching young people how to deal with stress and violent feelings. Most of the academic resources have been invested in the psychological characteristics of the offenders and, at the same time, teaching others methods of anger management. Psychologists and social workers have been involved in providing services to those who have been victims of school shootings and those who have had to live in the aftermath of school shootings. No study to this date has been conducted by social science researchers with regard to news coverage of school shootings. There is a void that exists in the academic study of news coverage of school shootings. 1.3 Key Terminology and Concepts The specific goals of this project are to determine the nature of media portrayal of school violence by domestically owned and published newspapers. School violence is classified as being of two types: traditional and new trend. Traditional school violence involves actors who are of a low socioeconomic status, unemployed, or working class. The target of the violence is generally other students and violence may be related to gang or drug activities. Types of weapons used vary from assault with hands to assault with guns that are not mass-casualty producing. The offender usually targets one or two persons, and violence seems to happen in more of a spontaneous nature. There also may be planned acts of violence in this category. New trend school violence is violence in which the offender is usually a white male of a middle class or higher socioeconomic status, and the target of the violence tends to be not only peers but also teachers, school officials, and parents. Violence is likely related to feelings of revenge or retribution and the types of weapons used are mass-casualty producing and tend to be excessively violent. The offender usually plans to cause many casualties but may actually kill fewer,

PAGE 11

4 and violence seems to be planned, premeditated, and organized. Some key distinctions between these two types of school violence offenders are their level of socioeconomic status and the relation and status of their targeted victims. This will not be measured at the individual level in this study due to lack of data. The SES data will be organized on the basis of the regional location in which the incident occurred. Another key concept in this project is the expected normal behavior of the offender. Expected normal behavior is applied in the following way: in general, white American news journalists and editors share common ideas about normal behavior that they expect to find in any given community. 2 In cases of school shootings that occur where the violent behavior is expected, the offender will be labeled as exhibiting normal behavior, and white American journalists will not likely give any special attention to these cases unless there is something unusual about them. For example, in a case where several victims were involved, if unusual weapons were used to commit the violence or some other anomaly along these lines, white journalists may exhibit special interest in a case. In cases of school shootings that occur where the behavior is not expected, the offender is labeled as exhibiting abnormal behavior and most white American journalists will give additional attention to these cases, covering them more thoroughly and are expected to show signs of unique interest. Some characteristics of unique interest that the white American news journalists are expected to exhibit in these abnormal cases are the following: an invocation of a psychological-medical model to explain the occurrence is possible. For example, in instances in which the offender is likely to be labeled as 2 This study uses the terms white American journalists and editors to convey the assumption that the majority of articles written in this study were written by the average white American journalist who is expected to present each case in a certain way. There are editorials in this

PAGE 12

5 psychologically ill or portrayed as an outsider from the normal student population at the school, words such as understand, sympathy, empathy, cope are expected to be used in this type of article. In instances in which readers assume that the violence was not supposed to happen in the community of the shooting, it is expected that the article will present a sense of urgency about finding a solution. The article may also describe steps that are being taken by authority figures in the community where the violence took place to rectify the situation, and there will likely be a description involving the aide that will likely be given to members of the community to help them cope with the act of violence. 1.4 Theoretical Perspectives Analysis of the literature shows that theory has not been applied to the issue of school shooting. Sociologists have not formally studied the ways newspapers present information and cover stories of school shootings. The following literature review is a compilation of texts written from the conflict perspective, labeling theory, power-threat hypothesis, with contributions from media criminology and critical-media perspectives. 1.4.1 Conflict-Labeling Criminology This section illustrates findings from two conflict-labeling theorists that describe the relationship between a criminal offenders status in society and the way that he or she is likely to be labeled by authority figures that preside within the criminal justice system. Both Turk (1964) and Quinney (1970) share similar ideas about the ways people are labeled as criminal. Both new trend and traditional school violence cases are labeled as criminal by newspapers. The difference is that traditional school violence (TSV) tends to sample which are not likely to have been written by white American journalists, but it is assumed that the majority of the editorials are written by white Americans.

PAGE 13

6 be labeled as more criminal relative to new trend school violence (NTSV). For both Turk (1964) and Quinney (1970), society is made up of a set of laws that are produced by people in power to maintain order in society. If a person violates those laws, the less power that individual has, the more likely he or she is to be punished and labeled in a negative way. In this argument, the critical component of power, as measured by SES or some other economic indicator, comes into play. School shooting offenders who have lower SES are typically labeled as criminal more readily than those of higher relative SES. How is this translated into the coverage of school violence by newspapers? One of the first points to realize is that the criminal justice system in the U.S. deals with the offenders of school shootings. In this system, police officers, attorneys, and judges are the authority figures who officially label the offenders in the legal sense. At the same time, the authors and editors of newspapers have an opportunity to further label the offenders by providing coverage of the case in the newspaper. Examining the work of Turk (1964) and Quinney (1970) might lead one to expect that the media would cover all school shooting cases. However, one would expect to find that the coverage of cases that are committed by offenders of low or working class SES will be more harsh toward the offender than those articles that cover the cases in which the offender is of middle or high SES. This differential coverage is likely because Turk (1964) and Quinney (1970) found that authority figures will treat criminals more harshly if they have less power. In this case, the lower class and working class school violence offenders are more likely to receive unforgiving coverage in relation to the middle class and upper middle class offenders.

PAGE 14

7 1.4.2 Power-Threat Hypothesis This section introduces two concepts that Hubert Blalock (1967) developed within the structure provided by the power-threat hypothesis ( Tolnay, Beck, & Massey, 1989). The two concepts offer further guidance in determining what type of coverage one would expect U.S. media sources to provide for the two types of school violence. Blalock (1967) identified two types of discrimination that can motivate the white majority to discriminate against the black minority: symbolic segregation and threat-oriented ideologies. He examined U.S. society during the 1960s, where he divided America into North and South and observed trends of discrimination in both regions of the country. Blalock (1967) developed the idea symbolic segregation in which white Americans create belief systems that promoted segregation between white Americans and minority groups, in particular black Americans. Blalocks (1967) idea of symbolic segregation would lead one to expect that the media may not present coverage of all school shooting cases, and that the news articles are likely to differ in their coverage of the two types of school violence. By presenting TSV cases as though they were normal occurrences relative to NTSV cases, or by not providing coverage of the TSV cases at all, the authors and editors of national newspapers seem to be symbolically segregating TSV offenders and their families from the rest of the society. By providing this type of coverage for TSV cases, the authors and editors keep TSV cases from entering the consciousness of most Americans. At the same time, newspaper coverage provides a sense that NTSV is abnormal which encourages the average American to think about these offenders and their families, the victims of such offenses, and the circumstances around which the offense occurred. It is deducted from Blalocks (1967) theory that the average American may develop

PAGE 15

8 sympathy in an attempt to understand what happened, and may try to determine what society can do to stop this new trend of school violence. This type of sympathy, compassion, and understanding is not extended to the TSV offenders, families, and victims since they are not within the realm of the average Americans consciousness. The symbolic segregation created by the editors and authors of newspapers in the coverage of TSV offenders leads to neglect and reminds the offenders and their families where they belong. The implied message is that they do not belong in the newspapers portrayed in a similar way as the NTSV offenders and that they do not deserve the same sympathy and respect that is given to the NTSV offenders, family members, and victims. These findings tie into the second concept that Blalock (1967) developed termed threat-oriented ideology. He argued that southern white Americans tended to exaggerate the threat that black Americans posed to the white community of recent history. Similarly, the average white American today tends to expect violence that occurs in the lower class and working class communities where TSV takes place. The average white American expects to see violent behavior to an exaggerated degree in these communities. This sentiment has been recorded by media interviews with many people after recent incidents of NTSV have occurred. The people who are interviewed state things like: How can this happen here? I cant believe this happened here. This is not supposed to happen here. These statements made many times by many, mainly white, people in America have a logical complementary statement: school violence is supposed to happen somewhere else. That somewhere is in communities unlike those middle and upper class communities where NTSV is occurring. It is supposed to happen in communities where TSV occurs. This line of reasoning leads to the culmination of the

PAGE 16

9 thought that there is a subconscious idea in the minds of the average white American that there are communities out there where violence is supposed to occur; and the newspapers do not cover this violence with the same care and understanding that is used in the coverage of NTSV because it is expected from those communities where those people live. This finding leads one to conclude that authors and editors of newspapers will cover NTSV cases with more emphasis placed on the abnormality of the occurrences relative to TSV cases. Additionally, only a portion of the TSV cases will be covered in the newspapers. A different way in which media organizations do portray TSV cases is introduced in the next section and is centered on the idea that really horrific cases of TSV that perpetuate the idea of threat-oriented ideologies will be published because it does reinforce thoughts that are currently socially acceptable with regard to TSV. 1.4.3 Media Criminology Media-criminologists conduct research on the proposed effects that news coverage has on the reader. Howitt (1998), Barak (1994), and McQuail (1992) all share similar ideas about the medias role in maintaining the social order that is made up of inequality and power struggles. The media present a world view which continually and pervasively regenerates the ideological structures required for the maintenance of existing power hierarchies ( Howitt 1998, p. 13). With this in mind, this paper now tries to explain why one should expect to find little national news coverage of TSV. Howitt (1998) found that the national news coverage of crime does not necessarily coincide with the number of occurrences that the crime took place. For example, lynchings were more likely to be reported if: (a) they were in areas with easy communications; (b) they involved rape or sexual assault; and (c) there were several victims (p. 25). This shows

PAGE 17

10 that the media covered only the sensational cases of lynchings and gives additional support to the idea that the national newspapers will cover TSV cases on a limited basis. Cases that are sensational will be presented in news articles. From this perspective, one would explain the heightened media coverage of NTSV cases because they are sensational. On average, the level of intended violence is sensational. It seems to be a relatively new phenomenon in which children of middle and higher SES are committing horrible acts of violence against fellow students and authority figures. This raises the question: Is the divide in the coverage of the two types of school violence simply a matter of the media going after the sensational big scoop? One of the goals of this project is to address this question. Another goal of this research is to examine the claim that the media present sensational articles, and those that maintain and reinforce current ideological structure. One could argue that the coverage of NTSV and the coverage of sensational TSV cases do hold true to this finding as well. The NTSV cases are covered in such a way as to separate the NTSV offenders from the normal group of students. It is expected that newspaper publishers will portray NTSV offenders as psychologically ill, or as part of some deviant sub-group of juveniles that committed these heinous acts of violence. This is a way in which the authors of these articles do maintain and reinforce current ideologies about the normal population. In a similar fashion, the TSV offenders who commit sensational acts of violence that seem to perpetuate current ideological beliefs will be covered in news articles. However, newspaper publishers will not attempt to separate the TSV offender from his or her peers because that would not reinforce and maintain current ideological beliefs about lower and working class SES communities. In

PAGE 18

11 both cases, though, the news articles tend to portray the offender in such a way to maintain status quo. However, the question still begs to be answered, is this simply a product of news journalists attempts to get the big scoop? One may argue that the journalists have gone after the big scoop. Indeed, they must approach their work in this manner in order to compete and succeed as journalists, but from the theory, one would counter that they have done so selectively in a way that perpetuated current socially accepted beliefs about both types of offenders. However, an answer from the data has not yet been established. Based upon the previously mentioned conclusion, let us now discuss Howitts (1998) and Baraks (1994) finding. They found that media sources tend to not only interpret stories but oversimplify stories into false dichotomies, either black and white. There is no middle ground. They also found that the media tend to focus on the individual for his or her situation rather than placing partial or total blame on social policy or some social institution ( Howitt, 1998; Barak, 1994). How is it that the newspaper tends to blame individuals? When the article presents the story in terms of either black or white, wrong or right, the reader may be left with the impression that the offender is either wrong or right for doing what he or she has done. There is not a middle ground. There is no room to introduce the idea that the offender did act a certain way, or that he or she may have acted differently if the contextual or situational characteristics of the occurrence where different. One reason to place total responsibility on the suspects involved is that it is easier for society to see these actions as an example of abnormal behavior rather than as a social problem.

PAGE 19

12 Barlow, Barlow, and Chiricos (1995) are a group of media-criminologists, as were Barak (1994), Howitt (1998), and McQuail (1992), who argue that the media rarely give coverage to cases involving working class and unemployed communities. This is primarily because these cases are not seen as noteworthy news. Again, this leads back to the previous finding that the average white American expects to see crime and violence in neighborhoods and communities where the working class and unemployed reside. An article on TSV is most likely to be published by the newspapers when the story is sensational and maintains the status quo ideologies of American society. 1.4.4 Critical-Media Perspective David Paletz and Robert Entman (1981) provide a general analysis of the media system in the U.S. In their analysis, they include discussion of both the sources of media and the factors that influence what is offered by media organizations and how it is presented. A key point described in their analysis is that the owners of media agencies are wealthy Americans with traditional American values . [such as] individualism, free enterprise, competitiveness, and materialism ( Paletz & Entman 1981, p. 10). These owners operate media agencies the same as any other business concerned about profits and prestige. Paletz and Entman (1981) note that the search for profit frames or places limits on the content that media agencies will present. Owners and producers of news coverage do not want to offend powerful and influential readers and advertisers because it is not good for business ( Paletz & Entman, 1981). Linking this in with findings from this literature review, one may state that this gives further evidence as to why news agencies are not likely to offer coverage that is counter to main-stream American ideologies. Not only would news producers potentially lose readers, but they could also lose advertisers, the support of powerful individuals in society, and potential investors.

PAGE 20

13 News agencies are businesses and must take into consideration the various consequences of the news that they provide before they offer coverage of certain topics. Paletz and Entman (1981) provide additional insight into the discussion of what makes something newsworthy. They state that journalists usually have a set of guidelines provided by editors and news executives that provide them with direction on how to discern what is newsworthy. According to Paletz and Entman (1981), some of the key components of newsworthy events are the following questions: does the event show some degree of drama and does it show some threat to American citizens either in a physical or a social sense? If the newsworthy event does threaten American citizens, it must have some resolution or solution so that the American public will be reassured of their security ( Paletz & Entman, 1981). The idea of threat is linked with cases of NTSV because these occurrences provide a potential physical threat for American children that are sent to school everyday and may come into contact with another student that is the next NTSV offender. The NTSV phenomenon is also a social threat. 3 News articles ask rhetorical questions about the future of American society pondering what is the matter with Americas youth? The model, provided by Paletz and Entman (1981), is not complete without the reassurance to American citizens that everything will be all right. Most news 3 Social threat refers more to the idea that the threat is societal in scale and may involve nonphysical threats to the average white American in general. For example, in the case of NTSV, the phenomenon obviously provides a physical threat that is presented as though it is exists on a national scale, but it also provides a sense that something is wrong with average white American youths and the average white American way of life in general. This is what generates a sense of newsworthiness in the case of NTSV. TSV, on the other hand, even if it is recognized as a nationwide threat of physical violence, the average white American does not necessarily feel a direct physical danger because the school system is segregated for the most part. The schools where the TSV is happening are not likely to have a direct affect on the children of average white Americans. Likewise, TSV does not provide any sort of social threat to the average white American. It is usually seen and presented as instances of black on black crime and has no bearing on the overall health and wellbeing of average white American youths.

PAGE 21

14 agencies that offer coverage of NTSV provide this reassurance with the help of government agencies in the form of expert advice about warning signs of disturbed youths, counselors for families of victims and survivors of NTSV attacks, political and police officials promise that everything is being done to resolve the problem and that the offenders have been apprehended. This model of human interest stories, as Paletz and Entman (1981) refer to them, can also be extended to include the TSV phenomenon (p. 18). Those cases which offer an opportunity for drama, yet also give an opportunity for resolution in a socially-acceptable way will be considered newsworthy and are more likely to receive coverage. The critical component in this model in the way that it explains the coverage of TSV cases is the idea of the reassurance portion of news coverage. In order to reassure readers that a particular case of TSV is no longer a threat to their physical or social wellbeing, journalists must provide a socially acceptable conclusion in their coverage of TSV cases. From this reading, I hypothesize that a socially acceptable resolution from the point of view of the average American will include a situation where the offender has been detained and locked away. This is the same conclusion that would be drawn in the coverage of NTSV cases. There too, one would expect to see that the offender has been apprehended and is in jail. The difference in the type of coverage that one would expect to see is that the TSV cases are not likely to be presented as a social threat to the average white American. In the case of NTSV, it is exactly the opposite. The average white American reader is likely to have a child who attends public school and who is potentially endangered by this recent form of school violence. This is a major explanatory factor in why NTSV is expected to be portrayed in

PAGE 22

15 a different way than TSV. News agencies are compelled to cover NTSV differently because of the affect that it has on their main audience, the average white American. In summary, the literature appears to suggest that news agencies should portray NTSV and TSV differently. NTSV provides an additional component of newsworthiness that TSV lacks, and that extra component is the threat to personal wellbeing of the average white American reader in both a physical and social way. When one considers news agencies from a business perspective, and considers that the average white American reader is the target audience of news producers, it becomes clear why news agencies would likely provide different coverage for TSV and NTSV. By providing detailed and resolute coverage of NTSV, news agencies are protecting their bottom line. 4 News producers are attempting to create solidarity among their readers and assure them that this new problem (NTSV) is under control; it is limited to a few sick and mentally ill young people who exist at the fringe of the population of young people in America. One could speculate that news agencies appear to provide the reassurance to maintain 4 By resolute coverage, I mean that the news article will likely portray the case of NTSV as being an abnormal occurrence as opposed to some new main-stream social movement by the youths of America. The offender is abnormal, he or she is psychologically ill, unable to cope as normal children are with the stresses of growing up. Blame may be placed on the parents of the children. Expert advice will be given along with reassurance from various officials that everything is being done that can be done to ensure the safety of Americas schools. The U.S. Congress and state legislative bodies debated various proposals with regard to the issue of gun control in the aftermath of NTSV in order to reassure the average white American that everything was being done to protect his or her wellbeing to include the safety of his or her family. Note also that this kind of dialog does not take place with regard to TSV. There are not officials on national media programs stating that they will do everything it takes to protect America against TSV offenders. Congress has not debated laws that concern the need for resolution to systemic problems within lower and working class communities that gives rise to TSV in schools. There appears to be a completely different approach in the way that new agencies handle TSV.

PAGE 23

16 current social beliefs that are necessary for the continuation of capitalist American pursuits. 5 1.4.5 Gaps in the Literature Specific gaps in the existing literature are as follows: No one has specifically studied the relationship between media portrayal of school violence and the factors about the school violence cases that affect the type of coverage given to the case. The conflictlabeling literature provides a foundation for the idea that laws are created by those who have power and those who do not have power are often labeled negatively with regard to those laws based upon their interactions with those who have power. Power-threat literature raises the idea that newspapers cover the different types of school violence differently to support the current social ideologies and to reinforce those ideologies. The differential coverage further segregates society symbolically, which perpetuates the existing system of stereotypical beliefs. This creates a self-perpetuating cycle of discrimination. The media criminology literature shows that media sources provide coverage of TSV cases that are sensational and, again, perpetuate the current social thoughts on TSV offenders. Finally, critical-media literature provides insight into what makes a particular event newsworthy. Ultimately, the criteria are linked to the financial prosperity of news agency. News producers are less likely to provide coverage that will inhibit their agencys ability to make a profit and gain prestige. 5 The reassurance was necessary to provide a sense that everything is going to be all right. One can easily see the ramifications of what a lack of reassurance and faith in the mainstream American way of life can do for the American capitalist economy in light of current lack of assurance and faith that the American public has in corporate America. In the face of current uncertainties and the apparent inability of news producers to reduce the anxiety in America, the economy is not doing extremely well. This is bad for those in power, those that have influence over the media; those that have much to lose if American and World economies do not start to move in a positive direction. Reassurance is a critical function of news agencies.

PAGE 24

17 The real gap is that none of the researchers have tested these ideas on cases of school violence directly. Turk (1964) and Quinney (1970) analyzed incidents in which youths were coming into contact with authority figures in society usually through some criminal justice system process, but not necessarily related to school violence. Blalock (1967) developed his ideas around civil rights issues and interactions that AfricanAmericans had with white Americans. Howitt (1998), Barak (1994), McQuail (1992), and Barlow, Barlow, and Chiricos (1995) studied the affects of media on the production of crime. Paletz and Entman (1981) were interested in the process of news production in general with no particular interest in media coverage of school violence. However, this study seeks to focus on the relationship between the way newspapers portray school violence and key factors about the offenders behavior that affect the way he or she is portrayed by domestic newspapers. The specific goals of this project are to determine (1) if domestic newspapers do portray the two types of school violence (traditional and new trend) in different ways, (2) if newspapers do differ in coverage of the two types of school violence, how do they differ in coverage, and (3) finally, why do the newspapers differ in coverage of NTSV and TSV? The differential coverage of school violence has something to do with expected normal behavior which is linked with the discussion of newsworthiness offered above. 6 It is hypothesized that the cases of NTSV will be portrayed as abnormal or deviant by the newspapers, where as TSV will be seen as less unusual. 6 Expected normal behavior is linked with newsworthiness in the following way: an event is considered newsworthy if it involves drama and is entertaining. According to Paletz and Entman (1981), drama as it relates to news coverage involves the idea that the event is threatening to the public, both physically and socially, and that there is some resolution to the problem. Expected normal behavior is related to the case of NTSV because these cases are sensation or involve drama. The behavior that is exhibited by NTSV offenders is not expected and is not seen as

PAGE 25

18 normal behavior for someone of their social position and race. In the case of TSV, the behavior exhibited by the offenders is considered to be normal for someone of their social position and race and is therefore considered to be expected normal behavior. Most likely, the only instance in which TSV will receive special interest is where the offenders behavior is unexpected, or put in the terms of Paletz and Entman (1981), where the behavior provides a social threat to the average white American in general.

PAGE 26

19 CHAPTER 2 METHODOLOGY The purpose of this paper is to study the portrayal of domestic school shootings by the newspapers in the U.S. using a compilation of theory and concepts developed from the structural functionalist and conflict perspectives, labeling-theory, power-threat hypothesis, media criminology, and critical-media perspective. There is no pre-existing data-source with the variables of interest for this study. It is, therefore, necessary to develop a dataset by using the Lexis Nexis search utility to find newspaper articles that covered school shootings in the U.S. The time frame for the search has been limited to include only articles which were published from 1 st May 1997 through 30 th April 2002. Lexis Nexis allows for a search to be conducted using keywords. The keywords used in this study were school shooting. These keywords were used as opposed to school violence because this study is specifically interest in school shootings rather than school violence in general. 1 Of the articles that were gathered for this study, the following list of twenty three newspapers contributed either full length articles covering the cases of school shooting, editorials, captions of a photograph at the scene of some school shooting or memorial 1 It is interesting to note, however, that the keyword search for school violence brought up a roughly similar number of articles on the Lexis Nexis search. This search utility searches for keywords in around fifty major newspapers worldwide. The main difference in using keywords school shooting as opposed to school violence is that the articles related to the keywords school shooting are primarily concentrated in the United States. While those articles found using keywords school violence had a larger percentage of articles that were written about incidents that occurred in other countries as well as the United States.

PAGE 27

20 service, or some form of a chronological listing of school shootings with brief details on each one: Rocky Mountain News Sacramento Bee The Washington Post Los Angeles Times The New York Times The Times-Picayune The Baltimore Sun The San Diego Union-Tribune Newsday St. Louis Post-Dispatch The Boston Globe Milwaukee Journal Sentinel The Atlanta Journal Constitution The Seattle Times USA Today The Hartford Courant The Plain Dealer St. Petersburg Times Chicago Sun-Times Information Bank Abstracts (Wall Street Journal) Star Tribune The Buffalo News and The San Francisco Chronicle Most of the newspapers offered either local or national coverage of school shootings. Some newspapers offered both local coverage of events that occurred within the community where the paper is published and coverage of school shootings in other communities around the country. 2.1 Summary of Findings News articles found using keywords school shooting for the five year period were analyzed. Initially, a total of 757 articles resulted from the search. Articles published by foreign news agencies or written about school shootings in other countries were deleted. Articles were removed from the sampling frame if they were duplicates. 2 In general, if there were duplicate articles, the more elaborate of the two was included in the sampling frame. This method of selection for inclusion was chosen simply to have some systematic way for choosing between duplicate articles. After this filtering process was complete, a total of 578 articles remained. Of these 578 articles, I randomly selected 100 articles for inclusion in the sample. Of these 100 articles, 27 cases of school shooting emerged.

PAGE 28

21 A few general trends emerged within these datasets: Table 2-1. Number of articles written per case. Case Name Number of Articles Albany HS (Albany, California) 1 Alva W. Dimmitt MS (Seattle, Washington) 1 Appalachian School of Law ( Grundy, Virginia) 1 Buell ES Mount (Morris, Michigan) 2 Burlington HS (Burlington, Wisconsin) 1 Carter G. Woodson MS (New Orleans, Louisiana) 2 Columbine HS (Littleton, Colorado) 22 Destrehan HS ( Destrehen, Louisiana) 1 Fort Gibson MS (Fort Gibson, Oklahoma) 1 Heath HS ( Paducah, Kentucky) 2 Heritage HS ( Conyers, Georgia) 10 John Batram HS (Philadelphia, Pennsylvania) 1 Lake Clifton Eastern HS (Baltimore, Maryland) 1 Lake Worth MS (Lake Worth, Florida) 1 Lew Wallace HS (Gary, Indiana) 1 Martin Luther King HS (New York, New York) 2 Parker MS ( Edinboro, Pennsylvania) 1 Pearl HS (Pearl, Mississippi) 1 Public School 18 ES (Buffalo, New York) 1 Ridgewood HS (New Port Richey, Florida) 1 Santana HS ( Santee, California) 12 Shawnee Mission South HS (Kansas City, Missouri) 1 Sumner HS ( Sumner, Washington) 1 Thurston HS (Springfield, Oregon) 2 Vigor HS ( Prichard, Alabama) 1 Westside MS (Jonesboro, Arkansas) 5 ? ES, (East Boston, Massachusetts*) 1 Total number of articles 77 The actual name of the school was not given in the article and other means of uncovering the name of the school were unsuccessful. Note: The names of the schools are given and the city and state are given. The abbreviations ES, MS, and HS indicate elementary school, middle school, and high school respectively. One can see from Table 2-1 that not all of the one hundred articles were written about one specific case. Only seventy seven articles are actually written about any one of the twenty seven cases of school shooting included in the sample. Of the thirty three 2 Duplicate articles were written about the same incident and published at two different instances either by the same newspaper or by different newspapers. The articles were listed under the same

PAGE 29

22 articles that are not written about a specific school shooting incident, one does not cover school shooting at all, there are three different articles that list several cases of school shootings that have occurred over a given period of time, and a majority of the twenty nine remaining articles are editorials that offer reactions to school shooting in general but that do not necessarily discuss a specific case of school shooting. Of the articles that do discuss specific incidents of school shooting, most of them are concentrated in the coverage of a few cases of school shooting. The vast majority of cases are only discussed in one brief article. 2.2 Discussion of Data Characteristics There are several limitations of these data that require discussion. However, I will first provide a listing of the variables that were common among almost all articles included in the sample. The following information was included in almost all articles: title of article, placement of article in newspaper, number of words in article, publication date of article, and the name of the school where the shooting occurred. The data used in this study includes a wide variety of articles. Some are full length descriptions of the events surrounding the shooting along with the reactions of school officials and police. Others are editorial descriptions and analyses of the shootings. Still others are very limited captions related to a photograph that was taken at the site of the school shooting or a memorial service. There are approximately three articles that list several incidents of school shootings that had taken place within the past few years, and these include a short description of each case. Additionally, there are a few articles that provide discussion of legislative and political issues that were debated regarding school shootings and school violence in general. title but the word count was slightly different.

PAGE 30

23 The task of coding for pertinent variables was complicated by a lack of description of these variables in a majority of the articles. Of the twenty seven cases included in the sample, three describe race and only eight describe socioeconomic status. These two variables are critical for this analysis. To supplement the lack of information about these variables offered in the articles, the U.S. Census American FactFinder was utilized to find both race and median family income figured for the cities where the school shootings took place. 3 These data were compiled at the case level of analysis along with other composite findings from the one hundred articles in the sample. Therefore, there are two data sets: (1) one with variables for each individual article, and (2) another for the composite variables that provide information on each case of the twenty seven cases. 3 The American FactFinder can be found at www.census.gov

PAGE 31

24 CHAPTER 3 FINDINGS AND ANALYSIS 3.1 Research Findings 3.1.1 Qualitative As described in the previous chapter, the vast majority of cases discussed in this study were only covered by one article. At the same time, there are a few cases that have a relatively large number of articles describing the events surrounding a given shooting. The articles in the sample vary greatly in their coverage of school shootings. Some of them give in-depth descriptions of the shooting and the events surrounding the offense. These articles tend to describe the reactions and plans of the school administrators as they deal with school shootings. Other articles briefly describe the shooting and the circumstances in which the shooting took place. A few articles were captions explaining the content of photographs, which were not provided by Lexis Nexis. Editorials included in the sample provide less descriptive coverage of the shootings, but they do provide insight about the way the authors of these editorials think administrators should deal with school shootings. Selective listings of school shootings are included in the sample. It is assumed that these listings are those cases that the news publishers considered to be newsworthy school shootings. The general tendency of the in-depth articles written on school violence was that they covered the more popular cases of school shootings, or put another way, they covered the more newsworthy cases of shootings. These detailed articles provided at least the basic information about each case, including the descriptions of the situation where the shooting

PAGE 32

25 took place, the age, name and gender of the offender, and the number of offenders and victims involved. The article would usually include a description of the surrounding community, the neighborhood where the offender lived. They were typically based on providing interviews with friends, teachers of the offenders, or survivors of the shooting and sometime all three. There is likely to be some description of the school administrators plans for dealing with the shooting, some description of aide to the victims and survivors, and possibly some discussion of legislation and gun policies. There are only a handful of these articles in the sample and they cover only the few cases which seem to have received the highest percent of coverage from the sample of articles used in this study. Another slightly different type of article described the basic information of a case briefly, and then, moved into a debate regarding legislation on gun control and school violence. 1 Some articles of this type discussed the political stances of the presidential candidates with regard to gun control policy as they were campaigning for the 2000 elections. Others discussed the implications of the age of the school shooting offenders and whether or not they would be tried by the judicial system as adults or juveniles. The underlying theme of this type of article is that it usually revolves around the ability of school officials to deal with the legal and political aspects of school shooting. A third type of article reviewed in this study briefly describes the circumstances of the shooting and provides some of the basic information for the case. Of the incidents in the sample that received coverage by one article, a majority of those articles were of this type. They were very simple descriptions of the case with limited description or discussion of any other facts.

PAGE 33

26 Captions that describe the content of a photograph were included in this sample. There are only a few of these captions in the sample itself. The captions provide a brief outline of the case to include most of the basic information and then a brief description of the photograph is given. Some of the captions only describe the photograph and do not describe the basic information about the case. Therefore, one can see that the information given in these captions varies greatly. Editorials included in the sample provide a plethora of information about school shootings. Most do not provide the detailed information about the case itself. They do not given descriptions of the offenders, victims, and crime scene. They do, however, provide information about the way that a variety of U.S. citizens see and interpret school shootings. In essence, they provide a crude indication of how newspapers and other media sources portray school shootings under the assumption that people formulate at least some of their opinion about school shootings based upon what they have learned from news coverage. The range of editorial opinions on this subject is so great that it is not feasible to describe each in detail. However, a few general tendencies do arise. Some editorials focus on reactions to school violence and tend to highlight a trend that the author feels is occurring in the U.S. These editorials discuss policy initiatives, gun control laws, and other alternatives for putting an end to school violence in the U.S. Another type of editorial, although rare, gives insight into the portrayal of school shooting and describes that school shooting is not necessarily a new phenomenon nor does the coverage of school shooting by the media give an accurate representation of what is actually happening in U.S. schools. Most editorials fall along these two types: (1) either providing a discussion on policy and legislation related to school 1 Basic information includes variables like, number of offenders and victims, age of offenders and victims, relation of offenders and victims, and name of school. Other than this the information

PAGE 34

27 shootings and gun control, or (2) providing information concerning the actual issue of school shootings in the U.S. and whether or not they are accurately described by media sources. Finally, there are three articles that provide listings of school shootings in the U.S. over certain periods of time. These listings are provided by journalists that work for a newspaper company. They are therefore the listings of shootings that are considered newsworthy by the news industry. The listings do not include all of the school shootings that occurred during the period covered by the article. The primary utility of these listings is that it allows for a comparison of findings from this study with regard to level of unique interest. In summary, the articles in this sample tend to vary greatly in their coverage of school shootings. Some of them offer in-depth descriptions of the shooting and the surrounding events. Additionally, they may describe the reactions and plans of the school administrators as they deal with school shootings. Other articles provide only a brief description of the shooting and the circumstances. Editorials included in the sample provide less description of the shootings, but more insight into the way that the authors of the editorials think about school shootings. Listings of school shootings are included in the sample. The listings, taken as a whole, provide a list of cases which the news industry believes to be newsworthy. 3.1.2 Quantitative The following list of variables has been compiled in the dataset for this study. Each article included in the sample has been coded for these variables. In the dataset that is nonaggregated, these articles are coded. Several of these variables are used to build an aggregate dataset that is coded for each case. Additional information has been collected for this aggregate level data. The sources for these data are both the articles provided by Lexis provided varied and even this basic information is not always given in each article.

PAGE 35

28 Nexis, the year 2000 data from U.S. Census Bureau American FactFinder utility, and the World Wide Web. 3.1.2.1 Non-aggregate Dataset Table 3-1. Basic figures for non-aggregate variables. Variable N Minimum Maximum Mean Std. Deviation section of the newspaper 99 1 4 2.48 0.98 number of words 100 20 2700 545 471 number of offenders 50 1 7 1.58 1.11 number of offenders killed 50 0 2 0.14 0.50 number of victims 47 0 45 8.70 9.63 number of victims killed 45 0 23 2.82 4.48 race of victim 5 1 3 1.80 1.10 race of offender 3 1 3 1.67 1.15 type of weapon 21 0 3 2.10 0.62 quantity of weapons 26 0 30 2.96 5.99 SES victim 10 1 2 1.80 0.42 SES offender 11 1 2 1.73 0.47 relation of victim to offender 47 1 2 1.28 0.45 unique interest 89 0 7 2.47 1.91 type of article 99 1 4 2.04 1.28 Section of the newspaper. Each article is published in a certain section of the newspaper and this variable refers to that section. The variable was broken down into several possible replies: (1) front page of news paper; (2) front page section but not on front page; (3) front page of a section other than the front page section; and (4) in a section other than the front page section not on front page of the section. On average, news articles were coded 2.48. The largest percent of articles were in the second category at 56.6%. Twenty five point three percent of the articles fell within the third category. The vast majority of articles do not make front page news. Of those articles that do appear on the front page, six out of ten, cases are related to NTSV shootings. Number of words. The number of words in the article is recorded as given by Lexis Nexis. The mean number of words for the set of articles was 544.7 with a standard deviation

PAGE 36

29 of 471.5. Of those articles which are in the top quartile of this variable, most of them are NTSV cases. Most of the articles in the top quartile tend to score above the mean on the variable that measures level of unique interest. Case of school violence. This variable is simply the name of the school where the school shooting took place and city and state where the school is located. Thirty three articles in the sample did not have any entry for this variable. This was due to the fact that they did not address a specific case, they did not focus on school shootings, or they covered several school shootings. Number of offender(s).* 2 The number of offenders involved in the case is taken directly from the article. Fifty out of the one hundred articles include information on this variable. Of the fifty articles that do include information on this variable, the average number of offenders is 1.58 as given per article. Number of offender(s) killed.* This is the number of offenders reported in a given article as being killed. Only the Santana High School and Columbine High School shootings were reported as having offenders that were killed during the act of committing the school shooting. Both of these can be classified as NTSV cases. Number of victim(s).* The number of victims involved in each case is taken directly from the article. Forty seven out of the one hundred articles included in the sample give information about the number of victims for a given case. The average number of victims involved in a school shooting is 8.70 as given per article. However, 63.8% of the forty seven articles including information about the number of victims include ten victims or less. The 2 The variables that are marked with asterisks were originally gathered at the article level of analysis but were later compiled at the case level of analysis. A description of the data for these variables will be given at both levels of analysis.

PAGE 37

30 Columbine High School incident makes up 23.4% of the forty seven articles with a total of forty five victims reportedly involved. Number of victim(s) killed.* The average number of victims that are reported a killed in the incident is 2.82. Of the forty five articles that include data on the number of victims killed in a school shooting, 91.1% of them report five victims or less. The two cases in which the number of victims killed was the highest are Columbine High School and Westside Middle School. Both of these cases are considered NTSV cases given their level of unique interest derived from the articles in this sample. Race of the victim and race of the offender.* The race variable is coded as follows: (1) white, (2) black, and (3) other. The race of the victim and the race of the offender are both inadequately discussed in the news articles. Of the one hundred articles in the sample, five gave information about the race of the victim. Three articles stated that the victim(s) were white and two articles gave very strong indications that the victim(s) race was not white. Three articles specifically state the race of the offender(s). Of these three articles, two reported the offender(s) as being white while one article gave an indication that the offender was not white. Type of weapon used in the shooting .* This variable is used to record the type of weapon that an offender is reported to have used in a school shooting incident. The variable is coded in the following manner: (0) no weapon is used, (1) a weapon other than masscasualty producing weapon is used, (2) a gun other than mass-casualty producing gun is used, and (3) a mass-casualty producing weapon is used. 3 3 Weapons other than mass-casualty producing weapons are defined as those weapons which can be used to assault and murder someone but this category does not include guns of any kind, examples include knives, brass knuckles. Guns other than mass-casualty producing weapons includes guns that can be used to kill people but do not use a high caliber ammunition and do not fire at a high rate of

PAGE 38

31 Of the one hundred articles in the sample, twenty one reported specific information about the type of weapon used in the school shooting. Of these twenty one articles, 76.2% reported that offender(s) used guns that were not mass-casualty producing to commit the shooting. A total of four out of the twenty one articles reported that the offender(s) used mass-casualty producing weapons. Three cases were associated with the reported use of mass-casualty producing weapons: Columbine High School, Thurston High School, and Heritage High School. Of these cases, only Columbine High School appears to be a NTSV case according to its level of unique interest. Quantity of weapon(s) used by the offender(s).* This variable simply indicates the number of weapons reported by an article as being used in the shooting. Thirty articles report information regarding the number of weapons used in a shooting incident. Seventy three point one percent of the articles report that one weapon was used in the offense. Columbine was reported to have the highest count of weapons at thirty weapons to include guns and bombs. The next highest count of weapons was nine for both Buell Elementary School and Thurston High School. Westside Middle School shooting came in with the third highest count reportedly involving the use of six weapons. Socioeconomic status of the victim(s) and offender(s).* This variable was coded as follows: (1) low, (2) medium, and (3) high. It was originally thought that the SES of the offenders would be the same as the SES of their parents. The SES of the victim(s) and offender(s) was to be recorded if it was reported in the news articles; however, only a small percent of the articles in the sample reported SES. speed, examples include .22 caliber handgun, and .22 caliber rifle. Mass-casualty producing weapons include weapons that can shoot large caliber ammunition and/or can fire at a high rate of speed. This category also includes explosive devices that can produce a large number of casualties examples include shotgun, machine gun, high-powered rifle, and bombs.

PAGE 39

32 Ten of the one hundred articles in the sample reported any information on the victims SES. Meanwhile, eleven articles in the sample reported information on the SES of the offenders. Of those articles that do report on the level of SES of victims and offenders, 80% and 72.7% of the articles list the SES of medium, respectively. Relation of victim to offender. This variable indicates the relation of the victim to the offender. The main relations that are theoretically significant are the following: (1) is the victim a peer to the offender or (2) is the victim an authority figure in relation to the offender? Forty seven out of the one hundred articles in the sample reported information on the relation between the victim and offender. Of these forty seven articles, 72.3% reported that the relationship between the victim and offender was a peer relation. The number reporting that that victim was an authority figure in relation to the offender was 27.7%. Of this 27.7% of articles reporting that the victim was an authority figure, there is no clear indication of whether the majority of these articles are NTSV or TSV in nature. Level of unique interest. The level of unique interest is a composite variable which is an overall indication of whether or not the author of the article showed a level of interest in the case that would be considered high with regard to this type of news coverage. There are seven variables that were used to determine the level of unique interest (UI). Each of these variables is coded (0) for a no response and (1) for a yes response indicating that yes the article does portray this variable. The seven variables are described below: (1) is the psychological-medical model used implying that the offender is somehow mentally unstable? (2 ) Is the offender labeled as an outsider or portrayed as an outsider from the majority of students? (3) Does the article used words such as empathy, sympathy, and understanding to

PAGE 40

33 describe the school shooting and the events surrounding the shooting? This variable has two potential positive answers: in one way, the article may be written in a way so that it makes an attempt to understand what has happened, in another way the article may express sympathy toward the victims and families of the victims. Either of these two attempts by the author would result in an affirmative response for this variable. (4) Does the article portray a sense that one should not expect a school shooting to happen in the community where it did? (5) Does the article convey a sense of urgency that the problem of school shootings in America is a critical problem that must be resolved right now? Also, is there a specific indication that the authority figures involved in the case under analysis reacted promptly to alleviate the shooting incident and any problems that may have arisen regarding the incident? Either of these two indications in the article would have given an affirmative response for this variable. (6) Is there a description of the actions that the school administrators and other authority figures involved in the case have taken to rectify the problem? (7) Does the article describe any aide that was provided for the victims and other students involved in the incident? Each of these seven variables was coded and the scores were tallied, and the resulting figure was the level of unique interest (UI) that was recorded for a given article. The range of score for the level of UI of any given article was zero to seven. If none of these variables were exhibited in the article, then a total score of zero UI would be recorded for the article. If all of the variables exhibited in the articles, then a score of seven would be recorded for the article. A total of eighty nine out of the one hundred articles in the sample were given a score on this variable. The actual range provided by the data for this variable was from zero to seven. The average score for any given article was 2.47. Of those articles that scored higher

PAGE 41

34 than the mean of 2.47, Columbine High School has eight hits, Santana High School has six hits, Carter G. Woodson Middle School has two hits, Westside Middle School has two hits, Heritage High School has two hits, Albany High School has one hit, Alva W. Dimmitt Middle School one hit, Burlington Middle School has one hit, Ridgewood High School one hit, Thurston High School one hit, and there are ten additional articles that scored above the mean on the UI scale but are not associated with any particular case of school violence. 4 Type of article in newspaper. This variable refers to the type of article as one of the following choices: (1) article related to one case of school shooting, (2) editorial, (3) citation of a photograph, (4) other which includes articles that do not fall into any of the other three categories. Ninety nine of the one hundred articles in the sample were coded for this variable. The one article that was not coded is the article that is not related to school shootings at all. Of these ninety nine articles, fifty two are full length articles that discuss one case of school shooting. There are a total of seventeen editorials in the data sample. Four photographic citations are included in the sample. A total of twenty six articles were listed as other. Of the articles listed in the other category, there appears to be no trend in the characteristics of these articles other than the fact that most of them are missing information on basic variables in this dataset and many of these articles have a small word count. 3.1.2.2 Aggregate Dataset Number of offender(s).* This variable refers to the number of offenders that are reported for each case. Twenty five of the twenty seven cases have reported figures on the number of offenders. Sixty four percent of the cases are reported as having one offender. 4 Hit indicates that there is one article written about the case that has a UI score above the mean of 2.47.

PAGE 42

35 Twenty four percent of the twenty five cases reported on were listed as having two offenders. Destrehan High School, Burlington High School, and Pearl High School were reported to have four, five, and seven offenders, respectively. Table 3-2. Basic figures for aggregate variables. Variable N Minimum Maximum Mean Std. Deviation number of offenders 25 1 7 1.76 1.48 number of offenders killed 25 0 2 0.12 0.44 number of victims 25 0 45 6.00 10.25 number of victims killed 24 0 13 1.58 2.81 race of victim 3 1 3 1.67 1.15 race of offender 3 1 3 1.67 1.15 revised race of victim 8 1 4 2.00 1.41 revised race of offender 13 1 3 1.38 0.65 type of weapon 15 0 3 2.07 0.70 quantity of weapons 18 0 30 3.78 7.10 SES victim 8 1 2 1.75 0.46 SES offender 9 1 2 1.67 0.50 number of articles per case 27 1 22 2.85 4.69 articles written within 1 year 27 0 20 2.41 4.38 unique interest within 1 year 23 0 5.0 2.24 1.49 median family income 27 $23,519 $66,870 $53,226 $11,599 percent white 27 11.90 95.90 66.87 24.83 percent black 27 0.4 84.50 23.18 25.55 percent other 27 1.30 29.30 9.74 8.67 total population of city 27 1,105 8,008,278 481,685 1,541,693 Number of offender(s) killed.* This variable records how many of the offenders were killed as a result of their involvement in the shooting. Columbine High School and Santana High School are the only two cases in this dataset that involve the murder of the offenders. No other offenders in this dataset were killed as a result of their involvement in the shooting. Number of victim(s).* Forty eight percent of the twenty five cases reported that the number of victims were one or less. A total of 20% of the twenty five cases reported no victim at all. Eighty percent of the twenty five cases did not have more than eight victims.

PAGE 43

36 Santana High School, Westside Middle School, Thurston High School and Columbine High School had a reported total of 15, 15, 26, and 45 victims, respectively. Number of victim(s) killed.* This variable refers to the number of victims that were reportedly killed as a result of the school shooting. Forty five point eight percent of twenty four cases report that no victims were killed. Twenty five percent of twenty four cases show that one victim was killed. Columbine surpasses all of the other cases in this dataset with a total of thirteen victims dead. Race of the victim and race of the offender.* The races reported in this dataset are recorded directly from each news article. In instances when the article did not explicitly state the race of the victim or offender, no race was recorded. A total of three cases in this dataset had information that indicated the race of the victim: two cases had white victims and one case had non-white victims. Twenty four cases in this set do not include information about the race of the victim because it is not explicitly given in any of the articles in the sample. Likewise, there are a total of three cases in this dataset that report the race of the offender: two cases reported white offenders, while one case reported non-white offenders. The race of offender and victim variables have been revised based upon additional internet searches performed for the explicit reason of identifying the races of offenders and victims. The new results of these searches are listed in Table 3-2 above. An additional race category had to be created for the revised variable. The new category is (4) multiple races which means that there were either victims or offenders of different races involved in a given case. For example, in the case of the Columbine shooting, two white offenders shot and killed several white victims and one black victim. The Santana shooting was the only one other case involved the used of the new code. The offender in the Santana shooting was

PAGE 44

37 white and he shot and killed one white victim and one Asian-American victim. Both Columbine and Santana are cases of NTSV. Type of weapon used in the shooting .* This variable is used to record the type of weapon that an offender is reported to have used in a school shooting incident. The variable is coded in the following manner: (0) no weapon is used, (1) a weapon other than masscasualty producing weapon is used, (2) a gun other than mass-casualty producing gun is used, and (3) a mass-casualty producing weapon is used. Fifteen cases reported information for this variable. Of these fifteen, one case reported that no weapon was used. Three cases reported that mass-casualty producing weapons were used, and the remaining cases were reported to have guns other than mass-casualty producing guns involved in the shooting. Heritage High School, Thurston High School, and Columbine High School reportedly involved the use of mass-casualty producing. Quantity of weapon(s) used by the offender(s).* A total of eighteen out of the twenty seven cases reported data describing the number of weapons used in the school shooting. Of these eighteen cases, one reported that no weapons were used. This case was actually one in which the offenders were planning an attack on the school but the police found out about the plan before the offenders were able to carry out the attack. The offenders had not been able to gather any weapons. Sixty six point seven percent of the eighteen cases reported that one weapon was used in the offense. Of the remaining cases, an unknown elementary school in Boston, Westside Middle School, Thurston High School, Buell Elementary School, and Columbine High School reportedly had two, six, nine, nine, and thirty weapons involved in the shootings, respectively.

PAGE 45

38 Socioeconomic status of the victim(s) and offender(s).* Eight of the twenty seven cases reported on the SES of the victims. Of these eight cases, the SES of the victims of two cases was reported to be low. The remaining six cases indicated that the SES of the victims was medium. Nine of the twenty seven cases in this dataset reportedly have data on the SES of the offenders. Of the nine cases with SES data on the offenders, three cases have offenders of low SES. Six of the nine cases, report that the offenders SES is medium. Number of articles per case. The findings of this variable were shown in chapter two. They will be described here briefly. Sixty six point seven percent of the cases in this dataset have one article included in the sample that corresponds directly with this case. Eighteen point five percent of the cases have two articles in the sample the directly correspond to the case in the dataset. Westside Middle School, Heritage High School, Santana High School, and Columbine High School have five, twelve, twelve, and twenty two articles, respectively, that are in the sample and that correspond directly with these four cases. Number of articles written within first year. This variable measured how many articles are written about a given case up to one year after the school shooting occurred. Three of the twenty seven cases in this dataset have no articles written corresponding to the case within the first year. Sixty six point seven percent of the cases in this set have one corresponding article in the sample that is written within the first year. Carter G. Woodson Middle School, Martin Luther King High School, Buell Elementary School, Heritage High School, Santana High School, and Columbine High School had two, two, two, ten, twelve, and twenty two articles, respectively, written about each case within the first year after the shooting occurred.

PAGE 46

39 Unique interest for articles written within first year. This variable refers to the average unique interest (UI) per case based upon the UI scores of all of the articles written on a given case within the first year. This method of recording UI per case for the first year may be misleading because those cases that have more articles written about them within the first year following the shooting tend to have disproportionately smaller UI levels even though some of the articles that were written about these cases do have high levels of UI. The range of UI for twenty three of the twenty seven cases in the set is from zero to five. Sixty five point two percent of the twenty three cases are at or below 2.2 on the UI scale. The mean UI for these twenty three cases is 2.41. Albany High School, Westside Middle School, Burlington High School, Thurston High School, Carter Middle School, Ridgewood High School, and Alva W. Dimmitt High School have 3.0, 3.0, 4.0, 4.0, 4.0, 5.0, and 5.0 levels of UI, respectively. Each of these cases has one or two corresponding articles in the sample. Median family income. The data for this variable were gathered from the U.S. Census Bureaus American FactFinder search utility on the www.census.gov website. 5 These data were organized on the basis of city name. All of the data are from the 2000 census updates. It is understood that all of the cases did not occur in the year 2000, but an assumption has been made that the data used from the census have not changed significantly over the period of time included in this study. The minimum family income recorded for any of the cities where the shooting occurred was $23,519 per year. The maximum median family income recorded for any of 5 The American FactFinder is a utility created by the U.S. Census Bureau that assists in finding demographic, economic, and various other forms of information that are gathered by the census bureau. These data are usually orders in geographically delimited regions. The user can search based

PAGE 47

40 the cities was $66.870. The median dollar amount for the median family income is $43,266 with a standard deviation of $11,599. Percent white. These data were also gathered from utilizing the U.S. Census Bureaus American FactFinder. The same qualification that applies above applies here as well. The minimum percent white for a given city where a shooting took place is 11.9% while the maximum percent is 95.9%. The mean percent white is 66.9% with a standard deviation of 24.8%. Percent black. These data were collected from the American FactFinder 2000. The smallest percent of the black population in a city where school shooting took place is 0.4% while the largest percent is 84.5%. The mean percent black for cities within the sample included in this dataset is 23.2% with a standard deviation of 25.6%. Percent other. Percent of those cities with a race other than white and black was also gathered using the American FactFinder utility. The minimum percent of other race recorded for this sample is 1.3% while the maximum percent other is 29.3%. The mean percent other race in this set of data is 9.7% with a standard deviation of 8.7%. Total population of city where shooting took place. The data for this variable originated in the American FactFinder utility. Of the cities that were included in the sample, the smallest city had a total of 1,105 people and the largest city had a total of 8.0 million people. The mean city population size was 481,685 people with a standard deviation of 1.5 million. If we exclude the outlier with the population of 8.0 million people the mean population is 139,187 with a standard deviation of 218,138 people. upon the limits for the region that he or she desires. In this study, I limited my search by city-level data.

PAGE 48

41 3.2 Analysis of Findings 3.2.1 Qualitative Analysis These findings will now be summarized and analyzed in the following qualitative analysis of the descriptive data given above. A majority of articles fall within front page section of the news paper but do not appear on the front page. Most of the articles are several hundred words in length although a few articles do exist in either extreme: very short or very lengthy. Carter G. Woodson Middle School, Columbine High School, Thurston High School, Santana High School, Westside Middle School, and Ridgewood High School appear in the top 25% of those articles with the highest word count. The majority of cases in this dataset have only one or two articles written on them that are included in this dataset. There are a few exceptions to this trend: Westside Middle School, Santana High School, Heritage High School, and Columbine High School all have five or more articles written about them that are included in this dataset. The vast majority of cases included in this study also had most of the articles written about them within the first year after the incident occurred. Again, there are exceptions to the rule: Heritage, Santana, and Columbine High Schools all have nine or more articles written about them within the first year after the incident occurred that are present in this sample. Of the cases included in this dataset, twenty three had articles written within the first year of the incident. A considerable percentage of these twenty three cases scored a level of approximately 2.0 on the UI scale. However, a considerable percentage scored 0.0 and 4.0 on the UI scale as well. The following list indicates the top 25% of cases that ranked the highest on the UI scale during the first year: Albany High School, Westside Middle School, Burlington, Thurston High Schools, Carter G. Woodson Middle School, Ridgewood High School, and Alva W. Dimmitt Middle School. A similar trend does appear in these results as

PAGE 49

42 it does in the previous results, but there are additional cases in these results that do not appear in the previous results. Most of the articles report no deaths of offenders and only a few deaths among the victims. However, there is a trend that starts to appear in this variable and will continue throughout the paper. Columbine High School had a total of two offenders and both killed themselves during the assault on the school. At Santana High School, the one offender involved in the case was reported as being killed during the shooting. Destrehan High School, Burlington High School, and Pearl High School were unique in that they had four, five, and seven offenders involved in the planning and or actually committing the shooting, respectively. The number of victims reportedly involved in a case range greatly from one through forty five. Here again, the trend develops: Pearl High School, Santana High School, Westside Middle School, Thurston High School, and Columbine High School are reported to have greater than ten victims involved in the shooting. While Pearl High School, Westside Middle School, Thurston High School, and Columbine High School all have the highest number of victims that are killed during the shooting as well. The only comment that can be made about the race of offenders and victims based upon the data given in the articles is that race is not discussed explicitly. One could make assumptions about the race of the offenders and victims based upon general knowledge of the racial makeup of the U.S. in addition to the language that is used to describe the school shooting, but an attempt has not been made to do so here. The methods used in this study do not allow for me to clearly state the race of the offender or victim, unless it is explicitly

PAGE 50

43 stated in the article. In summary, 97% and 95% of the articles do not give information about the race of the offenders and victims, respectively. When describing the type of weapon used in the school shooting, there is also a lack of information available from news articles in this sample. Seventy six percent of the sample does not include information about the type of weapon involved in the shooting. Of those articles that do present information about the type of weapon used in the shooting, the majority of the weapons that are used are guns other than mass-casualty producing guns e.g., a hand-gun. Heritage, Thurston, and Columbine High School shootings reportedly involved the use of weapons of mass-casualty producing capabilities. This gives additional evidence to the trend that appears to be manifesting itself in the data. It appears that of the twenty seven cases found in this dataset, eighteen have information about the quantity of weapons used in the assault on the school. Most of the offenders used one weapon. From the previous description on type of weapon, it is likely that this weapon is a handgun. However, there are several cases were the offenders used many weapons of varying types. The following list indicates cases in which the number of weapons used was in the top 25% reportedly used to carry out the offense: Unknown Elementary School in Boston, Westside Middle School, Buell Elementary School, Thurston High School, and Columbine High School. Again, a pattern is emerging from the data. A majority of the articles included in this sample are defined as regular length articles that describe a school shooting incident in sufficient detail. There are also, however, several articles that are editorials presenting a variety of views and discussions on school shooting. Columbine High School has the largest number of these normal length articles written with a total of ten. Santana High School has the second largest number of normal length articles

PAGE 51

44 written about the shooting that happened there with a total of eight articles. Heritage High School has the third highest number of these normal length articles written with a total of five. This supports the trend that has appeared in the study thus far. The SES of the victims and offenders is an integral part of the theoretical argument of this paper. There is a huge gap in this study left by the inadequate information that is provided about this variable on the individual level by the news articles. Of the twenty seven cases in this dataset, only eight and nine give explicit information about the victims and offenders SES respectively. The majority of offenders and victims appear to be middle class according to the vague information provided. Instead of using this information, the study utilized the American FactFinder utility of the U.S. Census Bureau to find data on a proxy measure of SES for the city where the school shooting took place. The proxy measure used is median family income. The range of the median family income of those cities were school shootings have taken place varies greatly. The names of the schools located in a town that is in the top 25% of median family income are listed below: Albany High School, Alva W. Dimmitt Middle School, Appalachian Law School, Burlington, Columbine, Destrehan High Schools, Parker Middle School, and Santana High School. The names of the schools in the cities that are in the top 25% for each race category are listed below. These schools that are in cities with the top 25% of whites in the population are: Santana, Thurston, Sumner, Columbine High Schools, Parker Middle School, Buell Elementary School, Ridgewood, and Burlington High Schools are in the top 25% of the white percent population variable. Shawnee Mission South High School, Heritage High School, Public School 18 Elementary School, John Batram Lake Clifton Eastern High

PAGE 52

45 Schools, Carter G. Woodson Middle School, Lew Wallace High School, and Vigor High School are all in the top 25% of the black percent population among the schools included in this dataset Santana John Batram High Schools, Lake Worth Middle School, unknown elementary school in Boston, Alva W. Dimmitt Middle School, Martin Luther King High School, Albany High School, and Fort Gibson Middle School are the top 25% of cases in the other race category which is made up of all other racial groups beside white and black. Many of the schools listed in the percent black and percent other categories have not been present in the other categories and variables that have been described thus far. This indicated that many of these other schools are not receiving the media attention that has been given to the schools that have been present in the previous listings. The schools that fall within the top 25% of cities with the largest populations are listed below: Public School 18 Elementary School, Shawnee Mission South High School, Carter G. Woodson Alva W. Dimmitt Middle Schools, unknown elementary school in Boston, Lake Clifton Eastern, John Batram and Martin Luther King High Schools. These schools appear to be the some of the same cases that have been showing up in the top 25% of black and other racial categories. In summary, the following general trends appear to be emerging out of the data that exist in this sample of articles and cases: (1) the majority of articles in the sample appear in the front page section of the news paper. (2) Most of the cases described in this dataset are only covered by one article that is included in the sample. (3) Twenty three of the cases included in this dataset had articles that were published within the first year after the occurrence happened. Of these, the average score on the UI scale was approximately 2.0. (4) The vast majority of articles reported no deaths among the offenders and an average of

PAGE 53

46 two deaths among the victims in any given case. (5) The actual number of victims involved in any given case varies significantly from case to case. (6) There is a complete lack of explicit information about the race of the victims and race of the offenders. (7) Many of the articles do not provide information about the type of weapons used in the offense. Of the approximately 25% of articles that do provide information about the type of weapon used, the primary type of weapon used was a handgun. (8) Slightly over half of the twenty seven cases described by these data offer information about the number of weapons used in the offense. In a majority of the cases that do provide information, the offender used one weapon to commit the offense. (9) A majority of the one hundred articles included in the sample provide full length coverage of any given case of school shooting. This is not to say that the full length articles are distributed evenly among the various cases. (10) Information concerning the SES of the victims and offenders is limited. Only several articles provide this information in the sample. Of these articles, the majority of them indicate that the victims and offenders are of middle class SES levels. (11) The range of the median family income varies greatly for this dataset. (12) The range of the actual population size of the cities included in this sample varies greatly. (13) The racial makeup of most of the cities included in the sample is primarily homogeneous. There are a few exceptions, but the bulk of cities in this sample consist of one racial group that consists of a large majority of the city population. Table 3-3 provides a visually representation of which cases appear to be exceptional. The cases that have a high incidence of checked boxes for variables one through twelve and fifteen should also be those cases that can be labeled as NTSV cases. Each time a case has a parameter that appears in the top 25% of a variable, there will be a checked box for that

PAGE 54

47 variable. The case with the highest number of checked boxes means that this variable has parameters which appeared in the top 25% of a lot of variables. The variables one through twelve and fifteen are those that, by definition, would indicate NTSV cases. Alternatively, those cases that have boxes marked for columns thirteen and fourteen are less likely to be labeled as NTSV cases. Table 3-4 provides a listing of those cases that have the largest number of markings in the boxes that would affirm NTSV status from lowest to highest. Columbine, Santana, Thurston, Westside, Burlington, Heritage, Albany, Alva W. Dimmitt, Pearl, and Ridgewood are within the top 25% of the list of cases on Table 3-4. Two high schools in the top 25% of cases are Columbine High School and Thurston High School. Two high schools in the lower 25% of cases are John Batram High School and Lew Wallace High School. To keep a more homogeneous group for comparison, the two bottom ranking and two highest ranking high schools will be analyzed in more detail in chapter four. Thr ee listings that were found by Lexis Nexis provide basic information on several cases of school violence (List of school shootings, 1998; School shootings, 1999; Other school shootings, 2000). Based upon an analysis of the listings, Pearl High School, Westside Middle School, Thurston High School, Heath High School, Lincoln County High School, and Parker Middle School are likely to be portrayed as NTSV cases. These findings are based on the assumption that news journalists and editors chose the cases for inclusion in the listings. It is more likely that NTSV cases will appear in the listings than TSV cases. However, it may be possible that TSV cases do appear in the listing, but it is likely that any TSV cases that appear in the listing will have some sensational component that makes them more newsworthy. According to the listings, the following schools are less likely to be

PAGE 55

48 NTSV cases, Bishop Neumann High School, Santana High School, Fort Gibson Middle School, a school in Deming, New Mexico, Heritage High School, a high school in Notus, Idaho, a school in St. Charles, Missouri, a school in Onalaska, Washington, Columbine High School, and Bethel Alaska High School. Only four of these schools have coverage in this sample. The others are not included in the sample. Therefore, conclusions cannot be drawn about whether or not these findings do convey that these cases should be less likely to be NTSV cases. However, in general, for the cases that do coincide in both the sample and the listings, there does not appear to be a strong correlation between what the listings predicted would be NTSV and what is actually provided from the data gathered in the sample. 3.2.2 Quantitative Analysis There are several key correlations between variables within this dataset. This discussion starts with the analysis of a composite variable that was first represented in Table 3-4. This variable is labeled NTSV vs. TSV (NVT) it is simply the total listed in column four of Table 3-4. The higher the score the more likely it is that the case is one of NTSV. For example, Columbine scored ten points on the NVT scale. The relationship between NVT and the number of articles written per case is strongly positive (0.724 ** ). Similarly, a moderately strong positive relationship exists between NVT and articles written within first year, unique interest for first year, median family income, percent white, number of victims killed, and quantity of weapons at the following correlations 0.691 ** 0.504 0.560 ** 0.610 ** 0.654 ** and 0.627 ** respectively. A moderately strong negative relationship exists between NVT and percent black (-0.559 ** ), and race of offender (-0.619 ). The number of articles written within the first year since the shooting occurred is positively correlated with the number of articles written about a given case in general, correlation is 0.981**. The number of articles written per case is directly correlated with the

PAGE 56

49 Table 3-3. Matrix of cases with exceptional qualities. Name 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 unknown elementary school Boston X X X albany high school X X X alva w dimmitt middle school X X X X appalachian school of law X buell elementary school X X burlington high school X X X X carter g. woodson middle school X X X X columbine high school X X X X X X X X X X destrehan high school X X fort gibson middle school X heath high school heritage high school X X X X john batram high school X X X lake clifton eastern high X X lake worth middle school X lew wallace high school X martin luther king high school X X parker middle school X X pearl high school X X X public school 18 elementary X X ridgewood high school X X X santana high school X X X X X X X X shawnee mission south high school X X sumner high school X thurston high school X X X X X X X vigor high school X westside middle school X X X X X X 1 Articles within top 25% of word count. 2 Cases in top 25% of number of articles. 3 Cases in top 25% of UI for first year articles. 4 Number of full length articles written on one case is five or more. 5 Cases in top 25% of number of offenders killed. 6 Cases in top 25% of number of offenders. 7 Cases in top 25% of number of victims. 8 Cases in top 25% of number of victims killed. 9 Cases in top 25% of that involve the use of mass-casualty producing weapons. 10 Cases in top 25% of number of weapons used. 11 Towns in top 25% of median family income. 12 Cities in top 25% of white race category. 13 Cities in top 25% of black race category. 14 Cities in top 25% of other race category. 15 Cities in top 25% of total population.

PAGE 57

50 Table 3-4. List of cases from lowest to highest affirmation of NTSV status. School Displayed Interest Factors Causing Interest Total NVT columbine high school 3 7 10 N santana high school 3 5 8 N thurston high school 2 5 7 N westside middle school 2 3 5 N burlington high school 1 3 4 N heritage high school 2 2 4 N albany high school 1 2 3 N alva w dimmitt middle school 1 2 3 N pearl high school 0 3 3 N ridgewood high school 2 1 3 N buell elementary school 0 2 2 N destrehan high school 0 2 2 N parker middle school 0 2 2 N unknown elementary school Boston 0 2 2 N appalachian school of law 0 1 1 T carter g. woodson middle school 2 -1 1 T sumner high school 0 1 1 T heath high school 0 0 0 T fort gibson middle school 0 -1 -1 T lake clifton eastern high school 0 -1 -1 T lake worth middle school 0 -1 -1 T lew wallace high school 0 -1 -1 T martin luther king high school 0 -1 -1 T public school 18 elementary school 0 -1 -1 T shawnee mission south high school 0 -1 -1 T vigor high school 0 -1 -1 T john batram high school 0 -2 -2 T Mean 1.89 Std Deviation 3.04 Median 2.00 Notes: 1. Displayed interest is the number of boxes checked in Table 3-3 columns 14 for each case. This indicates the actual level of interest that media sources have with regard to a case. 2. Factors causing interest are those factors in Table 3-3 columns 5 14. From the theory one expects that cases with columns 11 and 12 are more likely to gain special interest from the media. This column is computed in one of two ways: for those cases with no check in Table 3-3 columns 13 and 14, the sum of checks in columns 5 12 is added. For those cases with checks in Table 3-3 columns 13 or 14, checks in columns 5 12 are added. For those cases with checks in Table 3-3 columns 13 or 14 and no checks in columns 5 12, the checks in blocks 13 and 14 were subtracted. The theory behind this line of reasoning is that for cases that occur in mostly white populations and those that occur in more wealthy populations will be likely to gain the unique interest of the media. On the other hand, in communities where the population is mainly made up of minority groups, the media are less likely to show a unique interest in the case unless there is some thing sensational about the case yet still the case must allow for the maintenance of socially acceptable ideologies. 3. The T vs. NT column represents the relative status of the case: is it T or Traditional School Violence or is it NT or New Trend School Violence. This can only be determined on a relative scale.

PAGE 58

51 number of offenders that are killed in the school shooting, correlation 0.919 ** Also, the number of articles written per case increases directly as the number of victims involved in the case increases (0.806 ** ), and as the number of victims that are killed increases (0.761 ** ). The number of articles written per case increases when the race of the victim is either black or some other minority group (0.776 ). The number of articles per case also increases directly in relation with the quantity of weapons that are reportedly used or planned to be used by the offenders (0.754 ** ). The number of articles written within the first year of the shooting appears to be relatively weakly correlated with the median family income of the city where the shooting occurred (0.395 ). The number of articles written within the first year is strongly and directly correlated with the number of offenders that are killed in the school shooting (0.928 ** ). Likewise, the number of articles written within the first year increases directly as the number of victims involved in the case increases (0.745 ** ) and as the number of victims that are killed increases (0.682 ** ). Cases involving minority victims are likely to have a larger number of articles written about them within the first year after the incident (0.836 ** ). The number of articles written within the first year increases directly in relation with the quantity of weapons that are reportedly used or planned to be used by the offenders (0.712 ** ). The level of unique interest (UI) of the articles that are written within the first year appears to be correlated with the percentage of white people in the population of the city where the school shooting took place in a moderately positive direction (0.467 ). The level of unique interest (UI) of the articles that are written within the first year appears to be correlated with the percentage of black people in the population of the city where the school shooting took place in a moderately negative direction (-0.445 ). The level of UI of articles

PAGE 59

52 written within the first year is relatively strongly correlated with the race of the offenders in such a way that if the offender is white news agencies are more likely to display a greater level of UI (-0.730 ). The level of UI of articles that within the first year represents the average level of UI for all of the articles published about a given case within the first year of the occurrence. The level of UI is a broad indicator for type of school violence. A high level of UI indicates that it is likely that the case is a NTSV case. The lower the level of UI the more likely it is that the case is a TSV case. This finding indicates that cases with higher levels of UI tend to be more likely to have happened in communities where there are a higher percentage of white Americans living in the community. This finding also indicates that the higher the level of UI for a given case the smaller the percentage of black Americans in the community where the school shooting took place. The median family income is higher in those communities where a larger percentage of the population is white (0.490 ** ) and lower in those communities where a larger percentage of the population is black (-0.572 ** ). The correlation between the median family income of the community where the school shooting took place and the number of offenders killed in the shooting is 0.495 The percentage of white people in the community where the shooting took place is moderately correlated with the race of the offender in a way such that white offenders are more likely in communities with larger populations of white people (-0.668 ). The percent of white people in the population is positively and moderately correlated with the SES of the victim (0.717 ). This means that in communities where a larger percent of the population is white the victim is more likely to have a high SES. On the other hand, as the percentage

PAGE 60

53 Table 3-5. Correlations of aggregate variables. 1.000 .981 ** .057 .374 .242 -.209 -.068 -.073 -.070 .919 ** .806 ** .761 ** .470 -.352 .776 .754 ** .417 .526 .000 .794 .054 .224 .296 .735 .716 .741 .000 .000 .000 .077 .238 .024 .000 .305 .146 27 27 23 27 27 27 27 27 25 25 25 24 15 13 8 18 8 9 .981 ** 1.000 .047 .395 .198 -.180 -.026 -.048 -.055 .928 ** .745 ** .682 ** .438 -.290 .836 ** .712 ** .406 .456 .000 .832 .042 .323 .368 .896 .812 .794 .000 .000 .000 .102 .337 .010 .001 .319 .217 27 27 23 27 27 27 27 27 25 25 25 24 15 13 8 18 8 9 .057 .047 1.000 .289 .467 -.445 -.051 -.160 .143 .042 .124 .073 -.133 -.730 -.075 -.031 .636 .629 .794 .832 .182 .025 .034 .818 .466 .537 .855 .581 .746 .650 .011 .872 .906 .090 .069 23 23 23 23 23 23 23 23 21 21 22 22 14 11 7 17 8 9 .374 .395 .289 1.000 .490 ** -.572 ** .238 -.047 .343 .495 .304 .291 -.079 -.465 .665 .356 .254 .141 .054 .042 .182 .010 .002 .231 .815 .093 .012 .139 .167 .779 .109 .072 .147 .544 .718 27 27 23 27 27 27 27 27 25 25 25 24 15 13 8 18 8 9 .242 .198 .467 .490 ** 1.000 -.934 ** -.106 -.268 .203 .260 .366 .366 -.052 -.668 -.206 .408 .717 .536 .224 .323 .025 .010 .000 .598 .177 .330 .210 .072 .078 .855 .013 .625 .093 .045 .137 27 27 23 27 27 27 27 27 25 25 25 24 15 13 8 18 8 9 -.209 -.180 -.445 -.572 ** -.934 ** 1.000 -.255 .107 -.115 -.245 -.307 -.285 .025 .694 ** .182 -.338 -.612 -.510 .296 .368 .034 .002 .000 .200 .594 .585 .238 .136 .177 .928 .009 .666 .171 .107 .161 27 27 23 27 27 27 27 27 25 25 25 24 15 13 8 18 8 9 -.068 -.026 -.051 .238 -.106 -.255 1.000 .459 -.242 -.029 -.136 -.225 .082 -.143 .120 -.161 -.358 -.092 .735 .896 .818 .231 .598 .200 .016 .243 .890 .517 .291 .770 .641 .778 .522 .384 .814 27 27 23 27 27 27 27 27 25 25 25 24 15 13 8 18 8 9 -.073 -.048 -.160 -.047 -.268 .107 .459 1.000 -.129 -.080 -.129 -.167 -.037 -.094 .548 -.143 -.936 ** -.647 .716 .812 .466 .815 .177 .594 .016 .539 .705 .540 .436 .895 .759 .159 .571 .001 .059 27 27 23 27 27 27 27 27 25 25 25 24 15 13 8 18 8 9 -.070 -.055 .143 .343 .203 -.115 -.242 -.129 1.000 -.018 .031 .090 -.744 ** -.260 .300 .014 -.745 -.316 .741 .794 .537 .093 .330 .585 .243 .539 .932 .884 .684 .002 .439 .513 .959 .034 .407 25 25 21 25 25 25 25 25 25 25 24 23 14 11 7 17 8 9 .919 ** .928 ** .042 .495 .260 -.245 -.029 -.080 -.018 1.000 .805 ** .802 ** .325 -.271 .930 ** .798 ** .311 .354 .000 .000 .855 .012 .210 .238 .890 .705 .932 .000 .000 .257 .420 .002 .000 .453 .351 25 25 21 25 25 25 25 25 25 25 24 23 14 11 7 17 8 9 .806 ** .745 ** .124 .304 .366 -.307 -.136 -.129 .031 .805 ** 1.000 .942 ** .563 -.422 .630 .866 ** .458 .539 .000 .000 .581 .139 .072 .136 .517 .540 .884 .000 .000 .036 .172 .129 .000 .254 .134 25 25 22 25 25 25 25 25 24 24 25 24 14 12 7 18 8 9 .761 ** .682 ** .073 .291 .366 -.285 -.225 -.167 .090 .802 ** .942 ** 1.000 .462 -.308 .580 .903 ** .342 .388 .000 .000 .746 .167 .078 .177 .291 .436 .684 .000 .000 .096 .330 .173 .000 .408 .301 24 24 22 24 24 24 24 24 23 23 24 24 14 12 7 18 8 9 .470 .438 -.133 -.079 -.052 .025 .082 -.037 -.744 ** .325 .563 .462 1.000 -.447 -.408 .486 .548 .548 .077 .102 .650 .779 .855 .928 .770 .895 .002 .257 .036 .096 .374 .592 .078 .203 .203 15 15 14 15 15 15 15 15 14 14 14 14 15 6 4 14 7 7 -.352 -.290 -.730 -.465 -.668 .694 ** -.143 -.094 -.260 -.271 -.422 -.308 -.447 1.000 .000 .072 a -1.000 ** .238 .337 .011 .109 .013 .009 .641 .759 .439 .420 .172 .330 .374 1.000 .866 .000 13 13 11 13 13 13 13 13 11 11 12 12 6 13 8 8 5 6 .776 .836 ** -.075 .665 -.206 .182 .120 .548 .300 .930 ** .630 .580 -.408 .000 1.000 .370 a a .024 .010 .872 .072 .625 .666 .778 .159 .513 .002 .129 .173 .592 1.000 .540 . 8 8 7 8 8 8 8 8 7 7 7 7 4 8 8 5 3 4 .754 ** .712 ** -.031 .356 .408 -.338 -.161 -.143 .014 .798 ** .866 ** .903 ** .486 .072 .370 1.000 .312 .301 .000 .001 .906 .147 .093 .171 .522 .571 .959 .000 .000 .000 .078 .866 .540 .496 .469 18 18 17 18 18 18 18 18 17 17 18 18 14 8 5 18 7 8 .417 .406 .636 .254 .717 -.612 -.358 -.936 ** -.745 .311 .458 .342 .548 a a .312 1.000 .745 .305 .319 .090 .544 .045 .107 .384 .001 .034 .453 .254 .408 .203 . .496 .034 8 8 8 8 8 8 8 8 8 8 8 8 7 5 3 7 8 8 .526 .456 .629 .141 .536 -.510 -.092 -.647 -.316 .354 .539 .388 .548 -1.000 ** a .301 .745 1.000 .146 .217 .069 .718 .137 .161 .814 .059 .407 .351 .134 .301 .203 .000 .469 .034 9 9 9 9 9 9 9 9 9 9 9 9 7 6 4 8 8 9 Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N number of articles per case articles written within 1 yr unique interest for first year median family income percent white percent black other race composite total population of city number of offenders number of offs killed number of victims number of vics killed type of weapon race of offender race of victim quantity of weapons SES victim SES offender number of articles per case articles written within 1 yr unique interest for first year median family income percent white percent black other race composite total population of city number of offenders number of offs killed number of victims number of vics killed type of weapon race of offender race of victim quantity of weapons SES victim SES offender Correlation is significant at the 0.01 level (2-tailed). **. Correlation is significant at the 0.05 level (2-tailed). *. Cannot be computed because at least one of the variables is constant. a.

PAGE 61

54 Table 3-6. Table of significant correlations for aggregate variables. Variables Corr. Sign. N Variables Corr. Sign. N NTSV vs TSV (NVT)/number of articles per case .724** .000 27 unique interest 1 st yr/percent white .467* .025 23 NVT/articles written within 1 st yr .691** .000 27 unique interest 1 st yr/percent black -.445* .034 23 NVT/unique interest for 1 st yr .504* .014 23 unique interest 1 st yr/race of offender -.730* .011 11 NVT/median family income .560** .002 27 median family income/percent white .490** .010 27 NVT/percent white .610** .001 27 median family income/percent black -.572** .002 27 NVT/percent black -.559** .002 27 median family income/number of offenders killed .495* .012 25 NVT/number of offenders killed .676** .000 25 percent white/percent black -.934** .000 27 NVT/number of victims .768** .000 25 percent white/race of offender -.668* .013 13 NVT/number of victims killed .654** .001 24 percent white/SES victim .717* .045 8 NVT/race of offender -.619 .024 13 percent black/race of offender .694** .009 13 NVT/quantity of weapons .627** .005 18 percent other race/total population of city .459* .016 27 number of articles per case/articles written within 1 st yr .981** .000 27 total population of city/SES victim -.936** .001 8 number of articles per case/number of offenders killed .919** .000 25 number of offenders/type of weapon -.744** .002 14 number of articles per case/number of victims .806** .000 25 number of offenders/SES victim -.745* .034 8 number of articles per case/number of victims killed .761** .000 24 number of offenders killed/number of victims .805** .000 24 number of articles per case/race of victim .776* .024 8 number of offenders killed/number of victims killed .802** .000 23 number of articles per case/quantity of weapons .754** .000 18 number of offenders killed/race of victim .930** .002 7 articles written within 1 st yr/median family income .395* .042 27 number of offenders killed/quantity of weapons .798** .000 17 articles written within 1 st yr/number of offenders killed .928** .000 25 number of victims/number of victims killed .942* .000 24 articles written within 1 st yr/number of victims .745** .000 25 number of victims/type of weapon .563* .036 14 articles written within 1 st yr/number of victims killed .682** .000 24 number of victims/quantity of weapons .866** .000 18 articles written within 1 st yr/race of victim .836** .010 8 number of victims killed/quantity of weapons .903** .000 18 articles written within 1 st yr/quantity of weapons .712** .001 18 race of offender/SES of offender -1.0** .000 6 SES victim/SES offender .745* .034 8 of black people in the population of the community where the shooting took place increases the offender is more likely to be a minority (0.694 ** ). The smaller the population of the city where the shooting took place the lower the SES of the victim (0.936 ** ). White offenders are more likely to have higher levels of SES (-1.00 ** ). On the other hand, black offenders are more likely to have lower levels of SES. The SES of the victim is directly correlated with the SES of the offender (0.745 ). This indicates that offenders with higher SES are likely to assault victims with higher SES. Offenders with lower SES are likely to assault victims with lower SES. This also means that white

PAGE 62

55 offenders are more likely to shoot victims with higher SES and black offenders are more likely to shoot victims with lower SES. The number of offenders involved in a given case of school shooting appears to be relatively strongly correlated with the type of weapon used in a negative direction (0.744 ** ). Based upon the coding of the type of weapon variable that means that as the number of offenders involved in an offense increases it is more likely that these offenders will not use mass-casualty producing weapons. In other words, the larger the number of offenders the less violent the type of weapons used in the attack. As the number of offenders involved in an offense increases, the SES of the victim is likely to be less (0.745 ). Cases involving attacks on poor victims are likely to involved larger number of offenders. There appears to be a strong positive correlation between the number of offenders that die in an attack and the number of victims involved in the school shooting incident (0.805 ** ). Similarly, there is a strong positive correlation between the number of offenders killed and the number of victims killed in any given case (0.802 ** ). The number of offenders killed in the shooting incident is directly correlated with the number of weapons used in the shooting in a relatively strong way (0.798 ** ). The number of offenders killed is highly correlated with the race of the victim. When a larger number of offenders are killed, minority victims are likely to be involved in the case (0.903 ** ). The number of victims involved in any given case appears to be correlated in a strong positive direction with the number of weapons found to be used or planned to be used in each case (0.866 ** ). In situations where offenders seemed determined to kill a large number of people or to cause significant damage and thereby they used several

PAGE 63

56 weapons, it is more likely for these cases to involve many victims because that was the intent of the offenders. Similarly, the actual number of victims that are reported as dead increases as the number of weapons found to be associated with any given case increases (0.903 ** ). Another intuitive relation that comes from the data is that there appears to be a strong correlation between the number of victims involved in a case and the number of victims that actually die as a result of the shooting (0.942 ** ). It appears that the greater the number of victims involved in the case the more deadly the type of weapons used in the case (0.563 ). This means that when the offenders use weapons of mass-casualty producing capabilities there are more victims produced as a result of the shooting.

PAGE 64

57 CHAPTER 4 DETAILED DISCUSSION AND ANALYSIS OF SPECIFIC INCIDENTS According to Table 3-4 List of cases from lowest to highest affirmation of NTSV status, Columbine High School and Thurston High School are two of the highest ranking schools and John Batram High School and Lew Wallace High School two of the lowest ranking schools. To keep a more homogeneous group for comparison high schools will be analyzed here. A general description of these four cases follows: the number of articles written about each case is twenty two, two, one, and one corresponding to Columbine, Thurston, John Batram, and Lew Wallace, respectively. The article covering the Lew Wallace and John Batram cases are very short compared to the mean length. One of the articles describing the Thurston case is relatively long while the other is a brief article. Of the articles covering the Columbine case, there are many full length articles and editorials, and a few short descriptive articles. 4.1 Columbine High School Shooting The Columbine case is unique in that it ranks the highest on the scale provided by Table 3-4. The articles that provide information about the Columbine case describe it in such a way that Columbine appears in the top 25% of the following categories: number of words per article, number of articles per case, number of offenders killed, number of victims, number of victims killed, mass-casualty producing weapons, number of weapons, number of full length articles, median family income, and percent white population of city. These findings lead to the conclusion that it is likely that Columbine

PAGE 65

58 is a NTSV case and that the level of unique interest associated with the case is relatively high. According to the findings in this study, the actual level of unique interest for the articles written about Columbine within the first year of the occurrence was 2.2 which is around the mean UI for all the cases. One would not expect this given that Columbine is the ideal model for NTSV cases. However, given the apparent inadequacy of using the level of UI alone as an indicator of whether a case is NTSV or TSV, this finding is understandable. A summary of the findings from the twenty two articles is presented here. The article coverage is broken down into the following article types: ten of the twenty two articles are full length articles, six of them are editorials, and six of them fall into the other category which includes an article on poll results, SWAT team tactics, criminal investigation of the case, response of Catholic Church, and the television and movie industries response to Columbine. From the findings in this study, none of the TSV cases have elicited such a response. This is just one of the multiple ways in which NTSV cases are treated differently than TSV cases. One reason for the overall relatively low score on UI for Columbine is that there are so many articles related to this case. Of the twenty two articles covering the Columbine case, the full length articles contribute an average of 3.6 on the UI scale. The editorials have an average UI of 1.2. The various other types of articles have an average of 1.5 UI. The overall UI of 2.2 is significantly affected by the contribution from the editorials and other articles which do not necessarily provide adequate coverage of the Columbine case.

PAGE 66

59 The group of articles that falls into the other category provides little insight into the Columbine case other than the fact that these articles exist. By this, it is meant that with most other cases there were not a lot of articles written in response to the shooting. Columbine provided stimulus for a plethora of articles to be written in response to the shooting. Many of the articles that are written about other cases refer to Columbine at some point in the text. Columbine was a unique and powerful occurrence that shaped the way that society reacts to school shootings. Several television companies and a movie productions agency rescheduled certain movies and television programs so that they would not appear insensitive to the victims and survivors of the Columbine school shooting. Pope John Paul II ., made a speech in response to the Columbine tragedy letting the World know that his prayers and the prayers of the Catholic Church would be with those who were harmed by the shooting. A SWAT team video, which was produced before the Columbine shooting, to help SWAT teams and other law enforcement agencies work together to deal with school shootings was in high demand after the Columbine attack. Several polls were given by news and media agencies in response to the Columbine shooting in order to determine what Americas youths and parents thought about the attack and how to keep it from occurring again. Lastly, an article briefly provides information that authorities are looking into the link between two weapons that were used in the shooting and purchased at a local gun show in Colorado. Of all of these articles, only the last one would represent anything close to the type of coverage that one would expect to find in a TSV case. 1 1 By this I mean that TSV cases are not likely to be under special investigation by SWAT teams, nor are they likely to receive special address from the Pope, nor are news agencies likely to conduct various polls in response to TSV shootings.

PAGE 67

60 The six editorials can be broken down into three collections of two editorials each. The first two editorialists provide their opinion about how to solve school violence. There are no fundamentally moving findings provided by these two editorials. The second set of two editorials present the views of numerous authors about the Columbine school shooting in response to other editorials written by other authors. One set of responses mainly discusses the impact of gun control on America and whether or not gun control should be the focus of discussion about school shootings and school violence in general. Another editorial gives a summary of the general theoretical ideas behind this study and will be paraphrased here: people have become somewhat hardened to urban school violence; it has been thought of as being related to gangs and drugs. Recently, new shootings aimed at killing many people have occurred in rural and suburban communities. The offenders have been mainly middle class kids with no apparent reason for their rage. It is believed that the reason lies within the American culture (Looking for answers in wake of another school shooting, 1999). In one editorial, the author responses to an editorial that suggests, as this study does, that there are corollary statements which follow from the statements made in the aftermath of Columbine. Statements like this should not happen here lead to a conclusion that this violent behavior is expected in other locations. The author of the responding editorial does not think that these statements have anything to do with race overtly or covertly. He believes that coverage provided by news organizations is based solely on circumstance and newsworthiness of the case under scrutiny ( Rickerfor, 1999, p. B6). However, the author of the responding editorial fails to convincingly show that race does not affect circumstance and what journalists and editors consider to be newsworthy.

PAGE 68

61 Ishmael Reed made an argument based upon a 1994 statistic that found that school children in small U.S. cities, suburbs and rural communities were twice as likely to carry weapons as students in large cities (1999, p. 5C). Mr. Reeds argument based upon this statistic is that Americans should not be surprised to see shootings such as Columbine and other NTSV shootings. The findings from the 1994 report indicate that these shootings are a very real possibility, and yet white Americans still act as though NTSV cases are shocking. The argument being presented by Mr. Reed is a fairly strong case for gun control laws. While, at the same time, he points out that the critical component of racism is still a problem in America which permeates every aspect of our daily lives and yet is not discussed by most media sources. The editorial offers the idea that there is a general tendency to blame the nations youth violence problems on minority communities in inner cities, while perhaps it may be more accurate to take a look in the suburbs and rural communities in America to determine where the guns are coming from and where the desire to use them is coming from. The general tendency of these full length articles is to describe, in detail, the events that happened at Columbine and the response of the survivors, parents, and officials after the incident. Some particular issues that arise in the set of full length articles are: the Jefferson County Sheriffs office performed an investigation of the Columbine case in an attempt to prepare instructional material so that other law enforcement and rescue service personnel will know what to do in the event that a similar incident occurs ( Luzadder 1999). Another article describes the increased incidence of parents taking their children to the psychiatrist to be evaluated given the scare that has occurred in the aftermath of Columbine. Prior to the shooting, parents may have written-

PAGE 69

62 off certain behavior as being part of growing up, but now it is too risky for parents to simply allow certain behaviors to continue without further evaluation to determine the cause of the behavior (Jefferson, 1999). A third article describes how the reluctant experts at Columbine are helping other communities, in particular the community in Santee California with its school shooting. The article describes the information and advice that counselors provided for the officials at Santana High School. The article also describes the changes that occurred at Columbine as a result of the shooting: Columbine doubled the number of campus security guards to four, installed 16 surveillance cameras, instituted the use of identity cards, and equipped doors with key cards. . Two mental health counselors were added to the staff, bringing the school total to eight. A safe room was established, where distraught teachers or students could retreat for privacy and quiet. (Carter, 2001, p. A12) The very simple and yet utterly obvious conclusion from this article is that it is written as though Columbine is the epitome of school shootings. As though Columbine is the only school shooting and the experts that have dealt with the Columbine case can deal with any other case. There is an overtone in the articles that although NTSV is the only type of violence and the means by which officials deal with this type of case is the only real means of dealing with any case of school violence. It also appears from the Luzadder (1999) article that NTSV is the only type of violence that exists. There is a tendency to ignore the TSV cases that happen in America unless there is something that generates a unique interest in the case. It is then considered newsworthy and worthy of being published in non-local newspapers for a larger audience to read. One article from The Baltimore Sun describes the offenders as part of a gang that called itself the Trenchcoat Mafia. This article provides information about the circumstances that surrounded the case. One interesting factor that came out of this report was that the two offenders in this case had been convicted last year for first-

PAGE 70

63 degree criminal trespassing after they broke into a car. They completed a county program for troublemakers in February (Morgan, 1999, p. A1). The author asks: If these children had not been white middle class youths, how would their treatment had been different? Would they have been able to continue on from that conviction and plan and prepare for the mass murder of other students and a teacher? The offenders started planning for the shooting a year or more in advance. How would things have been different had they been working class youths attempting to plan this shooting after having been convicted of a serious offense? A parent of one of the surviving students stated that he was shocked after seeing the house that one of the offenders lived in. The parent asked, Who would have thought that a kid who lived in a home like that and had that kind of opportunity could do something like this? (Morgan, 1999, p. A1). This is a sentiment that has been expressed many times and in many ways in this case. There may be some truth to this sentiment but, at the same time, it still seems to imply that this behavior is expected of those who do not live in expensive houses and nice neighborhoods. Another article by OBrien, Gorov, and Zuckoff (1999) describes that the offenders had been planning the attack for at least a year and that they planned the attacks to coincide with Adolf Hitlers birth date. An interesting and important fact that distinguishes NTSV from TSV is that the offenders wanted to create as much harm and damage as possible. The police recovered more than 30 homemade explosives from the school, as well as two sawed off shotguns, a semiautomatic rifle, and a semiautomatic handgun, leading them to believe that others had to have been involved (OBrien et al., 1999, p. A1). This is not typical in TSV cases. The TSV offenders usually want to harm

PAGE 71

64 or kill the person that has lead them to perform the act of violence. The TSV offender does not want to kill as many people as possible and cause as much damage as possible. One theme that ran in the following two articles is the idea that the offenders were bullied and this is why they lashed out at others ( Yettick, 2002; Kenworthy, 1999). The Yettick (2002) article also describes the idea that the offenders and the Trenchcoat Mafia bullied other students as well. However, many of those who were killed by the offenders were football players responsible for bullying and picking on the offenders and the Trenchcoat Mafia. Frank DeAngelis, the principal of Columbine at the time of the shooting, said of the survivors, they were deprived of so much of their youth. They saw so much death. It scares me. I hope it made them stronger ( Yettick, 2002, p. 1B). While this statement is insightful, it did not find its way into any of the articles related to TSV that appear in this sample. It seems to be very difficult for white American journalists to portray TSV in a way that shows this much compassion and understanding, this much sympathy and respect for the victims and survivors of TSV. Certainly, those who experience TSV live in communities and experience situations which would allow them to be deserving of such a statement. 2 Additional curiosities that are described in the article by Kenworthy (1999) include the fact that President Clinton saw fit to address the Columbine incident in one of his speeches. It is unlikely that the president would address the country about school violence if Columbine had been a case of TSV. However, it is understandable that the, now former, president would make some kind of statement given that Columbine was one of the most violence cases of NTSV and TSV ever. Another 2 For examples of youths in America that have experienced violence and atrocities, see Code of the Street by Elijah Anderson. Anderson (1999) provides an explicit and detailed account of the conflicts that youths in inner city America face on a daily basis.

PAGE 72

65 interesting fact is the number of law enforcement officials that responded to the Columbine incident in total: Hundreds of heavily armed police officers, FBI and Federal Bureau of Alcohol, Tobacco and Firearms agents arrived on the scene, accompanied by dozens of ambulances and other rescue vehicles. . [ and] an armored personnel carrier. . [ were] dispatched to help remove the wounded. ( Kenworthy, 1999, p. A1) Again, it is highly unlikely that a TSV case would receive such resources. Not one TSV case in this study received such resources. At most there would likely be a squad of police officers and an ambulance or two. There would not likely be so many law enforcement resources utilized at one scene of violence in a TSV case. 4.2 Thurston High School Shooting There are two articles that describe the Thurston High School case. There are other cases that are ranked in the top 25% according to Table 3-4 and that have a larger number of articles per case that appear in the sample used for this study, but there are few that are ranked as high as the Thurston case. 3 The Thurston case ranks in the 25% percent of word count, unique interest, number of victims, number of victims killed, mass-casualty producing weapons, number of weapons, and percent white population of the city. The sum effect of these findings is that Thurston should be more likely to receive a higher degree of unique interest and should likely be labeled as a NTSV case. The first article that describes the Thurston High School shooting is entitled School shooting inspires one-act play that allows teens to explore violence (1999, p. A2). It describes a play that was written by an Oregon resident who decided to write it after watching the news about the Thurston High School shooting. The article describes 3 The fact that there are a certain number of articles, in this particular sample, on a given case does not mean that there are no other articles included in the population of articles that may be written about the case.

PAGE 73

66 some basic facts about the case such as it occurred on May 20 th 1998 in Springfield, Oregon. A fifteen year old white male student shot and killed both his mother and father and then went to school and shot and killed two classmates wounding twenty two others. It is almost the ideal description of a NTSV case: the offender is usually a white male who is of a middle class or higher socioeconomic status, the target of the violence tends to be not only peers but also authority figures, violence is likely related to feelings of revenge or retribution, types of weapons used are mass-casualty producing and tend to be excessively violent, number of targets are usually high, and violence seems to be planned, pre-meditated, and organized. This article scored a 4.0 on the UI scale. The article describes the opening scene of the play in which the offender is in his jail cell and is confronted by the ghosts of his victims. These ghosts ask the offender whyd you kill me? (School shooting inspires, 1999, p. A2). Eventually, in the play, the offender states why he killed the victims. He did it because he was wracked by the pain of his classmates laughing at him (School shooting inspires, 1999, p. A2). The playwright wrote the play in hopes that it would cause young people to face the problems associated with the culture of violence in America. It is interesting that there are no such plays written about TSV cases that appear in this sample. It seems that there is a need for such tools to combat violence that takes place in inner city and poor rural communities throughout America as well. The second article that describes the Thurston High School shooting case is a more detailed description of the case. The main theme of this article is the offenders mental status. His lawyers originally planned to have the offender plead not guilty by reason of insanity, but ultimately, the offender decided on his own that he would plead guilty to

PAGE 74

67 four counts of murder e.g., his parents and two students. This article indicates a relatively low score of 2.0 on the UI scale, but this is a result of the inability of this scale to provide a truly adequate representation of the level of unique interest that the author appears to be portraying in the case. However, after reading the article, one does realize how the portrayal of the offender and the victims in this case differs greatly from the portrayal of the victims and offenders in the TSV cases that follow. 4.3 John Batram High School Shooting One brief caption explains the situation at one school shooting that happened at John Batram High School in Philadelphia, Pennsylvania. The caption appears on page 12A in the front page section of The Boston Globe on October 5 th 1999. This caption had a total score of zero on the variable of UI. The caption simply describes the photograph which is of an ambulance that came to the school as a result of the shooting. Apparently, the assistant principal of the high school was shot in the leg when he confronted the seventeen year old offender about possessing the handgun. This case is described here because it ranks very low on the matrix in Table 3-4. The case does not appear in the top 25% on any of the key variables that would indicate that the case is likely to be classified as a NTSV case. The only variables that this case appears in the top 25% of is percent black, percent other race, and total population size. Philadelphia is large and has a high percent of black and other non-white racial groups when compared to the other cities in the sample. 4.4 Lew Wallace High School Shooting There is one article in the sample that provides coverage of the Lew Wallace High School shooting. This article is an editorial written in response to an article that appeared in the front page section of a newspaper on page 18A. The author of the editorial writes

PAGE 75

68 of the difference in the coverage of the shooting that happened at Lew Wallace High School and other recent shootings. The author states that there is no front-page story, no analysis of how a student was pushed to the edge by dire circumstances, no how could this happen here commentaries (Dixon, 2001, p. B8). The author attributes this lack of interest in the case to the fact that the offender and victims involved in the case are not white. The implication is that they are black American students that attended an inner city school in Gary, Indiana and this is why there is differential coverage of this shooting. On the scale of UI, this editorial received a score of zero. The Lew Wallace case appears to be off of the scale according to its ranking in Table 3-4. It does not appear in the top 25% of any of the variables except for the percent black population in the city. This indicates that the case should a TSV case as opposed to a NTSV case and there should be little unique interest in this case. Indeed, the evidence does seem to confirm the theoretical assumptions about the differences between the two types of school violence. The author of the editorial describes a case which is classified as TSV in this project. It gives support to the ideas generated from the theory that TSV is violence in which the offender may be white or of another racial category and is of a low socioeconomic status, the target of the violence is generally other students, violence may be related to gang or drug activities, types of weapons used vary from assault with hands to assault with weapons other than mass-casualty producing guns, the number of targeted persons are usually low, and violence seems to happen in more of a spontaneous nature although there may be planned acts of violence in this cases of TSV.

PAGE 76

69 CHAPTER 5 CONCLUSIONS The goal of this study was to investigate the way U.S. newspapers portrayed domestic school shootings from 1 st May 1997 to 30 th April 2002. Using Lexis Nexis online search utility, major newspapers in the U.S. were analyzed for content involving school shootings. A total of 100 articles were selected at random from a sampling frame of 578 articles within the specified time period. The following non-aggregate variables were coded in this analysis: title, case name, number of words, date of occurrence, date of publication, section of the newspaper that the article was published in, time difference between date of publication and date of occurrence, race of offender and victim, socioeconomic status of offender and victim, number of offenders and victims, number of offenders and victims killed, type of weapons used, number of weapons used, relation of victim to offender, and level of unique interest. Each articles qualitative characteristics were also described. Once the non-aggregate data were collected several of these variables were used to code for aggregate variables related to each school shooting in this study: NVT, case name, number of articles per case, number of articles written within the first year after the occurrence, socioeconomic status of offenders and victims, quantity of weapons used, type of weapons used, race of victims and offenders, number of offenders and victims, number of offenders and victims killed, level of unique interest controlling for time, racial makeup of the city where shooting occurred, median family income of the city where shooting occurred, and total population for the city where the shooting

PAGE 77

70 occurred. 1 The overall focus of this analysis is to determine the nature of news coverage and portrayal of school shootings. 5.1 Overview of Theories Utilized in this Analysis The perspectives used to address the news portrayal of school shootings have been the conflict and structural functional paradigms, with special interest in conflict-labeling criminology, power-threat hypothesis, media criminology, and critical-media perspectives. These perspectives view the topic of school shooting as one of a power struggle between people with lower levels of power and those with higher levels of power. News agencies in America are in a situation where they have the power to provide coverage of school shootings. They provide coverage of each case in a way that will perpetuate social ideals about those parties involved in each shooting. A key motivating factor for news agencies to continue to perpetuate socially acceptable beliefs about different cases is that the agencies rely on the business of their customers and the investments of many individual investors and advertising establishments. There are also political influences on the type of coverage that news agencies provide ( Paletz & Entman, 1981). The critical point to draw from the theory is that there are expected to be differences in the coverage of school shootings based upon the level of power that the various parties involved in an incident possess. 5.2 Overview of Important Findings This section will start with a summary of the findings in the relationship that exists between the chance that a case is NTSV and other key variables in the analysis. As the number of articles written per case, articles written within the first year, unique interest 1 The 2000 U.S. Census American FactFinder was used to find the racial makeup of the city where the shooting took place and the median family income of the city where the occurred.

PAGE 78

71 within the first year, percent white, median family income, number of offenders killed, number of victims involved and killed, and the quantity of weapons involved in a case increase there is also a greater likelihood that the case is one involving NTSV offenders. Cases of NTSV are less likely to happen in communities where a large percent of the population is black. The NTSV offenders are likely to be white and from the qualitative data more likely to be male. Now consider the major results involving the level of sensational characteristics about a case. As the number of offenders killed, the number of victims involved, the number of victims killed, and the quantity of weapons used in the case increases the number of articles covering a case increases. This seems to be due to the fact that when those factors are high the case is seen as being more newsworthy i.e., more sensational and dramatic. A finding which appears to be counter to the theoretically expected result is that number of articles written increased when the victims were likely to be minorities. This finding, and many others, may be different if one were to control for other factors such as number of offenders involved and killed, number of victims killed, number of weapons used, type of weapons used, and median family income. 2 A similar finding is that the number of articles written within the first year after the shooting increases when median family income, number of offenders killed, number of victims involved and killed, and quantity of weapons used increases. Another relatively unexpected finding was that the number of articles written within the first year increases when the victims involved are minorities. Again it would be interesting to determine if 2 One must remember that even though SPSS awards a certain level of significance to these relationships the relationships may warrant more or less strength based upon the actual number of cases involved in the relationship. However, these findings do give some idea of the relationships that exists in the population of articles that offer coverage of school shootings.

PAGE 79

72 this finding changed when holding constant several other key variables. Theory suggests that cases involving minority offenders are more likely when those cases allow journalists to perpetuate current socially acceptable ways of thinking about minority members ( Howitt, 1998; Barak, 1994; McQuail, 1992; Paletz & Entman, 1981). As the percentage of white people in the population increases, news agencies are more likely to provide coverage within the first year of an incident with a higher level of UI. Similarly, when the offender is white, the level of UI provided by articles written within the first year after the shooting is likely to be high. On the other hand, populations with larger percent of black people are less likely to receive coverage with a high level of UI. These findings do not deviate from what the theory would lead one to wait for. 3 The median family income is larger in communities where a large percent of the population is white. The median family income is smaller in communities where a large percent of the population is black. The median family income in communities where a large number of offenders were killed is likely to be higher. These findings can be expected based upon the literature. The percent of white people in the population is correlated with the race of the offender and the SES of the victim in such a way that as the percent of white people in the population increases the offender is more likely to be white and the victim is more likely to be of higher SES. Similarly, white offenders are more likely to have higher SES. Offenders are likely to attack those in a similar SES category. Therefore, white 3 The percent white and percent black figures were found using the Census American FactFinder utility and were found for each case in the sample. The previous findings related to the race of the offenders or victims were found using a world wide web search and the races of all offenders and victims were not founds for many of the cases. It is important to notice beside the impact this has on the validity of the finding that there is limited information available on any given case

PAGE 80

73 offenders are more likely to assault white victims and black offenders are more likely to assault black victims. In a similar way, populations with a large percent of black people are likely to have black offenders. Black offenders are more likely to have lower SES. Small cities are more likely to have victims with lower SES. Cases involving a larger number of offenders are likely to involve the use of less deadly weapons. When a large number of offenders are involved in a case the victims are likely to be minorities and have lower SES. The number of offenders killed in a shooting is directly related to the number of victims involved and killed, and the number of weapons used. The number of victims killed in a case is directly correlated with the number of victims involved in a case, which is also correlated with the number of weapons and the type of weapons used in a shooting. The larger the number of victims and the number of victims killed the more weapons that are involved and the more damage those weapons are capable of creating. A pattern is developing from these results. News agencies are more likely to provide coverage within the first year after the occurrence and beyond for cases that involve sensational behavior. Sensational behavior may be classified as behavior that is not expected within the norms provided by a society. In other words, cases that involve a large number of offenders and victims, large number of deaths, white offenders of middle and upper middle SES killing other white people in middle and upper middle SES communities with large percentage of the population being white is considered to be sensation. which usually varies with the likelihood that a case is NTSV or TSV. Put another way, if the case is NTSV, there is more information available on the case from various sources on the internet.

PAGE 81

74 The three questions posed at the beginning of this project are the following: Is there a difference in the type of coverage provided for the two types of school violence? If so, how does the coverage differ? Finally, why is there a difference in the coverage provided? For the first question, yes there does appear to be a difference in the type of coverage that news sources provide for the two types of school shootings. Table 3-4 shows a clear division in the coverage provided on the twenty seven cases included in this sample. This sample has been taken at random from a larger sampling frame and is therefore believed to be generalizable to the larger population of articles written on school shootings in the U.S. during the time frame of the study. Secondly, the coverage is different in several respects. Many of the results indicate that NTSV cases were given a great deal of attention by news agencies. For example, every NTSV case was in the top 25% for at least one of the following variables: word count, number of articles written overall and within the first year, and number of full length articles written. 4 At the same time, one out of thirteen of the TSV cases were in the top 25% of the variables listed directly above. The implication of this is that there exists a divide in the coverage provided for the two types of school shootings. Incidents of NTSV received most of the attention of news agencies during the time period studied. Finally, the key question that still remains is: Why is there such a divide in the coverage of the two types of school shootings? The answer seems to be found in the division that exists between factors causing interest presented in Table 3-3 and tallied in Table 3-4. Most of the NTSV cases within the top 25% of variables are likely to be seen by news producers as being newsworthy and sensational. The cases that provide media

PAGE 82

75 producers with a source of behavior that is not to be expected in the cases involving white offenders and behavior that is an exemplary instance of activity that is to be expected of minority offenders will be considered to be newsworthy. From this perspective most of the offenders of NTSV cases will be white, middle class or upper middle class, and the targets of their attack will be white, middle and upper SES. Likewise from this perspective most TSV offenders are likely to be either minority or white, of lower SES, and targeting others of lower SES who may be white or minority. The theoretical explanation for the differential coverage of school violence appears to be supported by this analysis. It is the level of sensational behavior with the combination of expected behavior that seems to be the determining factor behind the type of coverage any particular case will receive. 5.3 Critique and Suggestions for Future Research of Differential Newspaper Coverage Future research may include a dataset that will provide more information for a higher level of analysis. A rigorous analysis of more sophisticated data would better suit the topic. For example, one could analyze the unexpected findings further to determine whether they disappear after controlling for other factors as suggested above. In a similar fashion, one could analyze the many statistically significant findings of this project after controlling for the many variables to determine if any of these relations disappear. Additionally, future researchers may search for cases in a different way than this project has. A suggestion would be to compile a list of cases by school name and location and from this list pick a random sample of n cases. After compiling a list of cases to be included in the sample, search for articles on each case instead of picking a 4 See Table 3-3 Matrix of cases with exceptional qualities.

PAGE 83

76 random sample of articles which was the technique used in this study. This would provide a larger n when running various statistical analyses allowing for regression techniques to be used assuming that the case is the intended unit of analysis. Another area of potential research would be to further distinguish the affects of both sensational components of a case and the expected normal behavior of the case. They are closely linked, but from the theory one would expect that even though a TSV case is sensation, it may not receive coverage if it does not allow news producers to present it in a way that will further promote socially acceptable beliefs about the offenders and victims of the TSV case. Here the main limitation of this study becomes painfully clear. The data are not detailed enough to allow for more rigorous quantitative analysis e.g., linear regression analysis. Another critical relationship that must be analyzed further involves two key variables: race and SES. These variables are so intertwined in this and many research projects that it is difficult to separate the effects of each. In this study, the following appear to be the case: NTSV offenders are likely to be white males of middle to upper SES. While, TSV offenders seem to be either white or of a minority racial group but definitely seem to be of lower SES. There exists a basic idea that NTSV offenders would be mainly white and TSV offenders mainly black, but this would exclude a significant number of TSV offenders that are white and non-black minorities. The dynamics between sensational and expected normal behaviors becomes even more difficult to make sense of as one takes a closer look at race and SES. Some questions to guide research on this topic may include: Does differential coverage exist in cases that involve white TSV offenders and non-white TSV offenders? These offenders, by definition, would be of

PAGE 84

77 lower SES so that is one less variable to consider. Race becomes the prominent determinant here. The other key variable to consider besides race would be the expected normal behavior. A fundamental question that would arise is how is the expected normal behavior for white TSV offenders and non-white TSV offenders different, and how do those differences deviate from the expected normal behavior of white NTSV offenders? This study did not consider this topic in that much depth. This analysis is an initial attempt to determine the nature of news coverage with regard to school shootings in America. Depending upon future researchers intent, they may want to continue analysis of this topic with a critical eye on the media portrayal of school shootings in the U.S. Future researchers may want to analyze the racial components of media coverage more thoroughly. There are many different directions for future study of this topic. A big challenge for any future research is a general lack of data on the topic of media coverage of school shootings in America. Compiling the correct data for the correct purpose is especially difficult and time consuming but is the most critical task that any research faces.

PAGE 85

78 LIST OF REFERENCES Anderson, E. (1999). Code of the Street: Decency, Violence, and the Moral Life of the Inner City New York, NY: W. W. Norton and Company. Barak G. (Ed.). (1994). Media, Process, and the Social Construction of Crime: Studies in Newsmaking Criminology New York, NY: Garland Publishing, Inc. Barlow M. H., Barlow D. E., & Chiricos T. G. (1995). Economic Conditions and Ideologies of Crime in the Media: A Content Analysis of Crime News. Crime and Delinquency, 4 (1), 3-19. Blalock H. M. (1967). Toward a Theory of Minority-Group Relations New York, NY: John Wiley & Sons, Inc. Carter, J. (2001, March 7). Santee school shooting; Littletons reluctant experts try to help; Aid: officials who went through the columbine horror tell Santee what to expect. One warns that the trauma has just begun. Los Angeles Times p. A12. Dixon, D. (2001, April 3). Another school shooting. Los Angeles Times p. B8. Howitt D. (1998). Crime, the Media and the Law New York, NY: John Wiley & Sons. Jefferson, R. S. (1999, May 7). School shootings put parents on alert, therapist says. St. Louis Post-Dispatch p. 4. Kenworthy, T. (1999, April 21). Up to 25 Die in Colorado School Shooting; Two student gunmen are found dead. The Washington Post p. A1. List of school shootings. (1998, May 22). The Buffalo News p. 4A. Looking for answers in wake of another school shooting. (1999, April 22). The Baltimore Sun p. 19A. Luzadder D. (1999, July 11). Country will learn from Columbine Jeffco investigators create textbook on school shootings so other communities can prepare for such incidents. Denver Rocky Mountain News p. 5A. McQuail D. (1992). Media Performance: Mass Communication and the Public Interest London, UK: Sage Publications. Morgan, J. (1999, April 22). Dressed in black, fans of Hitler; Students were aloof, gave Nazi salutes, glorified death; Colorado school shooting. The Baltimore Sun p. 1A.

PAGE 86

79 OBrien, E., Gorov, L., & Zuckoff, M. (1999, April 25). School killings plotted in diary; Conspirators timed massacre to occur on Hitler birth date; Colorado school shootings. The Boston Globe p. A1. Other school shootings. (2000, March 1). The Seattle Times p. A3. Paletz D. L., & Entman R. M. (1981). Media, Power, Politics New York, NY: The Free Press. Quinney R. (1970). The Social Reality of Crime Boston, MA: Little, Brown and Company. Reed, I. (1999, May 30). The pundits misfire on school shootings; Causes: The media offer uninformed and often racist reasons for violence by U.S. students, a critic charges. The Baltimore Sun p. 5C. Rickerfor R. (1999, May 16). School shooting story has no racial undertones. The Times-Picayune (New Orleans) p. B6. School shootings. (1999, April 21). St. Petersburg Times p. 6A. School shootings. (1999, October 5). The Boston Globe p. 12A. School shooting inspires one-act play that allows teens to explore violence. (1999, February 9). St. Louis Post-Dispatch p. A2. Tolnay S. E., Beck, E. M., & Massey J. L. (1989). Black Lynchings: The Power Threat Hypothesis Revisited. Social Forces, 67 (3), 605-623. Turk, A. T. (1964). Prospects for Theories of Criminal Behavior. The Journal of Criminal Law, Criminology and Political Science, 55 December 1964, 454-461. Yettick, H. (2002, April 20). Echoes of tragedy; Effects of 1999 school shooting are shrouded yet unmistakable. Rocky Mountain News p. 1B.

PAGE 87

80 BIOGRAPHICAL SKETCH John Robert Bennett received his Bachelor of Arts degree from the University of Florida, Gainesville, Florida, in December of 2001 with a major in sociology. He has been enrolled in both the undergraduate and graduate sociology programs at U.F. since January, 2001, pursuing a Master of Arts degree in sociology. After receiving his M. A. degree in sociology, the author plans to continue at the University of Florida in pursuit of his Master of Science degree in business administration with a concentration in management. His main areas of interests are media, race, and criminology.


Permanent Link: http://ufdc.ufl.edu/UFE0000642/00001

Material Information

Title: Portrayal of school shootings by American newspapers
Physical Description: Mixed Material
Creator: Bennett, John Robert ( Author, Primary )
Publication Date: 2003
Copyright Date: 2003

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000642:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000642/00001

Material Information

Title: Portrayal of school shootings by American newspapers
Physical Description: Mixed Material
Creator: Bennett, John Robert ( Author, Primary )
Publication Date: 2003
Copyright Date: 2003

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000642:00001


This item has the following downloads:


Full Text












PORTRAYAL OF SCHOOL SHOOTINGS BY AMERICAN NEWSPAPERS


By

JOHN ROBERT BENNETT


















A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS

UNIVERSITY OF FLORIDA


2003















ACKNOWLEDGMENTS

This document is dedicated to my wife, Tosha L. Bennett. Without her support, I

would not have been able to complete this study. Special thanks go to Charles F.

Gattone, Ph.D. and Joe R. Feagin, Ph.D.; without their guidance, support, and providing a

committee for this thesis project none of this would be possible. I thank Terry L. Mills,

Ph.D., and Karen F. Parker, Ph.D., for their initial advice on how to go about developing

my ideas into a working project which ultimately became my thesis.
















TABLE OF CONTENTS
Page

A C K N O W L E D G M E N T S ......................................................................... .....................ii

LIST OF TABLES .................................. .. ... .... ................. .v

A B STR A C T................................................... vi

INTRODUCTION .......................... ........ .. ... .... ........ ...............

1.1 Statem ent of the R research Problem ..................................................................
1.2 B background ...................................................... 2
1.3 K ey Term inology and Concepts ........................................ ......... ............... 3
1.4 Theoretical Perspectives ............................................................................. 5
1.4.1 Conflict-Labeling Crim inology.................................................................5
1.4.2 Pow er-Threat H ypothesis ........................................ ......... ............... 7
1.4.3 M edia C rim inology............................................................ ............... 9
1.4.4 Critical-M edia Perspective ........................................ ...... ............... 12
1.4.5 Gaps in the Literature ................................................... 16

METHODOLOGY .................. ................................... ........... ............... 19

2 .1 Su m m ary of F in ding s................................................................. .....................2 0
2.2 Discussion of Data Characteristics .............. ........... .................................. 22

FIN D IN G S AN D AN ALY SIS.....................................................................................24

3.1 R research Findings .................. .................. .......................................24
3.1.1 Q ualitative .................................................. ...................... .... 24
3.1.2 Q uantitative ................................ ............. ........ .............. 27
3.1.2.1 N on-aggregate D ataset................................................................... 28
3.1.2.2 A ggregate D ataset...................................................... ............ 34
3.2 A analysis of Findings ........................... .... ................ ... .... .. ........... 4 1
3.2.1 Q ualitative A nalysis......................................................... ............... 4 1
3.2.2 Q uantitative A nalysis........................................... ........................... 48

DETAILED DISCUSSION AND ANALYSIS OF SPECIFIC INCIDENTS .................57

4.1 C olum bine H igh School Shooting .......................................................................57
4.2 Thurston H igh School Shooting................................................. ....... ........ 65
4.3 John Batram High School Shooting................................... ....................... 67

iii









4.4 Lew W allace High School Shooting................................... ...................... 67

CONCLUSIONS.................. ..... .. ....... .... ....................69

5.1 Overview of Theories Utilized in this Analysis........................................70
5.2 Overview of Important Findings........................ .......................70
5.3 Critique and Suggestions for Future Research of Differential Newspaper
C ov erag e ...................................... ................................................ 7 5

L IST O F R E FE R E N C E S ......................................................................... ....................78

B IO G R A PH ICA L SK ETCH ............................................................................ ........... 80









































iv
















LIST OF TABLES

Table page

2-1. Number of articles written per case. ........................................ ....... ............... 21

3-1. Basic figures for non-aggregate variables. .................................... .................28

3-2. Basic figures for aggregate variables................................... ......................... 35

3-3. M atrix of cases with exceptional qualities..................................... ............... 49

3-4. List of cases from lowest to highest affirmation of NTSV status.............................50

3-5. Correlations of aggregate variables. .............................................. ............... 53

3-6. Table of significant correlations for aggregate variables...............................54















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts

PORTRAYAL OF SCHOOL SHOOTINGS BY AMERICAN NEWSPAPERS

By

John Robert Bennett

May 2003

Chair: Joe R. Feagin
Cochair: Charles F. Gattone
Major Department: Sociology

The purpose of this analysis is to investigate the way American newspapers portray

domestic school shootings. To this date, there have been no investigations of media

portrayal of school shootings in America. School shootings make up a small percentage

of school violence in America; however, the damage inflicted on the victims of these

crimes is devastating. This study is concerned with the coverage and portrayal of these

shootings by the news publishers in America.

There are no current data sources listing all of the variables of interest to this study.

I relied on the Lexis Nexis database of newspapers to find a sample of articles that

address this issue. The time frame for the search has been limited to include only articles

which were published from 1st May 1997 through 30th April 2002.

These cases were analyzed using the structural functional and conflict paradigms

with a specific interest in the conflict-labeling criminology, power-threat hypothesis,

media criminology, and critical-media perspectives.









In conclusion, two key factors seem to explain the relationship between the

variables used in this analysis to determine if there is differential coverage of NTSV and

TSV. The two key factors are the level of sensational behavior exhibited in a given case

and whether or not the behavior exhibited was considered to be expected based upon

socially acceptable norms of American society.














CHAPTER 1
INTRODUCTION

1.1 Statement of the Research Problem

The purpose of this analysis is to investigate the way American newspapers portray

domestic school shootings. To this date, there have been no formal sociological

investigations of media portrayal of school shootings in America. School shootings make

up a small percentage of school violence in the United State; however, the damage

inflicted on the victims of these crimes is devastating. This analysis is not necessarily

concerned with the effects of each school shooting. It is more concerned with the

coverage and portrayal of these shootings by the news publishers in the U.S.

There are no current data sources listing all of the variables of interest to this study.

Therefore, I relied on the Lexis Nexis database of newspapers to find a sample of articles

that address this issue. The time frame for the search has been limited to include only

articles which were published from 1st May 1997 through 30th April 2002. I gathered and

analyzed articles that span a five year period of time. During this period, 578 articles

were published in U.S. newspapers on school shooting. Of these 578 articles, I selected

one hundred at random for inclusion in the sample and coded each for the following

variables: title, case name, number of words, date of occurrence, date of publication,

section of the newspaper in which the article was published, time difference between date

of publication and date of occurrence, race of offender and victim, socioeconomic status

of offender and victim, number of offenders and victims, number of offenders and

victims killed, type of weapons used, number of weapons used, relation of victim to









offender, and level of unique interest. I also developed qualitative descriptions of each

article.

Of the one hundred articles included in the sample, twenty seven cases emerged. I

organized these articles on the basis of case name so that key variables from the

individual articles could be represented in a combined form. The variables included in

the aggregate dataset are as follows: new trend vs. tradition, case name, number of

articles per case, number of articles written within first year after the occurrence,

socioeconomic status of offenders and victims, quantity of weapons used, type of

weapons used, race of victims and offenders, number of offenders and victims, number of

offenders and victims killed, level of unique interest controlling for time, racial make-up

of the city where shooting occurred, median family income of the city where shooting

occurred, and total population for the city where the shooting occurred. These cases

were analyzed using the structural functional and conflict paradigms with a specific

interest in the conflict-labeling criminology, power-threat hypothesis, media criminology,

and critical-media perspectives.

1.2 Background

Government agencies, criminal justice departments, and scholars in the U.S. have

not studied school shootings in the way that is proposed here. Among these institutions

the tendency has been to focus on data regarding the number of shootings and number of

victims, the types of weapons used in offenses, and components related to offenders'

psychological makeup. The Federal Bureau of Investigation has been involved in

compiling profiles for offenders in school shooting cases. Other agencies have developed


' The 2000 U.S. Census American FactFinder was used to find the racial makeup of the city
where the shooting took place and the median family income of the city where the occurred.









programs aimed at trying to prevent school violence and teaching young people how to

deal with stress and violent feelings. Most of the academic resources have been invested

in the psychological characteristics of the offenders and, at the same time, teaching others

methods of anger management. Psychologists and social workers have been involved in

providing services to those who have been victims of school shootings and those who

have had to live in the aftermath of school shootings. No study to this date has been

conducted by social science researchers with regard to news coverage of school

shootings. There is a void that exists in the academic study of news coverage of school

shootings.

1.3 Key Terminology and Concepts

The specific goals of this project are to determine the nature of media portrayal of

school violence by domestically owned and published newspapers. School violence is

classified as being of two types: traditional and new trend. Traditional school violence

involves actors who are of a low socioeconomic status, unemployed, or working class.

The target of the violence is generally other students and violence may be related to gang

or drug activities. Types of weapons used vary from assault with hands to assault with

guns that are not mass-casualty producing. The offender usually targets one or two

persons, and violence seems to happen in more of a spontaneous nature. There also may

be planned acts of violence in this category. New trend school violence is violence in

which the offender is usually a white male of a middle class or higher socioeconomic

status, and the target of the violence tends to be not only peers but also teachers, school

officials, and parents. Violence is likely related to feelings of revenge or retribution and

the types of weapons used are mass-casualty producing and tend to be excessively

violent. The offender usually plans to cause many casualties but may actually kill fewer,









and violence seems to be planned, premeditated, and organized. Some key distinctions

between these two types of school violence offenders are their level of socioeconomic

status and the relation and status of their targeted victims. This will not be measured at

the individual level in this study due to lack of data. The SES data will be organized on

the basis of the regional location in which the incident occurred.

Another key concept in this project is the expected normal behavior of the offender.

Expected normal behavior is applied in the following way: in general, white American

news journalists and editors share common ideas about normal behavior that they expect

to find in any given community.2 In cases of school shootings that occur where the

violent behavior is expected, the offender will be labeled as exhibiting normal behavior,

and white American journalists will not likely give any special attention to these cases

unless there is something unusual about them. For example, in a case where several

victims were involved, if unusual weapons were used to commit the violence or some

other anomaly along these lines, white journalists may exhibit special interest in a case.

In cases of school shootings that occur where the behavior is not expected, the offender is

labeled as exhibiting abnormal behavior and most white American journalists will give

additional attention to these cases, covering them more thoroughly and are expected to

show signs of unique interest. Some characteristics of unique interest that the white

American news journalists are expected to exhibit in these abnormal cases are the

following: an invocation of a psychological-medical model to explain the occurrence is

possible. For example, in instances in which the offender is likely to be labeled as


2 This study uses the terms "white American journalists and editors" to convey the assumption
that the majority of articles written in this study were written by the "average" white American
journalist who is expected to present each case in a certain way. There are editorials in this









psychologically ill or portrayed as an outsider from the "normal" student population at

the school, words such as "understand," "sympathy," "empathy," "cope" are expected to

be used in this type of article. In instances in which readers assume that the violence was

not supposed to happen in the community of the shooting, it is expected that the article

will present a sense of urgency about finding a solution. The article may also describe

steps that are being taken by authority figures in the community where the violence took

place to rectify the situation, and there will likely be a description involving the aide that

will likely be given to members of the community to help them cope with the act of

violence.

1.4 Theoretical Perspectives

Analysis of the literature shows that theory has not been applied to the issue of

school shooting. Sociologists have not formally studied the ways newspapers present

information and cover stories of school shootings. The following literature review is a

compilation of texts written from the conflict perspective, labeling theory, power-threat

hypothesis, with contributions from media criminology and critical-media perspectives.

1.4.1 Conflict-Labeling Criminology

This section illustrates findings from two conflict-labeling theorists that describe

the relationship between a criminal offender's status in society and the way that he or she

is likely to be labeled by authority figures that preside within the criminal justice system.

Both Turk (1964) and Quinney (1970) share similar ideas about the ways people are

labeled as criminal. Both new trend and traditional school violence cases are labeled as

criminal by newspapers. The difference is that traditional school violence (TSV) tends to


sample which are not likely to have been written by white American journalists, but it is assumed
that the majority of the editorials are written by white Americans.









be labeled as more criminal relative to new trend school violence (NTSV). For both Turk

(1964) and Quinney (1970), society is made up of a set of laws that are produced by

people in power to maintain order in society. If a person violates those laws, the less

power that individual has, the more likely he or she is to be punished and labeled in a

negative way. In this argument, the critical component of power, as measured by SES or

some other economic indicator, comes into play. School shooting offenders who have

lower SES are typically labeled as criminal more readily than those of higher relative

SES.

How is this translated into the coverage of school violence by newspapers? One of

the first points to realize is that the criminal justice system in the U.S. deals with the

offenders of school shootings. In this system, police officers, attorneys, and judges are

the authority figures who officially label the offenders in the legal sense. At the same

time, the authors and editors of newspapers have an opportunity to further label the

offenders by providing coverage of the case in the newspaper. Examining the work of

Turk (1964) and Quinney (1970) might lead one to expect that the media would cover all

school shooting cases. However, one would expect to find that the coverage of cases that

are committed by offenders of low or working class SES will be more harsh toward the

offender than those articles that cover the cases in which the offender is of middle or high

SES. This differential coverage is likely because Turk (1964) and Quinney (1970) found

that authority figures will treat criminals more harshly if they have less power. In this

case, the lower class and working class school violence offenders are more likely to

receive unforgiving coverage in relation to the middle class and upper middle class

offenders.









1.4.2 Power-Threat Hypothesis

This section introduces two concepts that Hubert Blalock (1967) developed within

the structure provided by the power-threat hypothesis (Tolnay, Beck, & Massey, 1989).

The two concepts offer further guidance in determining what type of coverage one would

expect U.S. media sources to provide for the two types of school violence. Blalock

(1967) identified two types of discrimination that can motivate the white majority to

discriminate against the black minority: symbolic segregation and threat-oriented

ideologies. He examined U.S. society during the 1960s, where he divided America into

North and South and observed trends of discrimination in both regions of the country.

Blalock (1967) developed the idea symbolic segregation in which white Americans create

belief systems that promoted segregation between white Americans and minority groups,

in particular black Americans. Blalock's (1967) idea of symbolic segregation would lead

one to expect that the media may not present coverage of all school shooting cases, and

that the news articles are likely to differ in their coverage of the two types of school

violence. By presenting TSV cases as though they were normal occurrences relative to

NTSV cases, or by not providing coverage of the TSV cases at all, the authors and editors

of national newspapers seem to be symbolically segregating TSV offenders and their

families from the rest of the society. By providing this type of coverage for TSV cases,

the authors and editors keep TSV cases from entering the consciousness of most

Americans.

At the same time, newspaper coverage provides a sense that NTSV is abnormal

which encourages the average American to think about these offenders and their families,

the victims of such offenses, and the circumstances around which the offense occurred. It

is deducted from Blalock's (1967) theory that the average American may develop









sympathy in an attempt to understand what happened, and may try to determine what

society can do to stop this new trend of school violence. This type of sympathy,

compassion, and understanding is not extended to the TSV offenders, families, and

victims since they are not within the realm of the average American's consciousness.

The symbolic segregation created by the editors and authors of newspapers in the

coverage of TSV offenders leads to neglect and reminds the offenders and their families

where they belong. The implied message is that they do not belong in the newspapers

portrayed in a similar way as the NTSV offenders and that they do not deserve the same

sympathy and respect that is given to the NTSV offenders, family members, and victims.

These findings tie into the second concept that Blalock (1967) developed termed

"threat-oriented ideology." He argued that southern white Americans tended to

exaggerate the threat that black Americans posed to the white community of recent

history. Similarly, the average white American today tends to expect violence that occurs

in the lower class and working class communities where TSV takes place. The average

white American expects to see violent behavior to an exaggerated degree in these

communities. This sentiment has been recorded by media interviews with many people

after recent incidents of NTSV have occurred. The people who are interviewed state

things like: "How can this happen here?" "I can't believe this happened here." "This is

not supposed to happen here." These statements made many times by many, mainly

white, people in America have a logical complementary statement: school violence is

supposed to happen somewhere else. That somewhere is in communities unlike those

middle and upper class communities where NTSV is occurring. It is supposed to happen

in communities where TSV occurs. This line of reasoning leads to the culmination of the









thought that there is a subconscious idea in the minds of the average white American that

there are communities "out there" where violence is supposed to occur; and the

newspapers do not cover this violence with the same care and understanding that is used

in the coverage of NTSV because it is expected from "those" communities where "those"

people live. This finding leads one to conclude that authors and editors of newspapers

will cover NTSV cases with more emphasis placed on the abnormality of the occurrences

relative to TSV cases. Additionally, only a portion of the TSV cases will be covered in

the newspapers.

A different way in which media organizations do portray TSV cases is introduced

in the next section and is centered on the idea that really horrific cases of TSV that

perpetuate the idea of threat-oriented ideologies will be published because it does

reinforce thoughts that are currently socially acceptable with regard to TSV.

1.4.3 Media Criminology

Media-criminologists conduct research on the proposed effects that news coverage

has on the reader. Howitt (1998), Barak (1994), and McQuail (1992) all share similar

ideas about the media's role in maintaining the social order that is made up of inequality

and power struggles. "The media present a world view which continually and

pervasively regenerates the ideological structures required for the maintenance of existing

power hierarchies" (Howitt 1998, p. 13). With this in mind, this paper now tries to

explain why one should expect to find little national news coverage of TSV. Howitt

(1998) found that the national news coverage of crime does not necessarily coincide with

the number of occurrences that the crime took place. For example, "lynchings were more

likely to be reported if: (a) they were in areas with easy communications; (b) they

involved rape or sexual assault; and (c) there were several victims" (p. 25). This shows









that the media covered only the sensational cases of lynchings and gives additional

support to the idea that the national newspapers will cover TSV cases on a limited basis.

Cases that are sensational will be presented in news articles. From this perspective, one

would explain the heightened media coverage of NTSV cases because they are

sensational. On average, the level of intended violence is sensational. It seems to be a

relatively new phenomenon in which children of middle and higher SES are committing

horrible acts of violence against fellow students and authority figures. This raises the

question: "Is the divide in the coverage of the two types of school violence simply a

matter of the media going after the sensational 'big scoop'?" One of the goals of this

project is to address this question.

Another goal of this research is to examine the claim that the media present

sensational articles, and those that maintain and reinforce current ideological structure.

One could argue that the coverage ofNTSV and the coverage of sensational TSV cases

do hold true to this finding as well. The NTSV cases are covered in such a way as to

separate the NTSV offenders from the "normal" group of students. It is expected that

newspaper publishers will portray NTSV offenders as psychologically ill, or as part of

some deviant sub-group of juveniles that committed these heinous acts of violence. This

is a way in which the authors of these articles do maintain and reinforce current

ideologies about the "normal" population. In a similar fashion, the TSV offenders who

commit sensational acts of violence that seem to perpetuate current ideological beliefs

will be covered in news articles. However, newspaper publishers will not attempt to

separate the TSV offender from his or her peers because that would not reinforce and

maintain current ideological beliefs about lower and working class SES communities. In









both cases, though, the news articles tend to portray the offender in such a way to

maintain status quo. However, the question still begs to be answered, is this simply a

product of news journalists' attempts to get the "big scoop"? One may argue that the

journalists have gone after the "big scoop." Indeed, they must approach their work in this

manner in order to compete and succeed as journalists, but from the theory, one would

counter that they have done so selectively in a way that perpetuated current socially

accepted beliefs about both types of offenders. However, an answer from the data has

not yet been established.

Based upon the previously mentioned conclusion, let us now discuss Howitt's

(1998) and Barak's (1994) finding. They found that media sources tend to not only

interpret stories but oversimplify stories into false dichotomies, either black and white.

There is no middle ground. They also found that the media tend to focus on the

individual for his or her situation rather than placing partial or total blame on social

policy or some social institution (Howitt, 1998; Barak, 1994). How is it that the

newspaper tends to blame individuals? When the article presents the story in terms of

either black or white, wrong or right, the reader may be left with the impression that the

offender is either wrong or right for doing what he or she has done. There is not a middle

ground. There is no room to introduce the idea that the offender did act a certain way, or

that he or she may have acted differently if the contextual or situational characteristics of

the occurrence where different. One reason to place total responsibility on the suspects

involved is that it is easier for society to see these actions as an example of abnormal

behavior rather than as a social problem.









Barlow, Barlow, and Chiricos (1995) are a group of media-criminologists, as were

Barak (1994), Howitt (1998), and McQuail (1992), who argue that the media rarely give

coverage to cases involving working class and unemployed communities. This is

primarily because these cases are not seen as noteworthy news. Again, this leads back to

the previous finding that the average white American expects to see crime and violence in

neighborhoods and communities where the working class and unemployed reside. An

article on TSV is most likely to be published by the newspapers when the story is

sensational and maintains the status quo ideologies of American society.

1.4.4 Critical-Media Perspective

David Paletz and Robert Entman (1981) provide a general analysis of the media

system in the U.S. In their analysis, they include discussion of both the sources of media

and the factors that influence what is offered by media organizations and how it is

presented. A key point described in their analysis is that the owners of media agencies

are wealthy Americans with "traditional American values [such as] individualism,

free enterprise, competitiveness, and materialism" (Paletz & Entman 1981, p. 10). These

owners operate media agencies the same as any other business concerned about profits

and prestige. Paletz and Entman (1981) note that the search for profit frames or places

limits on the content that media agencies will present. Owners and producers of news

coverage do not want to offend powerful and influential readers and advertisers because

it is not good for business (Paletz & Entman, 1981). Linking this in with findings from

this literature review, one may state that this gives further evidence as to why news

agencies are not likely to offer coverage that is counter to main-stream American

ideologies. Not only would news producers potentially lose readers, but they could also

lose advertisers, the support of powerful individuals in society, and potential investors.









News agencies are businesses and must take into consideration the various consequences

of the news that they provide before they offer coverage of certain topics.

Paletz and Entman (1981) provide additional insight into the discussion of what

makes something newsworthy. They state that journalists usually have a set of guidelines

provided by editors and news executives that provide them with direction on how to

discern what is newsworthy. According to Paletz and Entman (1981), some of the key

components of newsworthy events are the following questions: does the event show some

degree of drama and does it show some threat to American citizens either in a physical or

a social sense? If the newsworthy event does threaten American citizens, it must have

some resolution or solution so that the American public will be reassured of their security

(Paletz & Entman, 1981). The idea of threat is linked with cases of NTSV because these

occurrences provide a potential physical threat for American children that are sent to

school everyday and may come into contact with another student that is the next NTSV

offender. The NTSV phenomenon is also a social threat.3 News articles ask rhetorical

questions about the future of American society pondering "what is the matter with

America's youth"? The model, provided by Paletz and Entman (1981), is not complete

without the reassurance to American citizens that everything will be all right. Most news



3 Social threat refers more to the idea that the threat is societal in scale and may involve non-
physical threats to the average white American in general. For example, in the case of NTSV, the
phenomenon obviously provides a physical threat that is presented as though it is exists on a
national scale, but it also provides a sense that something is wrong with average white American
youths and the average white American way of life in general. This is what generates a sense of
newsworthiness in the case of NTSV. TSV, on the other hand, even if it is recognized as a
nationwide threat of physical violence, the average white American does not necessarily feel a
direct physical danger because the school system is segregated for the most part. The schools
where the TSV is happening are not likely to have a direct affect on the children of average white
Americans. Likewise, TSV does not provide any sort of social threat to the average white
American. It is usually seen and presented as instances of "black on black" crime and has no
bearing on the overall health and wellbeing of average white American youths.









agencies that offer coverage of NTSV provide this reassurance with the help of

government agencies in the form of expert advice about warning signs of disturbed

youths, counselors for families of victims and survivors of NTSV attacks, political and

police officials' promise that everything is being done to resolve the problem and that the

offenders have been apprehended. This model of "human interest" stories, as Paletz and

Entman (1981) refer to them, can also be extended to include the TSV phenomenon (p.

18). Those cases which offer an opportunity for drama, yet also give an opportunity for

resolution in a socially-acceptable way will be considered newsworthy and are more

likely to receive coverage. The critical component in this model in the way that it

explains the coverage of TSV cases is the idea of the "reassurance" portion of news

coverage. In order to reassure readers that a particular case of TSV is no longer a threat

to their physical or social wellbeing, journalists must provide a socially acceptable

conclusion in their coverage of TSV cases. From this reading, I hypothesize that a

socially acceptable resolution from the point of view of the average American will

include a situation where the offender has been detained and locked away. This is the

same conclusion that would be drawn in the coverage of NTSV cases. There too, one

would expect to see that the offender has been apprehended and is in jail. The difference

in the type of coverage that one would expect to see is that the TSV cases are not likely to

be presented as a social threat to the average white American. In the case of NTSV, it is

exactly the opposite. The average white American reader is likely to have a child who

attends public school and who is potentially endangered by this recent form of school

violence. This is a major explanatory factor in why NTSV is expected to be portrayed in









a different way than TSV. News agencies are compelled to cover NTSV differently

because of the affect that it has on their main audience, the average white American.

In summary, the literature appears to suggest that news agencies should portray

NTSV and TSV differently. NTSV provides an additional component of newsworthiness

that TSV lacks, and that extra component is the threat to personal wellbeing of the

average white American reader in both a physical and social way. When one considers

news agencies from a business perspective, and considers that the average white

American reader is the target audience of news producers, it becomes clear why news

agencies would likely provide different coverage for TSV and NTSV. By providing

detailed and resolute coverage of NTSV, news agencies are protecting their bottom line.4

News producers are attempting to create solidarity among their readers and assure them

that this new problem (NTSV) is under control; it is limited to a few "sick" and mentally

ill young people who exist at the fringe of the population of young people in America.

One could speculate that news agencies appear to provide the reassurance to maintain







4 By "resolute coverage," I mean that the news article will likely portray the case of NTSV as
being an abnormal occurrence as opposed to some new main-stream social movement by the
youths of America. The offender is abnormal, he or she is psychologically ill, unable to cope as
normal children are with the stresses of growing up. Blame may be placed on the parents of the
children. Expert advice will be given along with reassurance from various officials that
everything is being done that can be done to ensure the safety of America's schools. The U.S.
Congress and state legislative bodies debated various proposals with regard to the issue of gun
control in the aftermath of NTSV in order to reassure the average white American that everything
was being done to protect his or her wellbeing to include the safety of his or her family. Note
also that this kind of dialog does not take place with regard to TSV. There are not officials on
national media programs stating that they will do everything it takes to protect America against
TSV offenders. Congress has not debated laws that concern the need for resolution to systemic
problems within lower and working class communities that gives rise to TSV in schools. There
appears to be a completely different approach in the way that new agencies handle TSV.









current social beliefs that are necessary for the continuation of capitalist American

pursuits. 5

1.4.5 Gaps in the Literature

Specific gaps in the existing literature are as follows: No one has specifically

studied the relationship between media portrayal of school violence and the factors about

the school violence cases that affect the type of coverage given to the case. The conflict-

labeling literature provides a foundation for the idea that laws are created by those who

have power and those who do not have power are often labeled negatively with regard to

those laws based upon their interactions with those who have power. Power-threat

literature raises the idea that newspapers cover the different types of school violence

differently to support the current social ideologies and to reinforce those ideologies. The

differential coverage further segregates society symbolically, which perpetuates the

existing system of stereotypical beliefs. This creates a self-perpetuating cycle of

discrimination. The media criminology literature shows that media sources provide

coverage of TSV cases that are sensational and, again, perpetuate the current social

thoughts on TSV offenders. Finally, critical-media literature provides insight into what

makes a particular event newsworthy. Ultimately, the criteria are linked to the financial

prosperity of news agency. News producers are less likely to provide coverage that will

inhibit their agency's ability to make a profit and gain prestige.



5 The reassurance was necessary to provide a sense that everything is going to be all right. One
can easily see the ramifications of what a lack of reassurance and faith in the mainstream
American way of life can do for the American capitalist economy in light of current lack of
assurance and faith that the American public has in corporate America. In the face of current
uncertainties and the apparent inability of news producers to reduce the anxiety in America, the
economy is not doing extremely well. This is bad for those in power, those that have influence
over the media; those that have much to lose if American and World economies do not start to
move in a positive direction. Reassurance is a critical function of news agencies.









The real gap is that none of the researchers have tested these ideas on cases of

school violence directly. Turk (1964) and Quinney (1970) analyzed incidents in which

youths were coming into contact with authority figures in society usually through some

criminal justice system process, but not necessarily related to school violence. Blalock

(1967) developed his ideas around civil rights issues and interactions that African-

Americans had with white Americans. Howitt (1998), Barak (1994), McQuail (1992),

and Barlow, Barlow, and Chiricos (1995) studied the affects of media on the production

of crime. Paletz and Entman (1981) were interested in the process of news production in

general with no particular interest in media coverage of school violence. However, this

study seeks to focus on the relationship between the way newspapers portray school

violence and key factors about the offender's behavior that affect the way he or she is

portrayed by domestic newspapers.

The specific goals of this project are to determine (1) if domestic newspapers do

portray the two types of school violence (traditional and new trend) in different ways, (2)

if newspapers do differ in coverage of the two types of school violence, how do they

differ in coverage, and (3) finally, why do the newspapers differ in coverage of NTSV

and TSV? The differential coverage of school violence has something to do with

expected normal behavior which is linked with the discussion of newsworthiness offered

above.6 It is hypothesized that the cases of NTSV will be portrayed as abnormal or

deviant by the newspapers, where as TSV will be seen as less unusual.


6 Expected normal behavior is linked with newsworthiness in the following way: an event is
considered ne\\ s\\ orth\ if it involves drama and is entertaining. According to Paletz and Entman
(1981), drama as it relates to news coverage involves the idea that the event is threatening to the
public, both physically and socially, and that there is some resolution to the problem. Expected
normal behavior is related to the case of NTSV because these cases are sensation or involve
drama. The behavior that is exhibited by NTSV offenders is not expected and is not seen as

























































normal behavior for someone of their social position and race. In the case of TSV, the behavior
exhibited by the offenders is considered to be normal for someone of their social position and
race and is therefore considered to be expected normal behavior. Most likely, the only instance in
which TSV will receive special interest is where the offender's behavior is unexpected, or put in
the terms of Paletz and Entman (1981), where the behavior provides a social threat to the average
white American in general.














CHAPTER 2
METHODOLOGY



The purpose of this paper is to study the portrayal of domestic school shootings by

the newspapers in the U.S. using a compilation of theory and concepts developed from

the structural functionalist and conflict perspectives, labeling-theory, power-threat

hypothesis, media criminology, and critical-media perspective. There is no pre-existing

data-source with the variables of interest for this study. It is, therefore, necessary to

develop a dataset by using the Lexis Nexis search utility to find newspaper articles that

covered school shootings in the U.S. The time frame for the search has been limited to

include only articles which were published from 1st May 1997 through 30th April 2002.

Lexis Nexis allows for a search to be conducted using keywords. The keywords used in

this study were "school shooting." These keywords were used as opposed to "school

violence" because this study is specifically interest in school shootings rather than school

violence in general.1

Of the articles that were gathered for this study, the following list of twenty three

newspapers contributed either "full length" articles covering the cases of school shooting,

editorials, captions of a photograph at the scene of some school shooting or memorial


1 It is interesting to note, however, that the keyword search for "school violence" brought up a
roughly similar number of articles on the Lexis Nexis search. This search utility searches for
keywords in around fifty major newspapers worldwide. The main difference in using keywords
"school shooting" as opposed to "school violence" is that the articles related to the keywords
"school shooting" are primarily concentrated in the United States. While those articles found
using keywords "school violence" had a larger percentage of articles that were written about
incidents that occurred in other countries as well as the United States.

19









service, or some form of a chronological listing of school shootings with brief details on

each one: Rocky Mountain News, Sacramento Bee, The Washington Post, Los Angeles

Times, The New York Times, The Times-Picayune, The Baltimore Sun, The San Diego

Union-Tribune, Newsday, St. Louis Post-Dispatch, The Boston Globe, Milwaukee

Journal Sentinel, The Atlanta Journal Constitution, The Seattle Times, USA Today, The

Hartford Courant, The Plain Dealer, St. Petersburg Times, Chicago Sun-Times,

Information Bank Abstracts (Wall Street Journal), Star Tribune, The Buffalo News, and

The San Francisco Chronicle. Most of the newspapers offered either local or national

coverage of school shootings. Some newspapers offered both local coverage of events

that occurred within the community where the paper is published and coverage of school

shootings in other communities around the country.

2.1 Summary of Findings

News articles found using keywords "school shooting" for the five year period

were analyzed. Initially, a total of 757 articles resulted from the search. Articles

published by foreign news agencies or written about school shootings in other countries

were deleted. Articles were removed from the sampling frame if they were duplicates.2

In general, if there were duplicate articles, the more elaborate of the two was included in

the sampling frame. This method of selection for inclusion was chosen simply to have

some systematic way for choosing between duplicate articles. After this filtering process

was complete, a total of 578 articles remained. Of these 578 articles, I randomly selected

100 articles for inclusion in the sample. Of these 100 articles, 27 cases of school

shooting emerged.









A few general trends emerged within these datasets:

Table 2-1. Number of articles written per case.
Case Name Number of Articles
Albany HS (Albany, California) 1
Alva W. Dimmitt MS (Seattle, Washington) 1
Appalachian School of Law (Grundy, Virginia) 1
Buell ES Mount (Morris, Michigan) 2
Burlington HS (Burlington, Wisconsin) 1
Carter G. Woodson MS (New Orleans, Louisiana) 2
Columbine HS (Littleton, Colorado) 22
Destrehan HS (Destrehen, Louisiana) 1
Fort Gibson MS (Fort Gibson, Oklahoma) 1
Heath HS (Paducah, Kentucky) 2
Heritage HS (Conyers, Georgia) 10
John Batram HS (Philadelphia, Pennsylvania) 1
Lake Clifton Eastern HS (Baltimore, Maryland) 1
Lake Worth MS (Lake Worth, Florida) 1
Lew Wallace HS (Gary, Indiana) 1
Martin Luther King HS (New York, New York) 2
Parker MS (Edinboro, Pennsylvania) 1
Pearl HS (Pearl, Mississippi) 1
Public School 18 ES (Buffalo, New York) 1
Ridgewood HS (New Port Richey, Florida) 1
Santana HS (Santee, California) 12
Shawnee Mission South HS (Kansas City, Missouri) 1
Sumner HS (Sumner, Washington) 1
Thurston HS (Springfield, Oregon) 2
Vigor HS (Prichard, Alabama) 1
Westside MS (Jonesboro, Arkansas) 5
? ES, (East Boston, Massachusetts*) 1
Total number of articles 77
* The actual name of the school was not given in the article and other means of uncovering the
name of the school were unsuccessful.
Note: The names of the schools are given and the city and state are given. The abbreviations ES,
MS, and HS indicate elementary school, middle school, and high school respectively.

One can see from Table 2-1 that not all of the one hundred articles were written

about one specific case. Only seventy seven articles are actually written about any one of

the twenty seven cases of school shooting included in the sample. Of the thirty three


2 Duplicate articles were written about the same incident and published at two different instances
either by the same newspaper or by different newspapers. The articles were listed under the same









articles that are not written about a specific school shooting incident, one does not cover

school shooting at all, there are three different articles that list several cases of school

shootings that have occurred over a given period of time, and a majority of the twenty

nine remaining articles are editorials that offer reactions to school shooting in general but

that do not necessarily discuss a specific case of school shooting. Of the articles that do

discuss specific incidents of school shooting, most of them are concentrated in the

coverage of a few cases of school shooting. The vast majority of cases are only discussed

in one brief article.

2.2 Discussion of Data Characteristics

There are several limitations of these data that require discussion. However, I will

first provide a listing of the variables that were common among almost all articles

included in the sample. The following information was included in almost all articles:

title of article, placement of article in newspaper, number of words in article, publication

date of article, and the name of the school where the shooting occurred.

The data used in this study includes a wide variety of articles. Some are "full

length" descriptions of the events surrounding the shooting along with the reactions of

school officials and police. Others are editorial descriptions and analyses of the

shootings. Still others are very limited captions related to a photograph that was taken at

the site of the school shooting or a memorial service. There are approximately three

articles that list several incidents of school shootings that had taken place within the past

few years, and these include a short description of each case. Additionally, there are a

few articles that provide discussion of legislative and political issues that were debated

regarding school shootings and school violence in general.


title but the word count was slightly different.









The task of coding for pertinent variables was complicated by a lack of description

of these variables in a majority of the articles. Of the twenty seven cases included in the

sample, three describe race and only eight describe socioeconomic status. These two

variables are critical for this analysis. To supplement the lack of information about these

variables offered in the articles, the U.S. Census American FactFinder was utilized to find

both race and median family income figured for the cities where the school shootings

took place.3 These data were compiled at the case level of analysis along with other

composite findings from the one hundred articles in the sample. Therefore, there are two

data sets: (1) one with variables for each individual article, and (2) another for the

composite variables that provide information on each case of the twenty seven cases.


3 The American FactFinder can be found at www.census.gov.














CHAPTER 3
FINDINGS AND ANALYSIS

3.1 Research Findings

3.1.1 Qualitative

As described in the previous chapter, the vast majority of cases discussed in this study

were only covered by one article. At the same time, there are a few cases that have a

relatively large number of articles describing the events surrounding a given shooting.

The articles in the sample vary greatly in their coverage of school shootings. Some of

them give in-depth descriptions of the shooting and the events surrounding the offense.

These articles tend to describe the reactions and plans of the school administrators as they

deal with school shootings. Other articles briefly describe the shooting and the

circumstances in which the shooting took place. A few "articles" were captions explaining

the content of photographs, which were not provided by Lexis Nexis. Editorials included in

the sample provide less descriptive coverage of the shootings, but they do provide insight

about the way the authors of these editorials think administrators should deal with school

shootings. Selective listings of school shootings are included in the sample. It is assumed

that these listings are those cases that the news publishers considered to be newsworthy

school shootings.

The general tendency of the in-depth articles written on school violence was that they

covered the more popular cases of school shootings, or put another way, they covered the

more newsworthy cases of shootings. These detailed articles provided at least the basic

information about each case, including the descriptions of the situation where the shooting









took place, the age, name and gender of the offender, and the number of offenders and

victims involved. The article would usually include a description of the surrounding

community, the neighborhood where the offender lived. They were typically based on

providing interviews with friends, teachers of the offenders, or survivors of the shooting and

sometime all three. There is likely to be some description of the school administrator's plans

for dealing with the shooting, some description of aide to the victims and survivors, and

possibly some discussion of legislation and gun policies. There are only a handful of these

articles in the sample and they cover only the few cases which seem to have received the

highest percent of coverage from the sample of articles used in this study.

Another slightly different type of article described the basic information of a case

briefly, and then, moved into a debate regarding legislation on gun control and school

violence.1 Some articles of this type discussed the political stances of the presidential

candidates with regard to gun control policy as they were campaigning for the 2000

elections. Others discussed the implications of the age of the school shooting offenders and

whether or not they would be tried by the judicial system as adults or juveniles. The

underlying theme of this type of article is that it usually revolves around the ability of school

officials to deal with the legal and political aspects of school shooting.

A third type of article reviewed in this study briefly describes the circumstances of the

shooting and provides some of the basic information for the case. Of the incidents in the

sample that received coverage by one article, a majority of those articles were of this type.

They were very simple descriptions of the case with limited description or discussion of any

other facts.









Captions that describe the content of a photograph were included in this sample. There

are only a few of these captions in the sample itself. The captions provide a brief outline of

the case to include most of the basic information and then a brief description of the

photograph is given. Some of the captions only describe the photograph and do not describe

the basic information about the case. Therefore, one can see that the information given in

these captions varies greatly.

Editorials included in the sample provide a plethora of information about school

shootings. Most do not provide the detailed information about the case itself. They do not

given descriptions of the offenders, victims, and crime scene. They do, however, provide

information about the way that a variety of U.S. citizens see and interpret school shootings.

In essence, they provide a crude indication of how newspapers and other media sources

portray school shootings under the assumption that people formulate at least some of their

opinion about school shootings based upon what they have learned from news coverage. The

range of editorial opinions on this subject is so great that it is not feasible to describe each in

detail. However, a few general tendencies do arise. Some editorials focus on reactions to

school violence and tend to highlight a trend that the author feels is occurring in the U.S.

These editorials discuss policy initiatives, gun control laws, and other alternatives for putting

an end to school violence in the U.S. Another type of editorial, although rare, gives insight

into the portrayal of school shooting and describes that school shooting is not necessarily a

new phenomenon nor does the coverage of school shooting by the media give an accurate

representation of what is actually happening in U.S. schools. Most editorials fall along these

two types: (1) either providing a discussion on policy and legislation related to school


SBasic information includes variables like, number of offenders and victims, age of offenders and
victims, relation of offenders and victims, and name of school. Other than this the information









shootings and gun control, or (2) providing information concerning the actual issue of school

shootings in the U.S. and whether or not they are accurately described by media sources.

Finally, there are three articles that provide listings of school shootings in the U.S. over

certain periods of time. These listings are provided by journalists that work for a newspaper

company. They are therefore the listings of shootings that are considered newsworthy by the

news industry. The listings do not include all of the school shootings that occurred during

the period covered by the article. The primary utility of these listings is that it allows for a

comparison of findings from this study with regard to level of unique interest.

In summary, the articles in this sample tend to vary greatly in their coverage of school

shootings. Some of them offer in-depth descriptions of the shooting and the surrounding

events. Additionally, they may describe the reactions and plans of the school administrators

as they deal with school shootings. Other articles provide only a brief description of the

shooting and the circumstances. Editorials included in the sample provide less description of

the shootings, but more insight into the way that the authors of the editorials think about

school shootings. Listings of school shootings are included in the sample. The listings,

taken as a whole, provide a list of cases which the news industry believes to be newsworthy.

3.1.2 Quantitative

The following list of variables has been compiled in the dataset for this study. Each

article included in the sample has been coded for these variables. In the dataset that is non-

aggregated, these articles are coded. Several of these variables are used to build an aggregate

dataset that is coded for each case. Additional information has been collected for this

aggregate level data. The sources for these data are both the articles provided by Lexis


provided varied and even this basic information is not always given in each article.









Nexis, the year 2000 data from U.S. Census Bureau American FactFinder utility, and the

World Wide Web.

3.1.2.1 Non-aggregate Dataset

Table 3-1. Basic figures for non-aggregate variables.
Std.
Variable N Minimum Maximum Mean S.
Deviation
section of the newspaper 99 1 4 2.48 0.98
number of words 100 20 2700 545 471
number of offenders 50 1 7 1.58 1.11
number of offenders killed 50 0 2 0.14 0.50
number of victims 47 0 45 8.70 9.63
number of victims killed 45 0 23 2.82 4.48
race of victim 5 1 3 1.80 1.10
race of offender 3 1 3 1.67 1.15
type of weapon 21 0 3 2.10 0.62
quantity of weapons 26 0 30 2.96 5.99
SES victim 10 1 2 1.80 0.42
SES offender 11 1 2 1.73 0.47
relation of victim to offender 47 1 2 1.28 0.45
unique interest 89 0 7 2.47 1.91
type of article 99 1 4 2.04 1.28

Section of the newspaper. Each article is published in a certain section of the

newspaper and this variable refers to that section. The variable was broken down into several

possible replies: (1) front page of news paper; (2) front page section but not on front page;

(3) front page of a section other than the front page section; and (4) in a section other than the

front page section not on front page of the section. On average, news articles were coded

2.48. The largest percent of articles were in the second category at 56.6%. Twenty five

point three percent of the articles fell within the third category. The vast majority of articles

do not make front page news. Of those articles that do appear on the front page, six out of

ten, cases are related to NTSV shootings.

Number of words. The number of words in the article is recorded as given by Lexis

Nexis. The mean number of words for the set of articles was 544.7 with a standard deviation









of 471.5. Of those articles which are in the top quartile of this variable, most of them are

NTSV cases. Most of the articles in the top quartile tend to score above the mean on the

variable that measures level of unique interest.

Case of school violence. This variable is simply the name of the school where the

school shooting took place and city and state where the school is located. Thirty three

articles in the sample did not have any entry for this variable. This was due to the fact that

they did not address a specific case, they did not focus on school shootings, or they covered

several school shootings.

Number of offenderss).2 The number of offenders involved in the case is taken

directly from the article. Fifty out of the one hundred articles include information on this

variable. Of the fifty articles that do include information on this variable, the average

number of offenders is 1.58 as given per article.

Number of offenders) killed.* This is the number of offenders reported in a given

article as being killed. Only the Santana High School and Columbine High School shootings

were reported as having offenders that were killed during the act of committing the school

shooting. Both of these can be classified as NTSV cases.

Number of victimss)* The number of victims involved in each case is taken directly

from the article. Forty seven out of the one hundred articles included in the sample give

information about the number of victims for a given case. The average number of victims

involved in a school shooting is 8.70 as given per article. However, 63.8% of the forty seven

articles including information about the number of victims include ten victims or less. The



2 The variables that are marked with asterisks were originally gathered at the article level of analysis
but were later compiled at the case level of analysis. A description of the data for these variables will
be given at both levels of analysis.









Columbine High School incident makes up 23.4% of the forty seven articles with a total of

forty five victims reportedly involved.

Number of victims) killed.* The average number of victims that are reported a

killed in the incident is 2.82. Of the forty five articles that include data on the number of

victims killed in a school shooting, 91.1% of them report five victims or less. The two cases

in which the number of victims killed was the highest are Columbine High School and

Westside Middle School. Both of these cases are considered NTSV cases given their level of

unique interest derived from the articles in this sample.

Race of the victim and race of the offender.* The race variable is coded as follows:

(1) white, (2) black, and (3) other. The race of the victim and the race of the offender are

both inadequately discussed in the news articles. Of the one hundred articles in the sample,

five gave information about the race of the victim. Three articles stated that the victims)

were white and two articles gave very strong indications that the victimss' race was not

white. Three articles specifically state the race of the offenderss. Of these three articles,

two reported the offenders) as being white while one article gave an indication that the

offender was not white.

Type of weapon used in the shooting.* This variable is used to record the type of

weapon that an offender is reported to have used in a school shooting incident. The variable

is coded in the following manner: (0) no weapon is used, (1) a weapon other than mass-

casualty producing weapon is used, (2) a gun other than mass-casualty producing gun is used,

and (3) a mass-casualty producing weapon is used.3


3 Weapons other than mass-casualty producing weapons are defined as those weapons which can be
used to assault and murder someone but this category does not include guns of any kind, examples
include knives, brass knuckles. Guns other than mass-casualty producing weapons includes guns that
can be used to kill people but do not use a high caliber ammunition and do not fire at a high rate of









Of the one hundred articles in the sample, twenty one reported specific information

about the type of weapon used in the school shooting. Of these twenty one articles, 76.2%

reported that offenders) used guns that were not mass-casualty producing to commit the

shooting. A total of four out of the twenty one articles reported that the offenders) used

mass-casualty producing weapons. Three cases were associated with the reported use of

mass-casualty producing weapons: Columbine High School, Thurston High School, and

Heritage High School. Of these cases, only Columbine High School appears to be a NTSV

case according to its level of unique interest.

Quantity of weapons) used by the offenderss)* This variable simply indicates the

number of weapons reported by an article as being used in the shooting. Thirty articles

report information regarding the number of weapons used in a shooting incident. Seventy

three point one percent of the articles report that one weapon was used in the offense.

Columbine was reported to have the highest count of weapons at thirty weapons to include

guns and bombs. The next highest count of weapons was nine for both Buell Elementary

School and Thurston High School. Westside Middle School shooting came in with the third

highest count reportedly involving the use of six weapons.

Socioeconomic status of the victims) and offenderss)* This variable was coded

as follows: (1) low, (2) medium, and (3) high. It was originally thought that the SES of the

offenders would be the same as the SES of their parents. The SES of the victims) and

offenders) was to be recorded if it was reported in the news articles; however, only a small

percent of the articles in the sample reported SES.


speed, examples include .22 caliber handgun, and .22 caliber rifle. Mass-casualty producing weapons
include weapons that can shoot large caliber ammunition and/or can fire at a high rate of speed. This
category also includes explosive devices that can produce a large number of casualties, examples
include shotgun, machine gun, high-powered rifle, and bombs.









Ten of the one hundred articles in the sample reported any information on the victims

SES. Meanwhile, eleven articles in the sample reported information on the SES of the

offenders. Of those articles that do report on the level of SES of victims and offenders, 80%

and 72.7% of the articles list the SES of medium, respectively.

Relation of victim to offender. This variable indicates the relation of the victim to

the offender. The main relations that are theoretically significant are the following: (1) is the

victim a peer to the offender or (2) is the victim an authority figure in relation to the

offender?

Forty seven out of the one hundred articles in the sample reported information on the

relation between the victim and offender. Of these forty seven articles, 72.3% reported that

the relationship between the victim and offender was a peer relation. The number reporting

that that victim was an authority figure in relation to the offender was 27.7%. Of this 27.7%

of articles reporting that the victim was an authority figure, there is no clear indication of

whether the majority of these articles are NTSV or TSV in nature.

Level of unique interest. The level of unique interest is a composite variable which

is an overall indication of whether or not the author of the article showed a level of interest in

the case that would be considered high with regard to this type of news coverage. There are

seven variables that were used to determine the level of unique interest (UI). Each of these

variables is coded (0) for a no response and (1) for a yes response indicating that "yes" the

article does portray this variable. The seven variables are described below: (1) is the

psychological-medical model used implying that the offender is somehow mentally unstable?

(2) Is the offender labeled as an outsider or portrayed as an outsider from the majority of

students? (3) Does the article used words such as empathy, sympathy, and understanding to









describe the school shooting and the events surrounding the shooting? This variable has two

potential positive answers: in one way, the article may be written in a way so that it makes an

attempt to understand what has happened, in another way the article may express sympathy

toward the victims and families of the victims. Either of these two attempts by the author

would result in an affirmative response for this variable. (4) Does the article portray a sense

that one should not expect a school shooting to happen in the community where it did? (5)

Does the article convey a sense of urgency that the problem of school shootings in America

is a critical problem that must be resolved right now? Also, is there a specific indication that

the authority figures involved in the case under analysis reacted promptly to alleviate the

shooting incident and any problems that may have arisen regarding the incident? Either of

these two indications in the article would have given an affirmative response for this

variable. (6) Is there a description of the actions that the school administrators and other

authority figures involved in the case have taken to rectify the problem? (7) Does the article

describe any aide that was provided for the victims and other students involved in the

incident? Each of these seven variables was coded and the scores were tallied, and the

resulting figure was the level of unique interest (UI) that was recorded for a given article.

The range of score for the level of UI of any given article was zero to seven. If none of these

variables were exhibited in the article, then a total score of zero UI would be recorded for the

article. If all of the variables exhibited in the articles, then a score of seven would be

recorded for the article.

A total of eighty nine out of the one hundred articles in the sample were given a score

on this variable. The actual range provided by the data for this variable was from zero to

seven. The average score for any given article was 2.47. Of those articles that scored higher









than the mean of 2.47, Columbine High School has eight hits, Santana High School has six

hits, Carter G. Woodson Middle School has two hits, Westside Middle School has two hits,

Heritage High School has two hits, Albany High School has one hit, Alva W. Dimmitt

Middle School one hit, Burlington Middle School has one hit, Ridgewood High School one

hit, Thurston High School one hit, and there are ten additional articles that scored above the

mean on the UI scale but are not associated with any particular case of school violence.4

Type of article in newspaper. This variable refers to the type of article as one of the

following choices: (1) article related to one case of school shooting, (2) editorial, (3) citation

of a photograph, (4) other which includes articles that do not fall into any of the other three

categories.

Ninety nine of the one hundred articles in the sample were coded for this variable. The

one article that was not coded is the article that is not related to school shootings at all. Of

these ninety nine articles, fifty two are "full length" articles that discuss one case of school

shooting. There are a total of seventeen editorials in the data sample. Four photographic

citations are included in the sample. A total of twenty six articles were listed as other. Of

the articles listed in the "other" category, there appears to be no trend in the characteristics of

these articles other than the fact that most of them are missing information on basic variables

in this dataset and many of these articles have a small word count.

3.1.2.2 Aggregate Dataset

Number of offenderss)* This variable refers to the number of offenders that are

reported for each case. Twenty five of the twenty seven cases have reported figures on the

number of offenders. Sixty four percent of the cases are reported as having one offender.


4 Hit indicates that there is one article written about the case that has a UI score above the mean of
2.47.









Twenty four percent of the twenty five cases reported on were listed as having two offenders.

Destrehan High School, Burlington High School, and Pearl High School were reported to

have four, five, and seven offenders, respectively.

Table 3-2. Basic figures for aggregate variables.
Variable N Minimum Maximum Mean Std. Deviation
number of offenders 25 1 7 1.76 1.48
number of offenders killed 25 0 2 0.12 0.44
number of victims 25 0 45 6.00 10.25
number of victims killed 24 0 13 1.58 2.81
race of victim 3 1 3 1.67 1.15
race of offender 3 1 3 1.67 1.15
revised race of victim 8 1 4 2.00 1.41
revised race of offender 13 1 3 1.38 0.65
type of weapon 15 0 3 2.07 0.70
quantity of weapons 18 0 30 3.78 7.10
SES victim 8 1 2 1.75 0.46
SES offender 9 1 2 1.67 0.50
number of articles per case 27 1 22 2.85 4.69
articles written within 1 year 27 0 20 2.41 4.38
unique interest within 1 year 23 0 5.0 2.24 1.49
median family income 27 $23,519 $66,870 $53,226 $11,599
percent white 27 11.90 95.90 66.87 24.83
percent black 27 0.4 84.50 23.18 25.55
percent other 27 1.30 29.30 9.74 8.67
total population of city 27 1,105 8,008,278 481,685 1,541,693

Number of offenders) killed.* This variable records how many of the offenders

were killed as a result of their involvement in the shooting. Columbine High School and

Santana High School are the only two cases in this dataset that involve the murder of the

offenders. No other offenders in this dataset were killed as a result of their involvement in

the shooting.

Number of victimss)* Forty eight percent of the twenty five cases reported that the

number of victims were one or less. A total of 20% of the twenty five cases reported no

victim at all. Eighty percent of the twenty five cases did not have more than eight victims.









Santana High School, Westside Middle School, Thurston High School and Columbine High

School had a reported total of 15, 15, 26, and 45 victims, respectively.

Number of victims) killed* This variable refers to the number of victims that were

reportedly killed as a result of the school shooting. Forty five point eight percent of twenty

four cases report that no victims were killed. Twenty five percent of twenty four cases show

that one victim was killed. Columbine surpasses all of the other cases in this dataset with a

total of thirteen victims dead.

Race of the victim and race of the offender.* The races reported in this dataset are

recorded directly from each news article. In instances when the article did not explicitly state

the race of the victim or offender, no race was recorded. A total of three cases in this dataset

had information that indicated the race of the victim: two cases had white victims and one

case had non-white victims. Twenty four cases in this set do not include information about

the race of the victim because it is not explicitly given in any of the articles in the sample.

Likewise, there are a total of three cases in this dataset that report the race of the offender:

two cases reported white offenders, while one case reported non-white offenders.

The race of offender and victim variables have been revised based upon additional

internet searches performed for the explicit reason of identifying the races of offenders and

victims. The new results of these searches are listed in Table 3-2 above. An additional race

category had to be created for the revised variable. The new category is (4) multiple races

which means that there were either victims or offenders of different races involved in a given

case. For example, in the case of the Columbine shooting, two white offenders shot and

killed several white victims and one black victim. The Santana shooting was the only one

other case involved the used of the new code. The offender in the Santana shooting was









white and he shot and killed one white victim and one Asian-American victim. Both

Columbine and Santana are cases of NTSV.

Type of weapon used in the shooting.* This variable is used to record the type of

weapon that an offender is reported to have used in a school shooting incident. The variable

is coded in the following manner: (0) no weapon is used, (1) a weapon other than mass-

casualty producing weapon is used, (2) a gun other than mass-casualty producing gun is used,

and (3) a mass-casualty producing weapon is used.

Fifteen cases reported information for this variable. Of these fifteen, one case reported

that no weapon was used. Three cases reported that mass-casualty producing weapons were

used, and the remaining cases were reported to have guns other than mass-casualty producing

guns involved in the shooting. Heritage High School, Thurston High School, and Columbine

High School reportedly involved the use of mass-casualty producing.

Quantity of weapons) used by the offenderss)* A total of eighteen out of the

twenty seven cases reported data describing the number of weapons used in the school

shooting. Of these eighteen cases, one reported that no weapons were used. This case was

actually one in which the offenders were planning an attack on the school but the police

found out about the plan before the offenders were able to carry out the attack. The offenders

had not been able to gather any weapons. Sixty six point seven percent of the eighteen cases

reported that one weapon was used in the offense. Of the remaining cases, an unknown

elementary school in Boston, Westside Middle School, Thurston High School, Buell

Elementary School, and Columbine High School reportedly had two, six, nine, nine, and

thirty weapons involved in the shootings, respectively.









Socioeconomic status of the victims) and offenderss)* Eight of the twenty seven

cases reported on the SES of the victims. Of these eight cases, the SES of the victims of two

cases was reported to be low. The remaining six cases indicated that the SES of the victims

was medium. Nine of the twenty seven cases in this dataset reportedly have data on the SES

of the offenders. Of the nine cases with SES data on the offenders, three cases have

offenders of low SES. Six of the nine cases, report that the offenders' SES is medium.

Number of articles per case. The findings of this variable were shown in chapter

two. They will be described here briefly. Sixty six point seven percent of the cases in this

dataset have one article included in the sample that corresponds directly with this case.

Eighteen point five percent of the cases have two articles in the sample the directly

correspond to the case in the dataset. Westside Middle School, Heritage High School,

Santana High School, and Columbine High School have five, twelve, twelve, and twenty two

articles, respectively, that are in the sample and that correspond directly with these four

cases.

Number of articles written within first year. This variable measured how many

articles are written about a given case up to one year after the school shooting occurred.

Three of the twenty seven cases in this dataset have no articles written corresponding to the

case within the first year. Sixty six point seven percent of the cases in this set have one

corresponding article in the sample that is written within the first year. Carter G. Woodson

Middle School, Martin Luther King High School, Buell Elementary School, Heritage High

School, Santana High School, and Columbine High School had two, two, two, ten, twelve,

and twenty two articles, respectively, written about each case within the first year after the

shooting occurred.









Unique interest for articles written within first year. This variable refers to the

average unique interest (UI) per case based upon the UI scores of all of the articles written on

a given case within the first year. This method of recording UI per case for the first year may

be misleading because those cases that have more articles written about them within the first

year following the shooting tend to have disproportionately smaller UI levels even though

some of the articles that were written about these cases do have high levels of UI.

The range of UI for twenty three of the twenty seven cases in the set is from zero to

five. Sixty five point two percent of the twenty three cases are at or below 2.2 on the UI

scale. The mean UI for these twenty three cases is 2.41. Albany High School, Westside

Middle School, Burlington High School, Thurston High School, Carter Middle School,

Ridgewood High School, and Alva W. Dimmitt High School have 3.0, 3.0, 4.0, 4.0, 4.0, 5.0,

and 5.0 levels of UI, respectively. Each of these cases has one or two corresponding articles

in the sample.

Median family income. The data for this variable were gathered from the U.S.

Census Bureau's American FactFinder search utility on the www.census.govwebsite.5

These data were organized on the basis of city name. All of the data are from the 2000

census updates. It is understood that all of the cases did not occur in the year 2000, but an

assumption has been made that the data used from the census have not changed significantly

over the period of time included in this study.

The minimum family income recorded for any of the cities where the shooting

occurred was $23,519 per year. The maximum median family income recorded for any of


5 The American FactFinder is a utility created by the U.S. Census Bureau that assists in finding
demographic, economic, and various other forms of information that are gathered by the census
bureau. These data are usually orders in geographically delimited regions. The user can search based









the cities was $66.870. The median dollar amount for the median family income is $43,266

with a standard deviation of $11,599.

Percent white. These data were also gathered from utilizing the U.S. Census

Bureau's American FactFinder. The same qualification that applies above applies here as

well. The minimum percent white for a given city where a shooting took place is 11.9%

while the maximum percent is 95.9%. The mean percent white is 66.9% with a standard

deviation of 24.8%.

Percent black. These data were collected from the American FactFinder 2000. The

smallest percent of the black population in a city where school shooting took place is 0.4%

while the largest percent is 84.5%. The mean percent black for cities within the sample

included in this dataset is 23.2% with a standard deviation of 25.6%.

Percent other. Percent of those cities with a race other than white and black was also

gathered using the American FactFinder utility. The minimum percent of other race recorded

for this sample is 1.3% while the maximum percent other is 29.3%. The mean percent other

race in this set of data is 9.7% with a standard deviation of 8.7%.

Total population of city where shooting took place. The data for this variable

originated in the American FactFinder utility. Of the cities that were included in the sample,

the smallest city had a total of 1,105 people and the largest city had a total of 8.0 million

people. The mean city population size was 481,685 people with a standard deviation of 1.5

million. If we exclude the outlier with the population of 8.0 million people the mean

population is 139,187 with a standard deviation of 218,138 people.


upon the limits for the region that he or she desires. In this study, I limited my search by city-level
data.









3.2 Analysis of Findings

3.2.1 Qualitative Analysis

These findings will now be summarized and analyzed in the following qualitative

analysis of the descriptive data given above. A majority of articles fall within front page

section of the news paper but do not appear on the front page. Most of the articles are several

hundred words in length although a few articles do exist in either extreme: very short or very

lengthy. Carter G. Woodson Middle School, Columbine High School, Thurston High

School, Santana High School, Westside Middle School, and Ridgewood High School appear

in the top 25% of those articles with the highest word count.

The majority of cases in this dataset have only one or two articles written on them that

are included in this dataset. There are a few exceptions to this trend: Westside Middle

School, Santana High School, Heritage High School, and Columbine High School all have

five or more articles written about them that are included in this dataset. The vast majority of

cases included in this study also had most of the articles written about them within the first

year after the incident occurred. Again, there are exceptions to the rule: Heritage, Santana,

and Columbine High Schools all have nine or more articles written about them within the

first year after the incident occurred that are present in this sample.

Of the cases included in this dataset, twenty three had articles written within the first

year of the incident. A considerable percentage of these twenty three cases scored a level of

approximately 2.0 on the UI scale. However, a considerable percentage scored 0.0 and 4.0

on the UI scale as well. The following list indicates the top 25% of cases that ranked the

highest on the UI scale during the first year: Albany High School, Westside Middle School,

Burlington, Thurston High Schools, Carter G. Woodson Middle School, Ridgewood High

School, and Alva W. Dimmitt Middle School. A similar trend does appear in these results as









it does in the previous results, but there are additional cases in these results that do not appear

in the previous results.

Most of the articles report no deaths of offenders and only a few deaths among the

victims. However, there is a trend that starts to appear in this variable and will continue

throughout the paper. Columbine High School had a total of two offenders and both killed

themselves during the assault on the school. At Santana High School, the one offender

involved in the case was reported as being killed during the shooting. Destrehan High

School, Burlington High School, and Pearl High School were unique in that they had four,

five, and seven offenders involved in the planning and or actually committing the shooting,

respectively.

The number of victims reportedly involved in a case range greatly from one through

forty five. Here again, the trend develops: Pearl High School, Santana High School,

Westside Middle School, Thurston High School, and Columbine High School are reported to

have greater than ten victims involved in the shooting. While Pearl High School, Westside

Middle School, Thurston High School, and Columbine High School all have the highest

number of victims that are killed during the shooting as well.

The only comment that can be made about the race of offenders and victims based

upon the data given in the articles is that race is not discussed explicitly. One could make

assumptions about the race of the offenders and victims based upon general knowledge of the

racial makeup of the U.S. in addition to the language that is used to describe the school

shooting, but an attempt has not been made to do so here. The methods used in this study do

not allow for me to clearly state the race of the offender or victim, unless it is explicitly









stated in the article. In summary, 97% and 95% of the articles do not give information about

the race of the offenders and victims, respectively.

When describing the type of weapon used in the school shooting, there is also a lack of

information available from news articles in this sample. Seventy six percent of the sample

does not include information about the type of weapon involved in the shooting. Of those

articles that do present information about the type of weapon used in the shooting, the

majority of the weapons that are used are guns other than mass-casualty producing guns e.g.,

a hand-gun. Heritage, Thurston, and Columbine High School shootings reportedly involved

the use of weapons of mass-casualty producing capabilities. This gives additional evidence

to the trend that appears to be manifesting itself in the data.

It appears that of the twenty seven cases found in this dataset, eighteen have

information about the quantity of weapons used in the assault on the school. Most of the

offenders used one weapon. From the previous description on type of weapon, it is likely

that this weapon is a handgun. However, there are several cases were the offenders used

many weapons of varying types. The following list indicates cases in which the number of

weapons used was in the top 25% reportedly used to carry out the offense: Unknown

Elementary School in Boston, Westside Middle School, Buell Elementary School, Thurston

High School, and Columbine High School. Again, a pattern is emerging from the data.

A majority of the articles included in this sample are defined as regular length articles

that describe a school shooting incident in sufficient detail. There are also, however, several

articles that are editorials presenting a variety of views and discussions on school shooting.

Columbine High School has the largest number of these "normal" length articles written with

a total often. Santana High School has the second largest number of "normal" length articles









written about the shooting that happened there with a total of eight articles. Heritage High

School has the third highest number of these "normal" length articles written with a total of

five. This supports the trend that has appeared in the study thus far.

The SES of the victims and offenders is an integral part of the theoretical argument of

this paper. There is a huge gap in this study left by the inadequate information that is

provided about this variable on the individual level by the news articles. Of the twenty seven

cases in this dataset, only eight and nine give explicit information about the victims' and

offenders' SES respectively. The majority of offenders and victims appear to be middle class

according to the vague information provided. Instead of using this information, the study

utilized the American FactFinder utility of the U.S. Census Bureau to find data on a proxy

measure of SES for the city where the school shooting took place. The proxy measure used

is median family income.

The range of the median family income of those cities were school shootings have

taken place varies greatly. The names of the schools located in a town that is in the top 25%

of median family income are listed below: Albany High School, Alva W. Dimmitt Middle

School, Appalachian Law School, Burlington, Columbine, Destrehan High Schools, Parker

Middle School, and Santana High School.

The names of the schools in the cities that are in the top 25% for each race category are

listed below. These schools that are in cities with the top 25% of whites in the population

are: Santana, Thurston, Sumner, Columbine High Schools, Parker Middle School, Buell

Elementary School, Ridgewood, and Burlington High Schools are in the top 25% of the

white percent population variable. Shawnee Mission South High School, Heritage High

School, Public School 18 Elementary School, John Batram, Lake Clifton Eastern High









Schools, Carter G. WoodsonMiddle School, Lew Wallace High School, and Vigor High

School are all in the top 25% of the black percent population among the schools included in

this dataset. Santana, John Batram High Schools, Lake Worth Middle School, unknown

elementary school in Boston, Alva W. Dimmitt Middle School, Martin Luther King High

School, Albany High School, and Fort Gibson Middle School are the top 25% of cases in the

other race category which is made up of all other racial groups beside white and black. Many

of the schools listed in the percent black and percent other categories have not been present

in the other categories and variables that have been described thus far. This indicated that

many of these other schools are not receiving the media attention that has been given to the

schools that have been present in the previous listings.

The schools that fall within the top 25% of cities with the largest populations are listed

below: Public School 18 Elementary School, Shawnee Mission South High School, Carter G.

Woodson, Alva W. Dimmitt Middle Schools, unknown elementary school in Boston, Lake

Clifton Eastern, John Batram, and Martin Luther King High Schools. These schools appear

to be the some of the same cases that have been showing up in the top 25% of black and

other racial categories.

In summary, the following general trends appear to be emerging out of the data that

exist in this sample of articles and cases: (1) the majority of articles in the sample appear in

the front page section of the news paper. (2) Most of the cases described in this dataset are

only covered by one article that is included in the sample. (3) Twenty three of the cases

included in this dataset had articles that were published within the first year after the

occurrence happened. Of these, the average score on the UI scale was approximately 2.0.

(4) The vast majority of articles reported no deaths among the offenders and an average of









two deaths among the victims in any given case. (5) The actual number of victims involved

in any given case varies significantly from case to case. (6) There is a complete lack of

explicit information about the race of the victims and race of the offenders. (7) Many of the

articles do not provide information about the type of weapons used in the offense. Of the

approximately 25% of articles that do provide information about the type of weapon used,

the primary type of weapon used was a handgun. (8) Slightly over half of the twenty seven

cases described by these data offer information about the number of weapons used in the

offense. In a majority of the cases that do provide information, the offender used one

weapon to commit the offense. (9) A majority of the one hundred articles included in the

sample provide "full length" coverage of any given case of school shooting. This is not to

say that the "full length" articles are distributed evenly among the various cases. (10)

Information concerning the SES of the victims and offenders is limited. Only several articles

provide this information in the sample. Of these articles, the majority of them indicate that

the victims and offenders are of middle class SES levels. (11) The range of the median

family income varies greatly for this dataset. (12) The range of the actual population size of

the cities included in this sample varies greatly. (13) The racial makeup of most of the cities

included in the sample is primarily homogeneous. There are a few exceptions, but the bulk

of cities in this sample consist of one racial group that consists of a large majority of the city

population.

Table 3-3 provides a visually representation of which cases appear to be exceptional.

The cases that have a high incidence of checked boxes for variables one through twelve and

fifteen should also be those cases that can be labeled as NTSV cases. Each time a case has a

parameter that appears in the top 25% of a variable, there will be a checked box for that









variable. The case with the highest number of checked boxes means that this variable has

parameters which appeared in the top 25% of a lot of variables. The variables one through

twelve and fifteen are those that, by definition, would indicate NTSV cases. Alternatively,

those cases that have boxes marked for columns thirteen and fourteen are less likely to be

labeled as NTSV cases.

Table 3-4 provides a listing of those cases that have the largest number of markings in

the boxes that would affirm NTSV status from lowest to highest. Columbine, Santana,

Thurston, Westside, Burlington, Heritage, Albany, Alva W. Dimmitt, Pearl, and Ridgewood

are within the top 25% of the list of cases on Table 3-4. Two high schools in the top 25% of

cases are Columbine High School and Thurston High School. Two high schools in the lower

25% of cases are John Batram High School and Lew Wallace High School. To keep a more

homogeneous group for comparison, the two bottom ranking and two highest ranking high

schools will be analyzed in more detail in chapter four.

Three listings that were found by Lexis Nexis provide basic information on several

cases of school violence ("List of school shootings," 1998; "School shootings," 1999; "Other

school shootings," 2000). Based upon an analysis of the listings, Pearl High School,

Westside Middle School, Thurston High School, Heath High School, Lincoln County High

School, and Parker Middle School are likely to be portrayed as NTSV cases. These findings

are based on the assumption that news journalists and editors chose the cases for inclusion in

the listings. It is more likely that NTSV cases will appear in the listings than TSV cases.

However, it may be possible that TSV cases do appear in the listing, but it is likely that any

TSV cases that appear in the listing will have some sensational component that makes them

more newsworthy. According to the listings, the following schools are less likely to be









NTSV cases, Bishop Neumann High School, Santana High School, Fort Gibson Middle

School, a school in Deming, New Mexico, Heritage High School, a high school in Notus,

Idaho, a school in St. Charles, Missouri, a school in Onalaska, Washington, Columbine High

School, and Bethel Alaska High School. Only four of these schools have coverage in this

sample. The others are not included in the sample. Therefore, conclusions cannot be drawn

about whether or not these findings do convey that these cases should be less likely to be

NTSV cases. However, in general, for the cases that do coincide in both the sample and the

listings, there does not appear to be a strong correlation between what the listings predicted

would be NTSV and what is actually provided from the data gathered in the sample.

3.2.2 Quantitative Analysis

There are several key correlations between variables within this dataset. This

discussion starts with the analysis of a composite variable that was first represented in Table

3-4. This variable is labeled NTSV vs. TSV (NVT) it is simply the total listed in column

four of Table 3-4. The higher the score the more likely it is that the case is one of NTSV.

For example, Columbine scored ten points on the NVT scale. The relationship between NVT

and the number of articles written per case is strongly positive (0.724*). Similarly, a

moderately strong positive relationship exists between NVT and articles written within first

year, unique interest for first year, median family income, percent white, number of victims

killed, and quantity of weapons at the following correlations 0.691**, 0.504*, 0.560**, 0.610**,

0.654**, and 0.627**, respectively. A moderately strong negative relationship exists between

NVT and percent black (-0.559**), and race of offender (-0.619*).

The number of articles written within the first year since the shooting occurred is

positively correlated with the number of articles written about a given case in general,

correlation is 0.981**. The number of articles written per case is directly correlated with the









Table 3-3. Matrix of cases with exceptional qualities.
Name 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
unknown elementary school X X X
albany high school X X X
alva w dimmitt middle X X X X
appalachian school of law X
buell elementary school X X
burlington high school X X X X
carter g. woodson middle X X X X
columbine high school X X X X X X X X X X
destrehan high school X X
fort gibson middle school X
heath high school
heritage high school X X X X
john batram high school X X X
lake clifton eastern high X X
lake worth middle school X
lew wallace high school X
martin luther king high X X
parker middle school X X
pearl high school X X X
public school 18 elementary X X
ridgewood high school X X X
santana high school X X X X X X X X
shawnee mission south high X X
sumner high school X
thurston high school X X X X X X X
vigor high school X
westside middle school X X X X X X
1 Articles within top 25% of word count.
2 Cases in top 25% of number of articles.
3 Cases in top 25% of UI for first year articles.
4 Number of "full length" articles written on one case is five or more.
5 Cases in top 25% of number of offenders killed.
6 Cases in top 25% of number of offenders.
7 Cases in top 25% of number of victims.
8 Cases in top 25% of number of victims killed.
9 Cases in top 25% of that involve the use of mass-casualty producing weapons.
10 Cases in top 25% of number of weapons used.
11 Towns in top 25% of median family income.
12 Cities in top 25% of white race category.
13 Cities in top 25% of black race category.
14 Cities in top 25% of other race category.
15 Cities in top 25% of total population.









Table 3-4. List of cases from lowest to highest affirmation of NTSV status.
School Displayed Interest Factors Causing Interest Total NVT

columbine high school 3 7 10 N
santana high school 3 5 8 N
thurston high school 2 5 7 N
westside middle school 2 3 5 N
burlington high school 1 3 4 N
heritage high school 2 2 4 N
albany high school 1 2 3 N
alva w dimmitt middle school 1 2 3 N
pearl high school 0 3 3 N
ridgewood high school 2 1 3 N
buell elementary school 0 2 2 N
destrehan high school 0 2 2 N
parker middle school 0 2 2 N
unknown elementary school Boston 0 2 2 N
appalachian school of law 0 1 1 T
carter g. woodson middle school 2 -1 1 T
sumner high school 0 1 1 T
heath high school 0 0 0 T
fort gibson middle school 0 -1 -1 T
lake clifton eastern high school 0 -1 -1 T
lake worth middle school 0 -1 -1 T
lew wallace high school 0 -1 -1 T
martin luther king high school 0 -1 -1 T
public school 18 elementary school 0 -1 -1 T
shawnee mission south high school 0 -1 -1 T
vigor high school 0 -1 -1 T
john batram high school 0 -2 -2 T
Mean 1.89
Std Deviation 3.04
Median 2.00
Notes:
1. Displayed interest is the number of boxes checked in Table 3-3 columns 1- 4 for each case. This
indicates the actual level of interest that media sources have with regard to a case.
2. Factors causing interest are those factors in Table 3-3 columns 5 14. From the theory one expects
that cases with columns 11 and 12 are more likely to gain special interest from the media. This
column is computed in one of two ways: for those cases with no check in Table 3-3 columns 13 and
14, the sum of checks in columns 5 12 is added. For those cases with checks in Table 3-3 columns
13 or 14, checks in columns 5 12 are added. For those cases with checks in Table 3-3 columns 13 or
14 and no checks in columns 5 12, the checks in blocks 13 and 14 were subtracted. The theory
behind this line of reasoning is that for cases that occur in mostly white populations and those that
occur in more wealthy populations will be likely to gain the unique interest of the media. On the
other hand, in communities where the population is mainly made up of minority groups, the media are
less likely to show a unique interest in the case unless there is some thing sensational about the case
yet still the case must allow for the maintenance of socially acceptable ideologies.
3. The T vs. NT column represents the relative status of the case: is it T or Traditional School
Violence or is it NT or New Trend School Violence. This can only be determined on a relative scale.









number of offenders that are killed in the school shooting, correlation 0.919*. Also, the

number of articles written per case increases directly as the number of victims involved in the

case increases (0.806**), and as the number of victims that are killed increases (0.761**). The

number of articles written per case increases when the race of the victim is either black or

some other minority group (0.776*). The number of articles per case also increases directly

in relation with the quantity of weapons that are reportedly used or planned to be used by the

offenders (0.754**).

The number of articles written within the first year of the shooting appears to be

relatively weakly correlated with the median family income of the city where the shooting

occurred (0.395*). The number of articles written within the first year is strongly and directly

correlated with the number of offenders that are killed in the school shooting (0.928*).

Likewise, the number of articles written within the first year increases directly as the number

of victims involved in the case increases (0.745**) and as the number of victims that are

killed increases (0.682**). Cases involving minority victims are likely to have a larger

number of articles written about them within the first year after the incident (0.836*). The

number of articles written within the first year increases directly in relation with the quantity

of weapons that are reportedly used or planned to be used by the offenders (0.712**).

The level of unique interest (UI) of the articles that are written within the first year

appears to be correlated with the percentage of white people in the population of the city

where the school shooting took place in a moderately positive direction (0.467*). The level

of unique interest (UI) of the articles that are written within the first year appears to be

correlated with the percentage of black people in the population of the city where the school

shooting took place in a moderately negative direction (-0.445*). The level of UI of articles









written within the first year is relatively strongly correlated with the race of the offenders in

such a way that if the offender is white news agencies are more likely to display a greater

level of UI (-0.730*).

The level of UI of articles that within the first year represents the average level of UI

for all of the articles published about a given case within the first year of the occurrence. The

level of UI is a broad indicator for type of school violence. A high level of UI indicates that

it is likely that the case is a NTSV case. The lower the level of UI the more likely it is that

the case is a TSV case. This finding indicates that cases with higher levels of UI tend to be

more likely to have happened in communities where there are a higher percentage of white

Americans living in the community. This finding also indicates that the higher the level of

UI for a given case the smaller the percentage of black Americans in the community where

the school shooting took place.

The median family income is higher in those communities where a larger percentage

of the population is white (0.490**) and lower in those communities where a larger

percentage of the population is black (-0.572 ). The correlation between the median family

income of the community where the school shooting took place and the number of offenders

killed in the shooting is 0.495*.

The percentage of white people in the community where the shooting took place is

moderately correlated with the race of the offender in a way such that white offenders are

more likely in communities with larger populations of white people (-0.668*). The percent of

white people in the population is positively and moderately correlated with the SES of the

victim (0.717 ). This means that in communities where a larger percent of the population is

white the victim is more likely to have a high SES. On the other hand, as the percentage








53



Table 3-5. Correlations of aggregate variables.


IIIIIIIIIIIIIY~YYI








54



Table 3-6. Table of significant correlations for aggregate variables.
Variables Corr. Sign. N Variables Corr. Sign. N


NTSV vs TSV (NVT)/number of articles .724** .000 27


per case
NVT/articles written within 1styr
NVT/unique interest for 1st yr
NVT/median family income
NVT/percent white
NVT/percent black

NVT/number of offenders killed
NVT/number of victims
NVT/number of victims killed
NVT/race of offender
NVT/quantity of weapons

number of articles per case/articles
written within 1styr
number of articles per case/number of
offenders killed
number of articles per case/number of
victims
number of articles per case/number of
victims killed
number of articles per case/race of
victim
number of articles per case/quantity of
weapons
articles written within 1st yr/median
family income
articles written within st yr/number of
offenders killed
articles written within 1st yr/number of
victims
articles written within 1st yr/number of
victims killed
articles written within 1st yr/race of
victim
articles written within st yr/quantity of
weapons


.691**
.504*
.560**
.610**
-.559**

.676**
.768**
.654**
-.619
.627**

.981**

.919**

.806**

.761**

.776*

.754**

.395*

.928**

.745**

.682**

.836**

.712**


unique interest 1st yr/percent white

unique interest 1 styr/percent black
unique interest 1st yr/race of offender
median family income/percent white
median family income/percent black
median family income/number of
offenders killed
percent white/percent black
percent white/race of offender
percent white/SES victim
percent black/race of offender
percent other race/total population of
city
total population ofcity/SES victim

number of offenders/type of weapon

number of offenders/SES victim

number of offenders killed/number of
victims
number of offenders killed/number of
victims killed
number of offenders killed/race of
victim
number of offenders killed/quantity of
weapons
number of victims/number of victims
killed
number of victims/type of weapon

number of victims/quantity of weapons

number of victims killed/quantity of
weapons
race ofoffender/SES of offender

SES victim/SES offender


.025 23


of black people in the population of the community where the shooting took place


increases the offender is more likely to be a minority (0.694**). The smaller the


population of the city where the shooting took place the lower the SES of the victim (-


0.936 ). White offenders are more likely to have higher levels of SES (-1.00*). On the


other hand, black offenders are more likely to have lower levels of SES. The SES of the


victim is directly correlated with the SES of the offender (0.745*). This indicates that


offenders with higher SES are likely to assault victims with higher SES. Offenders with


lower SES are likely to assault victims with lower SES. This also means that white


.467*

-.445*
-.730*
.490**
-.572**
.495*

-.934**
-.668*
.717*
.694**
.459*

-.936**

-.744**

-.745*

.805**

.802**

.930**

.798**

.942*

.563*

.866**

.903**

-1.0**

.745*









offenders are more likely to shoot victims with higher SES and black offenders are more

likely to shoot victims with lower SES.

The number of offenders involved in a given case of school shooting appears to be

relatively strongly correlated with the type of weapon used in a negative direction (-

0.744*). Based upon the coding of the "type of weapon" variable that means that as the

number of offenders involved in an offense increases it is more likely that these offenders

will not use mass-casualty producing weapons. In other words, the larger the number of

offenders the less violent the type of weapons used in the attack. As the number of

offenders involved in an offense increases, the SES of the victim is likely to be less (-

0.745*). Cases involving attacks on poor victims are likely to involved larger number of

offenders.

There appears to be a strong positive correlation between the number of offenders

that die in an attack and the number of victims involved in the school shooting incident

(0.805**). Similarly, there is a strong positive correlation between the number of

offenders killed and the number of victims killed in any given case (0.802*). The

number of offenders killed in the shooting incident is directly correlated with the number

of weapons used in the shooting in a relatively strong way (0.798*). The number of

offenders killed is highly correlated with the race of the victim. When a larger number of

offenders are killed, minority victims are likely to be involved in the case (0.903**).

The number of victims involved in any given case appears to be correlated in a

strong positive direction with the number of weapons found to be used or planned to be

used in each case (0.866*). In situations where offenders seemed determined to kill a

large number of people or to cause significant damage and thereby they used several









weapons, it is more likely for these cases to involve many victims because that was the

intent of the offenders. Similarly, the actual number of victims that are reported as dead

increases as the number of weapons found to be associated with any given case increases

(0.903*). Another intuitive relation that comes from the data is that there appears to be a

strong correlation between the number of victims involved in a case and the number of

victims that actually die as a result of the shooting (0.942 ). It appears that the greater

the number of victims involved in the case the more deadly the type of weapons used in

the case (0.563*). This means that when the offenders use weapons of mass-casualty

producing capabilities there are more victims produced as a result of the shooting.














CHAPTER 4
DETAILED DISCUSSION AND ANALYSIS OF SPECIFIC INCIDENTS

According to Table 3-4 List of cases from lowest to highest affirmation of NTSV

status, Columbine High School and Thurston High School are two of the highest ranking

schools and John Batram High School and Lew Wallace High School two of the lowest

ranking schools. To keep a more homogeneous group for comparison high schools will

be analyzed here.

A general description of these four cases follows: the number of articles written

about each case is twenty two, two, one, and one corresponding to Columbine, Thurston,

John Batram, and Lew Wallace, respectively. The article covering the Lew Wallace and

John Batram cases are very short compared to the mean length. One of the articles

describing the Thurston case is relatively long while the other is a brief article. Of the

articles covering the Columbine case, there are many "full length" articles and editorials,

and a few short descriptive articles.

4.1 Columbine High School Shooting

The Columbine case is unique in that it ranks the highest on the scale provided by

Table 3-4. The articles that provide information about the Columbine case describe it in

such a way that Columbine appears in the top 25% of the following categories: number of

words per article, number of articles per case, number of offenders killed, number of

victims, number of victims killed, mass-casualty producing weapons, number of

weapons, number of "full length" articles, median family income, and percent white

population of city. These findings lead to the conclusion that it is likely that Columbine









is a NTSV case and that the level of unique interest associated with the case is relatively

high. According to the findings in this study, the actual level of unique interest for the

articles written about Columbine within the first year of the occurrence was 2.2 which is

around the mean UI for all the cases. One would not expect this given that Columbine is

the ideal model for NTSV cases. However, given the apparent inadequacy of using the

level of UI alone as an indicator of whether a case is NTSV or TSV, this finding is

understandable.

A summary of the findings from the twenty two articles is presented here. The

article coverage is broken down into the following article types: ten of the twenty two

articles are "full length" articles, six of them are editorials, and six of them fall into the

other category which includes an article on poll results, SWAT team tactics, criminal

investigation of the case, response of Catholic Church, and the television and movie

industries' response to Columbine. From the findings in this study, none of the TSV

cases have elicited such a response. This is just one of the multiple ways in which NTSV

cases are treated differently than TSV cases.

One reason for the overall relatively low score on UI for Columbine is that there

are so many articles related to this case. Of the twenty two articles covering the

Columbine case, the "full length" articles contribute an average of 3.6 on the UI scale.

The editorials have an average UI of 1.2. The various other types of articles have an

average of 1.5 UI. The overall UI of 2.2 is significantly affected by the contribution from

the editorials and other articles which do not necessarily provide adequate coverage of

the Columbine case.









The group of articles that falls into the other category provides little insight into the

Columbine case other than the fact that these articles exist. By this, it is meant that with

most other cases there were not a lot of articles written in response to the shooting.

Columbine provided stimulus for a plethora of articles to be written in response to the

shooting. Many of the articles that are written about other cases refer to Columbine at

some point in the text. Columbine was a unique and powerful occurrence that shaped the

way that society reacts to school shootings. Several television companies and a movie

productions agency rescheduled certain movies and television programs so that they

would not appear insensitive to the victims and survivors of the Columbine school

shooting. Pope John Paul II., made a speech in response to the Columbine tragedy letting

the World know that his prayers and the prayers of the Catholic Church would be with

those who were harmed by the shooting. A SWAT team video, which was produced

before the Columbine shooting, to help SWAT teams and other law enforcement agencies

work together to deal with school shootings was in high demand after the Columbine

attack. Several polls were given by news and media agencies in response to the

Columbine shooting in order to determine what America's youths and parents thought

about the attack and how to keep it from occurring again. Lastly, an article briefly

provides information that authorities are looking into the link between two weapons that

were used in the shooting and purchased at a local gun show in Colorado. Of all of these

articles, only the last one would represent anything close to the type of coverage that one

would expect to find in a TSV case.



'By this I mean that TSV cases are not likely to be under special investigation by SWAT teams,
nor are they likely to receive special address from the Pope, nor are news agencies likely to
conduct various polls in response to TSV shootings.









The six editorials can be broken down into three collections of two editorials each.

The first two editorialists provide their opinion about how to solve school violence.

There are no fundamentally moving findings provided by these two editorials. The

second set of two editorials present the views of numerous authors about the Columbine

school shooting in response to other editorials written by other authors. One set of

responses mainly discusses the impact of gun control on America and whether or not gun

control should be the focus of discussion about school shootings and school violence in

general. Another editorial gives a summary of the general theoretical ideas behind this

study and will be paraphrased here: people have become somewhat hardened to urban

school violence; it has been thought of as being related to gangs and drugs. Recently,

new shootings aimed at killing many people have occurred in rural and suburban

communities. The offenders have been mainly middle class kids with no apparent reason

for their rage. It is believed that the reason lies within the American culture ("Looking

for answers in wake of another school shooting," 1999). In one editorial, the author

responses to an editorial that suggests, as this study does, that there are corollary

statements which follow from the statements made in the aftermath of Columbine.

Statements like "this should not happen here" lead to a conclusion that this violent

behavior is expected in other locations. The author of the responding editorial does not

think that these statements have anything to do with race overtly or covertly. He believes

that coverage provided by news organizations is based solely on "circumstance" and

newsworthinesss" of the case under scrutiny (Rickerfor, 1999, p. B6). However, the

author of the responding editorial fails to convincingly show that race does not affect

circumstance and what journalists and editors consider to be newsworthy.









Ishmael Reed made an argument based upon a 1994 statistic that "found that school

children in small U.S. cities, suburbs and rural communities were twice as likely to carry

weapons as students in large cities" (1999, p. 5C). Mr. Reed's argument based upon this

statistic is that Americans should not be surprised to see shootings such as Columbine

and other NTSV shootings. The findings from the 1994 report indicate that these

shootings are a very real possibility, and yet white Americans still act as though NTSV

cases are shocking. The argument being presented by Mr. Reed is a fairly strong case for

gun control laws. While, at the same time, he points out that the critical component of

racism is still a problem in America which permeates every aspect of our daily lives and

yet is not discussed by most media sources. The editorial offers the idea that there is a

general tendency to blame the nation's youth violence problems on minority communities

in inner cities, while perhaps it may be more accurate to take a look in the suburbs and

rural communities in America to determine where the guns are coming from and where

the desire to use them is coming from.

The general tendency of these "full length" articles is to describe, in detail, the

events that happened at Columbine and the response of the survivors, parents, and

officials after the incident. Some particular issues that arise in the set of "full length"

articles are: the Jefferson County Sheriff s office performed an investigation of the

Columbine case in an attempt to prepare instructional material so that other law

enforcement and rescue service personnel will know what to do in the event that a similar

incident occurs (Luzadder, 1999). Another article describes the increased incidence of

parents taking their children to the psychiatrist to be evaluated given the scare that has

occurred in the aftermath of Columbine. Prior to the shooting, parents may have written-









off certain behavior as being part of growing up, but now it is too risky for parents to

simply allow certain behaviors to continue without further evaluation to determine the

cause of the behavior (Jefferson, 1999). A third article describes how the "reluctant

experts" at Columbine are helping other communities, in particular the community in

Santee California with its school shooting. The article describes the information and

advice that counselors provided for the officials at Santana High School. The article also

describes the changes that occurred at Columbine as a result of the shooting:

Columbine doubled the number of campus security guards to four, installed 16
surveillance cameras, instituted the use of identity cards, and equipped doors with
key cards. Two mental health counselors were added to the staff, bringing the
school total to eight. A safe room was established, where distraught teachers or
students could retreat for privacy and quiet. (Carter, 2001, p. A12)

The very simple and yet utterly obvious conclusion from this article is that it is

written as though Columbine is the epitome of school shootings. As though Columbine

is the only school shooting and the experts that have dealt with the Columbine case can

deal with any other case. There is an overtone in the articles that although NTSV is the

only type of violence and the means by which officials deal with this type of case is the

only real means of dealing with any case of school violence. It also appears from the

Luzadder (1999) article that NTSV is the only type of violence that exists. There is a

tendency to ignore the TSV cases that happen in America unless there is something that

generates a unique interest in the case. It is then considered newsworthy and worthy of

being published in non-local newspapers for a larger audience to read.

One article from The Baltimore Sun describes the offenders as part of a gang that

called itself the "Trenchcoat Mafia." This article provides information about the

circumstances that surrounded the case. One interesting factor that came out of this

report was that the two offenders in this case had been "convicted last year for first-









degree criminal trespassing after they broke into a car. They completed a county

program for troublemakers in February" (Morgan, 1999, p. Al). The author asks: If these

children had not been white middle class youths, how would their treatment had been

different? Would they have been able to continue on from that conviction and plan and

prepare for the mass murder of other students and a teacher? The offenders started

planning for the shooting a year or more in advance. How would things have been

different had they been working class youths attempting to plan this shooting after having

been convicted of a serious offense? A parent of one of the surviving students stated that

he was shocked after seeing the house that one of the offenders lived in. The parent

asked, "Who would have thought that a kid who lived in a home like that and had that

kind of opportunity could do something like this?" (Morgan, 1999, p. Al). This is a

sentiment that has been expressed many times and in many ways in this case. There may

be some truth to this sentiment but, at the same time, it still seems to imply that this

behavior is expected of those who do not live in expensive houses and nice

neighborhoods. Another article by O'Brien, Gorov, and Zuckoff (1999) describes that

the offenders had been planning the attack for at least a year and that they planned the

attacks to coincide with Adolf Hitler's birth date. An interesting and important fact that

distinguishes NTSV from TSV is that the offenders wanted to create as much harm and

damage as possible. The police "recovered more than 30 homemade explosives from the

school, as well as two sawed off shotguns, a semiautomatic rifle, and a semiautomatic

handgun, leading them to believe that others had to have been involved" (O'Brien et al.,

1999, p. Al). This is not typical in TSV cases. The TSV offenders usually want to harm









or kill the person that has lead them to perform the act of violence. The TSV offender

does not want to kill as many people as possible and cause as much damage as possible.

One theme that ran in the following two articles is the idea that the offenders were

bullied and this is why they lashed out at others (Yettick, 2002; Kenworthy, 1999). The

Yettick (2002) article also describes the idea that the offenders and the "Trenchcoat

Mafia" bullied other students as well. However, many of those who were killed by the

offenders were football players responsible for bullying and picking on the offenders and

the '"Trenchcoat Mafia." Frank DeAngelis, the principal of Columbine at the time of the

shooting, said of the survivors, "they were deprived of so much of their youth. They saw

so much death. It scares me. I hope it made them stronger" (Yettick, 2002, p. 1B).

While this statement is insightful, it did not find its way into any of the articles related to

TSV that appear in this sample. It seems to be very difficult for white American

journalists to portray TSV in a way that shows this much compassion and understanding,

this much sympathy and respect for the victims and survivors of TSV. Certainly, those

who experience TSV live in communities and experience situations which would allow

them to be deserving of such a statement.2 Additional curiosities that are described in the

article by Kenworthy (1999) include the fact that President Clinton saw fit to address the

Columbine incident in one of his speeches. It is unlikely that the president would address

the country about school violence if Columbine had been a case of TSV. However, it is

understandable that the, now former, president would make some kind of statement given

that Columbine was one of the most violence cases of NTSV and TSV ever. Another



2 For examples of youths in America that have experienced violence and atrocities, see Code of
the Street, by Elijah Anderson. Anderson (1999) provides an explicit and detailed account of the
conflicts that youths in inner city America face on a daily basis.









interesting fact is the number of law enforcement officials that responded to the

Columbine incident in total:

Hundreds of heavily armed police officers, FBI and Federal Bureau of Alcohol,
Tobacco and Firearms agents arrived on the scene, accompanied by dozens of
ambulances and other rescue vehicles. .. [and] an armored personnel carrier. ..
[were] dispatched to help remove the wounded. (Kenworthy, 1999, p. Al)

Again, it is highly unlikely that a TSV case would receive such resources. Not one TSV

case in this study received such resources. At most there would likely be a squad of

police officers and an ambulance or two. There would not likely be so many law

enforcement resources utilized at one scene of violence in a TSV case.

4.2 Thurston High School Shooting

There are two articles that describe the Thurston High School case. There are other

cases that are ranked in the top 25% according to Table 3-4 and that have a larger number

of articles per case that appear in the sample used for this study, but there are few that are

ranked as high as the Thurston case.3 The Thurston case ranks in the 25% percent of

word count, unique interest, number of victims, number of victims killed, mass-casualty

producing weapons, number of weapons, and percent white population of the city. The

sum effect of these findings is that Thurston should be more likely to receive a higher

degree of unique interest and should likely be labeled as a NTSV case.

The first article that describes the Thurston High School shooting is entitled

"School shooting inspires one-act play that allows teens to explore violence" (1999, p.

A2). It describes a play that was written by an Oregon resident who decided to write it

after watching the news about the Thurston High School shooting. The article describes


3 The fact that there are a certain number of articles, in this particular sample, on a given case
does not mean that there are no other articles included in the population of articles that may be
written about the case.









some basic facts about the case such as it occurred on May 20th, 1998 in Springfield,

Oregon. A fifteen year old white male student shot and killed both his mother and father

and then went to school and shot and killed two classmates wounding twenty two others.

It is almost the ideal description of a NTSV case: the offender is usually a white male

who is of a middle class or higher socioeconomic status, the target of the violence tends

to be not only peers but also authority figures, violence is likely related to feelings of

revenge or retribution, types of weapons used are mass-casualty producing and tend to be

excessively violent, number of targets are usually high, and violence seems to be planned,

pre-meditated, and organized. This article scored a 4.0 on the UI scale.

The article describes the opening scene of the play in which the offender is in his

jail cell and is confronted by the ghosts of his victims. These ghosts ask the offender

"why'd you kill me?" ("School shooting inspires," 1999, p. A2). Eventually, in the play,

the offender states why he killed the victims. He did it because he was "wracked by the

pain of his classmates laughing at him" ("School shooting inspires," 1999, p. A2). The

playwright wrote the play in hopes that it would cause young people to face the problems

associated with the "culture of violence" in America. It is interesting that there are no

such plays written about TSV cases that appear in this sample. It seems that there is a

need for such tools to combat violence that takes place in inner city and poor rural

communities throughout America as well.

The second article that describes the Thurston High School shooting case is a more

detailed description of the case. The main theme of this article is the offender's mental

status. His lawyers originally planned to have the offender plead not guilty by reason of

insanity, but ultimately, the offender decided on his own that he would plead guilty to









four counts of murder e.g., his parents and two students. This article indicates a

relatively low score of 2.0 on the UI scale, but this is a result of the inability of this scale

to provide a truly adequate representation of the level of unique interest that the author

appears to be portraying in the case. However, after reading the article, one does realize

how the portrayal of the offender and the victims in this case differs greatly from the

portrayal of the victims and offenders in the TSV cases that follow.

4.3 John Batram High School Shooting

One brief caption explains the situation at one school shooting that happened at

John Batram High School in Philadelphia, Pennsylvania. The caption appears on page

12A in the front page section of The Boston Globe on October 5th, 1999. This caption

had a total score of zero on the variable of UI. The caption simply describes the

photograph which is of an ambulance that came to the school as a result of the shooting.

Apparently, the assistant principal of the high school was shot in the leg when he

confronted the seventeen year old offender about possessing the handgun.

This case is described here because it ranks very low on the matrix in Table 3-4.

The case does not appear in the top 25% on any of the key variables that would indicate

that the case is likely to be classified as a NTSV case. The only variables that this case

appears in the top 25% of is percent black, percent other race, and total population size.

Philadelphia is large and has a high percent of black and other non-white racial groups

when compared to the other cities in the sample.

4.4 Lew Wallace High School Shooting

There is one article in the sample that provides coverage of the Lew Wallace High

School shooting. This article is an editorial written in response to an article that appeared

in the front page section of a newspaper on page 18A. The author of the editorial writes









of the difference in the coverage of the shooting that happened at Lew Wallace High

School and other recent shootings. The author states that there is "no front-page story, no

analysis of how a student was pushed to the edge by dire circumstances, no 'how could

this happen here' commentaries" (Dixon, 2001, p. B8). The author attributes this lack of

interest in the case to the fact that the offender and victims involved in the case are not

white. The implication is that they are black American students that attended an inner

city school in Gary, Indiana and this is why there is differential coverage of this shooting.

On the scale of UI, this editorial received a score of zero.

The Lew Wallace case appears to be off of the scale according to its ranking in

Table 3-4. It does not appear in the top 25% of any of the variables except for the percent

black population in the city. This indicates that the case should a TSV case as opposed to

a NTSV case and there should be little unique interest in this case. Indeed, the evidence

does seem to confirm the theoretical assumptions about the differences between the two

types of school violence.

The author of the editorial describes a case which is classified as TSV in this

project. It gives support to the ideas generated from the theory that TSV is violence in

which the offender may be white or of another racial category and is of a low

socioeconomic status, the target of the violence is generally other students, violence may

be related to gang or drug activities, types of weapons used vary from assault with hands

to assault with weapons other than mass-casualty producing guns, the number of targeted

persons are usually low, and violence seems to happen in more of a spontaneous nature

although there may be planned acts of violence in this cases of TSV.














CHAPTER 5
CONCLUSIONS

The goal of this study was to investigate the way U.S. newspapers portrayed

domestic school shootings from 1st May 1997 to 30th April 2002. Using Lexis Nexis

online search utility, major newspapers in the U.S. were analyzed for content involving

school shootings. A total of 100 articles were selected at random from a sampling frame

of 578 articles within the specified time period. The following non-aggregate variables

were coded in this analysis: title, case name, number of words, date of occurrence, date of

publication, section of the newspaper that the article was published in, time difference

between date of publication and date of occurrence, race of offender and victim,

socioeconomic status of offender and victim, number of offenders and victims, number of

offenders and victims killed, type of weapons used, number of weapons used, relation of

victim to offender, and level of unique interest. Each article's qualitative characteristics

were also described. Once the non-aggregate data were collected several of these

variables were used to code for aggregate variables related to each school shooting in this

study: NVT, case name, number of articles per case, number of articles written within the

first year after the occurrence, socioeconomic status of offenders and victims, quantity of

weapons used, type of weapons used, race of victims and offenders, number of offenders

and victims, number of offenders and victims killed, level of unique interest controlling

for time, racial makeup of the city where shooting occurred, median family income of the

city where shooting occurred, and total population for the city where the shooting









occurred.1 The overall focus of this analysis is to determine the nature of news coverage

and portrayal of school shootings.

5.1 Overview of Theories Utilized in this Analysis

The perspectives used to address the news portrayal of school shootings have been

the conflict and structural functional paradigms, with special interest in conflict-labeling

criminology, power-threat hypothesis, media criminology, and critical-media

perspectives. These perspectives view the topic of school shooting as one of a power

struggle between people with lower levels of power and those with higher levels of

power. News agencies in America are in a situation where they have the power to

provide coverage of school shootings. They provide coverage of each case in a way that

will perpetuate social ideals about those parties involved in each shooting. A key

motivating factor for news agencies to continue to perpetuate socially acceptable beliefs

about different cases is that the agencies rely on the business of their customers and the

investments of many individual investors and advertising establishments. There are also

political influences on the type of coverage that news agencies provide (Paletz & Entman,

1981). The critical point to draw from the theory is that there are expected to be

differences in the coverage of school shootings based upon the level of power that the

various parties involved in an incident possess.

5.2 Overview of Important Findings

This section will start with a summary of the findings in the relationship that exists

between the chance that a case is NTSV and other key variables in the analysis. As the

number of articles written per case, articles written within the first year, unique interest


1 The 2000 U.S. Census American FactFinder was used to find the racial makeup of the city
where the shooting took place and the median family income of the city where the occurred.









within the first year, percent white, median family income, number of offenders killed,

number of victims involved and killed, and the quantity of weapons involved in a case

increase there is also a greater likelihood that the case is one involving NTSV offenders.

Cases of NTSV are less likely to happen in communities where a large percent of the

population is black. The NTSV offenders are likely to be white and from the qualitative

data more likely to be male.

Now consider the major results involving the level of sensational characteristics

about a case. As the number of offenders killed, the number of victims involved, the

number of victims killed, and the quantity of weapons used in the case increases the

number of articles covering a case increases. This seems to be due to the fact that when

those factors are high the case is seen as being more newsworthy i.e., more sensational

and dramatic. A finding which appears to be counter to the theoretically expected result

is that number of articles written increased when the victims were likely to be minorities.

This finding, and many others, may be different if one were to control for other factors

such as number of offenders involved and killed, number of victims killed, number of

weapons used, type of weapons used, and median family income.2

A similar finding is that the number of articles written within the first year after the

shooting increases when median family income, number of offenders killed, number of

victims involved and killed, and quantity of weapons used increases. Another relatively

unexpected finding was that the number of articles written within the first year increases

when the victims involved are minorities. Again it would be interesting to determine if


2 One must remember that even though SPSS awards a certain level of significance to these
relationships the relationships may warrant more or less strength based upon the actual number of
cases involved in the relationship. However, these findings do give some idea of the relationships
that exists in the population of articles that offer coverage of school shootings.









this finding changed when holding constant several other key variables. Theory suggests

that cases involving minority offenders are more likely when those cases allow journalists

to perpetuate current socially acceptable ways of thinking about minority members

(Howitt, 1998; Barak, 1994; McQuail, 1992; Paletz & Entman, 1981).

As the percentage of white people in the population increases, news agencies are

more likely to provide coverage within the first year of an incident with a higher level of

UI. Similarly, when the offender is white, the level of UI provided by articles written

within the first year after the shooting is likely to be high. On the other hand, populations

with larger percent of black people are less likely to receive coverage with a high level of

UI. These findings do not deviate from what the theory would lead one to wait for.3

The median family income is larger in communities where a large percent of the

population is white. The median family income is smaller in communities where a large

percent of the population is black. The median family income in communities where a

large number of offenders were killed is likely to be higher. These findings can be

expected based upon the literature.

The percent of white people in the population is correlated with the race of the

offender and the SES of the victim in such a way that as the percent of white people in

the population increases the offender is more likely to be white and the victim is more

likely to be of higher SES. Similarly, white offenders are more likely to have higher

SES. Offenders are likely to attack those in a similar SES category. Therefore, white


3 The percent white and percent black figures were found using the Census American FactFinder
utility and were found for each case in the sample. The previous findings related to the race of
the offenders or victims were found using a world wide web search and the races of all offenders
and victims were not founds for many of the cases. It is important to notice beside the impact this
has on the validity of the finding that there is limited information available on any given case









offenders are more likely to assault white victims and black offenders are more likely to

assault black victims. In a similar way, populations with a large percent of black people

are likely to have black offenders. Black offenders are more likely to have lower SES.

Small cities are more likely to have victims with lower SES.

Cases involving a larger number of offenders are likely to involve the use of less

deadly weapons. When a large number of offenders are involved in a case the victims are

likely to be minorities and have lower SES. The number of offenders killed in a shooting

is directly related to the number of victims involved and killed, and the number of

weapons used. The number of victims killed in a case is directly correlated with the

number of victims involved in a case, which is also correlated with the number of

weapons and the type of weapons used in a shooting. The larger the number of victims

and the number of victims killed the more weapons that are involved and the more

damage those weapons are capable of creating.

A pattern is developing from these results. News agencies are more likely to

provide coverage within the first year after the occurrence and beyond for cases that

involve sensational behavior. Sensational behavior may be classified as behavior that is

not expected within the norms provided by a society. In other words, cases that involve a

large number of offenders and victims, large number of deaths, white offenders of middle

and upper middle SES killing other white people in middle and upper middle SES

communities with large percentage of the population being white is considered to be

sensation.


which usually varies with the likelihood that a case is NTSV or TSV. Put another way, if the case
is NTSV, there is more information available on the case from various sources on the internet.









The three questions posed at the beginning of this project are the following: Is there

a difference in the type of coverage provided for the two types of school violence? If so,

how does the coverage differ? Finally, why is there a difference in the coverage

provided?

For the first question, yes there does appear to be a difference in the type of

coverage that news sources provide for the two types of school shootings. Table 3-4

shows a clear division in the coverage provided on the twenty seven cases included in

this sample. This sample has been taken at random from a larger sampling frame and is

therefore believed to be generalizable to the larger population of articles written on

school shootings in the U.S. during the time frame of the study.

Secondly, the coverage is different in several respects. Many of the results indicate

that NTSV cases were given a great deal of attention by news agencies. For example,

every NTSV case was in the top 25% for at least one of the following variables: word

count, number of articles written overall and within the first year, and number of "full

length" articles written. 4 At the same time, one out of thirteen of the TSV cases were in

the top 25% of the variables listed directly above. The implication of this is that there

exists a divide in the coverage provided for the two types of school shootings. Incidents

of NTSV received most of the attention of news agencies during the time period studied.

Finally, the key question that still remains is: Why is there such a divide in the

coverage of the two types of school shootings? The answer seems to be found in the

division that exists between factors causing interest presented in Table 3-3 and tallied in

Table 3-4. Most of the NTSV cases within the top 25% of variables are likely to be seen

by news producers as being newsworthy and sensational. The cases that provide media









producers with a source of behavior that is not to be expected in the cases involving white

offenders and behavior that is an exemplary instance of activity that is to be expected of

minority offenders will be considered to be newsworthy. From this perspective most of

the offenders of NTSV cases will be white, middle class or upper middle class, and the

targets of their attack will be white, middle and upper SES. Likewise from this

perspective most TSV offenders are likely to be either minority or white, of lower SES,

and targeting others of lower SES who may be white or minority.

The theoretical explanation for the differential coverage of school violence appears

to be supported by this analysis. It is the level of sensational behavior with the

combination of expected behavior that seems to be the determining factor behind the type

of coverage any particular case will receive.

5.3 Critique and Suggestions for Future Research of Differential Newspaper
Coverage

Future research may include a dataset that will provide more information for a

higher level of analysis. A rigorous analysis of more sophisticated data would better suit

the topic. For example, one could analyze the unexpected findings further to determine

whether they disappear after controlling for other factors as suggested above. In a similar

fashion, one could analyze the many statistically significant findings of this project after

controlling for the many variables to determine if any of these relations disappear.

Additionally, future researchers may search for cases in a different way than this

project has. A suggestion would be to compile a list of cases by school name and

location and from this list pick a random sample of n cases. After compiling a list of

cases to be included in the sample, search for articles on each case instead of picking a


4 See Table 3-3 Matrix of cases with exceptional qualities.









random sample of articles which was the technique used in this study. This would

provide a larger n when running various statistical analyses allowing for regression

techniques to be used assuming that the case is the intended unit of analysis.

Another area of potential research would be to further distinguish the affects of

both sensational components of a case and the expected normal behavior of the case.

They are closely linked, but from the theory one would expect that even though a TSV

case is sensation, it may not receive coverage if it does not allow news producers to

present it in a way that will further promote socially acceptable beliefs about the

offenders and victims of the TSV case. Here the main limitation of this study becomes

painfully clear. The data are not detailed enough to allow for more rigorous quantitative

analysis e.g., linear regression analysis.

Another critical relationship that must be analyzed further involves two key

variables: race and SES. These variables are so intertwined in this and many research

projects that it is difficult to separate the effects of each. In this study, the following

appear to be the case: NTSV offenders are likely to be white males of middle to upper

SES. While, TSV offenders seem to be either white or of a minority racial group but

definitely seem to be of lower SES. There exists a basic idea that NTSV offenders would

be mainly white and TSV offenders mainly black, but this would exclude a significant

number of TSV offenders that are white and non-black minorities. The dynamics

between sensational and expected normal behaviors becomes even more difficult to make

sense of as one takes a closer look at race and SES. Some questions to guide research on

this topic may include: Does differential coverage exist in cases that involve white TSV

offenders and non-white TSV offenders? These offenders, by definition, would be of









lower SES so that is one less variable to consider. Race becomes the prominent

determinant here. The other key variable to consider besides race would be the expected

normal behavior. A fundamental question that would arise is how is the expected normal

behavior for white TSV offenders and non-white TSV offenders different, and how do

those differences deviate from the expected normal behavior of white NTSV offenders?

This study did not consider this topic in that much depth.

This analysis is an initial attempt to determine the nature of news coverage with

regard to school shootings in America. Depending upon future researchers' intent, they

may want to continue analysis of this topic with a critical eye on the media portrayal of

school shootings in the U.S. Future researchers may want to analyze the racial

components of media coverage more thoroughly. There are many different directions for

future study of this topic. A big challenge for any future research is a general lack of data

on the topic of media coverage of school shootings in America. Compiling the correct

data for the correct purpose is especially difficult and time consuming but is the most

critical task that any research faces.















LIST OF REFERENCES


Anderson, E. (1999). Code of the Street: Decency, Violence, and the Moral Life of the
Inner City. New York, NY: W. W. Norton and Company.

Barak, G. (Ed.). (1994). Media, Process, and the Social Construction of Crime: Studies
in Newsmaking Criminology. New York, NY: Garland Publishing, Inc.

Barlow, M. H., Barlow, D. E., & Chiricos, T. G. (1995). Economic Conditions and
Ideologies of Crime in the Media: A Content Analysis of Crime News. Crime and
Delinquency, 4(1), 3-19.

Blalock, H. M. (1967). Toward a Theory of Minority-Group Relations. New York, NY:
John Wiley & Sons, Inc.

Carter, J. (2001, March 7). Santee school shooting; Littleton's reluctant experts try to
help; Aid: officials who went through the columbine horror tell Santee what to
expect. One warns that the trauma has "just begun." Los Angeles Times, p. A12.

Dixon, D. (2001, April 3). Another school shooting. Los Angeles Times, p. B8.

Howitt, D. (1998). Crime, the Media and the Law. New York, NY: John Wiley & Sons.

Jefferson, R. S. (1999, May 7). School shootings put parents on alert, therapist says. St.
Louis Post-Dispatch p. 4.

Kenworthy, T. (1999, April 21). Up to 25 Die in Colorado School Shooting; Two student
gunmen are found dead. The Washington Post, p. Al.

List of school shootings. (1998, May 22). The Buffalo News, p. 4A.

Looking for answers in wake of another school shooting. (1999, April 22). The
Baltimore Sun, p. 19A.

Luzadder, D. (1999, July 11). Country will learn from Columbine Jeffco investigators
create textbook on school shootings so other communities can prepare for such
incidents. Denver Rocky Mountain News, p. 5A.

McQuail, D. (1992). Media Performance: Mass Communication and the Public Interest.
London, UK: Sage Publications.

Morgan, J. (1999, April 22). Dressed in black, fans of Hitler; Students were aloof, gave
Nazi salutes, glorified death; Colorado school shooting. The Baltimore Sun, p. 1A.









O'Brien, E., Gorov, L., & Zuckoff, M. (1999, April 25). School killings plotted in diary;
Conspirators timed massacre to occur on Hitler birth date; Colorado school
shootings. The Boston Globe, p. Al.

Other school shootings. (2000, March 1). The Seattle Times, p. A3.

Paletz, D. L., & Entman, R. M. (1981). Media, Power, Politics. New York, NY: The
Free Press.

Quinney, R. (1970). The Social Reality of Crime. Boston, MA: Little, Brown and
Company.

Reed, I. (1999, May 30). The pundits misfire on school shootings; Causes: The media
offer uninformed and often racist reasons for violence by U.S. students, a critic
charges. The Baltimore Sun, p. 5C.

Rickerfor, R. (1999, May 16). School shooting story has no racial undertones. The
Times-Picayune (New Orleans), p. B6.

School shootings. (1999, April 21). St. Petersburg Times, p. 6A.

School shootings. (1999, October 5). The Boston Globe, p. 12A.

School shooting inspires one-act play that allows teens to explore violence. (1999,
February 9). St. Louis Post-Dispatch p. A2.

Tolnay, S. E., Beck, E. M., & Massey, J. L. (1989). Black Lynchings: The Power Threat
Hypothesis Revisited. Social Forces, 67(3), 605-623.

Turk, A. T. (1964). Prospects for Theories of Criminal Behavior. The Journal of
Criminal Law, Criminology and Political Science, 55, December 1964, 454-461.

Yettick, H. (2002, April 20). Echoes of tragedy; Effects of 1999 school shooting are
shrouded yet unmistakable. Rocky Mountain News, p. IB.















BIOGRAPHICAL SKETCH

John Robert Bennett received his Bachelor of Arts degree from the University of

Florida, Gainesville, Florida, in December of 2001 with a major in sociology. He has

been enrolled in both the undergraduate and graduate sociology programs at U.F. since

January, 2001, pursuing a Master of Arts degree in sociology.

After receiving his M. A. degree in sociology, the author plans to continue at the

University of Florida in pursuit of his Master of Science degree in business

administration with a concentration in management. His main areas of interests are

media, race, and criminology.