Content-Type: text/html

                When the Shooting Stops/	






WHEN THE SHOOTING STOPS


A comparison of local, regional
and national newspaper coverage
of 1990s school shootings


*******************************
By Michael McCluskey
University of Washington
School of Communications
PO Box 713
Preston, WA 98050
<a href="/cgi-bin/wa?LOGON=A3%3Dind0101B%26L%3DAEJMC%26E%3Dquoted-printable%26P%3D630237%26B%3D--%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D_1771737%253D%253D_%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>
**********************************


Research has indicated that media in smaller communities cover events in their community differently than regional or national media, suggesting that the size of the community plays a role in how news events are covered. Breed (1958) asserted that newspapers in smaller communities are less likely to report on controversy in public affairs, and may omit or bury items which might jeopardize the socio-cultural structure and man's faith in it. Breed posited that a "chamber of commerce attitude" prevails, in which the newspapers try to portray the community's positive attributes. Olien, Donohue and Tichenor (1968) found that newspapers in larger communities were more likely to report on conflict than newspapers in smaller communities. The reason, they claimed, was that smaller communities have fewer mechanisms for protecting the social order against the total disruption that might result from uncontrolled public dispute, and the community may look to the press as an instrument for tension management.  Donohue, Tichenor and Olien (1972) claimed that the community press frequently reported the positive side of civic affairs and were more likely to stress community stability. Donohue, Olien and Tichenor (1992) claimed that media in larger more pluralistic communities, places with more formal channels for dealing with conflict, were more likely to report conflict in public life.  They found less reporting of conflict in smaller, more homogenous communities, in part because residents frequently had face-to-face contacts with each other.  Further, they claimed a "guard dog" press protected community power systems, including government institutions, from outside threats and would sacrifice deviant individuals to protect the system.
        Most of this research concentrated on reporting of local government and other public affairs, the types of stories that would generate little interest outside that community. For example, a typical city council meeting in a small town may be of great interest to local media, of minimal interest to regional media and no interest to national media. These types of stories fit within what Gans (1979) calls the routine activities of public officials. (p. 52-64) Perhaps a better assessment of differences could be found in the second type of news outlined by Gans, disorder news, especially analyzing events that small-town, regional and national media would all cover. Of particular interest is coverage of widely newsworthy events that took place in a small town, as it would explore the "chamber of commerce" Breed mentioned and explore whether small-town media would be more likely to promote community cohesion, such as by minimizing conflict. 
        A series of school shootings in small towns during the 1990s presented a prime opportunity to examine differences in the ways those stories were covered, or framed, by journalists. This study examines differences in news frames among national, regional and local newspapers during five school shootings in the late 1990s. Due to the nature of the shootings, those incidents can be considered local conflicts and challenges to community cohesion. Children murdering children in a public school may be of widespread interest. Nationally, those interests may not only include what happened, but why and potential causes. A regional community may need some specifics about what happened and who was involved. The local community may need details of what happened and information it needs to recover from the traumatic event. 
Variability is expected due to differences in the nature, roles and practices of newspapers in their communities. These differences may be explained by the audience and community needs of serving different types of readers, and/or by the organizational pressures embedded in the news-gathering process.

Audience and community needs
Newspaper coverage, except for the work of Donohue, Olien and Tichenor mentioned previously, has typically been studied as a monolithic phenomenon, in which the behavior of elite media came to represent behavior of media in general. Shaw and Sparrow (1999), examing that single-actor approach of the media in political coverage, found differences in the coverage patterns between inner-ring newspapers, including The New York Times, Washington Post and Los Angeles Times, and middle or outer-ring newspapers. While national, regional and local newspapers all contribute information, interpretation and entertainment to their readers, each has specific needs of readers and the communities they cover to take into account. Fuller (1996) claims that every newspaper is provincial, and to survive, it must reflect the needs of a specific audience. (p. 9) Those needs may be reflected in a bias toward coverage of news that happens close to the audience's community needs and interests. Those communities are not necessarily defined geographically, but can be bound by specific interests. A successful newspaper, according to Fuller (p. 92), must reflect its time and place. Thus, different communities of interest should be reflected in differences in emphasis in coverage of events like school shootings. 
A national newspaper, such as The New York Times, has a specialized audience and is edited to satisfy that typically national audience. In the school shootings, the societal implications may become critical and trends may be emphasized, all to explain why it matters to a national audience. 
Local newspapers, serving a community primarily of people who live near the shooting scene, may serve that community need primarily with an emphasis on information, with a secondary emphasis on topics that encourage community cohesion and healing, and limited mention of topics that could further harm the community's collective emotional state. A local newspaper covering a tragic event in its small community may pay less attention to depicting the event as a societal trend, for such an approach could hurt community cohesion by highlighting a shortcoming in the community structure. Instead, a local newspaper covering such an event may be more likely to depict individuals as deviant, as blaming the event on a societal trend would call into question the community itself.
Regional newspapers may be a hybrid of patterns expected in national and local newspapers, as it may have two distinct audiences. One audience includes those directly affected if the story happened in their hometown, perhaps involving friends, neighbors or family. For those readers, the regional newspaper may be an important source of information, including what happened and where, to which specific people or place, and what happens next. The other, broader audience may have interests closer to that of the national audience. This community may have less interested in specific information and more interest in interpretation and assessment of trends to serve these readers.

Organizational pressures
        Reporters from national, regional and local newspapers may share the traditional, underlying value of good journalism, objectivity, a traditional means of balancing a story. But that doesn't explain how journalists sharing the same values would cover an event in different ways. Gans (1979, p. 39-40) and Graber (1993, p. 21) place the news-making process into four models. The mirror model asserts that news is a reflection of reality, that reporters observe the world and report what they see, as accurately and objectively as possible. The professional model claims that journalism is the endeavor of highly skilled professionals who put together an interesting collage of events selected for importance, attractiveness to media audiences and balance among various elements of the news offering. The political model asserts that news is a product of the ideological biases of individual news people, as well as of the pressures of the political and economic environment in which the news organization operates. The organizational model claims that news selection emerges from the pressures inherent in organizational processes and goals, including interpersonal relationships, professional norms, and technical, economic and legal constraints. 
        The organizational model provides insight for understanding the differences in news frames among the local, national and regional newspaper journalists in covering the school shootings. Organizational theories emphasize that the sheer bureaucracy of news organizations constricts news judgment. Newsrooms have the principal attributes of a bureaucracy: a division of labor (editors, general assignment reporters, beat reporters), with specific substantive area (economics, sports, society, arts).  Sigal (1973, p. 4) suggests that the division of labor within news organizations creates conflicts, with journalists in one position viewing the world from a different perspective than others within the organization. Thus, conflicts often arise between reporters and editors, between reporters and between editors, the types of jurisdictional disputes built into many large organizations. Such conflicts can have important consequences for news content. For example, editors in a distant home office may dictate story emphasis to reporters in the field.
        Conflicts are part of the beat system, a division of labor that is perhaps the most important organizational structure in understanding news values. Beats offer a ready source of fresh information for reporters, but are also an ongoing tug-of-war between reporters and sources, a symbiotic relationship in which the two parties are forced to work together. Bargaining, notes Sigal, (p. 4) is continuously part of the relationship and can affect what information sources pass on to reporters.
        Among the most common newspaper beats are those devoted to crime coverage. To be effective, reporters forge a working relationship with the law enforcement sources, and a two-way trust may develop when reporters and police sources know each other and what to expect. Thus, reporters with an existing relationship, such as those from local newspapers in a school shooting, would have an edge in access to law enforcement sources over competing media. It may also mean local reporters would be more likely to follow the lead of their sources in framing a school shooting as primarily a crime story. In essence, the reporters may somewhat become spokesmen for their news sources rather than dispassionate observers. (Sigal, p. 9) They may become dependent on their regular sources (Sigal, p. 54) and less likely to publish information that could harm that relationship.
        By contrast, reporters from outside the area visiting a school shooting scene are not bound by the source sensitivities that the local reporters have to honor, as they are not as likely to be concerned with future access. Since they don't have a two-way trust or a previous working arrangement with those sources, visiting journalists may be less likely to frame the story from the perspective of the local officials. When such outside reporters are sent to cover a small-town story, said Gans (p. 141) "they can ignore local rules which obligate local reporters to minimize the disorder, either to prevent it from escalating or to preserve the positive image of the community."
        Journalists at smaller newspapers, like the local papers, may have different views about their roles than colleagues at larger newspapers. Much of that may be tied to their perception of where they fit within the smaller communities where the school shooting took place. Editors at small-town newspapers are much more likely to identify with the community and work to protect the community's image than editors or journalists at larger newspapers. (Qualter, 1989, p. 142) They look for consensus and avoid conflict (Edelstein and Contris, 1966). Such tendencies may cause those journalists to frame stories, such as the school shootings, as being caused by deviant individuals, rather than as part of a broader cultural problem. Such a societal approach may be seen as questioning the community's values.
        Rewards and motivations may differ as well. Small-town journalists are more likely to reflect community values, and more likely to be rewarded by their community contacts, such as praise in casual conversations with local residents at the grocery store. Writing stories that upset the public can have immediate and personal consequences, from subtle snubs to outright hostility from friends and neighbors. In contrast, journalists from regional and national newspapers have infrequent contacts with their readers and rarely would encounter small-town residents affected by their stories. Regular contacts for these journalists are primarily co-workers, sources, friends and families. Their rewards come within an occupational culture, in which they identify with other journalists and value competition, aggressive reporting and recognition within the industry.  The competition comes both from outside (television, radio and other newspapers) and inside (with co-workers for prominent play of stories). Investigative reporting, winning the Pulitzer Prize and autonomy of control rank high (Viall, 1992).	
        Logistics play a role in coverage, both in the proximity of the newspapers to the event scene and size of staff, making it difficult to cover the event as a local story. Visiting reporters not only have to deal with the technical problems of working on the road, but also have to quickly become acquainted with the town, who to speak with and how to find them. The small staffs of local newspapers mean they can't cover all of the angles that the larger papers can cover, which may mean a concentration on the law enforcement aspects of the crime. Meanwhile, the national and regional newspapers, with larger staffs, can look at a variety of issues, including the use of reporters in the home office to examine the cultural implications of the crime.

Problem
        Students were gathered around tables in the cafeteria of Thurston High School before classes started one morning in May 1998 when Kipland Phillip Kinkel strolled in. The 15-year-old freshman had been suspended from the Springfield, Oregon, school a day earlier for bringing a stolen handgun to campus. Now the boy was armed with three guns. Taking aim with a semiautomatic rifle, Kinkel began spraying gunfire around the crowded cafeteria. The carnage: Two dead and 22 wounded. As Kinkel paused to reload, five students tackled and subdued him. Authorities later found Kinkel's mother and father shot to death in the family home, two more victims of the boy's rampage.
        The school shooting was disturbingly familiar to Americans in the late 1990s. The Springfield shooting was one of three school shootings that day and the 17th school shooting in 28 months. All but one of the 12 violent school-associated deaths with multiple victims between 1992 and 1998 occurred in cities with a population fewer than 80,000, and nine of the 12 were in communities with population 52,000 or fewer, according to a preliminary report by the Centers for Disease Control and Prevention. (1) Two more school shootings in smaller communities occurred in the spring of 1999, the deadliest yet, with 15 killed, in Littleton, Colorado. And the media response was familiar as well: Local and national television crews focusing their cameras on distraught students, newspapers trying to probe how this could happen and how it could be prevented.
        The topic of school shootings reverberates with the American public. In response to the April 1999 Littleton shooting, a Gallup telephone poll of 1,073 adults found 55 percent had favorable opinions of media coverage and close to an even split whether the media coverage was about right or too much. (2) A further indicator of national interest in the school shootings was registered in newspaper coverage around the nation. A Lexis-Nexis search of six major U.S. newspapers, using the same incidents as in this study, showed growing attention to the school shootings. (3)
        This study examines differences in story type, emphasis and frequency among local, regional and national newspapers covering school shootings. It examines how differences in audience and community needs, and organizational pressures contributed to different types of newspapers framing the school shooting stories in different ways. Looking at the coverage patterns of the local, regional and national newspapers covering the emotions and dramas of the school shootings offer insight to how such media operate under pressure of a traumatic event. 
This study examines media frames as a dependent variable, putting the emphasis on factors that influence the newspapers and result in certain frames. 

Framing
Framing is the ways in which the media (in this study, newspapers) consciously or unconsciously present themes within the text of articles. According to Gitlin (1980, p. 7), frames act as central organizing ideas that "enable journalists to process large amounts of information quick [and to] package the information for efficient relay to their audiences." Gamson and Modigliani (1987, p. 143) see a frame is a "central organizing idea or story line that provides meaning." Thus, examining frames is an effective way to discover which aspects of news reality journalists deem most newsworthy. These frames may be found through analysis of text, such as through the presence or absence of key words or phrases, and the choice of sources of information.

Potential frames
        Review of literature identified differences that might be expected in how local, regional and national newspapers would cover events like school shootings. It's anticipated that content analysis will show local and national newspapers to differ greatly in the frames they select for emphasis. Three potential frames are expected: societal frame, crime and suspect frame, and tension management frame.
        Societal frame. National newspapers are expected to employ a societal frame, which emphasizes the societal implications of the school shooting. That frame includes potential reasons why the shooting occurred, the social causes and solutions. Local newspapers are expected to avoid societal frames, such as potential causes, blame and solutions, which could target the community's institutions as part of the problem and threaten community cohesion.
        Crime and suspect frame. Local newspapers are expected to focus on the crime aspects of the school shooting, rather than societal implications. This crime frame may focus primarily on suspects, the deviant individuals involved in the shooting. A secondary focus of local newspapers may be on victims, especially depicting them in humanizing terms, emphasizing the good and positive sides of their character.
        Tension management frame. Local newspapers respond to the community's needs in an emotionally trying time by emphasizing a tension management frame. This frame would be typified by providing extensive information to make the community aware of the details of what happened. The frame would appear in articles depicting ways for the community to cope with the tragedy, which helps reinforce community stability.

Hypotheses
        The local newspaper may recognize the needs of the community for facts about the school shooting by providing a great deal of information in its news coverage, part of the anticipated tension management frame.  This information would be presented in a variety of stories covering conceivably every angle that the journalists could compile within the limits of news space and time.  The story would undoubtedly be the most important news event in the small town at the time, thus any stories about events surrounding the shooting would be run in prominent places, especially on Page 1.  This description fits the community and audience needs theory mentioned previously.
        H1: In number of stories and prominent placement of stories, the local newspapers will have the most coverage and most prominent placement, while the national newspaper will have the least coverage and least prominent placement.

        To protect the community's unity, as predicted by Donohue, Olien and Tichenor, the local newspapers are likely to employ a crime frame.  This would be found in a focus on the individuals involved in the shooting, especially suspects and victims. This emphasis on the crime aspects and individuals is touched upon in both the audience needs and organizational pressures explanations.  In the audience needs section, it's found in an emphasis on deviant individuals - primarily suspects -- rather than societal implications.  The organizational pressures theory asserts that working on the crime beat influences journalists to see the shooting primarily as a crime event.  In addition, it would influence journalists to identify with the government officials on their beat and see the event through their eyes - as a crime, with suspects, victims and witnesses.
H2: Local newspapers will focus the highest percentage of stories and national newspapers the least on the individuals (including victims and suspects) reflecting the crime aspects of the shooting. 

        The local newspapers are more likely to describe victims of the shooting as good, innocent people.  That gives readers a last positive impression, which helps promote community goodwill and stability. 
H3: Local newspapers will focus the highest percentage of stories and national newspapers the least on humanizing aspects of the school shooting victims.

        Part of this crime frame among local newspapers may be reflected in the use of sources. Police, prosecutors and witnesses, what can be called government sources, typically provide basic information on crimes.  Another type of government source that local newspapers are expected to turn to is school officials, again related to an emphasis on crime coverage.  The use of government sources in articles points primarily to the organizational pressures theory, again an indicator that the local journalists identified the shooting as part of their crime beat and turned to typical criminal case sources.
H4: Local newspapers will use sources that depict the shooting as a criminal event, including law enforcement, prosecutors, witnesses and school officials, more frequently than regional and national newspaper. 

        While local newspapers are expected to have portrayed the school shootings as crimes, the national newspapers are expected to emphasize the societal frame.  The societal frame includes the reasons for the shooting and local newspapers de-emphasize those explanations.  The reasons for the shooting are expected to touch upon underlying problems in society that contributed to the shooting, including potential causes and blames, and possible responses.  It's expected that the national newspapers would link the shooting to specific societal problems, such as guns, the media and insufficient government action.  This emphasis on societal issues reflects the expected response of national newspapers under the community and audience needs theory.  Local newspapers are expected to play down societal explanations to help preserve community cohesion and stability, also as explained in the community and audience needs theory.
        The national newspapers' emphasis on a societal frame is expected to accelerate with each successive shooting, which fits with the community and audience needs theory.  All three types of newspapers are expected to mention social causes more frequently over time, but the national newspapers will still mention the social causes the most frequently.  Another link to societal causes anticipated from national newspapers is by explicitly linking one school shooting to previous school shootings.
H5: National newspaper will focus the highest percentage of stories and local newspapers the least on societal ills as an answer to why the shootings occurred. 

        Within that emphasis on societal frame, the national newspapers especially are expected to emphasize one potential cause - the weapon used.  This will be found in mention of the weapons used, related issues like guns usage, gun control and hunting.  Local newspapers, representing rural areas in which gun usage is often tolerated to greater degrees than metropolitan areas, would be much less likely to link the school shooting to the weapon.  Both the expected tendencies of the national and local newspapers on the gun issue are reflected in the audience needs theory.
H6: Regional and national newspapers are more likely to examine the school shooting in the context of the weapon used or related issues, like hunting or gun control.

        One expected attribute of local newspapers is to serve the community and its needs during a traumatic time.  That tension management frame will be found in the local newspapers providing information that the community needs to emotionally respond to the school shooting, such as where and how to get help.  This emphasis is reflected in the audience needs theory, as the local newspaper provides information and advice that will help the community recover from its collective trauma.
H7: The local newspaper will emphasize tension management, found in portrayal of community cohesion and healing, more frequently than regional or national newspapers.
	
        Content analysis. One way to detect frames is through content analysis, what Entman (1991, p. 49) sees as a way to probe for words or themes that consistently appear in a text.  This study looked for key words and phrases that typify the dominant frames to help determine if the frames existed, how those frames changed over time and how those frames differed among newspapers. Content analysis was used to quantify frequency and placement of articles; key words depicting suspects and victims; potential motives, causes or solutions for the shooting incidents; use of sources; linkage to other shootings; depiction of the community where the shooting took place; and what assistance was offered for readers dealing with the shooting. The unit of analysis was the article, which includes all news stories, editorials and opinion columns during the study period.
        Reliability. Questions used in the content analysis met acceptable standards for intercoder reliability. Approximately 6.6 % of the sample were tested. (4)

Sample
        This study examined five small-town school shootings using content analysis of newspapers, pared down from a list of 17 school shootings between February 1996 and May 1998. Shootings that occurred in cities or suburbs were rejected and several remaining small-town cases were rejected because they drew limited media attention (mostly single shootings of boyfriends or girlfriends.

Criteria
        Small towns. Communities with fewer than 60,000 residents as of 1996, the most recent available U.S. Census Bureau population statistics (5) at the time o study, were defined as small towns. To ensure these were truly small towns, all instances of suburbs -- communities within 20 miles of a major metropolitan area (500,000-plus population) -- were eliminated. 
        Local newspaper. The nearest daily newspaper (a paper that publishes at least five days per week) to the crime scene was picked as the local newspaper. 
        Regional newspaper. The daily newspaper of at least 100,000 circulation nearest to the crime area, aside from the local paper, was picked as the regional newspaper. (6) The definition takes into account the location of bureaus to determine which newspaper might be more likely to cover events in the crime town. Preference was given to availability of newspapers whenever there were slight differences in distances and circulation between two newspapers.
        National newspaper. The New York Times was selected as the national newspaper, due to its national scope, widespread respect as a leading newspaper in the United States, and availability.
        Study period.  The study examined one week of coverage for each newspaper, beginning with the day after the school shooting occurred.
        Study subjects. Following is a brief description of each school shooting, including the town where it occurred, the date and a narrative, along with the local and regional newspapers selected. The incidents are listed in chronological order.
        Moses Lake, Washington, February 2, 1996, algebra class shooting. A 14-year-old boy, dressed in a black trench coat and hat, walked late into his algebra class at Frontier Junior High School and fired a high-powered rifle. He killed two students and a teacher, and injured one other student, before a teacher subdued him. Local paper: (Moses Lake) The Columbia Basin-Herald. Regional paper: (Spokane) The Spokesman-Review. (7)              Pearl, Mississippi, October 1, 1997, murder conspiracy. A 16-year-old boy went to Pearl High School and shot nine students, killing two. Authorities later found the boy's mother stabbed to death. Six days later, five other teen-agers, alleged to be part of a satanic cult, were arrested in connection with the killings. Local paper: (Jackson) The Clarion-Ledger. Regional paper: Memphis Commercial Appeal (has a bureau in Jackson). (8)
        West Paducah, Kentucky, December 1, 1997, prayer group shooting.  A 14-year-old freshman opened fire on 40 students praying together before class in a West Paducah High hallway, killing three and wounding five. Local paper: The Paducah Sun. Regional paper: The St. Louis Post-Dispatch. (9)
        Jonesboro, Arkansas, March 24, 1998, fire alarm ambush shooting. Two boys, ages 11 and 13, pulled a fire alarm at Westside Middle School and then shot at students and teachers from nearby woods. One teacher and four students died and 10 were wounded. Local paper: Jonesboro Sun. Regional paper: The Memphis Commercial Appeal. (10)
        Springfield, Oregon, May 21, 1998, cafeteria shooting. A 15-year-old boy fired at students in the cafeteria at Thurston High School, killing two and wounding 22. Authorities later found his parents shot to death in the family home. Local paper: (Eugene) The Register-Guard. Regional paper: (Portland) The Oregonian. (11)

Results
        This study found evidence that local, regional and national newspapers differ in emphasis of coverage and in framing of the school shooting stories. The national newspaper gave a greater emphasis to the societal frame, depicting the school shootings as a social problem, than did the regional or local newspapers. All three types of newspapers placed more emphasis on the societal frame with each successive shooting case. The local newspapers adopted the tension management frame by giving the school shootings more total coverage and more prominent placement of articles, and more emphasis on community cohesion and healing, than did the regional or national newspapers.
        However, data didn't support the expectation that local newspapers would give more emphasis to the crime frame.
Frequency. Amount of coverage and prominent placement of articles helps assess how many resources (staff and news space) a newspaper devotes to a story and how important the story is considered. As reflected in the discussion of H1, the local newspaper may respond to a perceived community need for information by adopting a tension management frame. This frame could be found in the local newspaper running numerous stories covering a wide variety of potential story angles, and by prominent placement of articles.
        A raw count of articles by newspaper, as predicted in H1, shows that local newspapers gave the stories more coverage. (See Table 1) The local newspapers accounted for more than 63 percent of the 602 articles. The trend was repeated in each of the five shooting cases: the local newspaper had the most coverage, and the regional newspaper had more coverage than the national newspaper.
        Placement of articles was not significant different, although local newspapers had the highest proportion of Page 1 stories (28.3%), regional papers the next most (25.2%) and national newspapers the least (22.8%). In addition, local newspapers had the highest proportion of articles on other section fronts (8.9%), compared to 4.9% in regional newspapers and 1.8% in national newspapers.
        Focus on crime. This research considers whether the local newspaper will be more likely to use a crime frame, found by coverage of some individuals, especially portraying the school shooting as caused by individuals (suspects) engaged in deviant behavior.
There was no statistically significant evidence to support H2, which proposed that local newspapers would focus on individuals, including victims and suspects, more frequently than regional and national newspapers. Data in Table 2 contradicts H2, as the national newspaper was the most likely and local newspaper the least likely to mention the suspect by name. National and regional newspapers mentioned a suspect in more than half of the articles, while local newspaper less than half the time. Combined, the three types of newspapers mentioned suspects in 46% of the articles. National newspapers were most likely to mention the suspect's name first in an article, the suspect's age, appearance, participation in typical activities (such as sports) and previous threats of violence. In eight of the 10 measures, the local newspaper mentioned such attributes the least often.
        Data on coverage of victims in Table 3 partially supports H2, but the measures weren't statistically significant. Local newspapers were slightly more likely to mention a victim than a regional newspaper or a national newspaper. Victims were named in 41% of the articles. Local newspapers were more likely to mention a victim's name first in an article.
        Humanizing victims. Local newspapers were expected to be more likely to portray victims in humanizing terms, but no significant differences were found among the three types of newspapers. (See Table 3) Local newspapers were slightly more likely to note any hobbies or activities among victims, but regional newspapers mentioned victims' parents and relatives, and friends, the most frequently, with local newspapers the second-most frequent. National newspapers mentioned a victim's age most frequently. None of the articles connected a victim with any mention of deviant activity.
        Sources. Use of sources can indicate frames. Local newspapers were expected to use government sources, especially law enforcement, prosecutors, witnesses and school officials, to reinforce their framing of the school shooting as a crime event. 
The data showed no significant differences among the three types of newspapers in frequency of sources and little support for H4. All three types of newspapers depended primarily on law enforcement/elected officials (18.0% overall), witnesses/students (16.6% overall), and school officials (15.0% overall) as sources, the types of sources that would place the story within the context of a school crime. Local newspapers were more likely to use school officials (16.8%) as sources than the regional and national newspapers, while regional newspapers were more likely to use law enforcement/elected officials (20.0%) and witness/student (18.5%).
There were marginally significant differences (p<.10) in first sources cited in an article, but offered mixed support for H4. Local newspapers used school officials the most frequently and were just short of regional newspapers in using law enforcement sources, but used witness/student the least frequently.
        One indication that the law enforcement source-journalist relationship could affect newspaper coverage was found in an editorial page column by Mary Powell in The Columbia Basin Herald (12), responding to the Moses Lake shooting. Powell wrote that the paper's editor decided to cooperate with police and school officials, to delay interviews with children directly involved in the shooting and to stay away from the crime scene until things were sorted out. 
Societal ills. National newspapers were expected to more frequently emphasize the societal frame, especially reasons for the shooting. That emphasis was anticipated to highlight potential causes, assess sources of blame, and possible solutions to the school shooting. These differences can be explained by the community and audience needs theory, which anticipated the local newspaper playing down underlying causes to preserve community cohesion and stability, while national newspapers would highlight the causes to meet community and audience interests.
Evidence supports H5, which suggests that national newspapers would focus the most attention and local newspapers the least on societal ills as an answer to why the crimes occurred. It suggests that national and regional newspapers would depict the shootings as a trend and representative of a cultural phenomenon, perhaps as part of a broader social malaise.
        Such an approach was supported within one regional newspaper article. Columnist Michele McLellan wrote that exploration of causes should be a goal in The Oregonian's continuing coverage of the Springfield shooting, beyond "incremental developments, handicapping court strategies and offering recreations of events that add few new details. ... (N)ewspapers often produce them at the expense of devoting time to stories on the broader problems events may represent as well as potential solutions." (13)
        Evidence strongly supports H4. (See Table 4) On eight of the nine measures that were statistically significant (p < .05), national newspapers were the most likely and local newspapers least likely to ascribe a motive, potential cause or solution to the school shooting. The most common items mentioned (except for "other"), in order of frequency, were guns and gun culture; media, including television, movies, newspapers and magazines; and schools and school security. Media were mentioned as a potential motive, cause or solution in nearly one of four national articles, compared to just 7% in the local newspapers. Of the 14 measures (except for "other"), only counseling, was mentioned at least 10% in the local newspapers. Counseling fits well with another expected response of local newspapers, to provide information that helps the community heal (see H6) In comparison, both the regional and national newspapers reached 10% in six measures.
In addition, national newspapers were most likely and local newspapers least likely to mention other school shootings, to mention this shooting as part of a trend and to mention other crimes. (See Table 5) Two in five national articles mentioned other crimes, compared to one in seven articles in local newspapers.
Data strongly supported some expected chronological trends, as all three types of newspapers more frequently mentioned social causes with each successive shooting. The latter two shootings had a higher percentage of stories that mentioned a potential cause, motive or solution. (See Table 6) In six of the seven measures that were statistically significant (p<.05), the highest percentage of causes, motives or solutions occurred in either the March 1998 Jonesboro or May 1998 Springfield shooting, the last two in chronological order. Similar patterns could be found in the newspapers mentioning the latest school shooting in the context of other school shootings, as part of a trend, and other crimes. (See Table 7). On two of the three significant measures, the Springfield shooting had the most frequent mention and Jonesboro the second-most frequent mention.
        Raw story totals support H7 as well, especially when the national newspaper coverage pattern is taken into account. The national newspaper clearly had a greater emphasis on the social problems angle of the shootings stories and it published more articles on the latter shootings. (See Table 1) 
        Blaming guns. As part of that emphasis on social causes, the national and regional newspapers were expected to focus more on the weapons used in the shooting than the local newspapers. H6 suggested that regional and national newspapers would be more likely to examine the shootings in the context of the weapons used in the assaults, primarily guns. A report by The Freedom Forum (14) on media coverage of the Jonesboro shooting supported this hypothesis, as it found news magazines and national news accounts especially, and national newspapers less so, citing guns and gun culture as part of the reasons behind the shooting.
        Data from this study support that premise. Regional and national newspapers mentioned guns/gun culture and gun control (p<.01) as a possible cause, motive or solution in the shootings more than twice as frequently as did local newspapers. (See Table 4)
        One example of this can be found in a regional newspaper. The Oregonian, in its coverage May 22, 1998, the day after the Springfield shooting, devoted nearly half of its two editorial pages to a discussion of guns. Included were side-by-side guest columns under the banner heading: "Should tragedy spur gun control or more responsibility?" (15)
        In addition, guns/gun culture showed a gradual growth in mention as a potential cause, motive or solution, from no mention in the first shooting to mention in nearly 20% of the articles in the last shooting. (See Table 5) Gun control showed a similar trend, from no mention in the first shooting to more than 12% in the fifth shooting.
        Community cohesion, healing. Local newspapers were expected to provide information, fitting the tension management frame, that the community needed to respond to the school shooting and begin recovering from the trauma. Those needs may include places to get help, advice on dealing with the shooting and where to make donations.
Evidence mostly supports H7, which suggested that local newspapers would be more apt to emphasize community cohesion and healing. (See Table 8) The local newspaper more frequently provided information of help to readers, including places to get help and where to make donations. As mentioned previously, local newspapers cited counseling more frequently than any other potential cause, blame or solution, which fits well with this expectation of local newspapers providing information of help to its community. (See Table 4)
Even the labels placed atop the pages in the local newspaper's coverage reflected that emphasis on community reaction to the shootings. Three of the local newspapers used headers on some pages, each using the word tragedy: "Tragedy in Pearl" in The Clarion-Ledger, "Heath Tragedy" in The Paducah Sun and "Tragedy Hits Home" in The Register-Guard. By contrast, the regional and national newspapers used more prosaic labels: "Shootings in a Schoolhouse" (Jonesboro) and "Bloodshed in a Schoolyard" (Springfield) in The New York Times, "Springfield School Shootings" in The Oregonian and "Middle School Massacre" (Jonesboro) in The Commercial-Appeal. 
        Part of this difference in approach comes from a close identity between small-town journalists and the people they cover. Powell, in her The Columbia Basin Herald column, wrote that she reacted to the Moses Lake shooting as a community resident. "For us local media, the job of getting a story that we couldn't comprehend, about people we knew or worked with, seemed an insurmountable task. We wanted to know everything, too, not so much as a news reporter, but as a member of our community." (16)
        That sensitivity to the community was reflected elsewhere. In the Jonesboro shootings, reporter/photographer Curt Hodges didn't take photographs of victims shortly after the shooting, a reaction which John Troutt Jr., editor and owner of The Jonesboro Sun, said exhibited a typical sensitivity of his news staff (17). The paper won wide praise from Jonesboro citizens for its consideration shown to victims and families, its absence of sensationalism and accuracy. (18)

Discussion
        Results indicated that local, regional and national newspapers covered the school shootings in different ways. The evidence supported four of the seven hypotheses and two of the three expected frames. Looking at the hypotheses, the data offer solid support for the audience and community needs theory, and limited support for the organizational pressures theory. Overall, the results support the expectation that newspapers cover events of social conflict differently, depending on their community orientation.
The data supports the expectations that local newspapers would employ a tension management frame. This support was found in local newspapers having the most coverage and the most prominent placement (although not statistically significant), an indicator of the newspaper's attempt to meet the community and audience needs. The local newspapers also employed the tension management frame by providing information on where to get help and how to respond, emphasizing community cohesion and healing.
The societal frame was support as well. As expected, the national newspaper had the greatest emphasis and local newspapers the least on potential causes, blame or solutions to the shooting. As expected, the guns and gun culture were prominently mentioned as a part of the societal problems. Also as expected, the emphasis on the societal issues grew among all types of newspapers with each subsequent school shooting, but the national newspaper remained the most prominent in its societal issues coverage.
However, the data didn't support assertions that local newspapers would more frequently employ a crime frame. Instead of local newspapers mentioning suspects the most often, it was the national newspaper that focused more frequently on the suspect (on nine of 10 measures). None of the source items was statistically significant. Depiction of victims showed no significant differences, although the local newspaper had the most frequent mention in three of the seven humanizing victims measures.
The frame analysis offers stronger support for the audience and community needs explanation than the organizational pressures theory. As explained previously, local newspapers would respond to a perceived community and audience need by supplying information the community needs to respond to the traumatic event, while placing little emphasis on societal implications. The national newspaper, responding to a broader community of interest, would emphasize the societal implications and pay little attention to local information needs. The regional newspapers were expected to be somewhere between the local and national newspapers.
The data solidly supported the premise that community and audience needs influence how newspapers cover events. Local newspapers printed the most articles and gave the articles the most prominent placement, a reflection of the information needs of their readers, as noted in the tension management frame. National newspapers more frequently focused on societal ills to explain the shootings and emphasized the role of weapons used in the shootings, reflecting the audience's need for a broad societal interpretation and explanation, part of the societal frame. The local newspapers were more likely to emphasize community cohesion and healing, part of tension management and a response to the community's needs to maintain stability. However, an anticipated focus among local newspapers on the individuals involved in a crime event weren't found. Data backs up four of the six hypotheses supporting the community and audience needs theory.
Results offer little support for the organizational pressures theory, which expected the local newspapers to be more likely to portray the school shootings as a crime event. Data didn't show that local newspapers were more likely to focus on individuals, a key element of depicting the school shooting as a crime. Local newspapers, instead of being the most likely to name suspects, were the least likely. Local newspapers were only slightly more likely to name victims or present victims in humanizing ways. Instead of local newspapers being more likely to use law enforcement and other government sources, no significant differences were found.

Limitations
        While the results offer support for the overriding theory that newspaper coverage is a product of its community, several limitations deserve mention.
First is a wide spread in the numbers of stories among the three types of newspapers. Since these analyses were by frequency within stories, the differences in raw numbers (local newspapers accounted for 63% of the stories) could skew results. For instance, in looking at use of sources, on one day, the local newspaper may run 20 stories and use police sources in six (30%), the regional newspaper may run eight stories and use police sources in three (38%) and the national newspaper may run two stories and use police sources in one (50%). 
A second limitation is in use of the article as the unit of analysis. Thus, a fraction of a sentence and 2,000-word article devoted to a particular item would be counted equally. Analysis of the paragraph or sentence thus could result in different findings. 
A third potential limitation is with the sample, which drew one week's worth of newspaper articles. This gave newspapers little time to respond to the event. There may be different results by examining a longer coverage period. 
A fourth limitation may be in the newspapers used in the study. Perhaps the small number of newspapers used in the study (10) didn't provide an accurate view of journalism in the 1990s. Or perhaps The New York Times wasn't the best choice as the national newspaper. There may also have been significant differences in newspaper ownership, the sizes of the newspapers, or in the communities themselves.
        Another factor may be the nature of the events themselves. Perhaps there is something about youths as killers and/or victims that cause newspapers to frame the events differently than normal.
Footnotes

1. "Experts seek reasons for epidemic of youth violence," The Oregonian, 24 May 1998, sec. A, p. 
        14, col. 2.
2. Roper Center at University of Connecticut Public Opinion Online, Gallup Organization question 
ID USGALLUP.99AP26,R29 and R31, Lexis-Nexis, visited May 5, 1999.
3. Examined coverage in the New York Times, St. Louis Post-Dispatch, Washington Post, Atlanta 
Journal and Constitution, Los Angeles Times and USA Today. For each of the five 
shootings used in this study, one month of coverage in each newspaper was examined, and 
391 articles were found. The chronological progression: first incident, 1.8% of the total; 
second, 6.9%; third, 15.6%; fourth, 43.7%; and fifth, 32%
4. Four people coded 40 of the 602 articles. Questions met Cronbach's alpha of .8, except for 
questions in sections on cause/blame and solutions. Those two categories were compressed 
into one category for the actual coding.
5. Population estimates for 1996 from United States Census Bureau,         www.census.gov/population/estimates/metro-city
6. Newspaper circulation figures from Bacon's Newspaper Directory, 1997.
7. Moses Lake, Wash., population 13,984, 106 miles southwest of Spokane. Circulation: The 
Columbia Basin-Herald, 8,806 daily, no Sunday; The Spokesman-Review, 118,770 daily 
and 147,328 Sunday.
8. Pearl, Miss., population 21,175, located five miles east of Jackson, 213 miles south of Memphis. 
Circulation: The Clarion-Ledger, 107,876 daily and 127,393 Sunday; Memphis 
Commercial Appeal, 178,181 daily and 267,935 Sunday.
9. Paducah, Ky., population 26,601, located 173 miles southeast of St. Louis. Circulation: The 
Paducah Sun, 30,120 daily and 32,462 Sunday; The St. Louis Post-Dispatch, 323,374 
daily and 541,991 Sunday.
10. Jonesboro, Ark., population 52,656,  located 70 miles northwest of Memphis. Circulation: 
Jonesboro Sun, 26,965 daily and 29,530 Sunday; The Memphis Commercial Appeal, 
178,181 daily and 267,935 Sunday.
11. Springfield, Ore., population 49,430, located four miles southeast of Eugene and 110 miles 
south of Portland. Circulation: The Register-Guard, 75,881 daily and 79,428 Sunday; 
The Oregonian, 349,193 daily and 445,293 Sunday.
12. "Getting the story: Us versus them," Columbia Basin Herald, 8 February 1996, sec. 1, p. 4, 
        col. 3.
13. McLellan, Michele. "Coverage should have depth, restraint," The Oregonian, 24 May      1998, 
        sec. B, p. 1, col. 3 and p. 2, col. 1.
14. "Coverage should have depth, restraint," The Oregonian, 24 May 1998, sec. B, p. 1, col 3 
        and p. 2, col. 1.
15. "We need leadership to revise gun laws" and "Accountability, not ban key to cutting violence," 
The Oregonian, 22 May 1998, sec. D, p. 15, col. 1-3.
16. "Getting the story: Us versus them," Columbia Basin Herald, 8 February 1996, sec. 1, p. 4, 
        col. 3.
17. The Freedom Forum, Jonesboro: Were the Media Fair? (Arlington, Va.: The Freedom Forum, 
        1998), 5
18. The Freedom Forum, 13.

Bibliography

Bacon's Information, Inc. (1997). Bacon's Newspaper Directory. Chicago: Bacon's        Information, Inc.
Breed, Warren (1958) "Mass Communication and Socio-Cultural Integration," Social         Forces, 37: 109-116.
Donohue, George A., Olien, Clarice N. and Tichenor, Phillip .J. (1985) "Leader and    Editor Views of Role of Press in Community Development." Journalism     Quarterly, 62: 367-372.
Donohue, G.A., Olien, C.N. and Tichenor, P.J., (1972) "Gatekeeping: Mass Media
Systems and Information Control," in Kline, F. Gerald and Tichenor, Phillip J. (eds), Current Perspectives in Mass Communication Research. Beverly Hills, Calif.: Sage Publications.
Edelstein, Alex S. and Contris, Joseph J. (1966) "The Public View of the Weekly         Newspaper's Leadership Role." Journalism Quarterly, 43: 17-24.
Entman, Robert M. (1991). "Framing U.S. Coverage of International News: Contrasts      in Narratives of the KAL and Iran Air Incidents." Journal of Communication,         41 (4): 6-27.
Entman, Robert M. (1993). "Framing: Toward Clarification of a Fractured Paradigm,"     Journal of Communication, 43 (4): 51-58.
Fishman, Mark (1980). Manufacturing the News. Austin and London: University of     Texas Press.
The Freedom Forum (1998). Jonesboro: Were the Media Fair? Arlington, Va.: The    Freedom Foundation
Fuller, Jack (1996). News Values: Ideas for an Information Age. Chicago: University of
Chicago Press.
Gamson, W.A. and Modigliani, A., (1987). "The Changing Culture of Affirmative    Action." In Braungart, R.G. and Braungart, M.M. (eds) Research in Political   Sociology, v 3. Greenwich, Conn.: JAI Press, Inc.
Gans, Herbert J. (1979). Deciding What's News: A Study of CBS Evening News, NBC    Nightly News, Newsweek and Time. New York: Pantheon Books.
Gitlin, Todd (1995). The Twilight of Common Dreams: Why America is Wracked by     Culture Wars. New York: Henry Holt and Company.
Gitlin, Todd (1980). The Whole World is Watching: Mass Media in the Making &        Unmaking of the New Left. Berkley, Los Angeles, London: University of        California Press.
Graber, Doris A. "Media Impacts on the Political Status Quo: What is the Evidence?" in  Spitzer, Robert J. (ed) (1993). Media and Public Policy. Westport, Conn. and    London: Praeger.
Graber, Doris A. (1989). Mass Media and American Politics. Washington, D.C.:        Congressional Quarterly, Inc.
Herkov, Michael J. and Biernat, Monica (1997). "Perceptions of the Media in a      Community Exposed to Serial Murder," Journal of Clinical Psychology, 53, 8:
 909-915.
Keatley Garvey, Kathryn (1991). "Pack journalism invades rural community," Editor &       Publisher, 124, 8: 48.
Olien, Clarice N., Donohue, George A. and Tichenor, Phillip J. (1968) "The Community
Editor's Power and the Reporting of Conflict," Journalism Quarterly, 45: 243-252.
Olien, Clarice N., Donohue, George A. and Tichenor, Phillip J. (1992) "Conflict,
Consensus, and Public Opinion," in Theodore L. Glasser and Charles T. Salmon (eds) Public Opinion and the Communication of Consent. New York, London: The Guilford Press.
Parenti, Michael "News Media Bias and Class Control," in Margolis, Michael and
Mauser, Gary A. (eds) (1989) Manipulating Public Opinion. Pacific Grove, Calif.: Brooks/Cole Publishing Company.
Parenti, Michael (1986). Inventing Reality: The Politics of the Mass Media. New    York: St. Martin's Press.
Qualter, Terence H. "The Role of the Mass Media in Limiting the Public Agenda," in    Margolis, Michael and Mauser, Gary A. (eds) (1989) Manipulating Public    Opinion. Pacific Grove, Calif.: Brooks/Cole Publishing Company.
Romano, Carlin "The Grisly Truth about Bare Facts," in Manoff, Robert Karl and    Schudson, Michael (eds) (1986) Reading the News. New York: Pantheon Books.
Roshco, Bernard (1975). Newsmaking. Chicago and London: The University of Chicago        Press.
Scheufele, Dietram A. (1999), "Framing as a Theory of Media Effects," Journal of     Communication, 49 (1), 103-122. 
Schudson, Michael "Deadlines, Datelines and History" in Manoff, Robert Karl and    Schudson, Michael (eds) (1986) Reading the News. New York: Pantheon Books.
Shaw, Daron R. and Bartholomew H. Sparrow (1999). "From the Inner Ring Out: News
Congruence, Cue-Taking, and Campaign Coverage," Political Research 
Quarterly, 52: 232-351.
Sigal, Leon V. "Sources Make the News," in Manoff, Robert Karl and Schudson, Michael        (eds) (1986) Reading the News. New York: Pantheon Books.
Sigal, Leon V. (1973). Reporters and Officials: The Organization and Politics of    Newsmaking. Lexington, Mass. and London: D.C. Heath and Company.
Spitzer, Robert J. (ed) (1993). Media and Public Policy. Westport, Conn. and London:        Praeger.
United States Census Bureau, population estimates for 1996,        www.census.gov/population/estimates/metro-city
Viall, Elizabeth K. (1992). "Measuring Journalistic Values: A Cosmopolitan/Community         Continuum," Journal of Mass Media Ethics, 7: 41-53.
Wilkening, David (1992). "When the media circus came to town," Editors & Publisher,
125 (n10): 48.
Yahoo maps, maps.yahoo.com/py/dd/Results.py
TABLE 1
Article totals, by types of newspaper, for each school shooting, listed in chronological order.

Moses Lake
Pearl
Paducah
Jonesboro
Springfield
Totals
Local
4.7%
n=28
10.0%
n=60
11.0%
n=66
20.8%
n=125
17.1%
n=103
63.4%
n=382
Regional
3.0%
n=18
0.8%
n=5
1.7%
n=10
6.5%
n=39
15.1%
n=91
27.1%
n=163
National
0.2%
n=1
0.2%
n=1
1.2%
n=7
3.5%
n=21
4.5%
n=27
9.5%
n=57
Totals
7.8%
n=47
11.0%
n=66
13.8%
n=83
30.7%
n=185
36.7%
n=221
100%
n=602


TABLE 2
Frequency, in percentages, in which a suspect is mentioned, by type of newspaper.

National
Regional
Local
Total
P<
Suspect name
54.4
53.4
42.4
46.5
.05
Age
50.9
37.4
30.1
34.1
.01
First name in article
26.3
21.5
12.0
15.9
.01
Parents, relatives
15.8
11.7
11.8
12.1
NS
Violent threats
17.5
  9.2
  5.2
  7.5
.01
Appearance
14.0
  9.2
  5.2
  7.1
.05
Deviant behavior
14.0
  6.1
  6.6
  7.1
NS
Friends
  5.3
  7.4
  4.7
  5.5
NS
Hobbies, activities
19.3
  4.9
  3.7
  5.5
.01
Picked on
  7.0
  4.9
  2.1
  3.3
.10
n=57
n=163
n=382
n=602

NS: Not significant at p<.10



TABLE 3
Frequency, in percentages, in which a victim is mentioned, by type of newspaper.

National
Regional
Local
Total
P<
Victim name
36.8
41.1
42.7
41.7
NS
Age
33.3
27.6
23.6
25.6
NS
First name in article
15.8
15.3
18.8
17.6
NS
Parents, relatives
14.0
18.4
17.3
17.3
NS
Friends
14.0
16.0
14.4
14.8
NS
Hobbies, activities
12.3
12.3
12.8
12.6
NS
Deviance
  0.0
  0.0
  0.0
  0.0
NS
n=57
n=163
n=382
n=602

NS: Not significant at p<.10

TABLE 4
Frequency in which suggested motives, causes or solutions to the school shooting are mentioned, by newspaper type, in percentages.

National
Regional
Local
Total
p<
Guns, gun culture
19.3
18.4
  7.1
11.3
.01
Media (movies, TV) 
24.6
14.7
  7.1
10.8
.01
Schools, school security
  8.8
12.3
  9.2
10.0
NS
Counseling
  5.3
  9.8
10.2
  9.6
NS
Parents
15.8
14.7
  5.2
  8.8
.01
Government, laws
14.0
11.7
  6.5
  8.6
.05
Religion, prayer
  8.8
  4.3
  9.9
  8.3
.10
Gun control 
17.5
11.0
  2.9
  6.5
.01
Law enforcement 
  3.5
  8.0
  3.7
  4.8
.10
Dialogue 
  5.3
  6.1
  2.9
  4.0
NS
Gangs, cults 
10.5
  2.5
  2.4
  3.2
.01
Computers, video games 
  7.0
  4.9
  1.0
  2.7
.01
Incivility 
  0.0
  2.5
  2.9
  2.5
NS
Drugs, alcohol 
  7.0
  3.1
  1.3
  2.3
.05
Other 
24.6
26.4
14.1
18.4
.01
n=57
n=163
n=382
n=602

NS: Not significant at p<.10




TABLE 5
Frequency in which other school shootings, the latest shooting as part of a trend, other crimes, tragedies and disasters are mentioned, by type of newspaper, in percentages.

National
Regional
Local
Total
P<
Mention of other crimes
40.4
25.2
13.1
18.9
.01
Mention other school shootings
38.6
22.1
  9.9
15.9
.01
Latest shooting as part of trend
26.3
19.6
  5.8
11.5
.01
Mention other tragedies
  1.8
  1.2
  2.4
  2.0
NS
Mention other disasters
  0.0
  0.6
  1.6
  1.2
NS
n=57
n=163
n=382
n=602

NS: Not significant at p<.10


TABLE 6
Frequency in which suggested motives, causes or solutions to the school shooting are mentioned, by school shooting coverage period, in percentages.

Feb. 96
(Moses Lk)
Oct. 97
(Pearl)
Dec. 97
(Paducah)
Mar. 98
(Jonesboro)
May 98
(Springfield)

Total

P<
Guns, gun culture 
  0.0
  1.5
  2.4
11.4
19.9
11.3
.01
Media (movies, TV)
  8.5
  4.5
10.8
  7.0
16.3
10.8
.05
Schools, school security
12.8
  7.6
12.0
  7.6
  6.7
10.0
NS
Counseling 
10.6
  7.6
10.8
  7.6
11.3
  9.6
NS
Parents 
  8.5
  6.1
  7.2
11.9
  7.7
  8.8
NS
Government, laws 
  0.0
  3.0
  3.6
13.5
10.0
  8.6
.01
Religion, prayer 
  8.5
  7.6
19.3
  8.1
  4.5
  8.3
.01
Gun control 
  0.0
  0.0
  2.4
  4.9
12.7
  6.5
.01
Law enforcement 
  2.1
  3.0
  2.4
  0.5
10.4
  4.8
.01
Dialogue 
  6.4
  4.5
  6.0
  2.7
  3.6
  4.0
NS
Gangs, cults 
  0.0
  7.6
  2.4
  4.3
  1.8
  3.2
.10
Computers, video games
  2.1
  1.5
  0.0
  1.6
  5.0
  2.7
.10
Incivility 
  2.1
  3.0
  0.0
  3.2
  2.7
  2.5
NS
Drugs, alcohol 
  0.0
  3.0
  1.2
  1.1
  4.1
  2.3
NS
Other 
17.0
  7.6
16.9
15.7
24.9
18.4
.05
n=47
n=66
n=83
n=185
n=221
n=602

NS: Not significant at p<.10



TABLE 7
Comparison of other school shootings, the latest shooting as part of a trend, and other crimes, by school shooting coverage period, in percentages.

Feb. 96
(Moses Lk)
Oct. 97
(Pearl)
Dec. 97
(Paducah)
Mar. 98
(Jonesboro)
May 98
(Springfield)
Total
P<
Mention of other school shootings 
  4.3
  4.5
  8.4
14.6
25.8
15.9
.01
Latest shooting, as part of a trend 
  4.3
  7.7
  3.6
10.3
18.1
11.5
.01
Mention of 
other crimes 
10.6
25.8
16.9
13.0
24.4
18.9
.01
Mention of other tragedies
  0
  1.5
  2.4
  3.2
  1.4
  2.0
NS
Mention of natural disasters
  0
  0
  0
  2.2
  1.4
  1.2
NS
n=47
n=66
n=83
n=185
n=221
n = 602

NS: Not significant at p<.10

TABLE 8
Frequency of articles that provide information of potential help to readers, by newspaper type, in percentages.

National
Regional
Local
Total
P<
Places to get help 
0.0
2.5
8.6
6.1
.01
Advice to deal with problems 
0.0
5.5
5.8
5.1
NS
What to say, do for children
0.0
3.7
2.9
2.8
NS
Where to make donations
0.0
3.7
8.1
6.1
.05
n=57
n=163
n=382
n=602

NS: Not significant at p<.10