This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line,
send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").
An analysis of support and blame in the wake of two fatal shootings
Vincent F. Filak, Ph.D.
Assistant Professor, Ball State University
276 Art and Journalism
Muncie, IN 47306
Ph: (765) 285-8218
Fax: (765) 285-8248
Email: [log in to unmask]
Robert S. Pritchard, APR, Fellow PRSA
Captain, U.S. Navy (Ret.)
Assistant Professor, Ball State University
Office: (765) 285-9104
Fax: (765) 285-7997
E-mail: [log in to unmask]
Submitted for presentation in the Mass Communication and Society
Division for the Association for Education in Journalism and Mass
August 2005, San Antonio, Texas
An analysis of support and blame in the wake of two shooting deaths
Using Gilliam et al's (2000) theoretical framework of crime news as
script, this case study examined the response postings (n=389) left
on a newspaper's website regarding two fatal shootings. An analysis
of the postings found differences in placement of blame and support
based on whether the incident followed standard script patterns.
Postings regarding a shooting that fit the script were more likely to
attribute blame to the assailants as individuals and offer sympathy
to the victim's family. Postings regarding the shooting that didn't
follow the script were more likely to make broader attributions or
blame the victim. Theoretical and practical implications are discussed.
The line between entertainment and news continues to blur and media
products, especially news, tend with greater frequency to resemble
both genres (Johnson, 2004). Americans are increasingly unable to
distinguish among the various programming types. Is it entertainment,
tightly scripted, or is it news? The result is often that news
organizations, eager to fit their coverage into a mold individuals
are most familiar with, often take an episodic approach to their news
reports. These news reports then tend to take on the elements of a
television or movie narrative: predictable characters, dramatic
events and powerful emotions (Gilliam & Iyengar, 2000).
Nowhere is this scripting more prevalent than in the reporting of
crime news where both television (Gilliam, Valentino & Beckmann,
2002) and newspapers (Coleman & Thorson, 2002) rely heavily on
episodic or "horse-race coverage" that is more typically associated
with political news. From the first violent incident, the media
dutifully report each stage of the saga: the unfolding police
investigation, arrest of the suspect or suspects, in the case of
broadcast news, the "perp walk," and eventually the transition of
coverage to the courthouse, with opening statements, witness
testimony and all manner of courtroom intrigue. The typical saga
inevitably concludes with the jury's verdict and comment from the
families, both victim and criminal, serving as the denouement.
Researchers of this scripting approach to news coverage have noted a
number of deleterious effects on mass media consumers. By
consistently shaping the message to fit this news/entertainment
hybrid (Corso, 1992), audiences are learning more about stereotypes
than receiving news. News scripts often stereotype crime as violent
and random, with blacks often viewed as the perpetrators of those
crimes (Anderson, 2001). Furthermore, young criminals are typically
shown to be "super predators," a breed of violent thug that acts
without thought and lives without remorse (Gilliam & Iyengar, 1998).
In addition, illegal acts are viewed as individually perpetrated
actions that have no causal connection to societal issues (Coleman &
On occasion, an event occurs in which the script is changed and the
audience is forced to react to this change. We know from research
that, because of the structure of the human mind, individuals respond
best when incoming knowledge fits into a system of pre-existing
understanding (Rumehart & Ortony, 1977). In the case of a
well-scripted crime report, what we know and believe tends to fit
well together; we assign blame easily, dole out some perfunctory
sympathy and have a general sense that law enforcement efforts have
restored order. When the script is violated, prior knowledge becomes
far less useful in assessing new information and individuals actively
seek ways to make sense of the event.
This case study examines the responses of mass-media users following
the shooting deaths of two Ball State University students to compare
and contrast reader's responses when the scenario does and does not
follow the recognized format. One incident clearly follows the
stereotypical media crime script. A white student is robbed and
killed by three black men with whom he had no prior dealings. After
being arrested, the shooter said he killed the victim during the
robbery and a struggle, though the men had only intended to rob the
victim. One of the assailants also mentioned that his drug use led
him to rob and kill.
The second incident follows a portion of the script, but deviates in
several key ways. A white student is shot and killed in the backyard
of a home near campus. However, in this case, the shooter was a white
police officer. The student was unarmed and had mistakenly arrived at
the wrong house following a night of drinking. The officer, himself
only 24 years old, had not completed all his required training. He
was placed on paid administrative leave while an investigation was
conducted, but was never arrested and a grand jury refused to indict him.
Central to this study is the examination of blame and sympathy
afforded to the "players" involved in these scripts through a content
analysis of comments readers left on the Web site of the campus
newspaper, where the incidents were covered extensively.
We chose to examine these online posts because they afforded us a
unique opportunity to extend the research in this area. Studies
examining race and ethical reasoning (Coleman, 2003) and news as
script (Gilliam et al.,1996) have focused on print and broadcast as
the central media. While we expect that the use of online media would
not produce drastically different results, this study presents an
opportunity to test this supposition. Furthermore, online media users
are often viewed as more interactive and more educated (Kehoe, 1999)
than most traditional media users and might therefore deviate from
the stereotypical response seen elsewhere in research.
While various studies have examined the impact this type of coverage
has on individuals in the audience, much of the work has focused on
television news coverage (Campbell, 1995; Entman, 1992; Gilliam &
Iyengar, 2000). While it is generally agreed that television news is
the primary information source for most Americans (Papper, 2003), it
is also the medium that provides for the least amount of opportunity
for viewers to provide direct feedback. Newspapers often publish
letters to the editor. Web sites are beginning to host chat rooms or
message boards by which individuals can provide unsolicited reactions
to coverage or unfolding events, but television provides little
direct on-air time for this type of discussion. In this era of media
convergence, broadening the understanding of the effects of news as
script has an inherent value.
Most importantly, the individuals who have posted their thoughts in
this case study are reacting to a real event. Other studies rely on
recall or experimental lab situations in which individuals react to
hypothetical scenarios or news events that are unknown to them.
Individuals in this study have read actual news stories delivered to
them in the wake of the event and are issuing real-time
reactions. Their postings of feelings, thoughts and emotions are
genuine and speak to some of the issues raised by Gilliam and
colleagues. While previous research has set the stage extremely well,
this case study provides an opportunity to examine the theoretical
proposals in a more realistic setting.
In order to fully understand both the focus of this study and the
theoretical perspectives that provide the overarching framework for
it, we first need to outline the research pertaining to scripts,
schemas and other related theories. Then we will outline the scripts
that unfolded in each of the shooting incidents, thus allowing for an
assessment of the patterns in each script.
Scripts and local news coverage
Research into cognitive actions has revealed in a variety of arenas
that humans often rely on cues and images that prompt them to recall
previous encounters with past stimuli in order to properly react to
current experiences (Crockett, 1988). This natural desire to remain a
"cognitive miser," (Taylor, 1981) gives individuals the opportunity
to quickly review situations, make choices and move on without
expending unnecessary mental energy. It also allows an expeditious
facilitation of activities between individuals. Simple discussion of
"Hello, how are you? Good, how are you?" become rote actions, as
individuals understand the purpose of each request intuitively and
are allowed to complete the interaction quickly, without expending a
great deal of thought.
These "scripts" or schemas are based on stereotypical knowledge that
individuals acquire over time (Coleman, 2003). Individuals organize
knowledge and process situations based on prior interactions and move
forward based on expectations. Classical conditioning analyses, which
acted as the precursor to much of the latter schema theory studies,
outlined ways in which things like pictures of a sunny day can engage
positive affective states while negative or depressing images can
trigger exactly the opposite reaction in people (Petty & Priester, 1994).
Schank and Abelson (1977) found that a script can be recognized
within a few sentences, whereupon the individual's mind picks up the
story line and quickly comprehends what has taken place. The mind has
thus accessed the script and supplied the individual the responses
that are expected to occur subsequently. For example, if two
coworkers, Bill and Larry, pass each other in the hallway, Bill might
introduce an "acknowledgement script" that begins with, "Hey, Larry,
how are you doing today?" Larry's response is fairly standard and
perfunctory, since neither participant expects a protracted
discussion on the matter: "Good, how are you?" Bill then completes
the script with, "Doing fine, thanks." Very little information has
been exchanged, but the script is started and completed quickly and
meets the standard set of expectations of both participants. If
Larry breaks from this script, however, it becomes more difficult for
Bill to properly respond quickly. Therefore, if Larry didn't answer
with "Good, how are you?" and instead broke down crying that his wife
had left him, the script does not continue as planned. Bill's next
"line" no longer works in this script and Bill is forced to react to
a situation he's not ready for. He must then decide how to proceed
with this discussion by finding new cues for responses while trying
to forget about the responses he had expected to be able to use.
A number of studies (Entman, 1992; Gilliam, Iyengar & Simon, 1996;
Peffley, Shields & Williams, 1996) have found that the media rely on
these scripts to make news easier to comprehend. In studying
broadcast news reports of crime, Gilliam and Iyengar (2000) outlined
several key components in every media crime script: The anchor
outlines the crime that has recently occurred, the video takes the
viewers to the scene and then the anchor provides information about
the suspect, including identity and whether or not he remains "on the
loose." This type of coverage is standard fare for newscasts,
primarily because it contains elements common to story-telling
schemas. The story contains heroes (law enforcement), villains
(criminals) and victims. The temporal structure of the incident is
also common, as it has a beginning (a crime is committed), a middle
(police are working, victim is recovering, suspect is running) and an
end (police catch suspect, suspect is punished). The coverage is also
episodic in nature, moving from incident to incident.
In covering news in this fashion, journalists inadvertently create
other schemas for the viewing and reading public. Crime presented in
this fashion is viewed as violent and random (Gilliam et al., 1996).
Coverage of these events lacks context or a sense as to any
underlying rationale to explain why these events occur (Coleman &
Thorson, 2002). The crime is tied to individual actions, as opposed
to larger social forces (Coleman & Perlmutter, 2005).
Research into the "public health model" of journalism argues that
issues of violence should be approached in the same way journalists
approach diseases such as AIDS and cancer (Stevens, 1997). While this
model has successfully shifted discussions of traffic safety away
from random acts by individuals to systemic issues such as seatbelt
use and drunk driving laws, crime reporting remains tied to a far
A final piece of the script is that of the roles associated with
victims and assailants. Hacker (1995) argues that violent crime is
often painted in black and white, in which criminals are black and
the victims are white. He further explains that when Americans think
of "black crime" they think of violent acts, such as murder and
robbery as opposed to embezzlement or computer fraud. Race has become
a key component of the crime script, and this element of crime
coverage has become ingrained in the psyche of news users. A study by
Coleman and Thorson (2002) found that even when elements of race were
removed from stories, study participants overwhelmingly made
erroneous reports that painted crime in this manner. Peffley and
Hurwitz (1997) stated that one of the more largely held beliefs by
American whites is that blacks are "violent and aggressive," thus
further tying together the schemas of race and violent crime, even
though social reality doesn't mirror these images (Drummond, 1991).
In sum, individuals often process information based on schemas and
scripts that allow them to predict future events (Crockett, 1988) and
thus it prevents individuals from having to relearn everything each
time they engage in an activity. It would be nearly impossible to for
people to perform any more than a small series of basic actions if
they need to engage in extensive thought processes each time they got
dressed or saw a traffic signal. However, when people act as "lazy
organisms" (McGuire, 1969) in areas of social concern, incorrect
ideas and stereotypes can form and the scripts they rely on can be
problematic. Because scripts allow individuals to look beyond what
they have actually observed, they often have difficulty
differentiating between what they have experienced and what they
simply infer (Bower, Black & Turner, 1979). Therefore, repeatedly
relying on schemas for race (Anderson, 2001) and portrayals of
poverty (Gilens, 1996) can create oversimplified and incorrect views
of those individuals seek to categorize.
More than 100 stories were published by the Ball State Daily News,
the Muncie Star Press, the Indianapolis Star and the Associated Press
about the shooting deaths of Michael McKinney and Karl Harford. These
articles, garnered through searches of each newspaper's website and a
Lexis/Nexis search for "Michael McKinney" and "Karl Harford,"
provided the following information about each incident.
McKinney incident: In the early morning hours of Nov. 8, 2003, rookie
Ball State University police officer Robert Duplain shot and killed a
21-year-old student while investigating a possible burglary in an
off-campus neighborhood near the University.
The student, Michael S. McKinney, a junior marketing major from
Bedford, a small town in southern Indiana, was returning home from an
evening of drinking and mistook the back door he was pounding on for
a friend's home seven doors down.
Duplain arrived at the house southwest of the campus at
approximately 3:30 a.m. He was the first police office to respond to
the call in a neighborhood where Muncie and campus police share
jurisdiction. The officer and witnesses said McKinney lunged at the
officer from the home's back porch. After issuing several verbal
warnings, Duplain fired three shots into McKinney's torso. A fourth
bullet struck the student in the face. McKinney was pronounced dead
that Ball Memorial Hospital about an hour later.
Duplain was immediately placed on administrative leave with pay
pending the outcome of several investigations. The 24-year-old
graduate of Taylor University in Upland, Indiana, another small town
about 20 miles north of Muncie, was one of 31 sworn officers on the
campus police force at the time. He had been on the force about
seven months and had completed basic firearms and a law course
offered by the Indiana Law Enforcement Training Board along with 14
weeks of field training with University police. He had not yet
attended the Indiana Law Enforcement Academy, but was slated to
report in January, about two months after the shooting.
Friends of McKinney who had been with him hours prior to his death
stated that they had been drinking that day. McKinney had planned to
sleep at a friend's house that was within walking distance of "The
Village," an area of Muncie near Ball State that had a number of bars
and college hang outs. A blood test taken 10 hours after his death
found McKinney had a blood-alcohol content of 0.34 percent, more than
four times higher than the state's legal limit of 0.08 percent to
drive. The subject of McKinney's drinking habits was noted in a
large number of articles, including comments from his friends that
he'd been drinking all evening and was probably confused.
A Grand Jury, FBI investigation and internal Ball State investigation
found no evidence of wrong-doing by Duplain in the shooting. He has
since completed his training at the Indiana Law Enforcement Academy
and returned to the University police force.
The university announced significant changes to its new officer
training protocol in December following the shooting. The protocol
now requires officers who are not academy-trained to patrol with a
more experienced officer until they attend the Indiana Law
Enforcement Academy. If the officer completes field training before
space is available in the academy, the officer will respond to calls
with an experienced officer.
Field training was also expanded from 14 to 21 weeks to include two
weeks of core training, two weeks of investigation training, 15 weeks
of work on all three shifts, and two weeks of work on the first
assigned shift. During field training, all officers are trained in
the use of all equipment issued or approved by the police department,
including chemical defensive spray.
Police Chief Gene Burton was quoted often in the coverage of this
incident, stating that the department was reviewing its policies and
needed to rebuild trust. His media responses were generally
characterized by providing as much information as he was legally able
Harford incident: Approximately four months later, 20-year-old Karl
Harford, a sophomore from Carmel, Indiana, was found dead in an alley
behind an eastside Muncie home. Harford was the victim of a single
gunshot wound to the head and was found in the back seat of his car
Sunday afternoon, March 6, 2004. He was last seen leaving a party
near the Ball State University Campus.
The apparent motive of the murder was robbery. The killers got away
with $2, all the cash Harford had on him at the time.
Two suspects were arrested almost immediately in connection with the
shooting. Muncie resident Brandon Patterson, 18, was arrested at
7:15 p.m., Monday, March 7, 2004, and charged with the murder and
armed robbery of Ball State sophomore Karl Harford. A 14-year-old
male accomplice was arrested at the same time and charged with
assisting a criminal.
On Tuesday, details on the third suspect, Muncie resident Damien
Blaine Sanders, 21, were released. Sanders was subsequently arrested
on March 18 for murder and armed robbery. Bridget Brannon, 25, was
also charged with aiding a fugitive and faces up to four years in
jail. Muncie Police Chief Joe Winkle said Brannon helped Sanders
elude police for almost two weeks after Harford's death. He said
investigators believe she drove him to Louisville, Ky., and then
helped him rent a room at a Days Inn in Indianapolis.
Harford apparently met the three at a near-campus party and offered
them a ride home. According to an affidavit, Patterson told police
the three planned to rob Harford in his car. Sanders claimed Harford
was shot after a struggle.
Toxicology reports provided as part of his autopsy showed that
Harford had a blood-alcohol content of 0.16 percent, twice the legal
limit for driving in Indiana. Experts would note that Harford's
judgment was impaired by the alcohol at the time of the murder.
Sanders subsequently plead guilty to murder and armed robbery and,
following a plea agreement, was sentenced to 85 years in prison
during a short trial in March of 1995. Patterson is scheduled for
trial in May 2005. Authorities have charged the then 14-year-old
suspect with two counts of theft. While prosecutors dropped murder
charges against the youth in May 2004, they have indicated they may
yet charge him with murder after Sanders' and Patterson's trials.
Branson is scheduled for trial for assisting a criminal in May 2005, as well.
In these incidents, several elements fit within the script as
outlined. Both McKinney and Harford were white. The acts were both
violent and random, senseless and unforeseen incidents that ended
with death. In Harford's case, the script remains typical. The
assailants were black, unsavory characters who lived in the "bad
part" of town and had previous encounters with the police. Their
acts of violence were senseless; a robbery that turned into murder
for $2. At his sentencing, Sanders stated that he was addicted to
drugs and robbed to feed his habit. Harford fit the role of victim
and Sanders, Patterson and their juvenile partner were stereotypical
criminals. Sanders pled guilty to murder and armed robbery and
received an 85-year prison sentence. Patterson's trial is expected to
begin in mid 2005. While the charges against the juvenile remain
somewhat uncertain, his fate is expected to be decided by the end of
the 2005 summer. The episodic nature of the case fits the pattern of
the scripts outlined in previous research: Harford was killed, the
police determined who was responsible and arrested them, the guilty
individuals resemble truly "bad characters" and are being punished;
the case is being closed.
In McKinney's case, however, the script deviates strongly and sharply
from the norm. Duplain was not only white, but also a police officer,
which brings the expectation of protection from death, not death
itself. The role of the suspect in the McKinney case does not fit
with the schema that most individuals have come to associate with an
assailant. The police were not required to seek him out as a suspect
and, in the end, he was not tried, convicted or punished. He remains
on duty as a member of the University Police Department. The race and
the individuals connected with the case do not fit the standard
pattern of crime either.
The purpose of this study, then, is to examine individual reactions
to these two crime events: one that followed the script outlined by
Gilliam and colleagues and one that violates the basic tenets of that
script, namely race, roles and incident progression. This study poses
the following hypotheses:
H1: Placement of blame on the participants in these incidents will
significantly differ between postings made regarding the shooting
deaths of Karl Harford and Michael McKinney.
H2: Offers of support toward the participants in these incidents will
significantly differ between postings made regarding the shooting
deaths of Karl Harford and Michael McKinney.
This study examined the online feedback posts that were written in
response to articles written about the shooting deaths of Michael
McKinney and Karl Harford. An Internet archive search of the Ball
State Daily News, the campus newspaper for Ball State, was used to
obtain the articles used in this study. We chose to use the Daily
News because it provided the most extensive coverage on this issue,
when compared with the Muncie Star Press, the city paper of Muncie,
which was the only other media outlet directly tied to the Ball State
area. Furthermore, the website has a response function at the bottom
of each story, which allows readers to post their opinions about the
articles they've just read. When individuals post to the website,
their comments are added as a link at the bottom of the story, thus
allowing other readers to click on the link and view their comments.
This set up also provides the possibility of a continual discussion
about a given topic, much like discussion boards or online bulletin
Two searches were conducted, one for "Michael McKinney" and the other
for "Karl Harford." Postings that were attached to articles that
pertained to the shooting deaths were retained for examination.
We selected a period of time for each of the cases that spanned from
the shooting through to the point where an initial court hearing took
place. This allowed us to view a congruent script for each incident
that focused directly on the issue of crime. For the McKinney
incident, it began with the shooting and continued through the
coverage of a grand jury's unwillingness to indict the officer
accused of killing him. For the Harford incident, our examination
period began with the shooting and continued through a story of the
assailants pleading "not guilty" to the charges of murder. This gave
us a total of 35 stories with postings (21 McKinney, 14 Harford) and
a total of 389 posts (n=389).
The posts were examined by two independent coders and categorized
based on the subject of the article, the author of the post as well
as several other key elements. In assessing the subject variable,
articles were coded as being either about Michael McKinney or Karl
Harford. In assessing authorship, the coders examined any notations
the posters made as to connections they had to the incident. Often,
posters declared not only their position, but also group membership
or social affiliation (e.g. "I am a friend of Michael McKinney…"). To
better understand who was posting, we asked the coders to categorize
the posts as either coming from the family of the victim, the friends
of the victim, students at Ball State, alumni from Ball State,
faculty or staff from Ball State, family of others connected to the
campus (family of students etc.), friends connected to the campus
(friend of Ball State students), those connected to the assailant,
unknown and other. Coders were asked to track all "other" posts to
determine if any other categories would emerge. One additional
category did emerge: law enforcement officials. In neither of the
cases did an assailant speak out on the website.
The posts were then coded for any instances of support and blame.
Support was operationalized as sympathetic proclamations (e.g. "Our
hearts and prayers go out to the family.") and offers of positive
grieving including hyperbolic affirmations (e.g. "Rest in peace" or
"You were the best friend anyone could ever have;" Nager & deVries,
2000). Nager and deVries also include issues of "deserving" in these
cases. Positive affirmations, such as "You didn't deserve to die" are
coded here, where as negative emotions of this type (e.g. "Those
cold-blooded murderers should pay.") were coded in the blame
category. Other attributions coded as blame included finding fault
with an individual or a group and associating negative attributes
with a participant that sought to explain a poor outcome (e.g. "He
was way too drunk…" or "He was an untrained cop").
In dealing with both variables, the coders used the same seven
categories to assess the presence and object of either emotion:
victim, victim's family, assailant, victim's group, assailant's
group, other or none. To be coded as a victim, the post must have
directly mentioned the victim by name or with a specific reference
("that poor boy"). Victim's family requires a mention of family or a
specific family member ("My heart goes out to his parents…"). The
assailant category required a specific mention of the shooter or
those directly involved in the killing. The assailant could be
mentioned by name or indirectly (for example "that cop" or "that
thug"). The two "group" categories include references to social
categories (Turner, 1999) in which the victim and assailant belong.
This use of group identifiers interested us because episodic coverage
the follows that traditional crime script often produces attributions
that rest with the individual, while coverage that breaks from
tradition often leads to more systemic attributions (Coleman &
We used Cohen's Kappa to assess intercoder reliability due to its
applicability for variables that categorize responses across multiple
categories. Subject (1.0), author (.80), sympathy (.76) and blame
(.78) were all variables that qualified as substantial, as they were
above a level of .6. (Stemler, 2004). Discrepancies among coders were
evaluated and recoded by the researcher.
Descriptive statistics: Readers left a total of 389 postings on the
articles we examined, 238 on articles that dealt with McKinney and
151 on those that dealt with Harford. McKinney articles averaged 15.5
postings while Harford articles averaged 10. The most postings issued
on a McKinney article was 62 while 31 was the top number of posts for
a Harford piece.
The authors of the post ranged from family members of the victim to
those individuals who knew the assailant. The largest number of posts
that could be identified came from Ball State students who had seen
the article online and had weighed in with their opinions (See Table
1 for a full breakdown of the authors who posted). Due to the large
number of unknown posters and the number of cells that lacked
adequate numbers for a strong statistical analysis, we present these
numbers for descriptive purposes only.
Blame and sympathy analyses: These analyses examined whether
significant differences existed between the Harford and McKinney
shootings in regard to where individuals placed blame. A Chi-square
analysis of the posts (n= 168) was significant in regard to the
placement of blame (_2 = 15.6, p < .01), thus demonstrating
differences between the two cases. An examination of the cells that
contributed most heavily to this statistic found that the largest
difference was in the blame levied against the victim. While two
posts placed any blame on Harford or his family, there were 28 posts
that blamed McKinney for his own death and two that placed the blame
on his family. (The posts that blamed his family discussed what the
posters considered to be an unwillingness or inability on the part of
McKinney's family members to discuss alcohol use and abuse with their
son.) While blame was most heavily placed on the assailant in each
case, significantly more blame than was expected was placed on the
assailant in the case of Harford.
A final interesting finding in regard to this variable came from the
analysis of group blame in each incident. Fewer posts issued blame to
a collective or societal force in the case of the Harford shooting,
while a larger number of posts did so in the McKinney posts. A deeper
reading of the posts shows that posters were questioning the police
department's rules of force and how it could put a rookie officer in
a situation in which he might need to draw his gun. The Harford posts
that were categorized here looked at the "town/gown divide" but never
looked at the issue of poverty or drugs, which were discussed in the
articles that dealt with the arrest of the assailants. This bolsters
the propositions set forth by Gilliam and Iyengar and their
colleagues that crime according to news script is riddled with random
acts that have no way of being predicted and have no underlying causes.
Since larger matricies often result in significant Chi squares, we
collapsed the cells into three categories: Victim affiliates,
assailant affiliates and other. Again, the Chi square was significant
(_2 = 10.3 p < .01) with the cells that contributed most to the
statistic being those in which blame was conferred upon the victim.
Hypothesis 1 was supported.
Finally, we analyzed the posts for professions of support (n=217). A
Chi square again was significant (_2 = 31.6, p < .001) and a small
number of cells provided the majority of the differences. The most
interesting of these was in the category of assailant. No poster
offered any sympathy or support to the assailants or anyone connected
with them when it came to the Harford shooting. However, 29 posts
offered sympathy to the assailant in the McKinney case and two
additional posts offered sympathy to the police in general. In the
McKinney posts offering support to the police officer, statements
such as "the officer will have to live with this for the rest of his
life" accompanied sympathetic statements, "I'm praying for him." No
such statements were made about the assailants. Furthermore, far more
people posted sympathetic messages for the Harford family than did
the McKinney family.
Again, we collapsed the cells to verify the statistics and found a
significant result (_2 = 25.3, p < .001). Assailant affiliates was
the cell that contributed most heavily to these differences, with no
sympathy extended in the Harford case and 31 postings doing so in the
McKinney case. Hypothesis 2 was supported.
Discussion and conclusion
This case study examined a series of internet posting written in
response to articles related to the shooting deaths of Michael
McKinney and Karl Harford, two Ball State students. We used the "news
as script" framework established by Gilliam and colleagues and rooted
in schema theory to assess how individuals would react when the
incident did and did not follow the script. In examining the
postings, we found a number of differences in how readers reacted to each case.
In the case of Harford, the news followed the script. A white student
was randomly selected as a victim and brutally killed by three black
youths who appeared to offer no sense of remorse. The assailants were
truly "bad characters," their senseless act of murder netting them
all of $2. Individuals who responded to these articles offered far
more sympathy to the Harford family, no sympathy to the assailants
and wrote a large number of posts that placed blame on the
individuals charged with Harford's death.
The shooting of McKinney failed to follow the script in several
important ways. Again, a white student was killed, but in this case,
the assailant was a white police officer. While the act was random
and senseless, it appears far less so because of the cues or images
associated with police officers. As such, the incident produced far
less sympathy for the McKinney family, while offering sympathetic
messages to the assailant, a police officer. The posts also placed a
good deal of blame on McKinney himself and on his family.
Interestingly, no blame was ascribed to Harford in any of the posting
related to his death for drinking or having impaired judgment because
of drinking (his BAC was 0.16, twice the legal limit), while
extensive blame was cast on McKinney for being drunk (his BAC was
0.34, more than four times the legal limit, although he did not
drive). This tends to strengthen the findings of previous research,
specifically those of Bower, Black & Turner (1979) that scripts are
likely to cause people to make inferences; in this case that Harford
was an "innocent victim." That would also appear to be supported in
the blame ascribed to McKinney for being impaired and creating the
conditions that led to his death because of the extensive mention of
his behavior that day.
The variations on blame were interesting to us specifically because
both incidents involved the use of alcohol. Although it was far more
prominently discussed by posters in McKinney's case, given the
current mood of our society, we would have expected blame to be
equally apportioned. While McKinney walked home, Harford got behind
the wheel of a car while having a blood-alcohol content twice the
legal limit to drive. Yet posters seem to have given Harford a "pass"
on his abuse of alcohol.
Likewise, from a very early age, children in our society are
instructed to avoid strangers. Yet, oddly, there were no strong
negative statements about Harford giving a ride to three strangers.
Those posters who did mention the fact viewed it as a positive
affirmation of his character, stating that he was a "good guy" who
gave the trio a ride home. The hand full of posts that did attempt to
shift some blame onto Harford were sharply rebuked by subsequent
posters, and dissension on this matter did not arise again.
While we did not include an examination of the articles written on
these incidents in our analyses, as it was not our intent to discuss
the media's coverage of the incidents, a quick review provides some
interesting fodder for discussion. Episodic coverage dominated the
media's examination of both shootings. Of the 35 articles we
examined, 16 of the McKinney articles and 12 of the Harford articles
dealt with breaking news, such as the shooting itself, the arrest of
suspects and the release of the victims' toxicology reports. Of the
remaining seven stories, three were personality profiles of the
victim and one was a profile of an assailant, Officer Robert Duplain.
The remaining three stories were broader looks at overarching
concerns, such as the house-party culture on college campuses that
allowed Harford to meet the men who killed him or the University
Police Department's use of force policy, which did not require
officers to enact non-lethal methods before turning to deadly force.
Previous research has demonstrated that the media reports crime in an
episodic format. While that was not our intent here, it appears that
our work can offer support to that claim.
The value in this study is that it offers real-world support to a
continuing discussion about how individuals react to news based on
schemas. Previous work has established that mass-media users who
receive crime news as script are likely to make individuated
attributions of blame, ignoring societal or systemic causes for the
incident. This case study examines real reactions by audience members
to incidents in their community without the intrusion usually
associated with experimental manipulations. While we are not able to
predict causal relationships, we have been able to expand the
discussion of reactions to news as script and offer support to those
researchers who have studied these phenomena in more controlled environment.
Finally, this piece has practical implications for both journalists
and mass-media users. The repetitious nature of crime as script has
forged in the minds of media users schemas about crime that are often
distorted, especially when dealing with race, roles and underlying
causes. When crime deviates from a script and media outlets seek
answers that go deeper than assigning blame to an individual as a
root cause, better attempts can be made in seeking to solve systemic
problems. In the case of the McKinney shooting, the script deviated
from the norm and the media asked questions pertaining to larger
issues, such as the department's use of force.
Although the coverage of this topic was not exhaustive, it
nonetheless sought to examine issues that went beyond, as one poster
put it, the actions of "a trigger-happy rookie cop." Regardless
whether this discussion was prompted by the media coverage or not,
the fact remains that the campus police did review policies and look
for ways to improve their response to crime. If this approach to
reporting crime could be applied to violent crime as a whole, as
Stevens (1997) suggests, perhaps more discussion about the societal
forces that create the crime could begin.
Anderson, L. (2001). Presumed guilty: Or, I thought this was the
"Movie of the Week." The Black Scholar, 24, 4, 40-42.
Bower, G., Black, J. & Turner, T. (1979). Scripts in memory for text.
Cognitive Psychology, 11, 177-220.
Campbell, C. (1995). Race, myth and the news. Thousand Oaks, Calif.:
Coleman, R. (2003). Race and ethical reasoning: The importance of
race to journalistic decision making. Journalism and Mass
Communication Quarterly, 80, 2, 295-310.
Coleman, R. & Perlmutter, D. (2005). "Bullets as bacteria:"
Television news magazines' use of the public health model for
reporting violence. Journalism, 6, 1, 24-
Coleman, R. & Thorson, E. (2002). The effects of news stories that
put crime and violence into context: Testing the public health model
of reporting. Journal of Health Communication, 7, 5, 401-425.
Corso, P. (1992). Fact or fiction? Only your TV network knows. USA
Today Magazine, 121, 2570, 82-84.
Crockett, W. (1988). Schemas, affect and communication. In L Donohew,
H. Sypher & E Higgins (eds.) Communication, social cognition and
affect. Hillsdale, N.J.: Lawrence Erlbaum Associates.
Drummond, W. (1991). Not necessarily the news. Wilson Quarterly, 15,
Entman, R. (1992). Blacks in the news: Television, modern racism and
cultural change. Journalism Quarterly, 69, 2, 341-361.
Gilens, M. (1996). Race and poverty in America: Public misperceptions
and the American news media. Public Opinion Quarterly, 60, 4, 515-541.
Gilliam, F. & Iyengar, S. (1998). The superpredator script. Nieman
Reports. 52, 4, 45-46.
Gilliam, F. & Iyengar, S. (2000). Prime suspects: The influence of
local television news on the viewing public. American Journal of
Political Science, 44, 3, 560-573.
Gilliam, F., Iyengar, S. Simon, A. & Wright, O. (1996) Crime in black
and white: The violent, scary world of local news. Harvard
International Journal on Press/Politics. 1, 1, 6-23.
Gilliam, F., Valentino, N. & Beckmann, M. (2002). Where you live and
what you watch: The impact of racial proximity and local television
news on attitudes about race and crime. Political Research Quarterly,
55, 4, 755-780.
Hacker, A. (1995). Two nations: Black and white, separate, hostile
and unequal. 2nd ed. New York: Ballantine.
Hurwitz, J. & Peffley, M. (1997). Public perceptions of race and
crime: The role of racial stereotypes. American Journal of Political
Science, 41, 2, 375-401.
Johnson, P. (2004, Jan. 26). Worlds of politics, comedy converge. USA
Kehoe, C. Pitkow, J., Sutton, K., Aggarwal, G. & Rogers, J. (1999).
GVU's Tenth Annual World Wide Web user survey. Accessed 3/10/05
McGuire, W. (1969). The nature of attitudes and attitude change. In
G. Lindzey & E Aronson (eds.), Handbook of social psychology, 2nd
edition. Reading Mass.: Addison-Wesley.
Papper, R. (2003). 2003 local television news study of new directors
and the American public. RTNDF.
Peffley, M., Shields, T. & Williams, B. (1996). The intersection of
race and crime in television news stories: An experimental study.
Political Communication, 13,
Petty, R. & Priester, J. (1994). Mass media attitude change:
Implications of the Elaboration Likelihood Model of persuasion. In J.
Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and
research. Hillsdale, NJ: Erlbaum.
Rumelhart, D. & Ortony, A. (1977). The representations of knowledge
in memory. In R. Anderson, J. Spiro & W. Montague (Eds.) Schooling
and the acquisition of knowledge. Hillsdale, N.J.: Erlbaum.
Schank, R. & Abelson, R. (1977). Scripts: plans, goals and
understanding. Mahwah, N.J.: Lawrence Erlbaum Associates.
Stemler, S. (2004). A comparison of consensus, consistency, and
measurement approaches to estimating interrater reliability.
Practical Assessment, Research & Evaluation, 9(4).
Stevens, J. (1997). Reporting on violence: A handbook for
journalists. Berkeley, Calif.: Berkeley Media Studies Group.
Taylor, S. (1981). The interface of cognitive and social psychology.
In J. Harvey (ed.) Cognition, social behavior and the environment.
Hillsdale, N.J.: Lawrence Erlbaum Associates.
Turner, J. (1999). Some current issues in research on social
identity and self-categorization theories in N. Ellemers R. Spears &
Doosje (Eds). Social identity: Context commitment content. Oxford
Table 1: List of authors who posted on Harford and McKinney stories
Table 2: Offers of support by online posters in connection with the
Harford and McKinney shootings.
Table 3: Placement of blame by online posters in connection with the
Harford and McKinney shootings.