THE FIRST HOURS OF SEPTEMBER 11th : How Accuracy and Sourcing Fared in
Three Television Networks' Breaking News Coverage
By
Scott Abel
School of Media
Concordia International University Estonia
Kaluri tee 5
Haabneeme, Viimsi vald
74001 Harjumaa, Estonia
(+372) 527-7515
[log in to unmask]
&
Andrea Miller
School of Journalism
University of Missouri
10 Neff Hall
Columbia, MO 65211
(573) 256-8050
[log in to unmask]
&
Vincent F. Filak
School of Journalism
University of Missouri
10 Neff Hall
Columbia, MO 65211
(573) 447-0513
[log in to unmask]
Paper submitted to the Mass Communication & Society Division, AEJMC
Kansas City; July 31-August 2, 2003
THE FIRST HOURS OF SEPTEMBER 11th
9
THE FIRST HOURS OF SEPTEMBER 11th: How Accuracy and Sourcing Fared in
Three Television Networks' Breaking News Coverage
INTRODUCTION
Breaking news is a generally unstudied genre in television mass
communication research. The extant research on breaking news, or when a
current story or regular programming is interrupted for a story that must
be broadcast immediately (Miller, 2002), has focused on the newsgathering
process (Berkowitz, 1992). Criticisms of breaking news coverage after the
fact are common in trade and industry publications. However, few scholarly
studies, if any, exist that examine the breaking news product as the unit
of analysis.
Breaking news has grown exponentially in the past decade due to
competition and technological advances. The O.J. Simpson verdict, the
Oklahoma City bombing, and the final 2000 presidential election results
were brought to viewers live via breaking news coverage. The unprecedented
terrorist attacks of September 11th, 2001, have once again brought the
importance of television breaking news to the forefront. Billions of people
worldwide watched the events unfold live. Breaking news is valued
economically because when tragedies such as this happen on a national
level, continuous breaking news coverage significantly increases ratings
(Katz, 1999). It is important then to study this genre because of its
proliferation, saliency among viewers, and perceived importance as an
audience builder.
Timeliness is an important ingredient in news, especially
television news. However, breaking news coverage is often criticized for
emphasizing speed over accuracy (Seib, 2001). On September 11th, the events
and information changed minute by minute. In the case of such an important,
impactful news event, should emphasis on immediacy override the commitment
to accuracy?
This study seeks to use these tragic events to explore the untapped area of
breaking news by asking the following questions about the first hours of
September 11th: What was reported in the first hours? Was the information
accurate? Who and what were the sources of the information? Television
remains the most influential news venue during breaking news (Merli, 2000)
and the news audience grows considerably larger when a major story is
breaking (Seib, 2001). The September 11th attacks provide an excellent case
for studying television news accuracy and sourcing in three American
television broadcast networks' coverage of an American tragedy.
LITERATURE REVIEW
Accuracy
Patterson and Wilkins (2002) define accuracy as using the correct
facts and the right words and putting things in context. Despite one
study's finding that as many as one-third of the stories on TV news have
inaccuracies, only a sixth of the stations run corrections as often as one
a month (Smith, 1999). Many news directors believe admitting mistakes might
cause viewers to lose confidence in their news (Cremedas,
1992). Television's "quick-draw" on election night 2000 raised even more
questions about accuracy and credibility (Hickey, 2001). A correction was
made immediately and apologies were made for months after a similar
erroneous report in the San Antonio, Texas television market in November of
1999 (Pompilio, 2000). Just over six months after Columbine, two
television stations and one radio station broke into programming to report
a school shooting without confirming the story. The story was false.
Accuracy was sacrificed in order to accelerate the delivery of the story.
Context is an important element of accuracy. Context includes
couching the information if it is not yet confirmed. Disclaimers are
important in breaking news reporting because the coverage is in real-time
when facts change constantly. Breaking news often leaves journalists with
difficult choices. Judgments must be made to report relevant information
without being alarmist.
With the aid of the construction of a timeline of the facts and
information that were reported, we hope to answer the following research
questions concerning the first hours of September 11th coverage:
RQ1: What information was reported on the networks (ABC, CBS, &
NBC) and was
it accurate?
RQ2: Was information not yet verified accompanied by a disclaimer,
helping to place
it in proper context?
TV NEWS SOURCES
Herbert Gans (1980) defined news as information transmitted from
sources to audiences with the journalist as middleman – summarizing,
refining and altering that information. In this study, a similar
definition is used. Sources are individuals and organizations that
journalists depend on for information but may not actually be named or
appear in the news. In breaking news coverage, sources are often live or
quoted rather than being on videotape, which is common in traditional
newscasts. The journalist-source relationship has long been studied, but
not in the context of breaking news. This study seeks to fill this void.
In this study, we began with two main categories of sources,
official and non-officials. Berkowitz (1987) found a high reliance on
official and "affiliated" sources, although approximately 25-percent were
not identifiable by affiliation. We further categorized our sources by
named or anonymous and then identified the sources specific affiliation,
for example Pentagon or other media outlet.
Organizational Sources
Institutional or organizational sources are official sources. Like
the print media, television news in general relies heavily on
organizational sources. In particular, the networks have been oriented to
Washington or governmental sources, as would be the case in a national news
event. One of the most consistently replicated findings in American
journalism research is that government sources (state, local and national)
tend to dominate news coverage of national and international affairs (Gans,
1979; Hackett, 1985; Soley, 1989).
In this study, it is apparent that governmental sources will likely
dominate because the breaking news story is an issue of national security
and such sources in this domain would seem "natural" and inevitable
(Hallin, 1993). Other organizational sources coded in this study include
airlines (stakeholders and key players in this story) and other media
outlets such as the Associated Press. Altheide (1976) suggests that as much
as 70% of network news comes from wire services. Similar to a 1994 study by
Reese, Grant, and Danielian, we not only consider organizational sources
but also three other important categories that are not officially,
organizationally affiliated: experts, eyewitnesses, and journalists
themselves.
Experts
Former government officials or "experts" in particular fields such as
terrorism are likely to be used in breaking news situations for "color
commentary." This group of sources has enormous knowledge and experience
and is likely to be former government officials. One important aspect of
these sources is their perceived independence from the policies of the
day. Therefore, they may have more freedom to speak on the record and
presumably more frankly about the current situation than those currently
employed by the government (Hallin, 1993). Yet others argue such experts
only give the illusion of objectivity. When experts appear on news
programs to interpret unfolding events, they are often asked not only to
explain the policies and operations of certain organizations, but also to
make predictions (Steele, 1995). Soliciting predictions or speculation
undermines objectivity. Steele also argues these former public officials
often have difficulty separating themselves from the political positions
they once held.
Witnesses
Witnesses are also included in our sourcing analysis. They fall under
Gans' (1979) "unknowns" category where "ordinary people" come into news
only in extraordinary circumstances (pg. 15). In breaking news situations,
witnesses are an important element in telling the story. They hold key
information simply because they are eyewitnesses or are directly affected
by the news event. In the hours, days, and months after Columbine, there
was an endless parade of teen witnesses on news programs telling their
stories. Witnesses manifest what is touted as another strength of TV:
"broadcasting humanizes the story" (Brooks, Kennedy, Moen, & Ranly, 1999,
pg. 443). Once again, the September 11th attacks are thus an excellent
case for examining television news sourcing with its mix of government and
eyewitness sources.
Veiled Attribution
Another dimension of the use of sources is the frequency with which
they are cited anonymously (Culbertson, 1978). Unnamed or "veiled"
attribution not only fails to let readers know who it is, but how often a
given person is quoted and therefore, how important he or she is to the
total story (Culbertson, 1978). It could also reflect the importance of the
source to the reporter. The ability to insist on anonymity is often cited
as evidence of reporters' dependence on them (Hallin, 1993). Veiled
attribution also gives the audience the perception of an association or a
personal relationship between the reporter and the source. They believe
the reporter is connected and privy to important information (Culbertson,
1978). This could add a sense of intrigue that keeps the viewer watching.
The use of anonymous attribution can have positive and negative
results. An obvious positive example would be Woodward and Berstein's
"Deep Throat" source during their Watergate reporting. But anonymous
attribution and the unaccountability that accompanies it can result in
grievous error. One example is what some have described as the New York
Times' "witch hunt" of Los Alamos computer scientist Wen Ho Lee in 1999
(Patterson & Wilkins, 2002). The coverage was likened to a trial and
conviction of an innocent man propagated by unnamed sources whose
information was either not confirmed or discounted.
Journalists as Sources
Berkowitz (1987) found that television news often presented large
amounts of information without attribution during reports by
correspondents, field reporters and news anchors. In this study, they were
also coded as sources. Some reporters were also eyewitnesses to the tragic
events of that day. These journalists were coded as media witnesses. News
organizations need to be able to trust their reporters and correspondents
in such situations to report their observations independently rather than
expect them to act as a pack (Patterson & Wilkins, 2002).
After reviewing the above literature on television news sourcing,
we pose these final research questions:
RQ3: What sources did the networks rely on the most?
RQ4: Were these sources named or anonymous?
Television news content might not be completely comparable to
newspapers because TV journalists' trade practices seem to depart from the
stricter attribution policies of newspaper journalists (Berkowitz, 1987).
For example, TV news often cites the organizational affiliation of
officials and experts without giving specific names. Organizational
restraints such as time and the need for immediacy also affect the
newsgathering process in terms of source variety. Therefore, previous
literature on sources was adapted to fit television news parameters.
Again, this study seeks to use the tragic events of September 11th
to explore the breaking news coverage of three major television networks in
terms of accuracy and sourcing. This study lends itself to both
quantitative and qualitative interpretation. Descriptions, such as
powerful eyewitness recounts, are necessary to fully understand the
enormity and horror of this breaking news event. This study will also
offer insights into the coverage of breaking news using one of the ultimate
examples.
METHODS
We viewed three major broadcast networks' September 11th coverage that
included three network affiliates in Washington D.C. – WLJA-TV Ch. 7 (ABC),
WUSA-TV Ch. 9 (CBS), and WRC-TV Ch. 4 (NBC). All three stations were
providing national programming of the morning news shows when the day's
events began. The Television Archive (www.televisionarchive.org) provided
the content for analysis on its Web site.
The analysis was performed using the time period of 8:45 a.m-11 a.m.
Eastern Standard Time. This period was selected for a number of reasons.
First, this period encompasses all of the significant events that occurred
that day in terms of breaking news. The period includes the first two plane
crashes, President George W. Bush's statement regarding the attacks, the
Pentagon crash and the eventual toppling of the World Trade Center towers.
Second, in analyzing this as a breaking news event, we had hoped to measure
how well and to what degree each station reacted to a common series of
catastrophes. By limiting the scope in this manner, we sharpen the focus of
the research and examine the "breaking" nature of this event.
Statements of fact and observation in relation to the events of September
11th were coded in five different ways. First, the source of the statement
was coded as to being an official source or a non-official source. Official
sources are operationally defined as a privileged source, one that can be
quoted without fear of libel, or as an individual qualified in some manner
as an expert in the topic on which he or she is speaking. For example, an
FBI agent or a New York police officer would be deemed an official source,
as defined by U.S. journalism libel law, as they are officials acting in
such a capacity. A terrorism expert would also be coded as an official
source, as they are steeped in knowledge of the area in which they are
discussing. Non-official sources included media reporter, other media
outlets, witnesses and unknown sources. If ABC, for example, were to cite
an Associated Press report that one of the planes involved was hijacked
from Boston, the statement would be coded as a non-official source.
Second, the identity of the source was coded as either being a named
individual, named organization or unnamed. A named individual is
operationalized as either an official or non-official source who has been
identified by their full name. In this coding scheme, both witness Bill
Smith and Head of Security for American Airlines John Jones would be coded
as a named individual. Named organization is coded as agencies and groups
that are cited as a collective in the reporting of the event. An example of
this would be if a reporter said that FBI officials suspected terrorists
were responsible for the attacks. While a specific individual is not named,
the name of the agency offers a stronger source than a generic "sources
report" citation. Unnamed sources are coded as unattributed statements or
statements generically attributed to "sources" or "reports."
Third, the certainty of the statement made by the source was coded as
either containing a disclaimer or not containing a disclaimer. Disclaimers
were operationalized as any attempt by the network reporters and anchors or
witnesses to "couch" their statements as being unproven. This included the
use of phrases such as "This is all speculation" or "This has not been
confirmed." This also includes the use of words such as "possibly,"
"allegedly," "believe," and "might." The use of these and other words to
demonstrate that the information is still unconfirmed and that more
information needs to be collected allowed us to code the statements as
containing a disclaimer. The absence of these words of uncertainty allowed
us to code the statements as containing no disclaimer.
Fourth, the type of source cited in the broadcast was coded.
Pentagon/Department of Defense was operationalized as any member of the
Pentagon or Department of Defense or any citation that referred to the
agency as a whole. Airline was coded as anything having to do with the
airline officials. City, state and national were coded as citations of
city, state or national officials. Witnesses were coded as anyone who
witnessed the event who was not a member of the media and had called in to
the television station. Media witnesses were coded as reporters or anchors
who were making first-hand observations of events occurring in real time.
Experts were coded as officials who had been introduced by the interviewing
reporter or anchor as an expert within the field under discussion. Unknown
was coded as any statement attributed to "sources" or "reports" or that
went unattributed. Other was coded as any citation or statement that did
not fit into any of the other categories.
The fifth coding category dealt with the accuracy of the statements being
made. By reviewing newspaper and magazine accounts in the days and weeks
following the attack, the statements could be viewed as either accurate or
inaccurate. The information was also assessed in terms of the time in which
it was presented. For example, witness accounts that stated the plane that
struck the first tower was a small, propeller plane were coded as
inaccurate. However, the statement that President Bush was headed back the
White House following the first two plane crashes was coded as accurate,
given that the attack on Washington, D.C. had yet to occur and the intent
of the President to return to the White House was accurate at the time. The
fact Bush was diverted from Washington in the wake of the attack did not
render the statement inaccurate in this research.
Each author was responsible for coding one network's coverage. Intercoder
reliability was determined after each author coded the same segment. It was
calculated using Scott's Pi, which was computed at .92.
As a final step, the major events of the day and the time they were first
reported by each of the networks were entered into a table (see Appendix A)
to allow qualitative comparisons between the stations. A larger chronology
of events and reportage was constructed (see Appendix B) to help aid in the
examination of the day's events, and the overarching flow of the network
coverage.
RESULTS
We coded 347 statements across the three networks during the analyzed time
period (n=347). ABC had 107 coded sources, NBC had 128, and CBS had 112
during the 2 hour, 15 minute broadcast time examined in the study. Table 1
contains the descriptive statistics of the types of sources used in the
network broadcasts, either official or non-official, Table 2 shows the
identification of the source used, while Table 3 illustrates the type and
number of sources used by the networks:
TABLE 1
Official vs. Non-official Source
Source
Frequency
Percent
Valid Percent
Cumulative Percent
Non-official
237
68.3
68.3
68.3
Official
110
31.7
31.7
100.0
Total
347
100.0
100.0
TABLE 2
Description of Source
|
