Reporters as Sources: To What Degree Do Broadcast News Personnel Offer Expert Testimony in News Stories?


Rhonda Gibson & Joe Bob Hester

[WJMCR 5:1 December 2001]


Abstract|Introduction|Research Questions|Method|Results|Discussion


This study examines the extent to which broadcast news personnel have taken it upon themselves to interpret the news. Specifically, this study looks at the extent ofspeculation as to the outcomes of events covered in the news and the motivations for actions.

A content analysis was conducted of two weeks of evening network newscasts from ABC, CBS, NBC and CNN, a total of 483 news stories. In 32.7% of the stories, either the anchor or a correspondent made some type of prediction, generally an absoluteprediction with no type of qualifying language. Of the news stories containing predictions, 79.1% predicted consequences or outcomes of events, whereas 37.9% included speculation about motivations.


The role of sources in the production of news has been debated for as long as news has been produced. As Sigal1 notes, “News is not what happens, but what someone says has happened or will happen.” The use of expert and “person-on-the-street” sources to answer the questions of who, what, where, when, why and how is common journalistic convention, designed to produce accurate and objective accounts of events divorced from reporter bias. In the modern news world, objective reporting has meant avoiding the overtintrusion of the reporter’s personal values and minimizing interpretation in writing the story. This concept has been embraced by many in the field, dismissed by others, and modified by some.

In 1947, the Hutchins Commission issued standards for the news media that called on them to be more responsible and to provide a fair, balanced, and complete account of events, a reaction to the partisanship then in some newspapers. Many media scholars and practitioners read the Commission’s report as a specific demand for increasedobjectivity in news reporting, and one suggestion to accomplish this goal was to “balance” news stories by citing multiple, often competing, sources rather than relying on editor or reporter interpretation.

Other scholars, such as newsman Walter Lippmann, were not totally comfortable with the notion of objectivity. Lippmann had previously advanced a version of objectivity that challenged the assumption that reportable reality is independent of thereporter’s subjective states. He argued that the role of the observer/reporter isalways selective and usually creative, but that this creativity helps news consumers make better sense of their often confusing world. He proposed three press responsibilities: (1) “to make a current record,” (2) “to make a running analysis of it,” and (3) “on the basis of both, to suggest plans.”2

Finding even less merit in the concept of a press without “bias,” many journalism scholars fully scoff at the notion of objectivity, citing interpretation andassignment of meaning as among the field’s greatest contributions. Carey3 argues that journalists need not apologize for interpreting events for their readers or viewers and that all writing, even scientific writing, is a form of storytelling aimed at imposing coherence on the otherwise chaotic flow of events. Mayer likewise argues that well-respected news organizations are sometimes most valued for abandoning objectivity: “The greatness of the New York Times derives not from its efforts at creating the ‘record’ or maintaining an authoritative objectivity but from a century-long often interrupted yet eventually renewed struggle to acquire the expertise necessary to broaden, sharpen, and deepen the perceptualapparatus of the institution.”4

This exploratory study is designed to examine the sourcing and interpretation patterns of broadcast news personnel. Specifically, this investigation looks at the degree of personal speculation and/or prediction by broadcast news anchors and correspondents as to the outcomes of events covered in the news and the motivations for such actions. Rather than quoting sources who are selected for their expertise on a given subject or using anonymous sources with “veiled” identification, to what degree are broadcast news correspondents and anchors serving as sources themselves, giving testimony about why events have occurred or what will happen next? The answers to these questions have serious implications for those concerned about issues of media credibility and will be part of the ongoing debate between advocates of objectivity and those who argue for a more interpretative journalist role.

A number of academic studies have addressed the role of sources in broadcast news. Many of these studies suggest there is a disproportionate focus on elite orinstitutional news sources, while women, minorities, and the working class are often under-represented.5 Those of higher socioeconomic status are more likely to beused as sources of information on TV news than are “ordinary people.” Whitney, Fritzler, Jones, Mazzarella, and Rakow6 report that only about one-fourth of quoted sources in network news are private individuals.

Other studies have examined the use of anonymous attribution in the news. Although some in the journalism field link the use of unnamed sources to a decline in perceived industry credibility,7 one study has showed that readers perceive a controversialstory to be more accurate and fair when no source or an unnamed source is quoted than when a named source or two conflicting named sources are quoted.8 Wulfemeyer and McFadden9 argue that the frequency of anonymous attribution might help explain why readers seem to accept the practice. Culbertson10 and Wulfemeyer11 found that about 33% of newspaper stories and between 70% and 85% of newsmagazine stories contain some type of “veiled” attribution. In a study of broadcast news, Wulfemeyer and McFadden12 found that about 55% of stories in network television newscasts contained at least one anonymous quote. Thus, the frequency of anonymous attribution on broadcast was about 20% more than the frequency and about 20% less than the frequency in newsmagazines.

Although most journalistic codes of ethics caution against the overuse of anonymous sources, research into the acceptability of such practice has not been consistent. In one study, about 60% of the respondents disapproved of anonymous attribution,13 whereas in another, about 40% of the respondents disapproved of quoting an unnamed source, even in an investigative report.14 In general, readers give reasonably high credibility ratings to unnamed sources.15Industry practice generally dictates that whether cited by name or anonymously, news sources are chosen for their expertise on a given topic, their involvement with a particular event, or occasionally even for their usefulness as a”person-on-the-street” interview. But regardless of the type of source used in broadcast news, research shows that the duration of sound bites from these sources is shrinking while television news correspondents are taking up more airtime with increasingly longer appearances in news stories.16 As Grabe et al17 note, “Today’s television news is much more ‘mediated’ than the news of the sixties and early seventies.” Patterson18 argues that this enhanced role of the reporter, especially in the arena of political news, is a result of the public’s reduced trust of government officials and a resulting expectation that the news media are “obliged to inspect the candidates’ platforms, judge their fitness for the nation’s highest office, and determine their electability.” This tendency toward interpretation has transferred to coverage of issues other than politics. Hallin19 argues that reporters no longer view official sources as authorities to be taken at their word and instead believe thatreporting now requires the journalist to provide his own interpretation of the “facts.”

The proliferation of news magazine programs, 24-hour cable news channels, and talk programs may also have contributed to the tendency of reporters and anchors to view themselves as experts. With hundreds of hours of programming to fill each week, these shows are likely to call upon photogenic and camera-savvy news personnel to appear as analysts and commentators. Unlike the brief appearances of correspondents in nightly newscasts, the longer format of news magazine programs allows for more in-depth analysis on the part of the broadcaster, possibly contributing to the tendency of the reporter to put his or her own spin on information and provide a personal interpretation of events.

Source credibility theory suggests that the increased visibility of news anchors and correspondents on trusted news magazine programs and talk shows is likely to enhance their overall credibility. Source credibility is generally thought to be a function of expertise and trustworthiness.20 The news magazine genre of programming continues to be rated as more trustworthy than any other source of news in the United States. According to Sawyer,21 51 percent of the viewing audience trusts the information on news magazine programs, whereas 43 percent trusts the information on nightlynewscasts, and only 37 percent trusts national newspapers. Given their frequent appearance on shows that viewers consider to be trustworthy, these news personnel are likely to “absorb” some of this credibility themselves. Simply by being presented as experts on the various news magazine and talk shows, these correspondents may be likely to be viewed by the public as actual experts, regardless of whether they have actual training or experience in the topic at hand. And, although reporters are not always seen as unbiased,22 they nevertheless are not affiliated with any political or social cause, as many news sources are, so they may be considered neutral enough to be credible. As credible individuals who have access to millions of viewers who tuneinto the nightly news, these broadcasters could wield considerable influence if allowed to do more than simply report the news.

Given these compelling issues, this study attempts to begin to quantify the degree to which news personnel provide their own testimony in news stories. The followingspecific research questions were proposed.

Rsearch Questions

R1: How often do news anchors and/or correspondents speculate about or predict the outcome of news events without citing a source?

R2: How often do news anchors and/or correspondents speculate about or predict the motives of individuals in the news?

R3: Is news personnel prediction more likely to be presented as absolute fact or somehow qualified?

R4: In what types of news stories are news personnel most likely to predict outcomes and/or motives?


The broadcasts for the content analysis consisted of two random weeks (Monday – Friday) of evening network newscasts from ABC, CBS, and NBC, plus The World Today on CNN (5 p.m. central time). A purposive sample of consecutive-day broadcasts was created by videotaping each 30-minute newscast during the weeks of October 4, 1999 and July 26, 2000. Due to a technical malfunction, the CBS newscast on June 26 was replaced with the broadcast from the following week, July 3. A total of 40 newscasts was recorded and analyzed.

The unit of analysis was the individual news story. A news story could be an anchor alone reading copy about a topic or a package with one or more correspondents involved. Each story was coded by topic, name and gender of anchor and/or correspondent(s), and length in seconds. In addition, each news story was coded for the presence and length in seconds of prediction by either the anchor or thecorrespondent(s). Two types of predictions were coded: prediction of outcome/consequences and prediction of motive. Prediction of outcome/consequencesoccurred when the anchor or correspondent predicted what would happen next or sometime in the future. Prediction of motive occurred when the anchor or correspondent speculated about why something happened or someone did something. Predictions made by a news source were not coded.

Predictions were coded as either absolute or qualified. An absolute prediction was one in which no uncertainty was indicated by the predictor and no qualifying wordswere used. A qualified prediction was one in which some type of qualification was used to indicate the possibility of the prediction. For qualified predictions, qualifying words such as ‘possibly, probably, maybe, most likely, could be,’ etc. were recorded.

Predictions were also coded by catalyst for the prediction. For instance, does the anchor ask for speculation by the correspondent, or does the correspondent provide it on his/her own? Or, does the anchor simply state the prediction?

Finally, each story was coded by presentation as either a singular incident or part of some social phenomenon. Singular incidents were one-time occurrences that did not tie directly to any social issue. A plane crash or the death of a famous person would be examples of singular incidents. A social phenomenon story, on the other hand, addresses some larger social issue, even though it might be illustrated by a specific case study. For example, a story may open with information about a homosexual couple being denied insurance coverage but then address the larger issue of gay rights.

After a one-hour training session, the news stories were coded by three graduate students in mass communication. Each student coded approximately one-third of the news stories. Once the coding was complete, a random subsample of 50 stories (10.4% oftotal stories) was coded by one of the authors to determine intercoder reliability. Simple agreement for story topic was 98.0%. Simple agreement for all other codingcategories ranged from 94.1% to 100%. Scott’s Pi was then computed for the five dichotomous coding categories where simple agreement was less than 100%. With theexception one category, Scott’s Pi was greater than 0.85. For this category, absolute versus qualified speculation about motive, Scott’s Pi was 0.79 , which is still acceptable.23


A total of 483 news stories was analyzed for a total of 718.3 minutes of program time. The mean story length was 1.49 minutes. Table 1 indicates the totalnumber of stories, mean number of stories per half-hour broadcast, and mean storylength for each network. The news stories were most often presented as a package with one or more correspondents involved (57.3%, n=277). An anchor alone reading copy about a topic accounted for 42.7% of the stories (n=206).

Table 1.

Number of stories, mean number of stories per broadcast, and mean story lengthby network.

  ABC    CBS    CNN    NBC  
Total number of stories12612913692
Mean number per broadcast12.612.913.69.2
Mean length in seconds83.885.477.8118.9

The first research question asks how often anchors and/or correspondents speculate about or predict the outcome of news events without citing a source. Data in Table 2 show that in 32.7% (n=158) of the stories, either the anchor or a correspondent made some type of prediction. The anchor made the prediction in 7.7%(n=37) of the cases; a correspondent made the prediction in 20.3% (n=98) of the cases; and both the anchor and a correspondent made a prediction in 4.8% (n=23) of the cases. In the 60 cases where an anchor made a prediction, he/she did so on his/her own initiative. When a correspondent made a prediction, he/she did so on his/her own initiative 89.3% of the time. The mean length of the predictions was 9.6 seconds (SD=12.0).Table 2.

Prediction by anchors/correspondents.

   Predictor      N      %   
Anchor alone377.7
Correspondent alone9820.3
Both anchor & correspondent234.8

The number of predictions varied by network. As can be seen in Table 3, CBSaccounted for the most predictions (38.3%, n=23), while the NBC accounted for 35.0% (n=21) of the predictions. ABC and CNN each accounted for about one-third as manypredictions (13.3%, n=8).Table 3.

Prediction by network.

   Network      N      %   

The second research question asks how often news anchors and/orcorrespondents speculate about or predict the motives of individuals in the news. Of the 158 news stories in which a prediction was made, 79.1% (n=125) includedpredictions of consequences or outcome, while 37.9% (n=60) included speculation about motives.

The third research question asks whether news personnel prediction is more likely to be presented as absolute fact or somehow qualified. The majority were presented as absolute. Absolute predictions comprised 67.7% (n=107) of the stories in which a prediction was made, while 49.3% (n=78) were qualified in some fashion.

The fourth research question asks in what types of news stories are news personnel most likely to predict outcomes and/or motives. Of the 483 news stories in the sample, the majority 22.2% (n=107) covered international news or foreign affairs. Of these, 23.36% (n=25) contained some type of prediction. All categories of stories contained some type of prediction, ranging from 15.69% of stories about accidents and/or disasters to 100.0% of stories about immigration. At least one-half of the storiescontained predictions in these five categories: immigration, education, domestic politics, medical/health, and agriculture. Table 4 shows the percentages of stories containing predictions.Table 4.

Percentage of different types of news stories in which predictions are made.

   Story Type      N      Stories with Predictions       %   
Politics (domestic)16956.25
Medical / health432353.49
Science / environment13646.15
Federal government (legislative branch)351542.86
Courts / lawsuits20735.00
Federal government (judicial branch)21733.33
International news / foreign affairs1072523.36
Federal government (executive branch)18422.22
Accidents / disasters51815.69


Research indicates that news anchors and correspondents are being seen forlonger periods of time in network news stories, at the expense of experts and other types of sources.24 It is apparent from this study that the broadcast news personnel are not simply reporting the news during their additional air time; they are also interpreting the news. In about a third of the stories, there was some type of non-attributed prediction by the anchor or correspondent about why something had happened or what would happen next. Correspondents were more likely than anchors to offer non-sourced speculation, and these reporters overwhelmingly did so on their own, without being prompted by the anchor. Most often news personnel speculated about what the outcome of an event would be, and to a lesser degree what the motivations involved. When an anchor did make some type of prediction, it was more likely tooccur on CBS or NBC than on ABC or CNN.It appears from these data that, in a substantial number of cases, correspondents and anchors are providing their own expert testimony about news events without citing any specific source for their information. And they are often doing it in absolute terms. More than 60 percent of the time, prediction by news personnel was presented as absolute, that is, without any qualifying language such as “possibly,” “maybe,” “apparently,” etc. For example, a correspondent for ABC news, in a storyabout Chinese immigrants entering the United States through Canada, closed the package with the comment, “Every week ships of illegal immigrants leave China, someundoubtedly bound for Canada, with passengers heading south.” The correspondent provided no expert source for the information; he simply stated it as fact. Likewise, in a story about the popularity of hockey, Peter Jennings told viewers that one sports fan had suggested that hockey players donate a portion of their salaries to charity. “It wasn’t about to happen,” Jennings summarily said of the suggesting, citing nosource for the information.Less frequently, some type of qualifying language was used by anchors and correspondents who offered their own prediction in news stories. For example, Charles Gibson, serving as anchor for the ABC evening news, said of a court settlement involving State Farm Insurance, “The verdict could improve the quality of auto insurance, but it could also make it more expensive.” Other commonly used qualifiers were “maybe,” “apparently” and “likely.”Stories about politics/government/military, education, business, andmedical/health news were most likely to include some type of news personnelprediction. These are areas in which “beat” reporters tend to specialize and often cover for a long period of time, so it could be that the correspondents felt morecomfortable speculating about outcomes and motives for these subjects. Topics such as sports, entertainment and lifestyles, areas where “serious” journalists may only occasionally cover a story, were less likely to contain speculation. Economic news, however, did not follow this pattern.Overall, the data suggest a willingness on the part of news personnel to inject their interpretation of events into a newscast. It is certainly possible that this “testimony” by broadcast correspondents is in fact information they gathered fromactual sources and that the reporter is simply not stating the source of theinformation. It must be wondered, however, if broadcast journalists are not moving more toward an interpretive and analytical role in their reporting. Given their increased visibility on news magazine and talk shows, where they are presented asexperts on topics ranging from politics to health care to the environment, reporters may be more likely to view themselves as actual experts. And they are not likely to be discouraged from such practices by networks executives who are all too willing to promote their top anchors and correspondents as media “stars.”Such practices appear to go against the journalistic notion of objectivity,which asks reporters to prevent the intrusion of their personal values and biases into the news production process. Such objectivity, as Sigal25 notes, requires reporters to attribute information in a story, and especially any interpretation of what it means, to outside sources. Hespecifically states: “Keeping the reporter out of the news means relying onsources.”26Matusow,27 likewise, considers the growing power of on-air personalities to be dangerous: “Under the old system, the editorial product of network news emerged from consultations, or sometimes outright struggles, among anchors, producers, correspondents, and senior executives-a healthy state of affairs, since news judgments are seldom cut and dried.” She believes that “those in charge of running the television industry must examine the implication of giving up so much control to anchors and other star journalists and find ways to regain the old balance.”28Yet other news scholars do not see danger in journalistic interpretation orexplanation; in fact they applaud it. Carey29 says that interpretation, explanation, and thick description of an event are a reporter’s duties and that promotion of pure objectivity is na�ve. He argues that the most important descriptions and explanations of journalism are lost when they are sliced into daily fragments, thin tissue cultures of reality,disconnected from a narrative framework.Given the relatively small amount of airtime that is devoted to speculation by correspondents about why events have occurred or what will happen next, it is certainly too soon to sound an alarm about broadcast news personnel serving as newsmakers or attempting to tell their viewers what to think. The investigation of only two weeks of network newscasts and the consecutive-day sample used in this study limit the generalizability of the results to the two weeks in question. Future research should be conducted using a more representative sample. This could provide a more complete picture of the extent to which broadcast news personnel speculate as to the outcomes of events covered in the news and the motivations for actions. Such research should examine differences in speculation among various networks, among different news topics and between anchors and correspondents. In addition, it would be beneficial to conduct further research into the actual effects on viewers of suchinterpretative reporting by broadcast journalists. Are anchors and/or correspondents viewed as experts to the same degree that actual expert sources are? Do viewers understand that speculation by broadcasters is just that, or do they take such speculation as fact? Given that correspondents and anchors are taking up more andmore of network news time, scholars need to ask what effects these reporters are having on broadcast news consumers.

About the Authors:Rhonda Gibson and Joe Bob Hester are assistant professors at the University of North Carolina at Chapel Hill. An earlier version of this paper was presented to the News Division of the Broadcast Education Association.

Leave a comment

Your email address will not be published. Required fields are marked *