Skip to Content View Previous Reports

Content Analysis

Content Analysis

Despite continuously declining audiences, (see Audience) more people still gather around the three evening newscasts – more than 25 million each weekday – than for any other three news sources in America. What do they get there? How does it differ from morning news? What, if anything, distinguishes the content of PBS’s evening newscast? In a year that saw the CBS anchor under fire and the NBC anchor replaced, did network news change much?

To get answers, the project conducted a content analysis of all three network evening and morning newscasts as well as the NewsHour on PBS. The study built on the 2003 report and added several new areas of analysis. It encompassed a month of weekday newscasts (20), 110 hours of news programming, and examination of 1,756 separate stories (see Methodology).

Among the findings:


Storytelling Versus the Culture of Live

What news the networks do choose to offer has a thoroughness that is hard to find on cable. Much of that stems from the continuing reliance in evening newscasts on the taped, edited, correspondent package as the heart of the program.

The vast majority of the commercial evening news hole (86% of all time) is devoted to such pieces.1 Live reports, interviews and stand-ups, on the other hand, account for just 2%. About 12% of the time is made up of anchors reading short summary “tell stories” or narrating video.

The reliance on correspondents telling stories distinguishes the three network evening newscasts in the national TV news landscape. Much of what viewers get from TV news today (outside of their local newscasts) is characterized by a dependence on live, unscripted communication. Just as “reality” TV is replacing scripted drama and comedy on the entertainment side, news on TV is also becoming a more extemporaneous medium.

In network morning news, for instance, only a third of the time (32%) is made up of edited storytelling – and here we measured only the first hour, which is more hard-news oriented. The majority of time is made up instead of interviews (42% with outsiders and another 13% with in-house correspondents.) Live reporter stand-ups are rare (less than 1%).

Yet even morning news depends on more edited storytelling than cable. In the same 20-day sample of cable programming, just 24% of the time was correspondent packages, while 52% was live. (see Cable TV for a more detailed account).

The format of PBS’s NewsHour lies somewhere between commercial evening news, morning news and cable. At first glance, it most closely resembles morning news – about a third packages (31%) and slightly more than half (53%) interviewing of outside analysts.2 Yet the NewsHour sets up more of its interviews with introductory packages that offer viewers background. Another difference is that the interviews are often discussions with two or more analysts, rather than one guest or two opposing advocates in a debate format.

Incidentally, as video becomes more a part of the online universe, the taped package may become the basic unit of reporting in the interactive, video-streamed news of the future.

Depth of Reporting

To assess the thoroughness of the reporting, the study this year developed a series of new measures to break down the nature of the information viewers were getting. We measured:

Transparency of Sourcing

Overall, network nightly newscasts stand apart from the rest of TV news for how much they share with viewers about the sources they rely on. Half the stories on the three network evening newscasts (50%) had at least two fully transparent sources. And that number shot up to 81% for the stories that make up most of the broadcast time, correspondent packages. (The difference is the brief anchor-read items, which take up a small amount of time but account for a large number of stories.)

Source Transparency, Network Evening News

Commerical
PBS
Morning
No Sources
37%
36%
39%
1 Source
14
21
23
2-3 Sources
32
20
28
4+ Sources
18
23
11
Totals may not equal 100 due to rounding.

The NewsHour on PBS also distinguished itself for the transparency of its stories. It has three distinct formats – interviews, packages, and anchor-only items – and the level of sourcing is heavily influenced by the format. Even so, the NewsHour stands out. Its taped packages were the best sourced of all (79%), though they were a smaller part of the program than on the commercial networks. Furthermore, almost a quarter of the NewsHour’s stories (23%) had the highest level of transparency, four or more fully identified sources.

The morning shows on network news relied on fewer sources in their stories and shared less information about them with viewers. Just 39% of morning stories contained two or more fully identified sources and only 11% carried four or more. That is a function both of format, with fewer taped packages, and the kind of interviews the morning shows carry. The interviews are often one-on-one and focus on the interviewee’s personal life or experiences. Interviews on evening news programs tend to discuss issues that draw on outside information.

Still, all the broadcast news programs were well ahead of cable when it came to sharing information about sources. Television in general shared less source information than newspapers and online news, which have larger news holes.4

Television news also carries with it a natural added element of transparency – the visual image. An on-camera source offers the viewer additional audiovisual cues such as facial expressions, tone of voice, and even some sociological context. Such things may provide additional value, but they don’t always help make clear the sources’ connection to the story or the quality of their expertise.

Reporting: Breadth of Viewpoints

How many different aspects of the story did network newscasts tell? Here all three types of network newscasts fared well, and far better than cable. (Some stories, of course, simply present undisputed facts, such as describing an accident, and there isn’t necessarily a second side to the story. Those were excluded from the search for multiple viewpoints.)

Fully 87% of controversial stories reported on the NewsHour contained more than one side of the story. Morning news scored similarly; 86% of its stories told at least two sides of controversial stories. The nightly newscasts lagged behind. Only 72% of stories contained more than one side, meaning that 28% of stories that were in dispute carried, at best, only a passing reference to another side of the story.5

The morning-to-evening difference was even wider among packaged pieces. Nearly all, 98%, of morning packages with some inherent controversy had multiple viewpoints, compared to only 75% in the evenings. And the contrast was not due to the difference in overall available time (morning shows are two or three hours long, depending on the network). Evening news packages on average were five seconds longer than morning packages (146 seconds versus 141).6

One factor is story topic. About 25% of the morning packages with multiple viewpoints were about crime, accidents, celebrity, lifestyle and miscellaneous topics, compared to only 12% in the evening packages. As we’ve mentioned, the morning shows are fond of water-cooler controversy involving less than major news – Laci Peterson, or Michael Jackson.

In any case, all the viewpoints numbers stand out from cable news, where close to three-quarters (73%) of controversial stories told just one side or had only a passing reference to another point of view.

Reporting vs. Journalist Opinion

The third step in looking at reporting was to measure the extent to which stories contained opinion from the journalists themselves in ways that they do not attribute to any source or other reporting.

For the most part, journalists on the network evening news kept themselves out of their reporting. The vast majority stories (83%) did not contain any direct opinion from journalists.

Morning news, despite its heavy emphasis on interviewing, contained even less journalistic opinion (just 11% of stories).

Yet here PBS’s NewsHour again stood out from the rest. Only 3% of segments contained explicit journalistic opinion. That was half that even of front-page newspaper coverage over all (6%), and only about a fourth the coverage on front pages of the largest papers (13%).

In cable, we looked at three different hour-long programs on each network (a mid-day hour, the news roundup show and the highest rated prime-time talk show.) Both over all and looking just at their news digest programs, opinion from correspondents and anchors was more prevalent – 28% of all stories 26% of those on news digest shows.

Some topics on television either lend themselves to reporters’ offering their own assessments or are considered fairer game for journalists to weigh in – particularly politics. On the nightly commercial newscasts, for instance, 44% of all election stories carried some opinions from journalists themselves. That is markedly higher than for foreign affairs, domestic affairs and government stories (which ranged from 10% to 12%). In sharp contrast, PBS carried only two election stories with journalists’ opinions.

The spread of opinion was quite similar on the network morning news shows. Election stories were most likely to carry journalist opinion (20%); the figure was between 9% and 13% for stories about foreign affairs, domestic affairs and government.

Much of the journalistic opining appeared to be of the horse-race variety, particularly during the Democratic primary season. (see the Election section below). Still, with questions about partisanship on the rise (see Public Attitudes), the apparently looser standards about separating news and opinion in political coverage may carry even greater importance.

Journalist Speculation in News Topics
Network Evening News (Commerical and PBS)

No Opinion
Opinion
N/A
Government
87%
12
1
Elections
63%
34
3
Domestic Affairs
89%
10
1
Business
93%
5
2
Crime
98%
2
0
Foreign Relations
89%
11
0
Science Technology
88%
10
2
Celebrity
77%
23
0
Lifestyle
84%
14
2
Accidents/Disaster
97%
3
0
Miscellaneous
94%
5
1
Total
85%
14
1

Reporting: Anonymous Sources

Journalists’ reliance on anonymous sources has been debated for years. Various surveys have shown that the public tends to dislike anonymous sources. Journalists, on the other hand, consider them critical to gaining certain kinds of information – especially secret government and corporate activities, or indeed anything that gets beyond the “spin” of official talking points. Television personnel have also told us they sometimes drop identification of a source simply to save time in a report.

The level of anonymous sourcing also reflects a struggle for control between journalists and their sources. The more a journalist needs a source to “talk,” the more power a source has to demand anonymity.

A handful of erroneous reports over the last three years have led some news organizations to clarify their policies on anonymous sourcing. The Washington Post, for instance, now promises it will explain in every case why it agreed to allow a source this protection. And indeed in print we found the use of anonymous sourcing this year to have fallen (to just 13% of all front-page stories and 7% of all stories studied).

Was there evidence of a similar tightening of use of anonymous sourcing in network news? In a word, no.

In commercial nightly newscasts in 2003, we found that 43% of stories contained sourcing that to the audience was anonymous. In 2004, anonymity was up to 53%, more than half. The practice was slightly less prevalent at the NewsHour, but 47% of the stories contained at least one anonymous source, up substantially from 15% a year earlier. Part of the jump may be explained by the dominance in 2004 of internal, closely guarded government stories like Abu Ghraib and accusations surrounding the 9/11 hearings and report.

Package stories were more likely to have anonymous sources than anchor-read briefs, 68% to 31%, as were major running stories, which tend to be more controversial: about 60% of all the stories about Iraq, Afghanistan, Iran, Haiti, Israel/Palestine, the Madrid bombing, and the high-profile court cases had anonymous sources.

On the morning programs as well, 50% of all coverage included at least one anonymous source. The figure rose to 79% of the morning packages, roughly 10 percentage points higher than for the evening newscasts.

The Topic Agenda on Network News

As we noted last year, the time people choose for tuning in to network news changes markedly what world they will learn about. The topic agenda on the evening newscasts is very different from even the first hour of morning network news, and the PBS NewsHour is different still.

The Three Commercial Nightly Newscasts

The nightly news is the closest thing television has to a front page.

In 2004, nightly newscasts eased away from the intense coverage of foreign affairs that had built up over the previous three years. That was largely because the Iraqi war became something of a domestic story, as coverage of the war intertwined with the presidential election, especially in coverage of the torture of prisoners and the debates over U.S. intelligence. Some viewers might well consider those “government” stories as much about foreign affairs as about the Bush administration, and indeed the researcher Andrew Tyndall’s accounting of network topics categorized many such pieces that way. His data show that it is normal for foreign policy to be scaled back and campaign coverage to increase every four years in synch with the presidential election cycle.7

Commerical Nightly News Topics, Over Time
Percent of All Stories

1977
1987
1997
June ’01
Oct. ’01
2002
2003
2004
Government
37%
32%
18%
5%
7%
5%
16%
27%
Foreign Affairs
21
19
15
17
10
21
25
14
Defense/Military
1
1
3
6
29
16
3
1
Elections
9
Domestic
8
7
5
18
34
12
16
21
Crime
8
7
13
12
4
12
6
2
Business
6
11
7
14
5
11
12
8
Celebrity/Enter.
2
3
8
5
0
2
2
2
Lifestyle
4
11
14
13
1
17
6
5*
Science
4
5
6
4
11
2
2
3
Accidents/Disaster
9
5
10
4
0
3
10
4
Other
N.A.
N.A.
N.A.
3
0
N.A.
2
4

*Includes 2% sports
Totals may not equal 100 because of rounding.

Election coverage itself accounted for just 9% of stories on the nightly newscasts. Still, that was enough to drive down coverage of some other topic areas, namely accidents, disasters, crime, and business and the economy.8

Among the three newscasts, CBS was about 50% more likely than NBC and twice as likely as ABC to air stories about unexpected events like disasters and twice as likely to air feature stories not tied to breaking news (the ethics of using high-tech duck decoys, the passage of Venus between Earth and the sun, and the American struggle to pay credit card debt). Researcher Tyndall’s data for the entire year found a similar trend at CBS toward local or regional coverage, especially in a feature format, focused on topics like weather, animals, family and health.

The message, not a new one, is that the 30-minute “appointment news” format (news that is broadcast at a certain time every day.) is tightly tied to the headline stories. Indeed, the evidence suggests that the focus was even narrower in 2004 than before. The cutbacks in network resources, bureaus and reporters probably accentuated the narrowly focused nature of the nightly newscast. Or it might turn out that 2004 had an exceptionally narrow agenda because of a pair of massive preoccupations – Iraq and the presidential campaign – that sucked the oxygen out of all other coverage until the last week of the year, when the tsunami hit.

Topics on the NewsHour

The agenda of topics on the NewsHour with Jim Lehrer also changed in 2004. Government coverage, as a percentage of the number of stories, remained unchanged at 24%, but foreign affairs declined by nearly half to 20%, from 39% in 2003. Even so, foreign affairs was still a much bigger part of the PBS newscast than of the commercial networks or even newspapers’ front pages. In addition to an emphasis on Iraq in the feature-interview segments, the network usually also had one or two short anchor reads about Iraq in each program.

With foreign coverage declining, what replaced it? Largely, it was heavier coverage of the U.S. election and government affairs. Much of the government coverage involved stories about courts and law, and government agencies in addition to the executive branch. And as we saw last year, business and economics continued to make up a larger percentage of NewsHour coverage than we found in either network news or newspaper front pages. Fully 11% of NewsHour stories were about business and economics, 50% more than on network news and three times as much as on the front pages of newspapers, which, of course, have separate business sections for much of that news.

Topics on Morning News

Those who get their news in the morning, even during the first hour with its harder-news orientation, get a very different agenda. Morning news may be holding on to its audience while evening news is losing, but it is remarkably lighter fare, more focused on a handful of major crime stories and one or two big breaking-news events each day. What’s more, much of the coverage of Iraq, the year’s biggest story, consisted of so-called “Yellow Ribbon Journalism,” according to Andrew Tyndall – human-interest coverage of men and women in service or on the homefront, rather than military policy or diplomacy.

The most striking change in 2004, as was true in the evenings, was the large spike in government stories.

The picture may be somewhat misleading, though. While technically accurate, much of the uptick in government coverage came from two groups of stories – involving the war in Iraq and a series of high-profile crime cases – that might strike many as more fundamentally about Iraq, about crime and about celebrity.

The Iraq-related stories included the dispute over weapons of mass destruction and the Abu-Ghraib prison scandal, which made up fully a third of all the government stories.

The second big cluster involved three court cases – Martha Stewart, Scott Peterson, Kobe Bryant – which were shifted from being counted as celebrity scandals in 2003 to being legal stories in 2004 as they wended their way through the justice system. They accounted for another third of all the morning government stories.

If those two groups of stories are subtracted, the level of governmental stories in the morning would have been 7%, not statistically different from the 8% a year earlier. And if the war in Iraq had remained in the category of foreign stories, that category would have been 3% compared to 5% from the previous year.

Thus the shifts reflect changes in the news – and in particular changes in the nature of a handful of stories – but not, apparently, any major changes in the nature or focus of the morning programs themselves.

How Morning Shows Change Over the Hour

The study this year also reveals some nuances about how the nature of morning news changes as the programs progress each day. First, we can clearly see that the traditional notion about morning news – that the first hour is more hard-news oriented – is better understood as just the first half-hour. Hard-news topics on the morning shows in 2004 were usually concentrated into the first 20 minutes of the program (not including local news breaks or commercials).

In this first block of the programs, government topics led, accounting for 22% of all stories (and again celebrity crime stories made up a quarter of those). Domestic affairs, such as health care or domestic terrorism, accounted for another 16%, with foreign affairs at 9% and election news at 8%. In all, roughly seven out of ten stories in the first half-hour are so-called traditional hard-news topics.

Lifestyle features such as the Early Show’s story about what to do if someone sees a child getting into a car with an intoxicated adult and Today’s story about how “The Vagina Monologues” promotes awareness of violence against women rarely air in these first minutes – just 16 such stories in all for the 20 days studied.

In the second half-hour, the array of coverage offered is already quite different. Government stories drop by nearly half (to 12% of all stories), and foreign affairs and election news are largely absent. Lifestyle features increase (to 15%), and celebrity stories to 12%. The biggest category, though, is the miscellaneous event, accounting for a full 29% of all stories in this half hour. These stories often have little connection to other current events. They are “interesting features” such as a story about an attack by an escaped gorilla at a Dallas zoo or a story on ABC’s Good Morning America about the luxury liner Queen Mary 2.

Placement of Morning News Stories

1st 30 Minutes
2nd 30 Minutes
Government
22%
12%
Military/Defense
0
0
Foreign Affairs
9
3
Election
8
1
Domestic Affairs
16
15
Business
2
1
Crime
4
4
Celebrity
4
12
Lifestyle
3
15
Accidents/Disaster
6
2
Science
2
6
Other
24
29

Those findings match those of Tyndall Research, which uses a different methodology and somewhat different topic definitions. Excluding the news summary at the top of the hour, Tyndall found that 75% of the interview/segments are devoted to hard-news topics in the first half-hour, dropping to 39% in the second 30 minutes.


Television versus Newspaper Front Pages

TV journalists have long compared the nightly newscasts to a newspaper front page, and that analogy still holds. In 2004, the two media devoted a similar percentage of their story count (9% for both) and their news hole to the elections (at least up to October when this sample was completed) and the same proportion of their stories to foreign affairs (14% for both).

Topics in the News
Newspapers Versus Network Nightly News, 2004
Percent of All Stories

Commerical Nightly News
Newspaper Page A1 Only
PBS “NewsHour”
Commerical Morning
Government
27%
35%
24%
20
Foreign Affairs
14
14
20
7
Defense/Military
1
2
2
0
Elections
9
9
11
7
Domestic
21
14
19
16
Crime
2
4
2
4
Business
8
4
11
2
Celebrity/Enter.
2
1
2
6
Lifestyle
5
10
2
5
Science
3
3
2
3
Accidents/Disaster
4
3
2
5
Other
4
3
4
25
Totals may not equal 100 due to rounding.

The Election and the War in Iraq

The study in 2004 also looked more closely at coverage of the two major stories, the election and the war in Iraq. To do so, we grouped all the stories that related to those subject areas, even if they looked at a particular topic within them, say the environment and the election, so that we could get a clearer sense of election coverage overall. What did we find?

Election on the Nightly News

First, the tendency toward journalist opinion is even more apparent in campaign coverage than elsewhere.

In the three commercial evening newscasts, 44 % of election-related stories studied contained opinion, nearly triple the average for the evening news overall. The primary season was much more likely to feature journalists’ opinions (63% of stories) than the general election (32%). That raises interesting possibilities. One is that journalists are more comfortable offering horse-race opinions than other types – as in “candidate X needs a win here Tuesday or he is in big trouble.” Another is that journalists believe they have a greater responsibility to act as a referee and interject their own judgment because of the perceived level of spin or rhetorical license in modern campaigning. In fact-checking advertising or debate rhetoric, for example, journalists might consider it a strength to offer opinions. ABC News’s political director, Mark Halperin, even warned staffers not to fall victim to “he said he said” journalism that created false equivalencies. Halperin instructed his reporters to use their judgment to point out when one side was engaging in distortions more often than another.

PBS was much more deliberate about keeping itself out of the stories – just two election stories carried judgment from the correspondent. Among the three network evening newscasts, NBC Nightly News looked much more like PBS than the other two. Just 19% (four stories in all) contained overt journalistic opinion. ABC’s World News Tonight, on the other hand, included opinion in 68% of its election coverage, with CBS Evening News at 44%.

The morning news programs were more able than their evening counterparts to keep opinion out of their reporting, but less able than PBS. Roughly three-quarters, 76%, of their election stories were free of any journalist opinion, compared with 56% over all on commercial evening news and 84% on PBS.

Just as with the evening stories, the morning stories about the primaries had far more opinion. About 32% of 28 programs carried reporter speculation during the morning programs, compared to only 6% for the general election.


Tone of election coverage

Was the tone of the coverage more positive, more negative or reasonably balanced and neutral?

To try to get some answers, we created a way of quantifying the tone for stories about the war and the election. To derive tone, we first identified whether the story was about a particular newsmaker or issue. If so, each quote, innuendo, and assertion was counted as positive, negative or neutral for the story’s main newsmaker, or in the case of an issue story, about moving toward resolution of the central issue.9

For stories to be considered positive or negative, one position must dominate by at least a 2:1 ratio. For example, if a story contained four positive sentiments, it must then contain at least eight negative statements to be considered negative in tone and no more than two negative statements to be considered positive in tone. In all other cases, the story would be labeled as neutral.10

What did we find?

Over all, campaign stories on the nightly newscasts tended to be either neutral or positive, rarely negative, according to the data. They were nearly three times as likely to be positive as to be negative for the principal newsmaker or issue. ABC was clearly the most positive of the bunch. NBC was the most neutral, and CBS fell in between.

Tone of 2004 Election Coverage

Commerical
PBS
Morning
Positive
40%
46%
41%
Neutral
42
50
27
Negative
18
4
31

Totals may not equal 100 due to rounding

On PBS, positive coverage dominated even more – 46% positive versus just 4% negative (1 story) and 50% neutral.

Morning News

And what of morning news? Our earlier observation about those programs’ hope of provoking water-cooler debate was reinforced in the tone of their election coverage. Evenhanded neutrality was out of favor in the mornings. The segments were basically just as likely as their evening counterpart programs to carry a positive tone, 41% versus 40% for evening. But they were also nearly twice as likely to carry a negative tone, 32% to 18% for evening news. Basically, morning-news election stories were most likely not to be neutral. Many ascribe an upbeat tone to morning news programming. That is a misapprehension: lively controversy is closer to the mark.

Through the year, the Project did conduct two election-specific studies that allow some evaluation – one before the conventions and one during the debates – and both show President George W. Bush clearly getting the more negative coverage (see the PEJ studies on Character in the 2004 Campaign and the Presidential Debates).
Coverage of the War in Iraq

The other major story of the year was the war in Iraq. What stood out in the coverage here? First, Iraq coverage had the largest percentage of stories given any topic. The percentages for PBS and the commercial networks stories were almost equal, with 25% of commercial networks stories about Iraq and 27% of the NewsHour stories. The figure for morning news was 12%. (Tyndall Research, coding every weekday newscast for the entire year, similarly found 21% of the time on the nightly newscasts was devoted to stories about Iraq.)

On the war, network nightly news coverage carried much less journalistic opinion than did election coverage. Fully 84% of it was free of overt journalist opinion. On PBS, not a single Iraq-related story included journalist opinion.

What about the tone of coverage of the war? Administration officials and various conservatives argued in 2004 that the coverage was heavily critical of President Bush, focusing on U.S. casualties and other setbacks rather than on positive developments.11

Tone of Iraq Coverage

Commerical
Evening
PBS
Morning
Positive
16%
16%
31%
Neutral
44
18
36
Negative
28
26
19
Multi-Faceted/NA
13
40
13

Totals may not equal 100 due to rounding

In the weeks studied here, the bulk of network coverage about the ongoing war was neutral, but, indeed, commercial evening and PBS coverage was nearly twice as likely to be negative as positive. Roughly 27% of coverage on both carried a decidedly negative tone while just 16% carried a decidedly positive one. Still, most commercial evening news coverage was neutral (44%) and another 13% were multi-subject stories for which tone did not apply. The PBS NewsHour tended to do more multi-faceted stories (40%) while 18% were neutral.

Source Transparency of Iraq Coverage
Commerical and Public Network News

Evening
Morning
PBS
No Sources
29%
33%
36%
1 Source
20
28
21
2-3 Sources
36
26
20
4+ Sources
14
12
23
Total
100
100
100

Totals may not equal 100 due to rounding

This is quite different from the tone of newspapers which were pretty evenly divided between carrying a decidedly positive and negative tone. Whether that is a sign of bias or an accurate reflection of events on the ground is beyond the scope of this research.

Morning news programs adopted a more upbeat tone toward the war – 31% positive, 19% negative – in keeping with the penchant for the “Yellow Ribbon Journalism” cited by Andrew Tyndall.

When it came to journalistic opinion and the war, the networks were much more circumscribed than they were about the campaign. In the evenings, journalistic opinion was found in 16% of the 167 of the commercial-network stories about Iraq, which roughly equaled the overall average for all commercial network news stories. None of the 68 PBS Iraq stories had journalist opinion. In the mornings, stories about Iraq were even freer of journalist opinion than in the evening – 10% had it (another 2% were clear opinion pieces or commentary where opinion was expected).

We also looked at the level of sourcing in the war coverage. Looking at Commercial and PBS war reports in the evenings were less likely than the coverage overall to carry the highest level of sourcing, four or more transparent sources. Morning news looked similar. A third offered no fully identified sources, as against 29% for evening news, and less than half that (15%) contained four or more.

Footnotes

1. Andrew Tyndall’s research, which is based on every weekday broadcast of evening news, is remarkably similar. Tyndall counted 84% of time on evening newscasts made up of taped packages.

2. Unlike cable and the morning newscasts, PBS has few in-house experts as interview subjects.

3. Beyond identifying the source by name and title, did the story reveal the source’s level of knowledge about the events such as being an eyewitness, having a professional qualification that is identified beyond just title, or a personal experience? Did it make clear any likely biases or allegiances a source might have, such as noting the party or ideological affiliation of the source if relevant? In other words, could the average viewer judge for themselves the source’s authority?

4. Some in television argue that the compressed time of network news does not allow for prolonged explanation of sources. It is fine for newspapers to take up one more line of print, they argue, but for a network show to take up an additional 10 seconds in a 50-second story is much harder to justify. This argument, though, may not be as valid in 2004 and it was in, say, 1998. Technology now allows for other visual means of identification beyond the two-word chyron used in the past. MSNBC, for example, has used a technique in which, when a source appears on air, the screen splits in two for about 10 seconds and the second screen gives biographical information about the speaker.

5. For a detailed explanation of whether a story involved controversy and the breakdown of viewpoints, please see the section on Methodology.

6. Incidentally, the evening network newscasts tended to focus more on controversial matters than did most other TV news. Fully 59% of their stories were those involving clear disputes, compared with 46% on the NewsHour, 52% on morning news, and 54% on cable.

7. Tyndall’s research for the entire year found 20% to be devoted to foreign policy.

8. Our coding period ended in late September, before the debates and the home stretch of the campaign. Tyndall’s data for the whole year found 17% of the time on nightly newscasts devoted to the election.

9. The variable does not attempt to gauge the positive or negative nature of the event itself since that is often in the eye of the beholder. For example, candidate X’s winning a primary would be a positive event for some, a negative one for others and neutral for still others.

10. This is different from the tone coding done by some other researchers. The Center for Media and Public Affairs, for instance, codes every assertion separately and then tallies them up at the end to create percentages of positive and negative assertions. In coverage of a certain person or event, this suggests, the assertions were thus positive to this percentage point or negative to this percentage point. We believe the critical the unit of measurement is the story. This is how news is consumed, and impressions formed, story by story, not assertion by assertion. That is also closer to how journalists think of things – as whether a story is hard-hitting, down the middle or positive. This approach of ours also has a higher threshold. For a story to be positive or negative, it has to be decidedly and manifestly so – with statements in that direction outweighing all others by double.

11. James Kuhnhenn, “Press feels partisan pressure to perform,” Contra Costa Times, July 5, 2004. This story focused on the push by many Republicans and conservatives to discredit what they saw as a liberal media bias. The article quoted Senator Jon Kyl, an Arizona Republican; Deputy Defense Secretary Paul Wolfowitz, and Brent Bozell of the conservative Media Research Center.

Click here to view content data tables.