Are the stories on the web different than those in newspapers?
Contrary to the idea that the net is full of opinionated argument or unsubstantiated innuendo, campaign sourcing on the Internet was strong. More than one-in-five (21%) of all lead stories had more than seven sources. And overall, more than half had at least five sources.
Only 4% of all the stories had no sourcing. Most of that came from MSN and the National Review. In all, 21% of National Review lead stories were not sourced, as were 17% of MSN stories.
Nearly 90% of all stories led with a named source. A full 100% of first sources on Yahoo!, Go/ABC News, CNN, AOL News, the New York Times and the Washington Post were named.
And, as one might expect, the most common first source was a candidate. Half of all stories studied offered one of the candidates as the first source. Three-quarters of Go/ABC News, MSNBC and CNN stories named a candidate as the first source. Salon was the least likely site to use a candidate as the first source (17%), instead leading more often than other sites with named polls (25%) and with unnamed sources (17%). Still, Pathfinder had a far greater percent of unnamed lead sources than any other site-a full 38%.
Despite the strong sourcing, it was still not as solid as in the traditional newspapers. More than half, 52%, of the newspaper stories have more than seven sources. Only two stories, or 8%, did not have any sourcing at all. The remaining 40% had between 4 and 7 sources.
The newspapers did have a greater percent of unnamed first sources, 3 stories or 12% versus 6% among the web sites. But they were twice as likely to use an outside expert as their first source (12% versus 5%). They used a voter as the first source twice or in 8% of the stories, which the web sites never did. Further, the newspapers never relied on an outside journalist for the first source, which the web sites did 4% of the time.
Within the text of these lead stories, sites can use the Internet's potential to access more detailed information. For instance, a story about McCain's speech regarding Pat Robertson could contain a link to the verbatim text or a recording of the speech.
Few sites in the study took much advantage of this opportunity. Less than one quarter of all stories had any such links.
Only MSNBC had a majority of lead stories with at least one link to unfiltered information (67%). But its cousin, MSN, did so only 17% of the time. This was also the case for CNN and AOL News for 18%. Only two lead stories on the New York Times web site (8%) linked to additional unfiltered information, and Netscape and National Review only supplemented a single story during the period studied.
Interestingly, most of the links to unfiltered information were links to direct primary results.
The Internet is also supposed to be continuous and immediate. We downloaded lead stories four times a day to see how often the lead story changed or was updated. We found that things changed a fair amount. Forty-five percent of all the stories were completely new and 10% of the stories had been edited or added to in some manner. But we also learned that new news does not necessarily mean better or more complete.
Three sites that tended to update the most often seemed to do so purely because they could, rather than to exercise news judgment. Netscape offered the greatest percent of completely new stories (83%), but these were always the straight wire feed and required little if any effort on their end. Yahoo! and AOL News also offered new stories quite frequently (58% and 64%), but they too ran wire feed.
MSNBC was the most likely to simply edit or tack on a new paragraph or two to the previous story. It did this in half of its stories, often resulting in a jumble of various "pieces" of news under one headline by day's end.
Among sites that mostly offered original reporting, the New York Times had the greatest percent of new stories (46%), followed by the National Review (42%) and the Washington Post and Go/ABC News (38% and 37%, respectively). The least timely was Time Inc.'s Pathfinder, which updated only 13% of the time.
Stories were also much more likely to be new in the morning than they were at night, suggesting that perhaps, there is still some sequence to the news cycle. Sixty percent of all the 9 a.m. stories were completely new, while less than a third (31%) were new at 9 p.m. At that nighttime hour, 18% were edited or updated versions of earlier stories, compared with 4% at both 9 a.m. and 12:30 p.m. and 13% at 4:30 p.m.
Aside from the lead stories, most sites only did a moderate amount of updating featured stories. In all, nearly a third, 32%, of the front pages had no changes to their featured stories. Another 30% of the pages only had changes in one or two stories, which includes the lead.
The bylines of the lead stories largely mirrored the amount of original reporting found earlier on the political front pages. Portals without formal connections to journalism companies offered little that was unique. Every single Yahoo!, Netscape and AOL News lead political story was bylined by Reuters. That is quite different than Pathfinder, Salon and Washington Post, which produced 100% of their lead stories. The National Review, New York Times and Go/ABC News led with stories written by their own staffs a vast majority of the time (96%, 83% and 92%, respectively). CNN and MSNBC tended to combine the wire story with their own reporting, while half of all MSN lead stories were written by another news organization.
The clear majority of lead stories (85%) was purely political (that is they were about horse race, endorsements, staffing, etc. rather than policy, record, biography). This is no surprise given that the days were usually leading into or out of a primary. Still it is worth noting the lack of space given to content relating to the candidates as people or to policy issues. Only 6% of all stories primarily considered the candidate as a person. Only 2%–5 stories out of 72-were about policy. Another 1% dealt with social issues.
The political stories primarily looked at the battle ahead (28% of all stories), tactical maneuvering (21%), or primary results (9%) .
When we looked to see how the topics were treated-or framed-we found, as we have in other studies, that the most popular way to write stories is as straight news accounts-or use of the inverted pyramid (who, what, when, where, how, and why). This made up 46% of all lead stories.
When reporters did use a frame, they focused on the battle. Writing stories around tactics and strategy was the most popular frame (20%), followed by horse race (14%). Again, this is somewhat to be expected in the heat of the primary races. Reporters framed their stories around the political system 6% of the time and around a candidate's leadership ability another 6% of the time. There was not a single lead story written around the candidates' policy beliefs.
These numbers differ somewhat from the front page stories in the New York Times and Washington Post newspapers over the same six days. The newspapers were less likely to offer straight news (24% versus 46% online). They were twice as likely to frame the news around horse race (28% versus 14% online). The papers were less likely to frame the stories around tactics (12% versus 20%). More of their stories focused on the political system (16% versus 6%).
Looking at specific web sites, the Microsoft portal, MSN, was most likely to write stories around the candidates' leadership abilities (25%), followed by the National Review (17%) and the New York Times (17%).
The Washington Post and Pathfinder, on the other hand, ran the most stories framed as horse race (42% each). All of the remaining Pathfinder stories, 58%, were framed as tactics and strategies. Tactics and strategy was also a popular theme on the National Review site (46%) and on Go/ABC News (42%).
CNN lived up to its reputation as a video wire service that offers the news straight without a lot of spin on the ball. A full 92% of its stories were framed as straight news, even higher than the portals offering largely Reuters wire copy. For instance, 88% of Netscape stories and 73% of AOL News stories were straight news.
Salon carried the greatest variety of frames: 17% were framed around a candidate's current behavior, 17% around other internal politics, 17% around the political system, 17% as other frames, 12% around both tactics and horse race and 8% around a candidate's leadership ability.
What triggered the eventual the lead stories online? Even in the fast-paced world of the net, decisions by journalists, rather than external events, remain a major force. Fully 41% of the stories were triggered from within the newsroom, roughly the same number as those triggered by something the candidates said (39%). Eight percent of all stories were triggered specifically by election results.
Wire stories primarily followed what the candidates said and did. Among the sites that relied on wire copy, Yahoo! tended to run stories triggered by the candidate, 63%, as did AOL News, 50%. CNN and MSNBC, which often combined staff writing with wire service reports, also tended to carry candidate driven stories (71% and 63%, respectively). Netscape was a bit of an anomaly because nearly 4 in 10 lead stories (42%) did not relate to the election. Consequently, 42% of the stories had neither a campaign trigger nor a press trigger.
Pathfinder, Salon, and the National Review, which might be called the online magazines, wrote almost all of their own lead stories and initiated most of them as well. (Pathfinder 96%, Salon 71% and the National Review 83%).
The New York Times and the Washington Post had a greater mix of triggers than did the other sites. On the Times' web site, 42% of stories originated from within the newsroom, 21% were triggered by something the candidate said or did, 17% by a campaign statement or action, 12% by election results, 4% by independent polls and 4% by something else. On the Washington Post site, 46% were press driven, 42% were candidate driven, and 12% were driven by election results.
5 The difference between tactical maneuvering and the battle ahead is subtle. Tactical maneuvering stories clearly reported on the latest strategy or tactic a candidate was using or would use, while "the battle ahead" talked about an upcoming fight in general, but stopped short of focusing directly on the candidate's tactics.