Whether a citizen gets his or her news from the newspaper, the Internet or television, this study suggests, could have a great influence on the views they form about the candidates and the election as a whole.

Television tended to cover campaign themes more than any other medium—68% compared with 56% on the Internet and 53% in newspapers. Television was especially inclined to produce pre-and-post-debate analysis (32% versus 22% overall), suggesting that the made-for-TV event makes easy TV news. Television was much less inclined than other mediums to produce stories about the candidates' characters (5% versus 16% of Internet and 15% of print).

When it came to who television covered, it was much more likely than average to cover both candidates equally within a story (63% versus 47% overall). This suggests that the ever-increasing pressure to make stories shorter and to cut hard news time, leads journalists to squeeze coverage of both candidates into one short piece. Perhaps as a result of the minimal coverage, their stories were more likely than print or Internet to be neutral in tone (38% for TV versus 32% for Internet and 28% for print).

But when it came to how stories were framed and who they impacted, television fell into the same old trap of covering politics over policy. It was the most likely medium to frame stories around internal politics (50% versus 36% for print and 43% for Internet) and was the least likely to produce news that impacted citizens (20% versus 32% for print and 23% for Internet.)

Who Each Medium Impacted
 
Citizens 32% 20% 23%
Politicians 59 70 68
Interest Groups 6 4 6
No Impact 3 6 3
Total 100% 100% 100%

Newspapers, on the other hand, stood out as the medium most likely to write stories that impacted citizens (32%), though nearly six-in-ten print stories still mainly impacted politicians. Interestingly, a big reason is because editorials and Op-ed columns were more likely than other news accounts to connect the campaign directly to their impact on citizens.

In the newspaper coverage studied, stories about Gore were about twice as likely as those about Bush to make it one page one (14% for Gore versus 8% for Bush), though the two did receive roughly equal treatment in the editorial and op-ed pages (about a fifth each).

One interesting difference between the national and regional papers was in their coverage of character themes. Regional papers were almost three times less likely than the nationals to cover Bush's character (4% versus 11% in national papers).

Regional papers were much more likely than nationals to write stories that compared the two candidates (51% versus 39%) and were less likely to be negative in tone (43% versus 58%), with a quarter of their stories offering a positive tone. Finally, a third of all regional stories impacted citizens, compared with only a quarter of national print stories.

The Internet stood out in that it seemed to be more driven by stories that came from campaign operatives (12% versus 8% of television and 7% of print). Perhaps the Internet still sees itself as speaking to a more targeted audience, the "political junkie," and therefore provides more of an insider view.