On Friday, the Associated Press sent out a report, based on a study by the Center for Media and Public Affairs, claiming a pro-Obama bias in U.S. network news coverage of the presidential election.

I'm not so sure about the study's methodology.

From AP via TheStar.com (posted Oct. 31):

Comments made by sources, voters, reporters and anchors that aired on ABC, CBS and NBC evening newscasts over the past two months reflected positively on Obama in 65 per cent of cases, compared to 31 per cent of cases with regards to McCain, according to the Center for Media and Public Affairs.

ABC's World News had more balance than NBC's Nightly News or the CBS Evening News, the group said.

Meanwhile, the first half of Fox News Channel's Special Report with Brit Hume showed more balance than any of the network broadcasters, although it was dominated by negative evaluations of both campaigns. The center didn't evaluate programs on CNN or MSNBC.

"For whatever reason, the media are portraying Barack Obama as a better choice for president than John McCain," said Robert Lichter, a George Mason University professor and head of the center. "If you watch the evening news, you'd think you should vote for Obama."

The center analyzed 979 separate news stories shown between Aug. 23 and Oct. 24, and excluded evaluations based on the campaign horse race, including mention of how the candidates were doing in polls. For instance, when a voter was interviewed on CBS Oct. 14 saying he thought Obama brought a freshness to Washington, that was chalked up as a pro-Obama comment.

When NBC's Andrea Mitchell reported Oct. 1 that some conservatives say that Sarah Palin is not ready for prime-time, that's marked in the negative column for McCain.

ABC recorded 57 per cent favourable comments toward the Democrats, and 42 per cent positive for the Republicans. NBC had 56 per cent positive for the Democrats, 16 per cent for the Republicans. CBS had 73 per cent positive (Obama), versus 31 per cent (McCain).

Hume's telecast had 39 per cent favourable comments for McCain and 28 per cent positive for the Democratic ticket.

To my mind, coverage should reflect the reality of what's being said on the ground. If you talk to 100 people, and 55 say good things about the Democrats and 45 say bad things about the Republicans, reflecting that fact in your subsequent story is not biased.

If you construct your reporting to give 50-50 positive-negative coverage regardless of what people are actually saying, then that's a form of bias.

If a news organization is actively trying to shape public opinion, that's bias. If it's reflecting public opinion, sorry, that doesn't bother me too much.

Addendum

Added Nov. 10:

Here's the CMPA's Oct. 30 news release.

Here's a July 28 release: Media bash Barak (not a typo)