2004 Annual Report - Newspaper Content Analysis

Newspapers also stand out from other media for their originality. In all, 85 percent of the articles on newspaper section fronts were staff-generated. That compares to 32 percent of the lead stories on the Internet that were staff-reported. (For Internet this meant a staff person connected to the central news organization, i.e. a New York Times staff person who also reports in the print pages or a CNN correspondent who also appears on air.)

The comparison to television might be to stories with a correspondent involved or clearly some staff reporting, as opposed to brief stories or something explicitly identified as coming from an outside source. In the commercial evening news, 56 percent of stories were staff-produced. In mornings, 64 percent were staff-produced.

What about anonymous sourcing? Earlier studies have suggested the use of confidential sourcing in newspapers depends, not surprisingly, on the article. The level found here is basically consistent with what we found in 2001 in coverage of the response to the terrorist attacks. There, roughly 25 percent of all articles contained anonymous sources. In the broader 2003 study, 28 percent of all articles contained at least one anonymous source.

When confidentiality was granted to sources, there was almost always some attempt by the paper to describe the source's level of knowledge or potential biases, such as a police officer working on a case or a Republican operative. Only 2 percent of articles contained an anonymous source without some description of their relationship to events, such as "sources said."

At the other end of the spectrum, more than half of the articles studied (52 percent) contained the highest level of source description, that is four or more sources who were not only named, but of whom some attempt was made to describe their pertinent knowledge, expertise or potential biases.

How does newspaper sourcing compare to other media? Commercial network television news, evening and morning, was more likely to use blind anonymous sourcing (14 percent on commercial evening and 6 percent on morning versus 2 percent in newspapers) without any description of who the source might be. The commercial networks were about as likely to use at least one anonymous source that they tried to describe (29 percent on evening and 27 percent on morning versus 28 percent for newspapers).

The networks were also less likely than newspapers to contain the highest level of sourcing, four or more named and fully described sources (18 percent for commercial evening news and 8 percent for morning shows versus 52 percent for newspapers and 71% for A1 articles).