The Debate Effect
Thirteen media outlets-4 newspapers and 9 broadcasts-were monitored for 14 days. The monitoring period began Friday, October 1, 2004 and continued through the inclusion of Thursday, October 14, 2004.
Newspapers were selected to provide an understanding of coverage by both national press (New York Times, Washington Post) and regional publications (Columbus Dispatch, Miami Herald), representing both geographic and demographic diversity. Broadcast sources included both the flagship program and the morning news show from the three major over-the-air networks; the Newshour from Public Broadcasting; and CNN's News Night with Aaron Brown , as well as FOX's News with Brit Hume, representing content on cable all-news networks.
Inclusion and Screening
Both print and broadcast sources were monitored via the use of the NEXIS advanced search tool. Search criteria were designed to cast the widest net possible. Any appearance of the names Bush, Cheney, Kerry or Edwards in the headline or first third of the story, or any story indexed by NEXIS re: the presidential campaign was qualified and included in the original sample.
These criteria insured the inclusion of all pertinent stories; yet also produced an initial sample of several thousand articles or broadcast segments. The sample was next refined by eliminating duplicate stories, photo captions, letters to the editor and unedited transcripts of candidate debates or speeches.
Stories were then screened for agreement with the Project's inclusions rules. Stories less than 75 words long were excluded from the sample. Those stories that met the length requirement were next screened to identify cases where any designated Recurring Theme was referenced in the headline, subhead, or 3 lead paragraphs AND 1/3 or more of the text of the article was tied to any of the four major candidates or campaigns; OR any designated Recurring Theme constituted 50% or more of the story in a way that made direct reference to any of the four major candidates or campaigns.
The resulting project sample consisted of 817 stories, all of which were fully coded and are included in the final data analysis. For newspapers, there are some limited cases where NEXIS database will not deliver stories written by unaffiliated news services. However, all stories under the editorial control of these newspapers are included. All stories written by staff reporters, Op Ed pieces, and "specials to the news publication" are part of this analysis. For television, only weekday broadcasts of the pertinent shows are included, with the exception of the broadcast network evening news programs. (Note: the October 14th stories for the CBS Early Show have not yet appeared in NEXIS, and therefore are not included in this study.) While weekend evening news shows are often preempted by sporting events, those that did appear in the Vanderbilt Television News Archives for the dates of October 2-3, and October 9-10, are included.
Researchers worked with a detailed, standardized coding scheme. All stories were first coded for basic inventory variables-source, dateline, length, etc. Then, the project director coded for content variables — recurring themes, dominant candidate-and intent variables-story trigger, frame, tone, and impact. In all cases, coders worked with a defined set of rules per variable. Of particular note:
Tone: The 2 to 1 Rule: When calculating Tone, coders must quantify all the pertinent text that is positive for the Dominant Candidate, as well as all pertinent text that is negative for the Dominant Candidate. Additional weight is given to text within the headline of the story. In any case where the ratio between positive:negative equals or exceeds 2:1, the story is coded as positive tone for the Dominant Figure. Likewise, when the ratio between positive:negative equals or exceeds 1:2 the story is coded as negative tone for the Dominant Figure. All other stories are coded as neutral. In this study, stories determined to be straight news accounts were not coded for Tone.
Story Impact: The One-Half Rule: When calculating Story Impact, coders identify all text that implies which individuals or group's interests are at stake/affected by the events in the story. Only if 50% or more of the text makes this connection is the story considered for impact; all other stories are coded as "No impact implied at 50% or more."
Intercoder reliability measures the extent to which individual coders, operating independently of one another, reach the same coding decision. Tests were performed throughout the project: no systematic errors were found. As previously stated, senior project staff made all final decisions on both the content and intent variables.
The study of blog commentary on the Internet was drawn from five of the most popular and frequently updated political blogs: Andrew Sullivan's "Daily Dish," Atrios' "Eschaton," Glenn Reynolds' "Instapundit," and Joshua Michael Marshall's "Talking Points Memo." The four personal blogs studied were selected from the top-ten political blogs as of October 1st according to Truth Laid Bear's Blogosphere Ecosystem (http://www.truthlaidbear.com/ecosystem.php), a website that ranks blogs according to the number of links received per day (a measure of their influence on the political blogging community as a whole). Among the top ten, these four blogs were selected for their range of ideology and approach. In addition, we examined "The Note," a daily blog published by ABC News's political unit, to track a mainstream news organization's approach to political blogging.
Blog content was studied from 10:30 P.M. on September 30th (that is, starting at the conclusion of the first debate) through midnight, October 14th. Individual posts (designated by a time-of-posting marker) were used as the unit of analysis. In the case of "The Note," which is published once a day, headings created by the ABC News staff were used to divide each day's issue into individual posts.
Postings were selected for inclusion and coded if they were substantially about the presidential campaign. Posts that simply provided a link to an external site with no explanatory information or comment by the blogger were not included. However, posts that consisted solely of excerpts from outside writers (such as quotes from another person's blog or a newspaper article) were included within the study and coded.
The Coding Process: Researchers worked with a detailed, standardized coding scheme. All stories were first coded for basic inventory variables-source, date, time, etc. Then they were coded for content and intent variables-recurring themes, frame/focus and message. When coding for message, coders counted all assertions made by the individual blogger. In cases where the blogger simply re-ran narrative from an outside source, without any of personal comments, the assertions from the outside source were counted for message.