I also wanted to look at stories about the candidates.
Since I wanted to use more than one or two raters and I had no research money to pay them, I could not take expect to get the stories themselves read. Instead I looked at only the headlines of stories about McCain and Obama.
I included headlines when the articles were about the candidates, even those where the candidate wasn’t specifically mentioned, such as “Religious Youth Vote Could Tip Scales in Battleground States†about Obama. I had to in order to be certain that I noted all the headlines on the candidates. I took multiple headlines for the same story only if the headline changed throughout the pages. I did not include multiple headlines for the same story if the headline remained the same on, for example, both the main page and the elections coverage page.
Once I had a collection of four days’ headlines on the two candidates, I went through and copied the headlines. There were a total of four pages’ worth of headlines for these four days, 85 headlines in all. Of these 85 headlines, three sets were headlines which changed based on the page they were on. Two of those, including one with three different headlines, were about Obama. The other set was about McCain.
Looking at how the headlines changed across the pages is interesting, because the different headlines were, ultimately, for the same story. The McCain headline was fairly banal. The first version was a non-candidate identifying headline which read “Value-Based Leadership.†The second version said, “McCain Speaks at Colorado’s Aspen Institute.†If I saw these two headlines alone, I would not have connected them to each other. They do not seem to be closely related.
One of the headline changes for Obama was similar in terms of lack of clear connection. The first version read, “Class Over Race?†while the second was, “Questions Over What Form of Affirmative Action Obama Favors.†Again I would not have necessarily thought those two headlines went to the same article.
The other set of headline changes actually had different headlines on different pages. The headline on the political page read “Campaign Grudge? Obama Launches Another Attack in Defense of Wife.†The second headline, on the election coverage page, read, “Obama Launches Another Attack in Defense of His Wife.†These two headlines, though different, do appear to belong together. However, there was an interesting phenomenon on placement of this headline on the political page. The first line read “Campaign Grudge?†It was written in a different color and different font than the rest of the title. The second line said “Obama Launches Another Attack.†The third line said, “Fox News Channel in Defense of His Wife.†I wondered if people scanning would put the headline together correctly or if the “Fox News Channel†insert would cause them to skip the rest of the title. The designation “Fox News Channel†was written in the same font as the rest of the title. I elected to test the shortened version as well. Understandably, this was perceived negatively by the most raters and had the least number of positive ratings of the three, garnering a single positive. Because of the potential reading, I included the fragment as one of the 85 headlines, even though it is actually only a fragment and was not a headline on its own.
Aside from those three headlines, the article titles kept the same form from link to actual story. That left an additional 78 headlines to look at.
This paper was presented at the national conference of Popular Culture on 8 April 2009.