Content and Tone
In the end no consistent correlation between content or tone of a candidates press releases and their subsequent coverage in the media was found. This does not mean however that there were no relationships between content or tone and media coverage that were significant. The few examples where there was a significance relationship make perfect sense.
The only example where content had a significant relationship was the percentage of Joe Lieberman's policy releases. The percentage of Lieberman's policy releases was negatively correlated to his media coverage. This means that as the percentage of Lieberman's press releases that were policy oriented went up, his media coverage went down (See Predicted Values). This is easily explained. Despite his best efforts to galvanize"Joementum," Lieberman remains seen as a policy wonk, and, more generally a boring guy. It seems that the media was bored with this stereotype and saw no reason to cover Lieberman if he was not going to change.
One tone example fit our predicted results in terms of tone. The percentage of Wes Clark's negative releases was negatively correlated to his media coverage. This was true in overall coverage (See Predicted Values) and to coverage in newspapers (See Predicted Values). This means that as the percentage of Clark's releases became more negative, his coverage dropped. This is exactly what we had predicted. As candidates go more negative, the media pays less attention to them. The reason that Clark was the only candidate who was affected significantly could very well be that he issued more negative press releases than any other candidate per day.
Their was another instance of tone having an impact however and it did not meet our expectations at all. Howard Dean's percentage of negative press releases was positively correlated to his media coverage. This means, unlike Wes Clark, as Dean went more negative he got more coverage. There are number of possible explanations for this.One is that, since Dean was the front runner for most of the race, he had to issue a number of responses to other candidates attacks. The press may have picked up these responses since that can sometimes be one of the biggest test of the candidates. Another explanation is that it fit into the stereotype that the media was trying to create for Dean, that of the raging attack dog. In a campaign documentary that aired on CNN, Dean's campaign manager, Joe Trippi, complained time and time again that the media didn't pay attention to them when the campaign was positive but that they complained when they went negative. Our study may proven Trippi right.