At RJI, we’ve been working to improve how we share information with our readers.

Subscribe

The Twitter debate analyst team. From left: Brian Houston, Esther Thorson, Justin Wolfgang, Stephen Bock and Dhawal Nikam The first Twitter presidential debate analysis from the Oct. 3 debate was truly a learning experience for the University of Missouri and Reynolds Journalism Institute analysis team.

Stephen Bock, Microsoft Application Development Lab (MADL) manager and Dhawal Nikam, an MU computer science graduate student, helped the analysis team with its technology needs and created programs to computationally analyze the Twitter data.

MU Professors Mitchell S. McKinney and Brian Houston, who are heading the research project up, provided expert political communication and debate event insights and helped analyze the debate for key moments.

A handful of other staff and students searched for key moments in the debates that increased Twitter activity and wrote subsequent reports

RJI Research Director Esther Thorson coordinated efforts with the three newspaper partners and also helped with the analytical work.

This was the first time RJI has partnered with newspapers to compare tweets from local communities to a national sample and tried to understand the differences and similarities between the areas. According to team members, it was a challenging process.

“As such, this effort was difficult and we ran into some stumbling blocks on both sides of the process (national and local),” said graduate student Joshua Hawthorne. “We are working to address the issues that popped up and to make the overall process more efficient.”

One such issue included some incomplete data because of an error related to the software company providing the tweets afterward. Since the data was incomplete, other data tools had to be utilized. Team members said they’re planning to include a backup data provider for the next debate.

“Despite these issues, I am very happy with our results from our analysis and I believe that they provide interesting insights into the differences and similarities between the issues that matter to the local communities of the newspapers we partnered with and the national voice expressed through Twitter,“ said Hawthorne.

Response from MU researcher

McKinney has been studying presidential debates for decades and is internationally recognized for his research.

As the team began analyzing the tweets, McKinney was surprised at how difficult it was to try to make sense of so many tweets.

“We’re working on a way to more smoothly harvest the tweets, and we’ll be continuing to work on more sophisticated ways to try and analyze the content of the tweets,” said McKinney. “That’s not an easy thing when you’re trying to understand what tens of thousands of people are saying.”

Another aspect that surprised McKinney were some of comments received — quite a bit of “snark,” humor, profanity, etc. in the tweets. But there was also quite a bit of analysis, he said.

“While there’s still a lot of “entertainment” value to tweeting, people still tweet “serious” comments regarding the issues, “fact check” the candidates and what they’re saying, et cetera,” said McKinney.

Jennifer Nelson  
   
Senior Information Specialist



Share

Recommended for You

Related Stories

comments powered by Disqus