@Twitter Political Index

The next 100 days count – and Twitter is counting our sentiments and online interactions to gauge the potential outcome of the presidential campaign in the U.S.: Today, Twitter launched a new site called the Twitter Political Index. According to the Twitter blog, the @gov team is analyzing the +2 million tweets every week to understand how the nation’s Twitter users feel about President Obama and his challenger Romney.

The index represents:

a daily measurement of Twitter users’ feelings towards the candidates as expressed in nearly two million Tweets each week.

Twitter teams up with USAToday’s election team and Topsy to display the sentiment results in the newspaper’s Election Meter. The sentiments are measured on a 0 to 100 scale and everything above 50 is coded as positive sentiment. A sliding scale lets users go back in time and shows sentiments including their related historical events (such as important visits, or speeches):

Adam Sharp, Twitter’s head of news and social innovation (@gov), shared some of the ideas and analysis mechanisms on NYT’s Timescast on August 2. Asked how Twindex represents the American public, he responded on Twitter:

 

I was interviewed for our local channel 9 News to talk about #Twindex: PollstePollsters using social media sites to gauge popularity of candidatesrs

Related articles:

Advertisements

About Ines Mergel

I am Full Professor of Public Administration at the Department of Politics and Public Administration at University of Konstanz, Germany. Previously, I served as Assistant and then Associate Professor (with tenure) at the Maxwell School of Citizenship and Public Affairs, Syracuse University, NY. In my research, I focus on informal social networks in the public sector and the adoption and diffusion of digital service innovations in government organizations. I teach classes on social media management, digital government, public management, and social network analysis.

5 thoughts on “@Twitter Political Index

      1. Yes, but twitter users may not be a representative sample. I’d guess they are not. I’m not sure the direction of the bias, but I’d be interested in knowing before inferring anything from this experiment.

      2. Adam Sharp, Twitter’s Government lead, responded in a tweet to me the following (https://twitter.com/AdamS/status/231190305259532290): “@InesMergel #Twindex avg for 2 cands since 5/1 w/in 1 point of each other. That + frequent Gallup correlation gives us confidence in balance”. I still have difficulty understanding how they justify that Twindex represents the American public. Here is a link to the NYT timecast where he was asked the same question: http://thecaucus.blogs.nytimes.com/2012/08/02/timescast-politics-a-preview-of-the-jobs-report/

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s