About this video
Most of the time, a presidential debate can be hard to call one way or the other. It’s difficult to put a score to something that is ultimately decided in the minds of the vastly diverse American public.
However, computer science researchers at the University of Michigan have developed an algorithm that can analyze and measure the amount that one candidate linguistically matches their opponent and have found that matching your opponent in a debate leads to higher polling numbers.
Transcripts were collected and evaluated from debates dating back to 1976, and the researchers were able to determine that linguistically matching your opponent has an important effect on the perception of a third party viewer watching the interaction. Matching an opponent linguistically takes advantage of a concept called processing fluency, the act of using similar words as an opponent in order to make the information more digestible for the viewer. In the case of two candidates debating for the presidency, the ability to mimic your opponent could prove to be an important contributor to an eventual victory.
About the Professor
Daniel Romero is an Assistant Professor in Electrical Engineering and Computer Science as well as an Assistant Professor in the School of Information at the University of Michigan. His main research interest is the empirical and theoretical analysis of Social and Information Networks. His other focuses include understanding the mechanisms involved in network evolution, information diffusion, and user interactions on the Web.