Of those who have watched the GOP debates so far, about a third say seeing the candidates perform has changed their mind, so says a new report out from the fine folks at Pew.
But do debates really persuade? The results are mixed.
There is a huge body of work on political debates, and in that literature you'll find a lot of it devoted to the persuasive power of such debates on the public mind. To unfairly summarize all this work in a few sentences, the short answer is debates sometimes, but rarely, matter, at least at the presidential level. There are a few famous anecdotes (Nixon vs. Kennedy, or Ford freeing Poland), but for the most part debates don't move a lot of people one way or the other.
Most of the research focuses on the final big debates in a presidential election. Very little has been done to study debates at this stage, when members of the same political party are squabbling for the nomination.
The Pew report is really asking people whether they think their minds have been changed, not whether minds really were changed. That may seem more academic than practical, but you'd be wrong. There's a big difference between what people think they know and their actual knowledge. Yes, the correlation between the two at the aggregate level is positive and significant, but buried in those data are the folks who think they know a lot ... but don't. Also buried in the Pew data, I suspect, are those who think their minds were changed ... but weren't. As long as we're clear on what we mean by the questions we ask, no harm done.