MR. KOJO NNAMDIFrom WAMU 88.5 at American University in Washington welcome to "The Kojo Nnamdi Show," connecting your neighborhood with the world. A skilled politician can be like a magician on a debate stage. Ask a pointed question about a policy flip-flop and behold a slight of hand.

MR. KOJO NNAMDIThey know which hand your eyes will follow while they thank you for asking the question. You might watch closely from step to step, but you still don't notice how they subtly begin to change the subject as they slyly tiptoe around any admission of weakness or failure and ultimately answer a different question altogether.

MR. KOJO NNAMDITonight, Barack Obama and Mitt Romney meet on the campus of Hofstra University for the second of three presidential debates and if past precedent holds, viewers may find it very difficult to cut through the rhetorical smoke and mirrors of the candidates. But new research and new digital platforms are making it easier to detect and perhaps even discourage the worst examples of debate dodges and spins.

MR. KOJO NNAMDIJoining us in studio to talk about this is Philip Resnik. He is a professor of linguistics and advanced computer studies at the University of Maryland. Most recently, he is the creator of a new real-time polling app called React Labs. Philip Resnik, good to see you again. Thank you for joining us.

MR. PHILIP RESNIKThanks, great to be back.

NNAMDIAnd joining us from studios at Harvard is Todd Rogers. He's a professor of public policy at the Harvard Kennedy School. He was founding director of Analyst Institute LLC which uses behavioral science to understand and improve voter turnout. Todd Rogers, thank you for joining us.

MR. TODD ROGERSA pleasure to be here.

NNAMDITodd, savvy politicians have many weapons in their arsenal when it comes to avoiding tough questions. They can completely evade a question. They can re-interpret the question or they can just plain, well, lie. You've studied the full arsenal and you've actually broken down and analyzed the various rhetorical tactics politicians employ. But why do politicians dodge?

ROGERSWell, it's a good question. I mean, what my collaborator Mike Norton and I have been studying is whether people are able to detect when candidates dodge and under what conditions we can increase that dodge detection. And so, if you'll let me, I'm happy to describe the basic paradigm, Kojo.

NNAMDIPlease go ahead.

ROGERSSo what we'll do is we'll bring participants in and we'll ask them to watch one of three videos. In one of them, a moderator asks a question that says, what will you do about health care? The politician says, I'm glad you asked me that. There are so many important problems facing America today. We need universal health care because, and then they go on for two minutes about why we need universal health care. So that is a direct answer to the question.

ROGERSWe then, using video editing software, splice out the question and replace it with a different question. What will you do about the illegal drug use problem in America? The politician says, I'm glad you asked me that. We need universal health care because, the exact same response.

ROGERSAnd then a third group listens to, what will you do about the war on terror? And the politician says, I'm glad you asked me that. We need universal health care because. And so afterwards, after watching this whole thing, we'll ask the participants how honest, trustworthy and likable is the politician? And then we'll ask them what question was it the politician asked?

ROGERSAnd what we find is when the politician answers a question that is totally different than the one that they were asked, but could be seen as somewhat close, like illegal drug use, and they answer about health care, people think they're just as honest, trustworthy and likable as if they had answered about health care and they can't remember what the guy was asked.

ROGERSSo this artful dodge is answering a question that is different than the one you were asked, but is similar enough that people won't really notice. So the war on terror, everyone noticed and they thought that guy was a jerk. So the artful dodge, what we detect in this research is because of people's limited attention, they fail to notice these subtle dodges which really licenses being able to evade questions that they are specifically asked.

NNAMDI800-433-8850 is the number to call if you'd like to join this conversation. Do you find debates to be useful exercises for understanding the differences between the candidates or do you think the format itself ends up favoring style over substance? More importantly, do you feel debates reward candidates who answer honestly? 800-433-8850, because, Todd Rogers, how good are we as viewers at punishing people who dodge questions or rewarding people who actually answer them?

ROGERSSo when we detect it, we think they're jerks and we punish them. And what has been astonishing and basically motivated this whole line of research is how vulnerable we are to not detecting it. And so when we detect it, when we are explicitly watching it with the instruction to notice if the guy answered the question, people are great at it.

ROGERSWhen the text of the question is posted on the screen, people can absolutely detect dodges. But absent those two interventions, it's very difficult.

NNAMDIPhilip Resnick, in the era of Twitter and Facebook, for many people watching these debates, it's a social experience. We can follow along with the public conversation and weigh in when we detect spin or a dodge. You've developed an interesting platform called React Labs. Tell us about it.

RESNIKWell, sure, the idea actually began with very similar motivations to what Todd has described. I'm interested in the way that political language is used and people's reactions to it and how it affects them.

RESNIKIn my case, I'm a computer scientist and so the approach that I decided to take was building something that could actually collect data out in the wild and over the course of a year or. So I found myself collaborating with political scientists, notably Amber Boydstun, a professor at UC Davis and basically developing a real-time polling platform that runs on mobile devices, so smart phones, laptops, iPads.

RESNIKAnd what it does essentially is it allows the traditional sorts of demographic and survey questions, name and, are you male or female, and issue questions and so forth. But what it does differently is while you're watching the live event, it allows you to react in real-time by tapping buttons. Do you agree? Do you disagree? Do you see this as spin? Do you see this as a dodge?

RESNIKWe collect all of that data. We're able to chart it in real-time and able to analyze it afterwards. And so in collaboration with Professor Boydstun and other colleagues, we've actually created a program that's gotten thousands of students across the country reacting to the debates and giving us incredible data to analyze, as well as doing some things with viewers here in the D.C. area and again, collecting really interesting data.

RESNIKThe whole idea here is to shed light on something that, up to this point, has been, as you alluded to, something of a black art.

NNAMDIWell, tonight's presidential debate will be town hall style and CNN's Candy Crowley will be moderating. Let's talk about the role of the moderator. After the first debate, many supporters of Barack Obama complained about moderator Jim Lehrer, feeling that he allowed the candidates to seize control.

NNAMDIConversely, some conservatives were unhappy with the job that Martha Raddatz did, which was more forceful during the vice presidential debate. How does the moderator affect the viewer's perception of the debates? But before you answer that, allow me to have Ernest in Alexandria, Va. on because he would like to weigh in on the moderator issue also. So Ernest, you're on the air. Go ahead, please.

ERNESTHi, Kojo, thanks so much. Yeah, great comments on this. It seems to me the moderators are often pulling their punches because they don't want to lose access to the politician. But maybe they could put the question they've just asked at the bottom of the screen while the person is answering it so that the viewers would have a parity issue going on.

ERNESTAlso, the app sounds terrific, but I'm a little bit concerned about the difference between immediate reactions and reactions that may take place later. So I'd love to hear comments about that, okay, Kojo.

NNAMDII'll have Todd answer the first part of the question and Philip the second. Todd, you go, on the moderator.

ROGERSSo I think Ernest nailed it. So moderators have a very difficult task. Their task is to control two of the most powerful men in the world from saying whatever they want and walking all over the moderator, while also appearing likable and professional. And so Jim Lehrer got run over in the first presidential debate.

ROGERSHe's also moderated 12 presidential debates and he was invited to moderate that out of retirement because they like having him. So there's this delicate balance where they need to moderate in a proactive, professional way, but they can't be too aggressive because they have this professional reputation and audience.

ROGERSAnd so Martha Raddatz, I think, did what we should expect moderators to do, which is hold the politicians in an unbiased way to answering during finite windows, keeping it on task, asking questions, moving the debate along. She occasionally pushed the politicians on questions that they failed to answer, but not all the time.

ROGERSAnd so Ernest nailed it. Our research suggests there's an easy intervention that helps the moderators and will help viewers and hopefully make politicians answer questions, which is if you post the question on the screen, everyone can detect if the politician fails to answer the question. So it becomes completely transparent.

NNAMDILater, we'll hear a clip from the vice presidential debate that Martha Raddatz moderated. But since we're on the subject of moderators, some people might have felt that Jim Lehrer went a little bit too light on Barack Obama and Mitt Romney or that Martha Raddatz was too harsh in policing the vice presidential candidates, but the gold standard of moderators doggedly pursuing an answer from a politician may come from the U.K.

NNAMDIHere is an example of journalist Jeremy Paxman questioning Michael Howard, a former head of home security in the U.K.

MR. MICHAEL HOWARDMr. Marriott was not suspended. I was entitled to express my views. I was entitled to be consulted.

MR. JEREMY PAXMANDid you threaten to overrule him?

HOWARDI was not entitled to instruct Derek Lewis and I did not instruct him and the truth...

PAXMANDid you threaten to overrule him?

HOWARDThe truth of the matter is that Mr. Marriott was not suspended.

PAXMANDid you threaten to overrule him?

HOWARDI did not overrule Derek Lewis.

PAXMANDid you threaten to overrule him?

HOWARDI took advice on what I could or could not do.

PAXMANDid you threaten to overrule him?

HOWARDAnd I acted (word?) in accordance with that advice. I did not overrule Derek Lewis.

PAXMANDid you threaten to overrule him?

HOWARDMr. Marriott was not suspended.

PAXMANDid you threaten to overrule him?

HOWARDI have accounted for my decision to dismiss Derek Lewis.

PAXMANDid you threaten to overrule him?

HOWARDDetail before the House of Commons.

PAXMANI note you're not answering the question, whether you threatened to overrule him.

HOWARDThe important aspect of this which is very clear to bear in mind...

PAXMANI'm sorry, I'm going to be frightfully rude...

HOWARDYes, you can.

PAXMANI'm sorry. It's a straight yes or no.

HOWARD(unintelligible)

PAXMANDid you threaten to overrule him?

NNAMDITodd Rogers, what do you take from that exchange? Apart from amusement obviously...

ROGERSOh, my gosh, I love that clip. It's amazing. So Jeremy Paxman and Michael Howard both ended up being hurt by that interview. So Jeremy Paxman, we should all respect him and honor that, that is what we would love to see interviewers do, but people thought that he was aggressive and slightly mean-spirited, not everybody, but it hurt his reputation as particularly aggressive.

ROGERSAnd Michael Howard was like, wonderfully evasive. So yes, that would be -- I think we all have this vision that would be the ideal, is hold these guys to answer the question. But absent that, given the multiple competing pressures on these moderators and interviewers, the next best option is to let's just do a subtle tweak on the visual experience and post the question on the screen. And the research shows people can detect that and these politicians will realize that their dodges are being naturally detected.

NNAMDIAnd Michael Howard is home secretary. Philip Resnick, I never gave you the opportunity to respond to the second part of the question that our caller asked about and that is debate watchers plugging into social websites following the instant debate, the constant flow of information from several sources. Does that aid or inhibit the viewers' understanding of the debate?

RESNIKSure. So I mean, Ernest asks -- in both parts of his question he asked a great question. He was concerned about the idea of capturing instantaneous reactions versus delayed sorts of reactions...

NNAMDIYes.

RESNIK...and I think, you know, I'm going to steal a line from the Dean of Maryland's School of Public Policy who, when we were talking about this, talked about the idea of finding out what people think before the pundits tell them what to think. There is something different about looking and monitoring what people are doing in the moment as contrasted asking them after the fact.

RESNIKAnd so you can view what -- what we've been doing with React Labs is kind of the other side of the coin to what Todd has been doing. He's got research that's conducted in the lab that has, you know, shown some very compelling results. And what I've been focused on is, in effect, a measurement device, a way of reaching out so that you don't have to bring people into a lab or give them, you know, a piece of hardware with a dial on it in order to monitor what they're doing.

RESNIKAnd the goal here is really the best of both worlds. You want to be able to ask people specific questions the way pollsters do. So, for example, has your opinion of the candidate changed or are you likely to vote for them and so forth. But at the same time you want to capture things that they may not be thinking about quite as consciously in order to gauge the effects of the political language. So it would be -- and that's one of the reasons why Todd and I have started talking about doing experimentation together so that we can actually monitor what's happening out in the wild.

NNAMDISo to speak I was just thinking as you were saying that if we were able to either listen to -- and, Todd, is there a distinction to be made between people who exclusively listen to debates and people who watch debates if somebody heard a debate exclusively on the radio and saw no video at all, is their reaction likely to be different than if they saw it on television?

ROGERSI presume as much. I mean, I know there is research on the Kennedy debate -- the Kennedy/Nixon debate where apparently people who listened to it thought Nixon won. And people who viewed it thought Kennedy won. I presume the experience is very different but I -- you know, only from a common sense standpoint. I imagine that the demeanors of people are more important in the visual. So for example...

NNAMDIWhen you watch it on television.

ROGERSSo for example, Joe Biden's facial expressions, you know, some people loved it, some people hated it. I bet that that just got exacerbated by it being visual.

NNAMDIOkay. Speaking of Joe Biden, we'll get to him in a second, but first Mitt Romney. The best way to understand the difficulty of detecting dodgers might be to take a look at a few examples. So first, we'll listen to an example from an interview with Mitt Romney on 60 minutes.

SCOTT PELLEYDoes the government have a responsibility to provide health care to the 50 million Americans who don't have it today?

MITT ROMNEYWell, we do provide care for people who don't have insurance, people -- if someone has a heart attack they don't sit in their apartment and die. We pick them up in an ambulance and take them to the hospital and give them care. And different states have different ways of providing for that care.

PELLEYThat's the most expensive way to do it...

ROMNEYWell, the...

PELLEY...in the emergency room.

ROMNEYAgain, different states have different ways of doing that. Some provide that care through clinics...

NNAMDITodd, can you dissect that dodge for us? How does Mitt Romney move from the question that was asked to the question that he ultimately answers?

ROGERSYou know, it's amazing. It's like a magician where you misdirect attention from one space to another and then you can seamlessly move. He's asked specifically does the government have an obligation and then he answers. I guess if you were to listen to just the answer and try to infer what the question he was asked is you might think that he was asked, do people who don't have health insurance still get care and how do they get it? And then the moderator then -- then the interviewer then asks a follow-up question to Romney's response and they swiftly move on.

ROGERSAnd I'll bet that most listeners, unless they care passionately about that specific question about government obligation, most listeners thought it was a -- you know, an important topic and were just generally following along. And meanwhile Mitt Romney got to not answer explicitly whether the government has any such obligation.

NNAMDIAnd Philip Resnik, do people spot that immediately and say, hey he just dodged.

RESNIKIt depends on the answer. One of the things that you get out of the mobile devices allowing people to react in real time is you can actually see the flow of reactions so that you can see that at the beginning people may be reacting one way but as it continues they react a different way. And in fact when a candidate turns a question you can actually see evidence of that turning. I've also seen this happen with, you know, looking at people's reactions to campaign ads.

RESNIKSo people are actually often very aware of this. One of the things that it's very interesting to be able to do though is to break this down by the specifics of the viewers who are reacting. So there's a cost benefit analysis that was alluded to earlier. You know, should the candidate answer the question or can they get away with dodging the question? That cost and that benefit is going to differ depending on the particular audience you're reaching.

RESNIKAnd so one of the things that it's important to do, and one of the things that things like Twitter make it more difficult to do is to understand the properties of the people who are actually responding. So are they male or female? Are they Hispanic or Asian or African American? Are they predisposed to vote Democrat or Republican or are they independent? When you go for large quantities of data it gives you the opportunity to do that analysis. And what you might very well find is that something that looks like an overall wave of reaction once you break it down into the smaller demographic pieces gives you a lot more insight.

NNAMDILet's take another example, maybe more subtle of a dodge. This time it's Vice-President Joe Biden. This is a sound clip from last week's debate.

MARTHA RADDATZOne month ago tonight on the anniversary of 9/11 Ambassador Chris Stevens and three other brave Americans were killed in a terrorist attack in Benghazi. The State Department has now made clear there were no protestors there. It was a preplanned assault by heavily armed men. Wasn't this a massive intelligence failure, Vice-President Biden?

VICE-PRESIDENT BIDENIt was. It was a tragedy, Martha. It -- Chris Stevens was one of our best. We lost three other brave Americans. And I can make absolutely two commitments to you and all the American people tonight. One, we will find and bring to justice the men who did this. And secondly, we will get to the bottom of it. And whatever -- wherever the facts lead us -- wherever they lead us we will make clear to the American public -- because whatever mistakes are made will not be made again.

NNAMDITodd Rogers, Raddatz asked a very pointed question, was the recent Libya attack an intelligence failure and Biden escaped answering it. Based on his delivery how well do you think viewers would recognize that as a dodge?

ROGERSYou know, if you heard his question you might think that he was asked, what will you do about what happened in Libya? And I think that most people thought that that -- were interested in that question as well. So what we learned from our research is that by default when you are watching a speaker or a politician or a communicator or a CEO or a press conference, when you're watching someone you're by default evaluating first, do I like this person? Is he trustworthy, honest, likeable? That's the sort of default social evaluation we do.

ROGERSAnd so when we're watching people the first thing we're doing is paying attention to whether we like them. And then afterwards we're reflecting on what they're saying, and that's sort of secondary. If it's egregious then we can make the contrast. But it's really cognitively difficult. It's hard with our limited attention to be able to constantly ping back to, is this answering the narrow question that Martha Raddatz just asked?

ROGERSYou know, so for example, Kojo, what question did you just ask me?

NNAMDII asked you if it is possible to detect whether Joe Biden was dodging the question, but I have a better one than that, Todd. I got an email from Ed in Bethesda, Md. who says, "I just wanted to point out that when Kojo asked the first question about why do politicians lie, the guest actually didn't answer the question, Instead explaining the mechanics of his study and the results."

RESNIKYou know, I noticed that but I wasn't going to point to it.

NNAMDIAnd I suspect that Todd did it deliberately. Did you do it deliberately, Todd?

ROGERSI did and I've been trying to -- so what I wanted -- I did and then I realized it was the first question asked. It was too soon to say, Kojo -- or listeners, do you remember what question I was asked and did I answer it? I just thought it was too soon. But, yeah, Ed, very incisive.

NNAMDIVery astute, Ed. Would you care to comment, Philip?

RESNIKYeah, actually I want to go back to the Biden clip there a second ago.

NNAMDISure.

RESNIKOne of the things that I find interesting there, you'll notice that he didn't an immediate kind of one-word response and then go straight to what he wanted to talk about. But you notice that in that later part he refers to he says, "whatever mistakes were made" -- there's that famous phrase, mistakes were made in there -- this is an example of a particularly subtle kind of framing or spin.

RESNIKI refer to it as syntactic framing. It's a place where you're not using a beat-you-over-the-head kind of phrasing, like death tax versus estate tax or something. But rather you're actually using the syntactic -- the grammatical structure of what you're saying to subtly play down certain semantic aspects of what you're describing when you use a passive construction or there are other constructions like this. My four-year-old when he says, Daddy my toy broke, right, as opposed to Daddy, I broke my toy.

NNAMDIOf course.

RESNIKHe's -- it's not a passive. It's something call encoded construction, but he's doing exactly the same thing that Ronald Reagan did and that Biden did there. He is deemphasizing the causal of role, the (word?) role of the doer of an event. This is a particularly subtle kind of spin or framing. And in fact it's exactly this sort of thing that made me want to collect more data to figure out whether people are noticing it and what kind of effect it has.

NNAMDIAnd we're going to take a short break, but when we continue this conversation on the dark art of debate dodgers and spins, you'll hear Todd Rogers describe something close to what you just did -- or what your son just did. He calls it paltering. We'll find out what that is. But you can still call us at 800-433-8850 or send email to kojo@wamu.org. I'm Kojo Nnamdi.

NNAMDIAs we prepare for tonight's presidential debate, which will be carried live here on WAMU 88.5, the NPR broadcast or which you may be preparing to watch on TV and maybe listen to at the same time, we're having a conversation about the art of debate dodgers and spins with Philip Resnik. He is a professor of linguistics and advanced computer studies at the University of Maryland. Most recently the creator of a new real time polling app called React Labs.

NNAMDITodd Rogers joins us from studios at Harvard. He is a professor of public policy at the Harvard Kennedy School and founding director of Endless Institute which uses behavioral science to understand and improve voter turnout. We're interested in hearing from you. You can call us at 800-433-8850.

NNAMDIDo you find debates to be useful exercises for understanding the differences between candidates or do you think the format itself ends up favoring style over substance? 800-433-8850. Todd, when we went into that break, I promised that you would describe a deceptive strategy called paltering, which seems to be when you answer truthfully, but in a deceptive way, and it's my understanding that you have actually found or diagnosed yourself doing this in real life.

ROGERSI didn't realize it was going to be confessional.

NNAMDIYes, please. We're all confessing today.

ROGERSI -- so if you are asked a question you would rather not answer, there are a few things you could do. You can lie, which I think that the fact checkers and people who are knowledgeable and potentially moderators help protect us against a little bit in these debates. Another is to outright dodge the question. So to be asked a question about how will you create 12 million jobs, and to go launch into how if Romney is asked that question, and then instead of answering it he will launch into an attack on Obama's jobs policy.

ROGERSThat's a dodge. That's failing to answer the question, and listeners are not left with any sense of what the answer to the actual question was. Instead, they've mostly just forgotten what the question was.

NNAMDIMm-hmm.

ROGERSAnd they've moved onto the content of the response. And another strategy is to, what we call palter. And it's actually -- it's a word, an English word that is underused and I think is really a fascinating one. It is to say something that is factually true and knowingly lead your listener to a false conclusion. So...

NNAMDIMy toy is broken. But go ahead.

ROGERSRight. Right. Right. I found -- yes. Exactly. So you may again -- Romney may be asked, how will you create 12 million jobs, and he will say, economists agree that my plan will create this jobs by doing various things. What that certainly leads the listener to believe is that there may even be consensus that his jobs plan will create 12 million jobs, but what he may be saying truthfully is that there are a bunch of economists who agree, there may be more who disagree. But he is selectively reporting, knowingly leading the listener to a false conclusion.

ROGERSThe example that I use in this confessional is the...

NNAMDII can't wait.

ROGERSSo I -- yeah, exactly. And if the person is listening, he does live in D.C., so I -- an email was sitting in my inbox for about a month, and I -- it was asking me a serious question, and I just didn't get around to it, and then one day I replied to it, and I have to account for the fact that I ignored it for a month. And so what I did was I backed away from my computer, looked out the window, paused or about half a second, looked back at my screen and wrote, you know, I've been thinking about your email and proceeded.

ROGERSAs if suggesting that for the last month I've really been deeply ruminating on it, when in fact, all I was doing was honoring to myself that I wasn't saying something untrue, hopefully leading the respondent -- the target, who is a D.C. resident to believe that I had been thinking about it much more than the half second that I actually did. So in the debates, they can lie, they can dodge, they can palter. I think they dodge more often than they palter, but they palter too, and that's where the fact checkers come in and they start saying that they're saying untrue things when it's really just -- they're saying factually true things, but misleading people.

NNAMDIPhilip Resnick, hereinafter to be known as the respondent, or the target, were you the person that his email was actually sent to?

MR. PHILIP RESNICKYou know what, I don't want to put Todd on the spot. It's possible.

NNAMDIJust checking to see if you...

ROGERSYou know, Phil, I've still been thinking about the email.

NNAMDISee?

RESNICKThat makes me feel so much better.

NNAMDIThank you. Onto the telephones now. Here is E.W. in Silver Spring, Md. E.W., you're on the air. Go ahead, please.

E.W.Hey, great show today, gentlemen. I thought it was very insightful when one of the gentleman there said it's almost like a magician. Instead of sleight of hand, it's slight of tongue. What I thought was a good idea that the first called mentioned as far as having, you know, what the question was to scroll down at the bottom, but better yet, I think another good question is -- or should I say idea, would be, if they're going to know the questions anyway, have the candidates write out their answer and then when the question is posed, I mean, they can say what they want, but the answer would be pretty much there for us to see.

E.W.You know, if it's an elaborate one or, you know, paragraph, sentence, or what have you, you know, we would know if it's being answered or not, and I just wanted to know what your guests thought of the idea.

NNAMDIYou seem to be suggesting that the candidates know the questions beforehand?

E.W.Yes. I mean, their candidates -- should I say their campaigns -- the questions are ran through their campaigns, aren't they?

NNAMDINo. The questions are not run through their campaigns, not the questions coming from the moderator. Their campaigns run through presumably thousands of possible questions that could be coming to them, but the one thing that has remained constant throughout the campaigns throughout these debates is that the moderator never submits his or her questions to the campaigns or, for that matter, to anyone else beforehand. But I guess you're thinking that the candidates should have a form in which they can write answers to their questions. Have you ever heard of one such, Philip Resnick, writing the answers?

RESNICKOh, you know, I think there are some websites that have asked candidates to...

NNAMDIThis is true.

RESNICK...answer questions, and, you know, although people don't pay too much attention to party platforms...

NNAMDIThat's true.

RESNICK...the party platform is another kind of archival statement about what the candidate, or at least what the candidate's party stands for. I think that tonight's debate is -- in terms of E.W.'s comment, it's particularly interesting because it's a town hall format if I understand correctly and, therefore, more unpredictable, and there's a certain rhythm...

NNAMDIThat's true.

RESNICK...to the way that moderators and candidates interact, and this adds, you know a whole other dimension, that I think the candidates are going to need to deal with, and I suspect what you're going to see here, um, is very much the same sort of phenomenon that you see with a good improvisational artist, whether it's a magician or an actor, you know, reacting on the fly. You may see a lot more fine tuning on the fly than you see in the podium stand up sorts of debates, specifically because they actually need to adjust to the nature of the question that they weren't expecting.

NNAMDIAnd you might see some written answers from candidates in this area if you got to voterguide.wamu.org, onlinevoterguide.wamu.org. It is my suspicion that most candidates asked to answer questions in writing would rather write them in lead pencil because anything that is written in lead pencil can be erased, and a lot of candidates would like to do that. But Todd Rogers, since the moderator cannot possibly fact check every statement in a debate, do you think there's risk that candidates won't dodge questions, but will just respond untruthfully?

ROGERSSo first, Kojo, if I can share a classic quote that was related to the previous caller, Henry Kissinger used to approach the podium and speak to the press corps and say, okay, who has questions for my answers? Right? So these politicians may get -- have a thousand varieties of questions, but they're probably going to come with 15 or 20 answers, you know? And it's like a game trying to get as quickly as possible to the answer you wanted to give.

ROGERSYour question about the moderator brings to light this -- reminds me of this debate that's going on right now where if anybody has followed this in the news, the two campaigns are -- have their legal teams challenging Candy Crowley who is going to moderate tonight's debate. She's been saying that she's going to actively moderate the debate, and she will even ask questions and maybe even push people to answer questions if they dodge them.

ROGERSAnd the campaigns are protesting that she is going to take such a vigorous role because they want her to instead just call on people in the audience, and it's just a rare glimpse into the highly managed process of trying to prevent moderators from doing their job, and I like that Candy Crowley seems to be taking a pretty aggressive position on it.

NNAMDIWhether or not she will be able to accomplish that tonight is another story, but back to the telephones again. Here is Hassan in Alexandria, Va. Hassan, you're on the air. Hassan, you're on the air. Go ahead, please.

HASSANHello, Kojo. How are you?

NNAMDII am well.

HASSANAlmost four years ago, I called you and I said Obama would be the president, and I want you to introduce me to him because you have access to him.

NNAMDII remember that.

HASSANAnd he will get elected again, and this time you have to promise me.

NNAMDIWell, frankly I -- frankly, I don't remember the last time, so I can't make any promises this time. But what was the major point of your call?

HASSANOkay. My point is this. I watched the debate. I think he did well. Mr. Romney said he will do so many things, but he never said how. And then the media dramatized the situation. They declared the winner when we the elector declare the winner, and I wish the media can be a little bit patient and give the opportunity to the people to decide who did well and who did not well.

NNAMDIWhich is what I find, Philip Resnick fascinating about what you are doing. Because invariably when we watch debates they are preceded by and followed by the analysis of quote unquote "experts" who happen to be pundits, often very based in Washington, right here who weigh in on these debates. What you are doing is allowing everybody else -- all of the people that Hassan is talking about, to weigh in before even the pundits.

RESNICKAbsolutely. I mean, the idea here -- one of the ideas behind this is that traditional ways of tapping into public opinion are having trouble. You have landline telephone polling and, you know, lots of people don't even have landlines anymore, and they don't want to speak to somebody who's calling at dinner time at asking them questions. At the same time, exactly these people are using their mobile devices all the time, including a huge percentage while they're watching TV.

RESNICKAnd so there is an opportunity to take advantage of technology and tap into the energy. People want to be heard. It's not that they don't want to be heard, it's that they want to be heard on their own terms, social media, the overall sense of communicating through mobile devices, the two-way television experience. All of that is really going to change the way that we tap in to what people are thinking, and my political scientist collaborators and I are collecting whole bunches of data, both to answer traditional political science questions, but also to analyze this data and figure out what further insights we can get from it.

RESNICKGoing back for a second to the fact checking during the debates, I mean, this relates. There's a traditional flow. You think of is as a one-way communication, right? There's what's on the TV and it's coming back out to us, or the moderator asks a question and it's coming out to the candidates. But Kojo, here in the studio, you're sitting with a computer in front of you.

NNAMDIYep.

RESNICKAnd if something is amiss, somebody will point it out to you. I've seen it happen. There is no reason in principle why you could not have -- you wouldn't want necessarily the candidates getting immediate -- it'd be a little science fiction-y and scary for them to be tapping into, you know...

NNAMDIThat's true.

RESNICK...the population. But there's really no reason in principle why the moderator cannot be getting fact checking during the debate.

NNAMDIEven as it's going on.

RESNICKAnd why the pundits cannot be getting the reactions of the people at large instead of simply telling us what they thought.

NNAMDIRunning out of time very quickly, Todd Rogers. In 30 seconds or so, considering the last two debates, are there any political moves that you're expecting from tonight's debate?

ROGERSThat's a good question. I expect that there -- that we will find dodging. I expect that we will find paltering. What I am hoping is that networks will be posting the questions on the screen. These are real voters asking questions, and the politicians are trying to answer them, and CNN already does this, and my hope is that the other networks will also.

NNAMDIAnd I'm not paltering when I say I'd love to continue to talk to you, but we're really out of time. Todd Rogers is a professor of public policy at the Harvard Kennedy school. He was founding director of Analyst Institute which uses behavioral science to understand and improve voter turnout. Todd, thank you for joining us.

ROGERSIt's been fun, thanks.

NNAMDIPhilip Resnick is a professor of linguistics and advanced computer studies at the University of Maryland, most recently the creator of a new real time polling app called React Labs. Philip, good to see you again.

RESNICKSame here.

NNAMDIAnd thank you all for listening. I'm Kojo Nnamdi.