Tali Sharot: By the end of today, 4 mil­lion blogs will be post­ed, 80 mil­lion Instagram pho­tos uploaded, and 600 mil­lion tweets released into cyber­space. That’s more than 7,000 tweets per second. 

Why do you spend pre­cious moments every day shar­ing infor­ma­tion? There’s prob­a­bly many rea­sons, but it appears that the oppor­tu­ni­ty to impart your knowl­edge onto oth­ers is inter­nal­ly reward­ing. A study con­duct­ed at Harvard showed that peo­ple were will­ing to forego pay­ment in order to have their opin­ions broad­cast to oth­ers. Now, we’re not talk­ing well-crafted insights here. These were peo­ple’s opin­ions about whether Barack Obama enjoys win­ter sports, or whether cof­fee is bet­ter than tea. 

A brain imag­ing study showed that when peo­ple had an oppor­tu­ni­ty to share these pearls of wis­dom with oth­ers, their reward cen­ter in the brain was very strong­ly acti­vat­ed. You feel a burst of plea­sure when you share your thoughts, and that dri­ves you to com­mu­ni­cate. It’s a nifty fea­ture of our brain because it ensures that ideas are not buried with the per­son who first had them, and as a soci­ety we can ben­e­fit from the minds of many. 

But for that to hap­pen shar­ing is not enough. We need to cause a reac­tion in oth­ers. What then deter­mines whether you affect the way peo­ple behave and think, or whether you’re ignored? 

So, as a sci­en­tist I used to think that the answer was data. Get data, cou­ple with log­i­cal think­ing, that’s bound to change minds, right? So I went out to try and get said data. My col­leagues and I con­duct­ed dozens of exper­i­ments to try and fig­ure out what caus­es peo­ple to change their deci­sions, to update their beliefs, to rewrite their mem­o­ries? We peeked into peo­ple’s brains, we record­ed bod­i­ly respons­es, and we observed behavior. 

So you can imag­ine my dis­may when all of these exper­i­ments point­ed to the fact that peo­ple are not in fact dri­ven by facts. People do adore data. But facts and fig­ures often fail to change beliefs and behav­ior. The prob­lem with an approach that pri­or­i­tizes infor­ma­tion is that it ignores what makes us human. Our desires, our fears, our emo­tions, our pri­or beliefs, our hope. 

Let me give you an exam­ple: cli­mate change. My col­leagues Cass Sunstein, Sebastian Bobadilla Suarez, Stephanie Lazzaro, and I want­ed to know whether we could change the way peo­ple think about cli­mate change with sci­ence. So first of all we asked all the vol­un­teers did they believe in man-made cli­mate change? Did they sup­port the Paris Agreement? And based on their answers we divid­ed them into the strong believ­ers and the weak believ­ers. We then told every­one that experts esti­mat­ed that by 2100 the tem­per­a­ture would rise by six degrees, and please give us your own esti­mate. So, the weak believ­ers gave an esti­mate that was low­er than the strong believers. 

Then came the real test. We told half of all the par­tic­i­pants that the experts have reassessed their data, and now con­clude that things are much much bet­ter than pre­vi­ous­ly thought and the tem­per­a­ture would only rise by one to five degrees. We told the oth­er half of par­tic­i­pants that the experts have reassessed their data, and now con­clud­ed the things are much much worse than pre­vi­ous­ly thought, and the tem­per­a­ture would rise by sev­en to even degrees, and please give us your own estimate. 

The ques­tion was, would peo­ple take this infor­ma­tion to change their beliefs? Indeed they did. But most­ly when the infor­ma­tion fit their pre­con­ceived notions. So when the weak believ­ers heard that the experts are say­ing that actu­al­ly things are not as as bad as pre­vi­ous­ly thought, they were quick to change their esti­mate in that direc­tion. But they did­n’t budge when they learned that the experts are say­ing that actu­al­ly things are much worse than pre­vi­ous­ly predicted. 

The strong believ­ers showed the oppo­site pat­tern. So when they heard that the experts are say­ing that things are much more dire, they changed their esti­mate in that direc­tion. But they did­n’t move that much when they learned that the experts are say­ing that things are not that bad. 

When you give peo­ple infor­ma­tion, they are quick to adopt data that con­firms their pre-notions, but often will look at coun­terev­i­dence with a crit­i­cal eye. This will cause polar­iza­tion, which will expand and expand as peo­ple get more and more information. 

What goes on inside our brain when we encounter dis­con­form­ing opin­ions? Andreas Kappes, Read Montague, and I, invit­ed vol­un­teers into the lab in Paris. And we simul­ta­ne­ous­ly scanned their brains in two MRI machines while they were mak­ing deci­sions about real estate and com­mu­ni­cat­ing those assess­ments to one another. 

What we found was that when the pair agreed about a real estate, each per­son­’s brain close­ly tracked the opin­ion of the oth­er, and every­one became more con­fi­dent. When the pair disagreed, the oth­er per­son was sim­ply ignored, and the brain failed to encode the nuances of that eval­u­a­tion. In oth­er words opin­ions are tak­en to heart and close­ly encod­ed by the brain most­ly when it fits our own. 

Is that true for all brains? Well, if you see your­self as high­ly ana­lyt­i­cal, brace your­self. People who have bet­ter quan­ti­ta­tive skills seem to be more like­ly to twist data at will. In one study, 1,000 vol­un­teers were giv­en two data sets: one look­ing at skin treat­ment, the oth­er at gun con­trol laws. They were asked to look at the data and con­clude: Is a skin treat­ment reduc­ing skin rash­es? Are the gun laws reduc­ing crime? 

What they found was that peo­ple with bet­ter math skills did a bet­ter job at ana­lyz­ing the skin treat­ment data than the peo­ple with worse math skills. No sur­prise here. However, here’s the inter­est­ing part. The peo­ple with bet­ter math skills? They did worse at ana­lyz­ing the gun con­trol data. It seems that peo­ple were using their intel­li­gence not nec­es­sar­i­ly to reach more accu­rate con­clu­sions but rather to find fault with data that we’re unhap­py with. 

The ques­tion then becomes why have we evolved a brain that is hap­py to dis­re­gard per­fect­ly good infor­ma­tion when it does­n’t fit our own? This seems like poten­tial­ly bad engi­neer­ing, leav­ing errors in judg­ment. So why has­n’t this glitch been cor­rect­ed for over the course of evolution? 

Well, the brain assess­es a piece of data in light of the infor­ma­tion it already stores, because on aver­age that is in fact the cor­rect approach. For exam­ple, if I were to tell you that I saw pink ele­phants fly­ing in the sky, you would con­clude that I’m delu­sion­al or lying, as you should. When a piece of data does­n’t fit a belief that we hold strong­ly, that piece of data, on aver­age, is in fact wrong. However, if I were to tell my young daugh­ter that I saw pink ele­phants fly­ing in the sky, most like­ly she would believe me, because she has yet to form strong beliefs about the world. 

There are four fac­tors that deter­mine whether a piece of evi­dence will alter your belief: your cur­rent belief; your con­fi­dence in that cur­rent belief; the new piece of evi­dence; and your con­fi­dence in that piece of evi­dence. And the fur­ther away the piece of evi­dence is from your cur­rent belief, the less like­ly it is to change it. This is not an irra­tional way to change beliefs, but it does mean that strongly-held false beliefs are very hard to change. 

There is one excep­tion, though: when the coun­terev­i­dence is exact­ly what you want to hear. For exam­ple, when peo­ple are told that oth­ers see them as much more attrac­tive than they see them­selves, they’re hap­py to change their self-perception. Or if you learn that your genes sug­gest that you’re much more resis­tant to dis­ease than you thought, you’re quick to change your beliefs. 

What about pol­i­tics? Back in August, 900 American cit­i­zens were asked to pre­dict the results of the pres­i­den­tial elec­tion by putting a lit­tle arrow on a scale that went from Clinton to Trump. So if you thought Clinton was high­ly like­ly to win you put the arrow right next to Clinton. If you thought it’s a 50‍/‍50, you put it in the mid­dle. And so on and so forth. They were also asked, Who do you want to win?” 

So, half of the vol­un­teers want­ed Trump to win, and half want­ed Clinton to win. But back in August, the major­i­ty of both the Trump sup­port­ers and the Clinton sup­port­ers believed that Clinton was going to win. 

Then a new poll was intro­duced, pre­dict­ing a Trump vic­to­ry, and every­one was asked again, Who do you think is going to win?” Did the new poll change their pre­dic­tions? Indeed it did. But most­ly it changed the pre­dic­tions of the Trump sup­port­ers. They were elat­ed to hear that the new poll was sug­gest­ing a Trump vic­to­ry and were quick to change their pre­dic­tions. The Clinton sup­port­ers did­n’t change the pre­dic­tions as much, and many of them ignored the new poll altogether. 

The ques­tion then is how do we change beliefs? I mean, sure­ly opin­ions do not remain sta­ble, they do involve. So what can we do to facil­i­tate change? The secret is to go along with how our brain works, not against it. 

So, the brain tries to assess any piece of evi­dence in light of the knowl­edge it already stores. And when that piece of evi­dence does­n’t fit, it’s either ignored or sub­stan­tial­ly altered. Unless of course it’s exact­ly what you want to hear. So, per­haps instead of try­ing to break an exist­ing belief, we can attempts to implant an new belief alto­geth­er and high­light the pos­i­tive aspects of the infor­ma­tion that we’re offering. 

This all sounds very abstract, I know. Let me give you an exam­ple: vac­cines. So, par­ents who refuse to vac­ci­nate their kids because of the alleged linked to autism often are not con­vinced by sci­ence sug­gest­ing that there’s no link between the two. What to do then? A group of researchers, instead of try­ing to break that belief offered the par­ents more infor­ma­tion about the ben­e­fits of the vac­cine. True infor­ma­tion. How it actu­al­ly pre­vents kids from encoun­ter­ing dead­ly dis­ease. And it worked.

So, when try­ing to change opin­ions, we need to con­sid­er the oth­er per­son­’s mind. What are their cur­rent beliefs? What are their moti­va­tions? When some­one has a strong motive to believe some­thing, even a hefty sack of evi­dence to the con­trary will fall on deaf ears. So we need to present the evi­dence in a way that is more con­vinc­ing to the oth­er per­son, not nec­es­sar­i­ly in the way most con­vinc­ing to us. Identify com­mon motives and then use those to implant new beliefs. Thank you.