0:00:00.000,0:00:05.855 (intro music) 0:00:05.855,0:00:07.370 My name is Laurie Santos. 0:00:07.370,0:00:10.370 I teach psychology at Yale[br]University, and today 0:00:10.370,0:00:12.549 I want to talk to you about anchoring. 0:00:12.549,0:00:16.119 This lecture is part of a[br]series on cognitive biases. 0:00:16.119,0:00:19.429 Let's do a math problem.[br]really quickly, and you've 0:00:19.429,0:00:20.760 gotta do it in your head 0:00:20.760,0:00:21.649 Ready? 0:00:21.649,0:00:27.529 First, multiply the following numbers:[br]eight times seven times six 0:00:27.529,0:00:32.460 times five times four times three times[br]two times one. 0:00:32.460,0:00:35.250 OK, that's it. 0:00:35.250,0:00:36.840 What's your guess? 0:00:36.840,0:00:37.840 A thousand? 0:00:37.840,0:00:39.820 Two thousand? 0:00:39.820,0:00:43.030 When the psychologists Danny Kahneman[br]and Amos Tversky tried this with 0:00:43.030,0:00:45.390 human subjects, subjects on average 0:00:45.390,0:00:47.990 guessed about two thousand[br]two hundred and fifty. 0:00:47.990,0:00:49.379 Seems like an OK guess. 0:00:49.379,0:00:53.109 But now, let's suppose I gave you[br]a different math problem. 0:00:53.109,0:00:54.650 What if I gave you this one? 0:00:54.650,0:00:56.050 Ready? 0:00:56.050,0:01:00.030 One times two times three times four 0:01:00.030,0:01:04.770 times five times six times[br]seven times eight. 0:01:04.770,0:01:06.310 What's your answer? 0:01:06.310,0:01:08.150 If you're like Kahneman and Tversky's 0:01:08.150,0:01:11.040 subjects, your answer might[br]be a bit different here. 0:01:11.040,0:01:13.880 For this question, their subjects[br]guessed a lot lower. 0:01:13.880,0:01:17.450 On average they said the answer[br]was about five hundred and twelve. 0:01:17.450,0:01:19.509 The first amazing thing[br]about these similar 0:01:19.509,0:01:23.620 mathematical estimates is that people get[br]the answers really, really wrong. 0:01:23.620,0:01:25.219 In fact, the real answer? 0:01:25.219,0:01:29.339 Well, for both, its forty thousand[br]three hundred and twenty. 0:01:29.339,0:01:31.939 People are off by an order of magnitude. 0:01:31.939,0:01:35.259 But the second, even more amazing [br]thing is that people give 0:01:35.259,0:01:39.600 different answers to the two problems,[br]even though they're just different ways 0:01:39.600,0:01:42.020 of asking exactly the same question. 0:01:42.020,0:01:44.060 Why do we give completely different answers, 0:01:44.060,0:01:47.079 when the same math problem[br]is presented differently? 0:01:47.079,0:01:49.500 The answer lies in how we make estimates. 0:01:49.500,0:01:51.590 When you have lots of time to do a math 0:01:51.590,0:01:55.810 problem, like eight times seven times six[br]times five times four times three times 0:01:55.810,0:01:58.559 two times one, you can multiply all of 0:01:58.559,0:02:01.139 the numbers together and get an exact product. 0:02:01.139,0:02:02.719 But when you have to do the problem 0:02:02.719,0:02:05.279 quickly, you don't really have time to finish. 0:02:05.279,0:02:07.309 So you start with the first numbers. 0:02:07.309,0:02:10.188 You multiply eight times[br]seven, and get fifty-six. 0:02:10.188,0:02:12.700 And then you've gotta[br]multiply that by six, 0:02:12.700,0:02:16.670 and, well, you're guessing the final[br]number's gotta be pretty big, bigger than 0:02:16.670,0:02:19.660 fifty-six, like maybe two thousand or so. 0:02:19.660,0:02:22.450 But when you do the second problem, you start 0:02:22.450,0:02:26.880 with one times two, and, well, that's only[br]two, and two times three's only six. 0:02:26.880,0:02:28.530 Your answer's gonna be pretty small, 0:02:28.530,0:02:31.310 maybe only like five hundred or so. 0:02:31.310,0:02:33.780 This process of guessing based on the first 0:02:33.780,0:02:36.110 number you see is what's[br]known as "anchoring." 0:02:36.110,0:02:37.680 The first number we think of 0:02:37.680,0:02:39.730 when we do our estimate is the anchor. 0:02:39.730,0:02:41.830 And once we have an anchor in our head, 0:02:41.830,0:02:44.730 well, we sort of adjust[br]as needed from there. 0:02:44.730,0:02:48.750 The problem is that our minds are biased[br]not to adjust as much as we need to. 0:02:48.750,0:02:51.580 The anchors are cognitively really strong. 0:02:51.580,0:02:54.530 In the first, problem you probably[br]started with fifty-six, and 0:02:54.530,0:02:57.720 then adjusted to an even[br]bigger number from there. 0:02:57.720,0:03:00.999 And in the second problem, you started[br]with six, and then adjusted from there. 0:03:00.999,0:03:05.849 The problem is that starting at different[br]points leads to different final guesses. 0:03:05.849,0:03:10.970 Like real anchors, our estimated anchors[br]kinda get us stuck in one spot. 0:03:10.970,0:03:14.709 We often fail to drag the anchor far[br]enough to get to a correct answer. 0:03:14.709,0:03:17.709 Kahneman and Tversky discovered that this 0:03:17.709,0:03:19.889 sort of anchoring bias happens all the time, 0:03:19.889,0:03:22.480 even for anchors that are totally arbitrary. 0:03:22.480,0:03:25.480 For example, they asked[br]people to spin a wheel with 0:03:25.480,0:03:28.320 numbers from one to a hundred,[br]and then asked them to estimate 0:03:28.320,0:03:31.739 what percentage of countries in[br]the United Nations are African. 0:03:31.739,0:03:34.739 People who spun a ten on[br]the wheel estimated that 0:03:34.739,0:03:36.769 the number was about twenty-five percent. 0:03:36.769,0:03:39.769 But people who spun a[br]sixty-five estimated that 0:03:39.769,0:03:41.650 the number was forty-five percent. 0:03:41.650,0:03:46.340 In another experiment, Dan Ariely[br]and his colleagues had people 0:03:46.340,0:03:49.300 write down the last two digits[br]of their social security number. 0:03:49.300,0:03:50.770 They were then asked whether they would 0:03:50.770,0:03:54.119 pay that amount in dollars[br]for a nice bottle of wine. 0:03:54.119,0:03:58.250 Ariely and colleagues found that people[br]in the highest quintile of social security 0:03:58.250,0:04:02.799 numbers would pay three to four times[br]as much for the exact same good. 0:04:02.799,0:04:04.939 Just setting up a larger anchor can make a 0:04:04.939,0:04:07.099 person who would pay eight[br]dollars for the bottle 0:04:07.099,0:04:10.649 of wine be willing to spend[br]twenty-seven dollars instead. 0:04:10.649,0:04:14.619 Sadly for us, sales people use[br]anchors against us all the time. 0:04:14.619,0:04:18.399 How many times have you noticed[br]a salesperson or an advertisement 0:04:18.399,0:04:21.100 anchoring you to a particular price, or 0:04:21.100,0:04:23.890 even to how much of a particular[br]product you should buy? 0:04:23.890,0:04:26.430 Whether it's buying a car, or a sweater, 0:04:26.430,0:04:30.270 or even renting a hotel room, our[br]intuitions about what prices 0:04:30.270,0:04:34.510 are reasonable to pay often come[br]from some arbitrary anchor. 0:04:34.510,0:04:38.490 So, the next time you're given an[br]anchor, take a minute to think. 0:04:38.490,0:04:40.139 Remember what happens when you 0:04:40.139,0:04:42.300 drop your anger too high, and then 0:04:42.300,0:04:45.330 consider thinking of a[br]very different number. 0:04:45.330,0:04:49.129 It might affect your final estimate[br]more than you expect.