Jeff and Rebecca discuss Chapter 16 of THINKING FAST AND SLOW
United States


00:00:08alright we're back chapter 16 of thinking fast and slow in a crazy week so far it's only been 1 chapter to get one of the can so we can move on she keeps on the one that we were like embarrassingly excited about statistics so that we can get two more actual statistics I guess and chapter 17 but we were just saying before we started recording that this kind of weird chapter it builds on the previous stuff about base rates and what having additional information related to a base rate does to our intuitive answers about something and he starts to get into what stereotypes due to our guesses about frequency and causality of things but basically this is I like what's the point
00:01:05moment of dread first teachers of psychology that we like knowing what we know about Danielle Cohn a man and his tendency towards like depression and a cloudy Outlook it kind of makes sense so you can imagine how to get the hundred and seventy pages into writing this book and he's like why am I even doing it so I think I mean as a teacher having been a teacher at one point for a while I think there's always those Dark Knights of the pedagogical so love like even know what is this what is happening to hear it does it add up to a hill of beans or not so it is a moment about why why you got to that it's not just sort of a dark days is
00:01:54observation like there's a real statistical study that leads him to say it I wonder what it means and it comes out of this condition in which the chapter the title of the chapter is causes Trump statistics which is a nice easy way of remembering it but it's a little more complicated than that it basically in a wreck I'm going to need you to be extra vigilant about checking I'm not screwing this up cuz I I I have a tenuous grassed on a myself is that basically there are two kinds of Base rates that we think of one is the Bayesian base rate and what is a causal base rate and it's difficult a little bit to distinguish between them because you freeze them differently they can be interchangeable but the phrasing is what matters so statistical base rates are okay you expect that a coin will be heads 50% of the time you expect that a normal dice will be three one out of six times so there
00:02:54as to what you know I rolled the dice nine times how many times would you expect it to be 3 you can do the math as well you would expect one out of six to be three and it's pretty simple algebra from there the cause of Base rates is a little bit different because it is information about a general condition so rather than it's not a percentage it's sort of a condition so the one example examples of gets here is interest in cycling is widespread in France so it's more of a description of a category then it is an actual Bayesian base rate that doesn't describe it didn't say how many people in France have bikes right people in France describe themselves as being interested in cycling it's just this is a thing this is a statement about the universe is a car so then if you say
00:03:52give me some information on Sand what is the likelihood that they're going to be interested in cycling you can't ask you can't answer probably Stickley you can save 14% you can only say well there French so I guess they're probably interested in cycling that's all you can do there and the in the conflict between those two things into that you different uses of those two kinds of Base rates is basically the subject of this chapter do I have that essentially right yeah and he gets into some of the details like that the statistical base rate describes a population or a group as a whole so like you know it what's the one about the taxi cabs earlier like if a cat was involved in a hit-and-run at night and there's two cab companies the green and the blue that operate in the city if 85% of the cabs are green and 15% are blue like an absence of any other information you know was it a green cab or a Blue Cab that was in the accident while the base rate about this population is that you know 85% of the
00:04:52abs are green so most likely feel like you have an 85% chance that a cab in an accident like in the absence of any other info but this the console stereo like the causal base rate Taps into how we explain one instance of a thing not the population as a whole so if you're trying to explain like this one or make a guess about who's involved or what's involved in one instance in this one accident you have these base rates about the Green Cabs in the blue tabs but if they add some causal really that information like that Green Cabs are involved in 85% of the accident then you go beyond you know just the base right about the existence of things in the group but what is the likelihood about this one particular instance and it gets a little more grounded I think like this chapter chapter is not very grounded in general as as condiments work goes this is a little more like it floats around a little bit more but he starts talking about stereotypes and I think if we
00:05:52if you think about it as stereotypes instead of causal versus statistical base rates it becomes easier to think about that like a stereotype consider that were to be value-neutral for these purposes but a stereotype describe the trends in the population of a group as a whole so like here's what we know among the women 85% do this thing 10% do this other thing here is one woman you can you can make your guesses about what this one woman will be like based on what you know about base rates but when you're in a sensitive situation like judging one individual woman we want to judge based on something else other than the assumption that an individual fit The Stereotype on 169 we consider it morally desirable for bass race to be treated as two facts about the group rather than presumptive facts about individuals in other words reject causal base rates the social Norm again stereotyping including the opposition to profiling has been highly beneficial
00:06:52in creating a more civilized more equal Society is used to remember however that neglecting valid stereotypes in early results and sub optimal judgments resistance is stereotyping is allowed Memorial position but the simplistic idea that the recent is Costless is wrong that's so interesting I think I got this is valid then that means something different than if it's invalid because that's the one who sings stereotype as a condition is wrong or less accurate basically as much as he's fighting in this book to get people to make their judgments and decisions based on statistics there are cases where we set those things aside and the name of like a greater Society at all good and you're not going to judge like one black person you're not going to make guesses about the one black person what they'll be like based on stereotypes or base rates of information that you have about
00:07:52black people as a whole because we have decided that that's not a fair and Equitable way to operate Society even though the statistical reasoning of it would make sense that's not the way that we want our society to to work so we set aside those that we look at the end we try to look at the individual that that's where we get into that's where we get into trouble because our system one he's I can deal with stories in which elements are causally link but it is weak and statistical reasoning so as soon as you have a stereotype that's activated and you have some explanation about why you know people by an individual in this one group might do or be a thing the storage is just going to start like running away with itself that we should be stereotyping except when we might make mistakes otherwise if you don't have any other kind of information stereotyping can make sense but it's a different kind of reasoning that
00:08:52the upshot is though that if we have both consoles at 6 and Bayesian to Texas base race at 6 available to us we will prefer in general causal statistics. That's I guess that's the mistake that were talking about here if we know what the base rate is but we're also told something about the group to which the individual belongs we will prefer to our error the causal reasoning that we like reasons better than just attention because I think that's the thing that's helps us understand statistics don't want to know why I think that's fair I mean that's something that commonly didn't save like why are 85% of Green Cabs in the accidents on the street are green well then they just make them at the same rate per unit
00:09:43but there's more of them right so it's It's Tricky it's a very slight difference but I think that's like this is why we screwed up psychology is in that that we will not go to the base rates we will not go to the statistical information if the fact conflict with something else that we believe I'm so the classic experiment that he describes is one in which they teach psychology students that in the presence of a bunch of other people if you hear someone in distress and there are other people present who might help you or any individual are less likely to go help like this is the thing that we know from social psychology it's referred to as the bystander effect that if there are multiple bystanders anyone bystander is less likely to act because they think that someone else is going to take care of it I'm so they can they tell people this is what happens like these are the statistical facts but we like to believe about end
00:10:43jewels that we know or an individual that you read a paragraph about in this in these kinds of studies that this sounds like a good person who would help like despite what we know about statistics it sounds like this is a good person who would help so when they put in front of those students a case study of like imagine that you know this bad thing is happening Jeff the guy in the paragraph that you just read is one of the bystanders who's witnessing it there were six other people as well you know How likely is it that Jeff offered help they overestimate the degree to which Jeff would offer help in spite of knowing the stats and because it conflicts with the other belief about who Jeff is like the story that they told themselves about who Jeff is and there's a really famous example and that I remember learning about in my psycho by social psych classes about a woman named Kitty Genovese see who is killed like I think this is where we started learning about the bystander effect really wisdom in 1964 she was killed outside her apartment bill
00:11:43King and Queens and there were like 37 Witnesses who either saw or heard the attack but none of them called the police and researchers started getting into what like what the hell is happening that 38 people heard this woman getting murdered or saw her out their window getting murdered and no one did anything and as they dug into it it was like everyone thought will surely someone else has called the police so in the I guess I'm kinda means like a real grounded nihilistic since that thing that he's hoping would happen or that we hope would happen by teaching people that this is a thing is like if you're out somewhere and you see a bad thing happening you should take action because you should assume that everyone else around you is subject to the bystander effect could they are and that his own nihilism becomes well even if you tell people what the base rate for helping a stranger in distress is they still will
00:12:43overestimate the likelihood of Indian anyone individual helping like they're resistant because their belief there stereotype about people as being you know helpful general in general people who aren't terrible like that I guess our cause was stereotypes right people wouldn't hear a woman getting this right not do anything in the people aren't terrible than Shirley it's impossible that 37 people would 38 people wouldn't do anything and yet that's how we think about it so I don't know what you can get my thinking is always there there's always Sanctuary upper level of maybe upper-level if you know that you're resisting to this kind of thinking does that inoculate you could listen to suck itself doesn't inoculate you maybe you need to know the base rate of people believe in Psychology but I think that's what's difficult is it's not enough to say to your mom you know 48% of people who smoke
00:13:43develop lung cancer
00:13:45right right back I think that's one reason that kind of argumentation doesn't work because she and we whatever it is that we do this isn't say we shouldn't we have causal stereotypes as well but I smoke but I eat well you know I'm not right or you know what are workout but or I can eat a bunch of junk because I work out or I can eat a bunch of junk food because I don't smoke or do all these different things where we're looking for a way to affirm our world you and our world you in this case is we still want to drink or smoke and eat pizza every night and so we come up with a reason or we find reasons we ever believe the reasons why in this case is people who work out or healthier than people who don't include information about people who smoke or people who don't eat vegetables or heat transfer whatever else
00:14:45irrelevant that stereotype mean that the problem stereotypes themselves is they will because of their seeming totality overwhelm whatever secondary information might be available to counter man that's their types and Bayesian rates are one of those things that I can come from countermand
00:15:03yeah and he he gets into like if you want to teach psychology effectively you have to surprise people like this Kitty Genovese story is surprising I remember it from school like almost 20 years ago up like it's surprising that this is a true thing that you know 37 people and I could I could be misremembering but I think that there's also a proportional relationship like the larger the group the less likely it is that anyone will act because there are more people to diffuse that bystander I mean responsibility someone get hit by a car you like there is no one is 100% of people will be a bystander if I'm a bystander here and there's two people you know well there's a 50% like in the chances of you being the bystander go down in your mind the more people there are because if everyone has a 90% chance of responding well shit 37 and 9.9 * .937
00:16:03is a very low percentage but it actually doesn't work that cuz everyone's like Game Theory everyone else is doing the same math themselves or not accounting for other people knowing there's a bunch of people around that might do something about it so you know again also so I guess one lesson is if you see someone they're getting her to do something that's another funny thing is I guess love you in danger I guess that's a that's a calculation I could be in the case but like there's some sort of weird downside were trying to avoid as a cop on my God what if we both help so I don't know what's so strange about that particular case scary that someone is getting attacked outside my window I don't want to get involved I don't want to be in danger either like I saw this at work last year I was getting off a flight that we're like two passengers had some kind of altercation in the process of the whole you know like everyone rushes to stand up and get their bags out of the overhead bin
00:17:03get off the flight and one of the flights are one of the passengers was like a middle eastern man wearing a turban like he had on a turban or some sort of I don't know exactly what it was which religion he was representing but some marker of a cultural identity and the other passenger was shouting slurs at him as we were like exiting the plane on that little bike ramp way thing between the plane in the airport and everyone just froze and you could see it's like a full playing there's like 200 people you're in between the space where the airport exist and the plane where the flight attendants are like no one official was witnessing this to do anything and you could like I froze everyone around me froze and I was running the like should I say something and then like one woman next to me said something and went and got a flight attendant but then everybody else just like ran away it was like that guy handled I don't want to be late for my next flight I don't want to get involved like they're all of these you know you can imagine all the
00:18:03that people might have had for not being the one to speak up but I thought about it after the fact if she had spoken up would I have done anything and if I hadn't like it's not because it didn't occur to me that someone should do something but like maybe I was waiting to somebody else was going to maybe I just did the bystander saying anything in this chapter about like this is really the nihilistic challenging thing about knowing this thing but also about really knowing all the things that you learn from Thinking Fast and Slow is we can understand we can have the knowledge in our brain but it doesn't necessarily change the way that we view the world because to shift our beliefs about how life Works takes more than a statistic even a very surprising statistic and here because this is where he ends is that the thing basically our mental model Will Will trump to tistics that would make us change our mental models
00:19:03also I think it helped thing you think about political argumentation but it's surprising personal experience Alters the mental model so it's attacking the thing that he's altering which statistics cannot do so I bet if you're one of those people that was born at 37 that didn't do anything I'm guessing a lot of them the next time they do if they were having a situation like that again I wonder you know if you seen the store I'm going to be in a crayfish five people insulting someone getting a fist fight I'm going to go try to break them up I'll be more likely to do it because my mental math that's surprising result the surprising story The surprising thing says to me that your mental model is wrong but for some reason your mental model is immune to mirror statistical attacks on it
00:20:03I think that's I think that's a good by talking about that one that actually reading it by the end of little bit makes me glad that we're doing this whole exercise cuz like 75% of the time I read the chapters and I'm like there's not that much to talk about this chapter and then here we find ourselves it is so juicy but we'll see we'll see next time

Transcribed by algorithms. Report Errata
Disclaimer: The podcast and artwork embedded on this page are from Jeff O'Neal, which is the property of its owner and not affiliated with or endorsed by Listen Notes, Inc.


Thank you for helping to keep the podcast database up to date.