Skip to main content
Loading Explore guides…
FTC Privacy & Heuristics · Acedra
Home
Explore
Law
FTC Privacy & Heuristics
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
77 segments
Copy
Hide video
0:01
so first let's talk about this article that you all read online you privacy plan questions loom this article discussed current regulation that is under consideration at the federal trade commission and if you remember from the lecture last week about advertising law federal trade commission is one of those bodies that is tasked with regulating businesses in the united states and so there are two competing groups this article talks about one of those groups let's at this working here
0:43
right so one of these groups uh... as privacy advocates and they are lobbying the FTC to create new regulations if you remember from the lecture from last week when we talk about advertising law that is one of the things that progressive groups will tend to do is lobby for increased regulation and so their privacy advocates are saying we need this increased regulation there needs to be away that people can change a setting in their internet browsers so that
1:12
when they visit websites their behavior is not tracked this is very common - a lot of advertisers i think the article says about eighty percent of web advertisements use some kind of behavioral tracking how many of you have been on amazon.com and you'll buy an item or you'll be looking at an item, and the next time you log in to amazon dot com it pops up and says these are the items you recently looked at that uses tracking cookies on your computer to track the behavior that you have
1:45
done in previous browsing sessions and then it uses that information to target you with advertisements in the future and so these privacy groups are saying no, we should be able to opt out of this on the other side advertisers are saying you know what we can do a pretty good job of policing ourselves and you'll remember from that lecture last week that that is the common stance of the advertising industry they don't want regulation regulation is expensive for them
2:16
it's hard for them to keep up with it requires them to dedicate a lot more people to just making sure that they're in compliance with regulation, and it distracts them from the primary mission of their business which is to make money by selling advertisements to customers so that these two competing of factions if you will that this article talks about some saying we want this ability to opt-out we don't want advertisers to track what we do on-line and use that
2:44
to send us relevant advertisements other saying look will police ourselves will give consumers the ability to opt-out but don't let them do it in a really easy way and they provide some reasons why they think it should not be really easy to opt out of that and only that to the article for you to read now it's important to know that do not track doesn't mean you're not gonna see advertisements anymore it just means that those advertisements are not going to be based on
3:10
your previous browsing behavior another thing is each impression is less valuable to an advertiser if they're not able to track what you've done in the past and the reason is we know this from all of that classes that we've had before this we've been talking about the importance of of customizing advertisements the more you can identify your target market and deliver relevant ads to your target market the more likely it will be that that abt will be effective
3:42
and sell for advertisers if they can say what this person is already looked at product similar to the when you're trying to sell show them this advertisement that kind of impression that's what it's called in uh... web advertising world is called an impression whenever a person views your ad or has it displayed on your screen that impression more valuable to them if they know that you are already likely a part of their target market soreness in key questions that you should
4:10
be able to answer from this article for the exam first what is the difference between a contextual at in a behavioral add on the internet all give you the answer to this one a conceptual that is one that is displayed it next to content that you are viewing at a given time if you go and our reading an article about hemorrhoids and you have day hemorrhoidal ad on the screen right next to you that is eight contextual act to advertisers are assuming that if you're reading about about hemorrhoids
4:45
it's probably because you're experiencing some discomfort in the posterior region and humane be inclined to purchase some kind of heroic point in a behavioral ad on the other hand actually looks at all of your behavior that you have exhibited on eight given sellers website or perhaps even across multiple websites and their delivers ads that you are likely to be interested second question you should be able to answer for the exam
5:14
why our behavior lads worth more advertisers touched on that a little bit already and third what are the negative outcomes for consumers that advertisers claim might happen if a large number of consumers opt out and there's one ky one at the article talks about the should be able to answer that as well for the exam let's move on to the lecture for today this is lecture seven b if you have jake's version of the slides i put anew version of the slides up on blackboard
5:45
they don't differ very much so if you just have jake's that's fine you be able to follow along if you wanna pull down the new version that's available on blackboard right now today we're gonna be talking about your respects this is a really interesting topic i think and it's one that has implications that reach far beyond the world of advertising but starr was a little bit of background information so psychologists have identified two primary modes of decision making or two
6:18
primary models by which people make decisions the first model is one in which an individual considers an issue at length and makes a rational decision based on that existing evidence you can see this uh... this finding new uh... lenovo thinkpad laptop right up here i just got this this week and i did a lot of research before i bought this computer i was waiting for the new models to come out and i've spent probably three months looking at reviews in
6:53
uh... you know pre pre product reviews in reviews of the processor that was in the reviews of some of the other items that were related to this particular model of laptop computer spent a lot of time i considered all the alternatives i looked at other brands of laptop that were competing in this particular segment and ultimately i made the decision to purchase the thinkpad t four twenty that was aviary what we call systematic way of processing that information okay so systematic
7:30
decisionmaking occurs when you look at all of the available evidence you evaluate it and you alternately make a decision using more formalize rules of logic in inference the other way that we often make decisions we do this a lot is where we make a quick decision based on some principle okay this is called churry stick decision making my wife for example uh... when she gets ready to buy a new laptop she will just by whatever i recommend
8:03
she doesn't care she knows that i know a lot about the gears and so whatever i recommend so i heard that she purchased that's what she'll get that is that heuristic way making the decision now we'd like to believe that most of the time we make decisions based on systematic processing we want to think that with thoughtful people in we care a lot about the decisions that we make and sell consider all of the evidence and we evaluated in wayan and make our decisions in a very systematic way
8:35
in reality a lot of judges that a lot of judgments that we make on a daily basis are based on your wrist x this year is sticks are basic principles that help us make sense of the world's let's talk about those a heuristic principle i'll give you a clear definition it's a simple decision rule so for example some common here s sex and why is the term common here is declining heuristics that most people use one of those if you remember back to the lecture on balance theory
9:06
one common year is sick that balance theory sort of export case a little bit is the idea that people highlights usually have correct opinions c remember that triangle people i like to think like me that's a heuristic we assume that people we like have the correct opinions and so we were faced with a situation where a person that we like has a certain opinion were more inclined to agree with that opinion because we think that people we like should have correct opinions
9:37
another calming heuristic is uh... visual clarity reveals spatial distance so the more sharply you can see an object the more clearly it it appears to be or the closer its appears to be if you've ever been in a car you'll notice that the passenger side window always has this little warning on it objects in your are closer than they appear why is that well the reason is we have this heuristic that if object is smaller it is a farther away
10:11
and those mirrors on the passenger side in order give you a reasonable field of vision functionally shrink the image down and so if you look in that mere you see the small in the gym i now realize that the semi that looks like a tiny little toy car in the mir is actually very close to you that is a situation where the here is thick that we use can lead to systematic errors that could actually kill you so in this particular case we have this morning to remind you
10:41
what you can't use this heuristic in this particular case where you might get flattened by summer there are also what i like to call idiosyncratic your istics these are heuristics that an individual person might use to make decisions so for example my wife when she will buy whatever computer i suggest she buys that's an idiosyncratic heuristic its specific to her there maybe other people who also will buy whatever computer i recommend but probably not nearly as many as people who would
11:15
who had use the heuristic people i like usually have correct opinions so for example sony tvs are reliable that might be a heuristic that someone might use when deciding what type of television to purchase all right before i got married and and had a wonderful wife to buys my clothes for me uh... used to have this year is stick that uh... the closing combinations on store displays represents sound fashion choices i'll be honest by and not a great dresser
11:45
and so when i would have to go to the stores and buy clothes i would often just look at what the display manikins were wearing and i would assume that if they put on the display americans was probably a good combination i could buy that outfit and wear it without any problems one for julie for me this was a bad heuristic loaded up with some really ridiculous looking outfits that uh... once i got home and my friend sunny weather too so you cannot wear that now it's horrible
12:14
so we have in situations where here s sticks well often lead us down the wrong path we're gonna talk today about three specific situations where heuristics can lead to poor decision making heuristic store useful and i know what you get the idea during this lecture that you were sticks or a bad thing the truth is that teristics are very useful it would be impossible to consider at length and carefully evaluate the evidence to come to a systematic decision
12:44
about every issue our choice that we know it so here is sick simplify the task of assessing probabilities predicting values can make dick making decisions and there are a key reasons why we used your istics first we use here is six because of limited time in two thousand seven cornell university did a study where they found that on average people make two hundred and twenty one decisions everyday just related to food now say that again so two thousand seven they did this study
13:21
they found that people make tick an average of two hundred and twenty one decisions everyday just related to food so you can imagine if that's all the decisions that we make any given day about food that if you do anything more than eat during the day you're obviously going to be making potentially thousands of decisions we simply don't have the time to systematically process information for every decision we met there's also an issue out energy you've heard of this term mental drain
13:50
cognitive effort actually is effort and you will run out of energy if you're trying to systematically process every decision that you make the third reason we use your istics is because of interest some decisions just art very interesting if you find yourself systematically processing every decision you make he'll get bored really quickly so uh... for example one of the decisions i made this morning uh... by loaded my dishes from breakfast into the dishwasher and i had to decide
14:22
if i was going to run the dishwasher now or if i was going to wait intel i'd put in the dishes from dinner suvarna systematically process this okay so on the one hand yes i watch it now than when i get home while although this is an awfully click and they'll be really nice but on the ardent there's still some room in there i could probably put my dinner dishes in there and i'm bored already cdc there is a factor of interest if it's not interesting we're not going to want
14:50
to elaborate and make that systematic decision all of the time so decisions are often made on your istics and over the alta millie i decided that i would wait to do the dishwasher because i like to empty it in the morning so awake until dinner or put in the dinner dishes and do it in the morning so that's my heuristic i'd like to empty the dishwasher in the morning we also this issue of capacity our cognitive resources are limited and if you'll remember jake likes to talk
15:16
about this librarian in our brain who is constantly running around in this tax trying to get the information that we need to do with the various things that are brain does and that librarian is overworked you're only going to be providing more work for that librarian for some reason we were systematically processing every decision that we make these for reasons we use your essex 'cause we have limited time when we did energy limited interest unlimited capacity so here's where we are generally
15:52
heuristics are a good thing they're useful and we couldn't function in the world without them there are times and there are certain cases where teristics lead two severe errors in judgment so for example let's return to this idea of visual clarity reveals spatial distance billy pilots in the room but he's a pilot ernest byron pilot case a pilot answer question for me what do you do but how do you party behave differently when visibility is poor
16:32
user s rates you have to rely on the instruments why do you have to rely on the instruments you can't see anything even if you can see a little bit you can't trust what you can see so is there is a like a uh... what's that distance that you have to be able to see in order to fly with visual cues rather than your instruments one-mile so you have to be able to see a mile ahead of you or you must use your instruments are saying beacause
16:58
dot the heuristics that we would use the process that visual information could lead to serious errors in judgment could ultimately kill you could kill the people on your point today we're gonna talk about three specific heuristics that matter representative misspoke bela beli an adjustment in anchoring and we'll talk about the basic idea of the year is sick and specifically some problems this specific predictable there is that people make when using each of these hearings sticks
17:32
so let's start with the representative next year and stick the basic idea of this is if they can be or similar in one case they're similar in all cases let me give you some specific examples stereotyping is one of the problems that can occur when we use the representative ness heuristic and let's say that usaa picture of this man then you know that a former neighbor said this about so someone who live next door and said this eric is very shy and withdrawn he's helpful
18:11
but he has little interesting people or the world of reality he's meeting very tight he has a need for order and structure and a passion for detail then they asked what is the probability that eric is our farmer a salesman a pilot a librarian er physician when he said we think eric s i'm here lots of different things whatever you decided here's how you've done you've used a stereotype of what you think of farmers salesperson a pilot or a librarian should be like
18:48
and you've compared this description of eric to this stereotype and so people assess the degree to which a description fits a stereotype for each occupation people forget that librarian from this list is probably the least likely cancer even though in a study that use this exact scenario it was found that most people decided that eric was a library most likely though he's not and here's why if you look at the number of people in the u_s_ that hold each of these positions
19:23
white gray and is way way way down on the lips and this description is wholly based the characteristics and descriptors that could be could describe it anyone in any of these professions so stereotyping is one thing when the systematic error that we can engage in as a result of curious tix the second one is misconceptions of regression this occurs where original fall results are thought to be completely predictive of future results i'm a star by teaching a little bit about scientific history so
20:05
now finally looking gentleman in the photograph here is sir francis called you've heard of sir francis called in probably not many of you serve as a sculpting was a geneticist and in his studies of genetics heat discovered several statistical principles that are the foundation of modern statistical analysis techniques now sir francis called in as you can probably tell by the friendly expression on his face spent a lot of time studying sweet peas and he did that because it was did adinet assist
20:39
and he could look at the influence of successive plantings of sweet peas very quickly because they grow quickly maybe and so one of the studies that he did wise he would go a crop of sweet peas he would measure the size of the peace that he would plant those peas go another copy measure the size of the peas planted those peace grow another crop measure the size of those peas is over and over and over and over again and what i discovered was this idea that certain if you have a large
21:12
a particularly waar jp generation of peas the following generation of peas planted from those seats would tend to be a little bit smaller than the large peaks they found that he had a generation of peace that were particularly small the children if you will of those pieces would tend to be eliminate bigger then the small peaks and he called this phenomenon regression towards the mean and that means that in any given situation things will tend the average demean the mathematical average
21:49
he did this with height as well and he found that extremely tall parents we'll tell you have to call children but they're taught sheldon will be slightly shorter in the extremely tall parents they're the same way extremely short parents we'll have short children but they'll be somewhat tali philby closer to the average of the population as a whole this is not just true though in genetics in biology this is true in a lot of different areas
22:18
so for example numerous studies have been done in the education arena and it turns out that if you give uh... a large group of individuals an exam and look at their scores and take the top students and the bottom students in terms of their performance on the test and give them a second exam that covers the same material and is of equal difficulty the best students from the first exam will tend to do somewhat worse on the second exam and the were students from the first exam
22:52
we'll tell you somewhat better on the second exam curtain and the reason for this is this phenomenon of regression toward than me people and situations and events and any kind of situation intends to go toward the average cell you'll see this warning a lot past performance is not completely predictive of future results and that was though ultimate finding of this idea of regression toward the knee just because you do really well on one exam does not mean you will do just as well
23:28
on a similar exam given at a later time in fact regression toward the mean would say that you will probably do works you-all move toward the average score the third misconception or problem with the representative nis heuristic are misconceptions of chance when this occurs when people expect that a sequence of random events will represent the essential characteristics of that process even when the sequence is short so one of the essential characteristics of a coin flip first
24:07
if the quiz fair then its random there should be no relationship between one flip in the next the second is that it should be half heads and have tails on average if it's a fair court so let's say we flipped a coin six times and we got heads heads has details tales tales it turns out that we have this your estate even a short sequence of events should replicate the essential characteristics of that event so we feel like this is not a very likely outcome heads eds eds tales tales tales
24:45
'cause that violates our expectation that the process israel and that seems like a pattern same thing with this resolved if we got heads heads had said sense tales hats that while it's our expectations it should be about half had been have tails and so we think that this would not be a very likely results were much more comfortable or something like heads tales heads tales tales heads that seems like hits fulfilling the characteristics that we know to be present if we were to flip a coin
25:21
than infinite number of times but the problem is the first the third sequence here heads tails heads tales tales heads deviate systematically from chance expectation because it contains too many alternations into fuel rods it's true that if you flip a coin in infinite number of times it will be heads half the time entails half the time but if you're only looking at six flips characteristics of the process itself is not necessarily going to be duplicated are very short sequence
26:01
we also have what's called the gambler's fallacy if you've ever been to vegas go watch the roulette wheel while you'll hear people saying things indicate they believe in the gambler's fallacy and that is this so you're playing roulette and you see ret come up six times in a row we have this almost d internal belief once there's been a long run of red the next one is got to be black because it's got to balance out in some way this is the gambler's fallacy because you'll know
26:36
if you've ever taken a class in probability or elementary statistics that each of those rule that runs red or black is an independent event it doesn't matter what happened before it's a fifty percent chance of red or black coming up every time you throw the ball in the roulette wheel what happened previously has no impact on the outcome of each subsequent trial there's a fifty percent chance of red and a fifty percent chance of black on every spare so let's talk about the second heuristic
27:10
and that is availability this is a really coming here a stick and this is talking about the used to which it instance can be brought to mind is indicative of its probability of frequency so then back to the library that runs your brain the belly breaking easily accessed instances of a given event we assume we have this year restricted that means that the event is more common unfortunately that's not always the case for example a couple of problems first the retrieve ability error
27:44
there are all out of things other than that actual representative mr the actual availability uh... a given item that determine how retrieval bulletins so for example news television uh... n television programming increase the retrieve ability of certain types of situations even though less sensational events are much more common so for example we see that situational lethal events in one study like botulism poisoning tornado flood
28:19
homicide vehicle accidents were rated as much more likely than quiet killers like asthma or diabetes or stomach cancer or heart disease in spite of the fact that into in the real world if you look at actual rates of death through each of these things stomach cancer for instance is five times more likely than homicide but it's really easy when you think about think about you also and have a million images in your head you may be no uh... you've seen emilie episodes of c_s_ side
28:50
murder happens every week on c_ sa is very rich removal and so he was so best that it's more common stomach cancer you may not have as many retrieval bull instances where you can think of someone who suffered from stomach cancer we also have this issue of false correlations if you're told that a relationship exists you will see that relationship whether it actually exists or not tuners as really interesting study that was done uh... an illusory correlation study and basically
29:28
what they did was they present in subjects with a set of drawings that looks something like this supposedly this is a drawing done by a mental patient i actually do this one i'm about this one overt this over here is and evaluation from a clinical psychologist so this is what they told the subjects they given actually forty five of these different pictures different evaluations p so here's a picture drawn by a mental patient and hear are two statements
30:00
that a clinical psychologist has written that's their diagnosis of this patient so they had to look through forty five of these and later they asked subjects to estimate the frequency that each diagnosis had been accompanied by various features of the drawing features of the john would be things like uh... atypical allies or distinctive hair or emphasis of the mouth or uh... a large header emphasis of the head and they set up these forty-five
30:31
pieces of paper with the drawing and uh... statements so that there was absolutely no relationship between drawing features and the diagnoses further actually was no relationship when asked people go ahead and uh... tell us for example uh... many of the uh... subjects that you saw are worried about how intelligent they are which characteristics are were most present in those particular drawings and what they found was many would say that uh... had emphasis
31:04
they have a bigger head because they're worried about this or if people were worried that they weren't very manly then they would be bigan muscular in the drawing those relationships didn't exist but because people were told that some relationship existed between the quarter the drawing itself and these statements they infer that those relations existed that's another problem with v availability heuristic what we found from subsequent research is that the brain essentially stores
31:40
things that support the theory and tosses aside contradictory information so if we're told that their maybe some relationship between a drawing in i'd diagnosis of from a clinical psychologist they were going to infer that that relationship exists whether it actually does or not the third is what's known as anchoring an adjustment the basic idea is that we extrapolate our final position according to an initial position even if the latter is arbitrary let me give you an example so
32:20
the problem that happens with anchoring an adjustment is what's called insufficient adjustment and so no matter what that initial value is people don't just enough to come to a true a true value so here's the example there is this estimating study that was done they would bring in the group of people it would spin the wheel know we all had numbers from zero to one hundred or so they spin this we'll that comes up with a number and they said object so here's our number
32:55
this is the governor says the anchor subjects didn't know that split the anchor so this is the number i want to tell me is that number cobb african countries that are in the united nations the percentage of african countries that are in the united nations higher or lower than this number and then i wanted to estimate that number by moving up or down this speaker may explain that again so they spend it it's a totally random we'll comes up with a number and then they said
33:29
estimate the percentage of countries in that africa that are members of the united nation tell me if this is higher or lower than this random number that we generated and then go up or down from that number to come up with your actual aston well here's what they found if people are word given the random number of ten their average estimate was that twenty five percent of countries in africa were members of the united nations if they were given a random number of sixty-five
34:02
they estimated that forty five percent of african countries were in the united nations on average this is an average of a lot of people and so what this what this means is that even forgiven and arbitrary number week anchored to that number and we're afraid to stray too far from that number i do a lot of research and consulting in the legal field nicest attorneys with making the best arguments they can for the particular case that they're doing and one persistent finding in the field of
34:36
legal decision making is that if a plaintiff attorney asks foreign damage award in court it doesn't matter how high they go the more they asked for the more they will get if the jury awards averred and that's because of this principle of anchoring an adjustment if i passed for fifty million dollars even if you don't give me fifty million dollars you're gonna give me more than if i asked for ten million dollars because of this principle anchoring and adjustment and what's even weirder is
35:06
there's no ceiling effect i could ask for all ha but jillion dollars and all the reasons that has been done is that it doesn't matter there's no backlash if we set the anchor doesn't matter how ridiculous it is you or your people to give amounts than if the anchor as well so let's bring it all back what does this mean for scholars of communication what does this mean for advertisers people who were interested in advertising advertising industry so diverse can come in if you're interested in
35:47
in this uh... idea this article from nineteen seventy nine they note that people are susceptible to heuristic bites and as practitioners we need to be wary of how very simple message features org message changes can predict decisions
36:12
so diverse team canada came up with this idea of what they call prospect theory within this idea are these ideas of risk aversion and risk seeking so a person who is risk obverse were in a state of risk aversion is going to of void whatever the risk is if they are of risk seeker then they're going to guo after that risk they're going to be engaging more risky behavior we think of rescue in terms of prospect theory as probability so high risk might be ninety percent chance of failure
36:46
low-risk might be one percent chance of failure so what is prospect theory say the basic idea prospect theory is that people are risk averse it they will avoiding risk when they are confronted with games and they will seek a risk when they are confronted with losses let's use your alicia are you up at short video clip they'll sort of explain this idea of prospect theory in a really clearing easy way
37:34
expletives images
37:43
he confused misdirected and bamboozle so fifty one official forty-one discussed appleton evidence support sticking out for today pizza loved institute fields market working for us and he wants to give away inside the pockets quilts twenty words twenty pounds date forty fifty pounds they'll accept cheque you can walk away with that all he wants to looking over the last fifty-fifty alts you cannot that's fifty planes with the lolz you get nothing i get twenty that decision to take two twenty
38:32
all camel is there's a lot of acceptable don't know is that pete is manipulating anything else now that's not true if you want simply to walk through the way he presents the proposition is subtly different company tickets twenty points in the first case apology was presented with the freeway so my surprise this is called gains six aircraft consent that we can walk away yards gobind everyone today gamble and that you would think swiping the second place upon the data needed by subsequent
39:16
will make it fifty quit fifty quick but you also check mothers then almost immediately presented with what is called lost sonar hundreds that fit to give it back you wanna when that campbell the outcome both scenarios is identical keep twenty ok for fifty we want to see if the decision to go could be affected by the way he proposition his name gavel there are rigorously scientific study institutions market those who felt the ones you got were more likely to count to win the monitor
40:05
redstone pecans contempt citations not supposed to those partisan presented with a gain is twenty right alright who decided to stick with their own expected good fortune naked without realizing it of volunteers have decided on risk sorry for safety simply because of perception
40:49
so you saw on the video clip that just by changing the way the scenario was presented ultimate outcome with the same you get twenty bucks twenty quid twenty pounds has done in britain so twenty pounds and you can choose to gamble and either win fifty or when nothing if you have the twenty in your hand you are likely to just walk away with it if on the other hand you were given fifty and then thirty was taken back and you had a gamble to get that thirty back you're more likely to gamble
41:25
this is the idea of lost framing verses gain frame so people can react to the same information in very different ways depending on what ultimately are very subtle aspects of the message now it may be and some researchers suggested that risk in message framing in iraq so if there's a high risk that is being framed we're going to be of verse to that risk got forty bucks in your hand you might lose that if you gamble used to walk away with it on the other hand
42:01
if you had fifty and now you only have twenty you seem like you've lost that thirty you're going to be more likely to seek a risk in that situation what's the big picture today we look at three here wrist x and some of the predictable errors that they cost but we know that we're often guided by heuristics we look first at the idea of representative nissen we saw how uh... we might engage in stereotyping where were extrapolating incorrectly based on one or more features
42:36
we talked about misconceptions of regression another example of the representative heuristic gone wrong original results are thought to be completely predictive of future results you know that's not the case because of things like regression toward the knee we talked about misconceptions of chance where we often believe that a short sequence represents the essential characteristics of a process we also talked about the availability heuristic
43:06
underneath that to retrieve ability error and false correlations where we see relationships just because we're told they exist or because something is kisi to recall we think that is more com or more likely event and then we talked about the destin and anchoring where we will extrapolate our final position according to our initial position even if that initial position is arbitrary three questions about the contents of that alright have a great weekends we will see you again on tuesday
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat