Sherlock Holmes -- Science Study Break

Uploaded by utaustintexas on 04.01.2012

>> Roxanne Bogucka: We want to thank you all for coming out to the program this evening and we want to thank our generous sponsor University Federal Credit Union,
who supports the program's Science Study Break and makes possible the wonderful programming that we put on here and also the refreshments that you're all enjoying.
So, a big hand for UFCU.
[ Applause ]
If you've been with us in the past, then you've seen some of our other programs.
For example, we've looked at the biology in the movie Avatar.
We had a program where they looked at zombie epidemiology around Halloween a couple of years ago.
Many of these programs were recorded and are available on the Science Study Break section of the UT Austin YouTube page.
So, please check them out.
Also if you would like to suggest a TV show, a movie, a topic or presenter for future Science Study Break treatment, please leave your suggestions on the feedback form.
If you don't have a form right now, there are people who will be passing out the feedback forms.
It's a little half sheet.
Go ahead and fill it out.
We've got some golf pencils if you don't have an implement so that you can fill it out and leave that behind for us so that we can use that for future programming.
I can tell you that Sherlock Holmes is one of the most requested programs that we've had in the several years of Science Study Break and so I'm so happy
that we're finally getting to do this one this evening.
So your suggestions will not go unheard so please leave them for us.
If you are following us on Twitter, we're at scistudybreak or @utlsl.
You can follow Life Science Library on Facebook and if you're tweeting this event, the hashtag is #ssbsherlock.
Also, you might have noticed that there are all sorts of recycling containers all around the building both inside and out.
So please use the containers for your cans.
Please no cans in the trash.
Thanks for your cooperation.
It takes some diligent people to put on the Science Study Break and here are some of the folks who make it happen for us.
So, a big hand for all the people whose unseen efforts make this program happen.
[ Applause ]
Remember it's brought to you by the UT Libraries.
We have great resources and great people here to help you all the time so let's get right to it.
There's plenty of content here so let me first introduce Dr. James Bryant, who has a BSci from the University of the West of England and a Ph.D. from UT Southwestern Medical Center.
He's been a lecturer in the biology department since 2005 teaching classes in biostatistics and immunology.
His research interests include embryonic stem cell models, assessing 3-D structures of astereological techniques, mouse knockout technologies and the Bright transcription factor.
He is a PADI dive master and a campanologist.
Next we have Dr. Sam Gosling who has a BA from the University of Leeds and a Ph.D. from the University of California Berkeley.
He's been in the psychology department since 1999 teaching classes in introductory psychology and personality assessment.
He's an associate editor of the Journal of Individual Differences and Personality and Social Psychology Science.
His research interests include fundamental issues and impression formation; how research on animals can inform theories of personality and social psychology;
and using empirical indices to track trends in the history of psychology.
He's the author of the book, "Snoop: What your stuff says about you."
And now please welcome to tell us all about the statistics and observations of personality used by Master Sleuth Sherlock Holmes,
please welcome Dr. Jim Bryant and Dr. Sam Gosling.
[ Applause ]
>> Dr. Sam Gosling: Hello, okay, very good.
It's working.
Hello, it's Sherlock Holmes and they got two English people to talk about it because obviously we know a lot about Sherlock Holmes because we're English.
>> Dr. Jim Bryant: Technically British.
>> Dr. Sam Gosling: British.
>> Dr. Jim Bryant: There we go.
>> Dr. Sam Gosling: Okay, so anyway so how we thought we'd do it is we are going to, Jim is going to talk first about some stuff about 20 minutes and then I'm going to talk
for a little bit and then we would like to open it up because we both think it would be much more kind of interesting to make this more into some kind
of discussion so we can address your questions.
I beg your pardon?
>> Roxanne Bogucka: We're going to move the mic down into the middle of the aisle there.
So when you do have questions, please come to the mic.
>> Dr. Sam Gosling: Yeah, what she said.
All right.
It gives me great pleasure, over to you, Jim, hit it.
>> Dr. Jim Bryant: Okay.
So what I thought we would do first is we'd talk about Sherlock Holmes and how he used critical thinking and statistics
to reason back or forwards to the perpetrator of crimes.
My name is James Bryant and this is Sam Gosling.
So first I thought I'd give you a little introduction to Sherlock Holmes' little routine.
Let's see if we can get this to work.
[ Silence ]
[ Video clip inaudible ]
Okay so that was Sherlock Holmes and that was the first meeting with Sherlock Holmes with Dr. Watson.
Sherlock Holmes is accredited with being able to make from small observations very big inferences.
So what we'll do is we'll talk about basically the mechanism of how he does that.
Sir Arthur Conan Doyle was the author of Sherlock Holmes.
He was born in 1859 and died in 1930.
Born in Edinburgh, studied medicine in the University of Edinburgh, authored a lot of fiction books in addition to Sherlock Holmes and quite a number of non-fiction.
He conceived Sherlock Holmes, who was first published in 1887.
He killed off Holmes in 1893 due to he pretty much wanted to publish other things and write other stories and he felt very constrained by writing about Sherlock Holmes.
Over the next 8 years there was a lot of public opinion and demonstration against him and then he then actually ended
up bringing Sherlock Holmes back in 1901 in the Return of Holmes.
So, the real Sherlock Holmes that was accredited by Conan Doyle was actually Dr. Joseph Bell.
He was accredited by Conan Doyle as we said.
He was born in 1837, died in 1911.
He was a lecturer at the University of Edinburgh at the time that Conan Doyle was studying there and studying his medical degree.
He was actually a pioneer forensic scientist.
He actually was a real life criminologist in the 19th Century and he emphasized the key of observation.
If you can't observe it, there's no way you're going to understand it.
So he proposed observation and questioning for diagnosis in medical field.
Sherlock Holmes was classified as one of the world's best known fictitious detectives.
He featured in the Strand Magazine and Lippincott's Monthly Magazine in 1887.
He's characterized as the world's only consulting detective.
He employed impartial observations so he collected information without bias and then he evaluated the information either abductively or deductively.
Deductively is once I've removed everything that could explain something then whatever is left, however, improbable is the cause of the event.
Abductive is if I collect a series of facts, I can piece them together into an understanding or model of the world by creationism.
So, he used these processes to make inferences and in that case it was the solution of mysteries and identifying criminals.
So he was basically using critical thinking, which is basically statistics and there's a couple of branches of statistics we'll talk about later.
So, when do we need critical thinking?
So, first off let's talk about what is critical thinking.
It's the mental process of actively skillfully collecting, summarizing, conceptualizing which we would call descriptor statistics, analyzing, sensing and evaluating,
which we would call inferential statistics, information in order to reach a conclusion or an understanding about the real world.
So when do we need it?
If we were learning a new subject, we could structure how we learned to use critical thinking to make that learning process easier;
we could answer exam questions; we could determine the appropriate answer.
We can think of a lot of answers to questions and usually it's the most, if we have an ability to look at potential answers, we can then chose the best answer
and that's usually going to be the right answer.
Writing a book you have to organize, collect information, organize it and then organize the plot, add in the character definition and stuff like that.
So you can see basically we're going to need critical thinking all the time and especially in our lives when the chips are down.
We want to critically think our way out of problems and have the best solution to those problems.
So there are a number of steps to promote critical thinking and so critical thinking isn't just something that happens in a vacuum.
So to promote critical thinking, the number one is to understand the world we basically have to question.
The times you don't question, it's probably going to be the time where that information is unreliable and you're going to make a mistake.
So you want to question then you want to adopt this clean sheet of paper approach.
What you want to do is you want to empty your mind of preconception, you want to empty your mind of what you think is going on
and then fill a piece of paper with conceptual steps.
That's not looking for an answer; that's looking for an understanding.
Those steps to conceptualization are defined as definition, define the terms; draw, and that's things like schematics, figures, concept maps and graphs to help you understand.
Then redefine the question, what is really being asked in the question, and then you can then proceed, if necessary, to locate appropriate information.
Most of our problem solving doesn't need additional information than what we probably bring to the table.
If it was a bigger problem, we need to make a game plan.
We might need to order reagents.
If we're building a house, we might need a plan of the house, we might need the materials to pour the foundation, the materials for the roof and stuff like that.
Then we have to execute the game plan and stick to the game plan.
So, in the case of most simple problem solving, let's restate the question as part of your answer then it's really, really,
really state the absolute obvious and work from obvious to not quite so obvious.
Most of us when we problem solve, we tend to jump to a solution or jump to a statement and we tend to over-focus on detail
and so we miss the connecting steps to link ideas together.
Then finally most of us are human so most of us make a considerable amount of mistakes so we want to evaluate and we want to check
and double check our answers before they leave our desk.
So how does statistics work?
How do critical thinking of statistics work?
We ask questions in order to understand the world and that works through three steps.
We collect information and we want to collect that information so that our information is impartial.
It's not representing what we think is going on but what is actually going on.
Then we organize the information and that pretty much is tidying up.
So, pretty much tidying up everything in life is 80% of tidying up and the rest of it is just a little bit of chucking things together.
Thanksgiving meal 80% 2 hours preparing the meal, 20 minutes eating it, 2 hours cleaning up.
So most of the world is tidying.
Then we have to summarize the information and Sherlock Holmes said if you can't see it, you can't understand it.
So, we have to have some mechanisms of visually or summarizing the information so we can see what's going on.
With a lot of information it's just too much for us to take in so we have to have some structure to that.
Then the very last step, which is really the most unimportant of the three steps.
Collection, the most important because if what you collect isn't representative of reality, it's never going to tell you about reality.
If you can't summarize it, you can't understand your perception of that reality.
The last step is called inferential statistics and that's basically from what I've collected how reliable it is.
If I choose two people to describe the IQ of people in this room, I could choose the two wrong people and then I could get an overestimate or an underestimate.
If I add to my sample, I'm going to get a better estimate.
So, obviously things like sample size are going to lead to an error and I have to be able to measure my errors or imprecisions and my perceptions
to be able to build up an accurate view of the world.
So, let's see how Sherlock Holmes relates to all of those three steps.
So sampling and observation.
So, a problem is simplistic.
"When I hear you give your reasons," said Watson, "the things always appear to me to be so ridiculously simple that I could easily do it myself though
at each successive instance of your reasoning I am baffled until you explain your process.
And yet I believe my eyes are as good as yours."
And Holmes says, "Quite so.
You see but you do not observe.
Distinction is clear."
For example, you see the steps that lead up to this room but you fail to observe or count them.
So, two points come out here.
Watson could easily make the same inferences and deductions that Sherlock Holmes does, yet he doesn't follow the process in order to achieve that mastery.
Then the process begins with observation.
If you don't collect information, if you don't question the world, there's no way you're going to understand the world.
So let's have a look at a couple of instances.
Impartial observation.
"You don't seem to give much thought to the matter at hand," said Watson.
"No data yet," answered Holmes.
"It is a capital mistake to theorize before you have all of the evidence.
It biases the judgment."
This was from "A Study in Scarlet."
Let's flick over a moment for a little segment from Sherlock Holmes.
[ Video clip inaudible ]
So you can see here that a lot of people will respond to a little bit of information.
That response could be erroneous, it could be over responsive, under responsive and typically can be emotional, but that response doesn't necessarily bring an understanding
of reality or the ability to shape that reality in a favorable way.
So biasing assumption.
"There's nothing more deceptive than an obvious fact," said Sherlock Holmes and that's a lot of times we think things are obvious but that doesn't make them real.
So, we have to collect impartial information and then evaluate it.
Incorrectly fitting observations to theories.
"This is indeed a mystery," said Watson.
What do you imagine that this means?
I have no data yet, it is capital mistake to theorize before one has data.
One begins to twist the facts to suit theories instead of the theories to suit the facts and there's a lot of statements here that like the police they want
to arrest someone but that doesn't mean that they're arresting the person who perpetrated the crime.
So a lot of times they will have a theory that this person committed a crime and then they will look for facts to justify that theory instead of going from the other way
and collecting information and see who did commit the crime.
So how do we collect information impartially?
There's a number of tools in statistics for us to do that.
We can randomly select cases.
What that means is if we were to choose people in this room to describe IQ or height or any characteristic we would label people and then we would pull
from a random number generator, numbers that would correspond to people who were going to be in the sample and that ensures that every case
in the population has the same chance as being selected.
So that removes me as the observer from who is going to be in the observational sample.
Then we can do things like double-blind studies.
These are the gold standard for drug testing and things like that.
So, in double-blind study, the patient doesn't know what treatment they're receiving and the subject, sorry,
the patient doesn't know the treatment and the doctor doesn't know the treatment.
That insulates the information from things called placebo and nocebo effects.
If I give you a pill and you think that's going to make you smarter and you say it makes me smarter, that's the placebo effect or positive effect.
If I gave you a pill and say that's going to make you dumber and then you think you're dumber, that's a nocebo effect; that's a negative effect.
These effects although subtle are very powerful.
If we compare placebo effect to untreated, if I give you a pill and say this works and you believe it works compared to untreated, you can get response rates between 30-60% based
on your faith and belief system and that can be manifest not just as a psychological belief but as a physical manifestation.
For instance, things like hypnosis.
If I was to put a copper coin on some people and to hypnotize them to believe that they were allergic, they could actually mount an allergic reaction.
So, psychological belief can be incredibly powerful and influential not just in a psychological realm but in a physical realm.
If we put that into the context of drugs, most drugs that are on the market work 30-60% better than placebo.
Placebo works 36% better than not treatment.
So you can see the benefit of the drug compared to placebo is comparable to the placebo compared to untreated.
So, placebo effects or belief of the experimenter can be highly influential and belief of the experimentee can be highly influential.
So we have to insulate information from these preconceptions then the other thing is things like blocking and stratification.
If we believe if we treat with a drug that the drug affects age groups differently or genders differently, we can stratify the data so that we can compare control
to treatment of same age and so that's stratification to remove extraneous variables and those all help us to have impartial observation and to have an ability to collect data
that is going to represent reality and not some sort of other variable or other explanation or a preconception or bias.
So, the other parts of statistics, descriptive statistics and Sherlock Holmes said I see it, I deduce it and this was from "A Scandal in Bohemia."
Another part of statistics was conceptualization.
Holmes underwent frequency analysis in "The Adventure of the Dancing Men."
There were all these little characters of dancing men drawn on flowerpots and things like that from one of the characters in the plot and these were,
it was a cryptography and so he did an analysis.
He drew out all of these characters and then he made an analysis based on the frequency of the characters to which letter of the alphabet these represented.
So, basically he's taking information and he's doing some way of visualizing or summarizing information in order to have a handle on it.
So there are various steps to visualizing information.
It's essential to see, visualization is essential to see and understand information.
It is 90% of the process of correctly understanding the world.
It's a big, big chunk.
We don't need complicated things in order to understand the world.
Graphs and tables get us most of the way.
So, graphs.
These are pictures paints a thousand words and they are qualitative representations of information.
So they summarize information pictorially and then they're not truly quantitative.
Tables of summary indexes like mean, standard deviations and medians these are quantitative representations of information.
So if we give a table and a graph together, we can see visually what's kind of happening in the rough direction and then we can supplement it
with quantitative values of how much is happening.
So let's have a little exercise beating cancer.
Survival rates of 5 years of cancer range from 10%-90%.
Here we have various types of cancers ranging from thyroid cancer to pancreatic cancer.
Obviously what's your view if you go to the doctor and you're told you have cancer, what's your first thing you're going to say?
What type of cancer.
So if we look at breast cancer and testicular cancer, your survival rate is pretty good and then if compare that to lung and brain cancer the survival rate isn't that great.
Why do you think there's a discrepancy?
So now we visualize that we see a discrepancy try to visualize in your head, your body and articulate why there is a discrepancy between the various types
of cancers testicular, breast compared to brain and lung.
Go ahead?
>> [Inaudible].
>> Dr. Jim Bryant: So the less vital organs have the highest survival rate, why?
>> [Inaudible].
>> Dr. Jim Bryant: Right.
So, treatment.
If you have a non-vital organ, breast and testes, although you guys might not like to lose them, they can be removed and you can survive, but brain and lung,
any removal of those tissues is going to affect some sort of function.
Even a cubic millimeter of brain tissue could severely affect function.
So that affects prognosis.
What else?
Okay. So, we have later diagnosis in brain and lung cancer, why?
Why do you think it's later in those tissues compared to breast?
>> [Inaudible].
>> Dr. Jim Bryant: You can't look at them, you can't see them.
Your brain is in a bony cavity the same with your chest cavity is hiding your lungs and protecting those vital organs from damage.
So by the time you realize something is wrong, you're probably coughing up blood, you're probably dribbling blood from your ears or your nose if it was a brain.
You're probably falling over by which time the diagnosis is much later and the cancer is much more severe, okay?
The other thing that is also important and this is not intuitively obvious lung and brain don't have internal pain receptors.
They have reporting of stretch for the lungs so you can breathe correctly, but there's no pain sensors.
Same thing for the brain.
So you can have development without the body knowing it.
So these tissues have a cancer state.
In the case of breast and testes, they're external and you can palpate them and observe them and then you can lead that to early diagnosis.
So basically the ability to observe gives you power and affects a change on the probability of being a recipient of a bad event.
So, observation and understanding is power.
All right so let's have another little go at it.
What do you think the distribution of criminal tendency is looking like?
If we plotted the frequency on the Y-axis of criminal tendency against the X-axis of how criminal you people are, what would it look like?
What sort of shape would it be?
What do you think about criminals?
What do you think about yourselves?
Are you all criminals or you're not criminals?
>> [Inaudible].
>> Dr. Jim Bryant: We're talking about as we go this way on the X-axis we're getting more and more criminal.
>> [Inaudible].
>> Dr. Jim Bryant: Anything.
So, you know, low degree of criminal tendency, I steal erasers.
High degree of criminal tendency, I'm sticking the pencil in your eyeball before I steal the eraser.
[laughter] so, how does criminal tendency, how do the numbers of people with criminal tendency change with severity of criminal tendency.
>> [Inaudible].
>> Dr. Jim Bryant: You think it's a normal bell curve.
So most people have an average degree, quite a high average degree of criminal tendency.
>> [Inaudible].
>> Dr. Jim Bryant: So everyone is middling criminal?
>> [Inaudible].
>> Dr. Jim Bryant: Right.
So there's probably more low-criminal tendency than high-criminal tendency.
If we look at it, that's pretty much what we see.
So, we understood already criminal tendency, but if we didn't understand criminal tendency, what does this tell us?
There's a way to observe to understand.
So, what does this distribution tell us about criminal tendency?
In order to do that, we want to observe what is factually in front of our face.
So what would you say first off?
Look at the graph and tell me what the graph tells you.
As we go this way, criminal tendency is getting more extreme.
So what do we say about criminal tendency?
It's skewed so there's a tail to the right.
What else?
Hugo, what can you tell me?
>> [Inaudible].
>> Dr. Jim Bryant: Anybody?
You're telling me what you told me before.
You observed there are more low criminal people with low criminal tendency than high criminal tendency.
So, what does that tell us about the system of criminality?
So we observe the distribution of criminal tendency is right skewed, it has a tail to the right.
We observe from the data that low-criminal tendency is more common than high-criminal tendency and, therefore, from that we would speculate
that there must be a strong incentive not to be criminal, all right?
What would we normally expect the distribution of something to be?
What do you think?
Bell shaped.
So if we perturb the bell shape to a skew this way, so we've pushed the data this way, there must have been predominate factors to push criminal tendency down
and that's social pressure, it's not socially acceptable to be criminal so there are social pressures and obviously tendencies towards it's simpler to be honest.
So those factors predominate over criminal tendency.
Okay so the third part of statistics is inference.
So let's talk about Sherlock Holmes and inference.
Deduction and inference.
"On the contrary, Watson, you see everything.
You fail, however, to reason from what you see.
You are too timid in drawing your inferences," said Holmes.
We've got another little clip of this.
I think this is actually the clip.
[ Video clip inaudible ]
So Holmes then goes on to make quite an extensive critique of the owner of the hat.
So, he basically notes observations and then pieces those observations to describe the characteristics of the owner of the hat.
So, we have deductive or frequentist inference.
There are two branches of inference.
There's frequentist and Holmes typifies both of these.
Once you eliminate the impossible whatever remains no matter how improbable must be the truth.
So this is deductive reasoning and then Holmes also practices a process called Bayesian inference, but I have heard, "all of you have heard," said Watson,
"without however the knowledge of pre-existing cases which serve me so well," said Holmes.
So, Holmes is using two types of inferential statistics to piece together his understanding of the world and I think this is one of our last little clips.
Let's have a quick look at this.
[ Video clip inaudible ]
So, this was a demonstration of I guess more frequentist statistics.
He took all of the information, he built an abductive model of what was going on from observation.
That model was counter to the model the police had but, however, it explained all of the observations and the facts.
In that case, there was a tendency for misdirection by the criminal to pretend that this crime was actually a suicide when it wasn't.
All right.
So, we talked just briefly and I'm not going to go into detail about frequentist statistics because this is where actually the math and number crunching comes in.
So, we have two types of statistics.
This is the statistics of impartiality.
If your doctor was frequentist, you'd go to your doctor and he would test you for every disease under the sun and then he[1] would eventually diagnose you
but it could be quite slow, but he would always get there.
This statistics was proposed by Fisher in 1935 and it was actually quite smart.
What Fisher said is we can't actually observe the real world.
We can only observe a sampling of that real world and that sample by its nature has to have imprecision or has to have error.
So, if we can determine whether that sample is due to, the differences in samples are due to error, we can't make an inference about the population
because we can ascribe those differences in the sample to an error.
However, if we can ascribe differences in sample to error and we drew those samples from the population, they're telling us what's going on about reality
and so he focused this all on this principle of the null hypothesis that there is no real difference and can I overcome a threshold of being able to explain the difference
by sampling error and if I can overcome that threshold that I can't explain the difference by sampling error then there must have been a real error.
Sorry, a real observation.
In the other type of inferential statistics called Bayesian statistics, this is where you'd go to the doctor, you've got goop coming out of both ends and he would think, well,
you've got either enteric, an enteric situation either bacterial or viral and then he would run a number of tests to see if it was bacterial or viral.
If it was bacterial, he'd give you antibiotics and tell you to drink plenty and go home and suffer.
If it was viral, he'd tell you to drink plenty and go home and suffer.
So, but the problem with that is it can be very sufficient but say you had a cancer that was causing you to be nauseous and have diarrhea?
If it was out of his purview, out of his experience, he may never actually diagnose and you could die.
So, for a Bayesian statistics it has this prior assumption.
It can be incredibly efficient and get to the answer very, very quickly, but when it's wrong, it can be very wrong.
Frequentis is the statistics of impartiality.
It could be plodding but it will always get there in the end.
Bayesian statistics was proposed by the Reverend Thomas Bayes in 1763.
So some of these conceptual ideas have been around for quite a while.
So, in summary, why are statistics so vital to our lives?
It helps us perceive complex information; it helps us to interpret information; it's one way, not the only way, but it's one important way of seeing the real world;
it helps us to view and understand the gray world.
The real world is not all good or bad; it's a degree of good and a degree of bad.
So, our world is gray and statistics helps us to perceive and to understand and evaluate that; it helps us to put the gray world into a perspective that is reliable;
it allows us to weigh the costs and benefits of certain actions; and it provides us with choices; it provides us with ways
to make those choices once we understand what's going on.
Once we can visualize, we can usually see solutions.
Important choices are never easy.
Statistics can provide an avenue to work things out and to improve our lives if we choose to use them, but if we don't learn them,
we don't even have the choice to use them to enrich our lives.
So that pretty much summarizes my little discussion of what is critical thinking and statistics and now I want to open the floor up to Sam.
>> Dr. Sam Gosling: Okay, thank you.
[ Applause ]
Thanks. I don't think he's done.
He's hanging out.
I've told him to interrupt me if he wants to say anything.
I'm just going to quickly go through -- can I turn this stuff off and turn some lights on?
Would that be possible because then I can stand here?
Obviously not possible.
All right, okay.
[laughter] So, I'm a psychologist and what I do is I study how people leave traces of their personality, the various characteristics, in their spaces.
So it's very much like what Holmes does in terms of trying to make inferences about people from snooping around in their stuff is exactly what I do.
I am being blinded by that thing.
Can you hit B?
If you hit B, doesn't it turn it off?
There you go.
All right.
I don't mind that it's the light.
Okay. So who here believes you can if you were to like, for example, snoop around somebody's space, their bedroom, who believes that you can learn something about people?
Raise your hand if you think you can.
All right most of us do.
That's very consistent with our research.
Most people believe that after music preferences the next best place to learn about somebody is essentially to look around their space
because people's personalities is reflected in those spaces.
Okay so now who believes you can do it like Sherlock Holmes does which is go in and say, oh, this means this, this means this, this means this?
Who believes you can do that?
Yes. Well, the truth is it's much harder to do that, but many, many people do believe that's the case.
So, and so I do a little bit of consulting for the FBI and the FBI they believe that.
They believe that you go into a place and you do much like Sherlock Holmes does.
Oh, look, here's a bowler hat and it's got, you know, little worn around the rim and all those sorts of things.
And the truth is you can learn some things in general, but it's very, very hard to learn things to say X means Y and the reason it is is
because there are many reasons, the world is just complicated.
I wish it wasn't complicated but the world is complicated and the fact is that there are actually five different reasons why you might have a worn edge of your bowler hat.
It might be because, you know, you always put it up on the car.
It might be because, you know, somebody gave you a worn bowler hat.
Who knows what.
It might be because, you know, you always sit on the left hand side of the carriage or whatever you had in his days and, you know,
the thing is there are multiple different reasons why you might have these things.
So, behavior, so first of all objects and things are multipally determined.
There are many legitimate reasons why you might have a bowler hat with a worn edge.
The other side of the coin is also is that a given trait, say a certain personality trait whether it's extroversion or whatever, that can be manifested in many different ways
and so there isn't, there just simply isn't a one-to-one relationship between objects and things.
I wish there was.
So what you have to do is you have to do many of the things that Jim was talking about is you have to build up a story piece-by-piece.
So, for example, if you go, so our research shows that you can learn a lot about some traits from people by snooping around their stuff.
So, for example, some of the things you can learn about, one of the things you can learn about is a trait known as openness to experience.
So do people like to try new things?
Are they sort of philosophical and abstract?
Do they go on exotic adventures?
You know, would they go to this sort of the avant garde, a new restaurant that mixes sort of Himalayan cooking with southern cooking or something like that?
Or do they want to go to the Olive Garden, right?
So, which of those appeals to them?
Neither one is better, but people high in openness want the avant garde, people low in openness want the Olive Garden.
So it turns out you can learn a lot about this trait from snooping around people's stuff, but it's a great mistake to look at single clues.
So, for example, even one of the strongest [inaudible], one of the most diagnostic things for somebody being high on openness is do they have original art on the walls, okay?
But even that clue alone doesn't work.
You have to combine it with other things.
So, if you go into somebody's space, you should look at this thing and say, okay, I'm going to build up a theory.
So, this person could be open but I need to look for other things.
So any single clue can be very, very misleading.
So, in terms of going into people's spaces, we recommend that what you do is a couple of things.
If you want to go and be Sherlock Holmes and try and figure out, you know, you just meet someone and you go on one date with them then they go to the bathroom and you have three
and a half minutes to run around and see if they're crazy.
[laughter] I wouldn't do this but I have friends who would.
[laughter] So this is what you do.
First of all what you need to do is you need to look for certain clues which signal personalities but you really need to look for themes.
You need to look for themes because this helps you remove the possibility something is there by accident.
One of the biggest mistakes people make when they go into people's spaces is they look for things that are very distinctive.
Now something that is distinctive is doubly bad.
So, if you go into somebody's space and you see something that really sticks out those are really, really dangerous things to look at when you're looking around people's spaces
and the reason is as I said you want to look for clues, you want to look for themes, things that are consistent.
And so distinctive things by definition, by definition they are inconsistent with what's said.
That is what makes them distinctive, okay?
So first of all they are the worst clues.
They're probably the least likely to be telling something and they are most likely to grab your attention, too.
One of the spaces we were looking at in our research was just like this.
It was somebody's space and it was clearly a very responsible person, quite a sort of traditional, conservative person, somebody who engages in a lot
of civic activity, probably isn't much of a, you know, much of a crazy person, you know, calm person and everything in the room said that, right?
Except over in the corner of the room there was a blue plastic crate with a whole bunch of drug paraphernalia in it.
It was like a bong and all kinds of weird stuff in the room and, of course, I sent my research team in there and they kind of look around the room and go boring, boring, boring,
oh, that looks interesting and they go and look and they spend ages looking at this bong and all that stuff in the corner of the room.
Now, that stuck out in a way that it would not have stuck out in somebody who was more rebellious and less socially conventional.
It wouldn't have stuck out but it really stuck out in this room, right?
It really stuck out and grabbed their attention and they weighed it very heavily when they were trying to figure out what this person is like.
So, it was misleading.
Now I later asked the occupant because I knew the occupant I said so tell me what was that doing in the corner of your room?
She said oh, yeah, yeah, I completely forgot about that because my housemate she was going to travel around the world and she asked me to look after all her drug stuff.
So being the responsible kind person she was she put them all in a crate and put them in a corner.
So they did tell you about the person but not for the reason that you thought they did.
So, anyways that's one of the tips.
Be very, very careful about looking at these distinctive clues.
So, generally what is it that you want to do when you go, I only have about 2 minutes, is that right?
How are we doing on time?
I definitely want to leave time for questions because I thought that will be most interesting.
So I'm just going to spend a, you know, sort of tips about what you can do when you're snooping around people's places.
It's really the idea here is you have to try and think how these things got in this place, right?
You have to think of the connection between the person, the person's personality and what I call the behavioral residue
and so really there is the three broad ways people leave traces of their personalities in their spaces.
Two of them are deliberate and one of them is inadvertent.
So, so the first way is what I call identity claim.
So these are deliberate statements we make to the world about ourselves.
So I see many of you wearing things on your t-shirts, right?
These are deliberate statements you're making.
You chose to have these things on your t-shirts.
You are making deliberate statements to others, you're broadcasting your values.
Now, that doesn't mean you're disingenuous.
So people think because you're doing it deliberately they're just messing with you.
No, there's lots of research in psychology that shows that people are happy, healthy and more productive when they can bring others to see them as they see themselves.
So that's a lot of what's going on here.
Just because people have taxes you probably really do feel strongly about taxes and want to broadcast that to others.
You're not just doing that to mess with people.
So, first of all people deliberately say things about themselves in their spaces.
Another thing they do deliberately to their spaces is they try and affect them to make themselves feel a certain way.
The most obvious way, of course, is by playing different sort of music, playing lively music to make yourself excited or relaxing music to help you calm down,
but there are many things people put in their spaces which kind of reflect the sorts of emotions they habitually want to evoke.
So, sometimes you go into someone's bathroom and there's brightness and stars and they want to get up and get going.
Other people have a bathroom which is relaxing, there's nice smelly things they put in the bath and candles and magazines and music.
So you can tell so people are doing things deliberately to their spaces here.
They're doing things deliberately to their spaces but the goal here is not to send a signal to others; it's really to affect how they feel.
Now just because they're not sending signals to others deliberately that doesn't mean we can't learn about them.
So, as a snooper, you can still go in and notice they do this even though it wasn't intended for you; it was intended for them.
So these are what I call thought and feeling regulators.
People are regulating what they think about and how they feel are affecting the environment in which they live.
Those are the two things.
The third thing is what I call behavioral residue.
Behavioral residue is just the fact that in our spaces we engage in many activities, all kinds of activities and there's some set of those activities,
not all of them by any shot, but a subset of those activities leave a physical trace in the world.
So, for example, in olden days people used to have these things called CDs.
You may have heard of them, but if you look at somebody's CD collection and in the olden days they had these things called books, too, right?
There were books and CDs and in olden days you could go in and you could look at somebody's CD or book collection and see how organized it was
and how organized it was it doesn't organize itself, right?
That is a residue, can only get organized as a result of you organizing it and persistently organizing it, right?
I did that.
I thought, okay, I'm going to be an organized person, I'm going to organize my CD collection.
I'm going to be one of those people who has an organized CD collection and I do.
I spent a whole afternoon sorting them into genres and alphabetizing them and doing all this stuff and it was fantastic.
I had an organized CD collection for a day, right?
Because you can't just do it once.
You have to live that life, you have to live that life and that's one of the reasons why people's spaces are so informative.
They're so informative.
If I wanted to try and fool you about what my personality was like in an interview, I might be able to pull it off because I would just have to focus really, I'd have to focus
and just maintain the act for that half hour interview, but it's almost impossible to do in my space.
It would be almost impossible to do because I can't just suddenly change my whole space.
Even if I knew what to do, you know, I'd have to organize everything and keep it organized and buy new things and throw things out and these spaces become receptacles
of persistent and consistent behavior and that's why they are so valuable when trying to learn what somebody is like.
So, with that as background I'm not going to go into any detail and we can talk about it in Q&A later if you want, but here are the three things you sort of,
the procedures you need to do when you go into somebody's space.
You go in and, of course, you'll look around the space and try and get the themes as I mentioned earlier.
So you need to ask yourself three questions about any object.
So first of all you need to say what is that object?
What is the thing they have there?
That is significant.
It is significant that in your office space you have a desk calendar versus the person who doesn't have a desk calendar.
Okay, that's a start.
You've got a desk calendar.
That means you had at least aspirations to be organized at some point.
At least that, right?
But then you need to go beyond that, of course, you need to look at the state of it.
So, what is the object, first question.
Second question what is the state of it?
The state of it is very important because that tells you how it's used.
So, look at the calendar.
Is it meticulously filled out with birthdays in blue and appointments in yellow and little stars meaning various different things?
That means they're engaging in this, they're really using it and it tells you about their behavior.
Or as I've seen in many, many desk calendars bought by people on some deluded day when they think buying a calendar is going to make them organized, you'll see, oh,
it's a calendar which there is dust all over it and it hasn't been used for three years.
So the state of it is very important because that tells you how these things are used.
It doesn't mean it's not important but even in the case where it's dusty at least you know something about their aspirations.
You know they at least care about being organized.
Many messy offices people just don't care about it.
So, there's messes that people care about and messes that people don't care about.
Okay, so, we have the object, we have its state and then the next question to answer really is the location.
The location is really important because that tells you about psychological function.
That tells you what the purpose is of this space.
One of the best places you can do that is looking at the people, if you do into people's office spaces, look at the location of the photographs they have.
Photographs can serve multiple purposes.
They're particularly useful if they happen to have photographs on their desk.
So you go and see if the desk, now where are the photos facing?
So are they facing them?
Are they facing inwards, look, here's all the family photos and me shaking hands with somebody important and okay so it's for the occupant themselves.
In that case, it's like a thought and feeling regulator.
It's to make them feel while they're working away and then they can, oh, there's my wife and puppy and whatever it is and they can look
at that and I feel better now; that's for them.
Or if the photo is facing out then it's for you, then it's an identity claim.
Then it's making a statement.
Now, again, it doesn't need to be, it doesn't mean this is a disingenuous statement but what it means is it's serving quite a different function.
It's there for other people.
So those are the three things we can do.
I'm happy to talk about lots more but I definitely want to leave at least a couple of minutes for questions.
We have five and a half minutes for questions.
So, if you have any questions to either of us we will both take them.
Is that right, Jim?
You're up with that?
So, questions for us.
Maybe we can try and integrate our approaches.
Anyone have a question?
I can repeat them.
No you don't.
When do we go home seems to be the main question.
[laughter] Oh, yeah, there's a question.
If you yell it, I'll yell it back.
>> [Inaudible].
>> Dr. Sam Gosling: Okay, the question was is [inaudible] Holmes is it unrealistic?
I think yes and no.
I think a lot of it is realistic in the sense that objects and their state and location and so on do tell you about the activities
and potentially characteristics of the person who left those traces there.
However, I think generally it's very, very unrealistic I think generally because for the reasons I mentioned earlier it's because you simply, they simply don't have enough data
to make the inferences they claim to be making.
It's just each of the things.
Like in the TV show, they have Holmes saying, oh, well, I see this going on and, you know, I'm making these 15 different assumptions, you know, about this person.
Well, in the TV show, you can make that work but in reality that's not true.
In reality, you could have ended up with that exact same state of things for many different reasons and it's true if you think about your own space.
If you think about your own space, your own bedroom or your own office or your own whatever and you look at an object there I mean I do this all the time.
So, I study people's spaces and I look at objects in my space and I think look that really would give completely the wrong impression about me
and it would give completely the wrong impression about me because it's there for some silly reason like, you know,
I was going to take it out of the office and I forgot or blah, blah, blah.
So I think it is unrealistic in the sense that the world is much more complicated than it would be.
>> Dr. Jim Bryant: Well, having said that the obvious statistic is to understand our worlds.
>> Dr. Sam Gosling: I agree.
>> Dr. Jim Bryant: And you go into an environment and you try to understand that person from that environment, but what you're saying is from one or two very small pieces
of information if there are other ways to attribute a cause to that information then you can't ascribe what's really going on.
So you then have to collect more data and then you have to question if I can deduce or if I can say that this piece of information can only attribute
to this result then I can piece together a picture and an explanation.
So, in the case of the bong, so your collectors then went in and they honed in straight away to the bong and then they said this person is a drug user instead of looking
at the general messages that were more representative.
So, instead of being impartial to their environment, your sample was biased because it selected the --
>> Dr. Sam Gosling: -- right, but there were 10 possible reasons --
>> Dr. Jim Bryant: -- of course --
>> Dr. Sam Gosling: -- that could have explained that combination.
>> Dr. Jim Bryant: For that one bong.
>> Dr. Sam Gosling: The combination of things.
>> Dr. Jim Bryant: You've also said that person was fairly stable and calm.
>> Dr. Sam Gosling: Apparently, apparently from the belongings around.
>> Dr. Jim Bryant: And then when you questioned further that existence of the bong you then said, okay, well that bong was attributable to someone else bringing into that space
and it conformed to their personality trait because they were conventional and tried to please and tried to help others.
>> Dr. Sam Gosling: Well, yeah, in that case, that is what happened.
>> Dr. Jim Bryant: We're not really arguing against each other.
We're saying, no, we're not.
[laughter] No we're not.
We're saying can you understand your world?
Can you use systems to understand your world?
Can we?
>> Dr. Sam Gosling: Well, what I'm saying, my perspective, is that sure but there are so many parameters and the world is so complicated.
There's so many different things going on that it would be very, very hard to be able to collect enough data to say that this configuration of the organized responsible room
and the bong in the blue crate in the corner really means that because you could have that exact same configuration because somebody was cleaning the other room and put it
in here or this person had just moved into a new room and, in fact, the bong was their thing.
>> Dr. Jim Bryant: Right.
>> Dr. Sam Gosling: There are 20 different things.
>> Dr. Jim Bryant: But if you looked at some total of that person's room, you would say, okay, well, this one fact is not representative of all facts.
So if you would have said let's collect data impartially, you would have said, okay, 90% of the information is saying this person is fairly conventional;
one to ten percent is saying this person is not conventional.
So then if you then piece all of the information together as a story you're saying 90% of this person is conventional and then there might be some tendency not to be conventional
and that's your worst diagnosis, which is pretty damn close to what the person was.
>> Dr. Sam Gosling: yes, you're right.
You're absolutely right.
So, what I'm saying is you should absolutely make a thing I think this is going on but that's not how I see Sherlock Holmes working.
Sherlock Holmes doesn't come in and say, well, I think there's a 20% chance this happened and --
>> Dr. Jim Bryant: -- well, that's not quite true.
That's what he did do.
[laughter] That is exactly what he did do.
He said, so he saw all these weird little characters that people were writing on water fountains or whatever and he said, okay, let's look at them,
let's get enough data to look at these messages and then he said, okay, this figure is standing out so this probably represents E and then he worked through that
and the probabilities are, well, he could have been wrong at each step, but he knew eventually he was right because the model worked
because he abductively pieced the pieces together and built a whole and that whole was resonant because he could translate all of the messages.
>> Dr. Sam Gosling: Right.
>> Dr. Jim Bryant: So, although at each step he had a probability of being wrong, he appreciated his probability being wrong and then could assemble it into a truth.
>> Dr. Sam Gosling: Sounds right.
[laughter] Any other questions?
>> Roxanne Bogucka: This microphone is live now so if you can make your way to this mic for questions you'll be richly rewarded if you come here to ask your question.
>> Dr. Sam Gosling: They've put a barrier to asking questions now.
[laughter] Go ahead.
>> In your normal everyday life when you meet people, do you do this kind of thing?
Infer their personalities based on what they do and what their space is like ?
>> Dr. Sam Gosling: Yes, I think I do but -- was the question directed to me?
>> Yes.
>> Dr. Sam Gosling: Yeah, I think absolutely I do, but I think we all do, but the thing is it's so natural for all of us to do it that we don't notice it most of the time.
I think because, and you only notice it when you mess up.
So you think you know someone and you go, oh, this person is really interesting and then you go to their house and you see this huge collection of moose heads.
[laughter] You just see something that surprises you and you say whoa.
So, but the fact that you're sometimes surprised by it what that does is that indicates that you're doing it all the time.
So, but what I tend to do is where it's more useful to me is knowing the sorts of mistakes that people habitually make.
So, I know, for example, that our research has showed that people make systematically they make some wrong inferences about people
and I make those too but then I will stop myself.
I'll say, oh, you know, that person seems like a jerk to me but I know that we form impressions about whether our people are jerks on the basis of things,
on the basis of clues they're actually diagnostic or something else so I try to reel back in those sorts of things, but I think we all do it.
I think how I might be, how I might do it differently is I'm perhaps more sort of systematic in trying to think
about putting the information together may, but I bet you do it, don't you?
>> I guess without even knowing it.
>> Dr. Sam Gosling: Yeah, I think you do.
>> Dr. Jim Bryant: I would agree.
I think we all are looking at our world and trying to understand our world and we're all trying to assess what's going on.
Who is this person?
How do they behave?
And without training I don't know that we do it terribly well.
We would tend to have our pre-conception of what is going on and what would tend to influence our beliefs.
So you said you always have to check yourself.
You have to say, okay, well, I think this guy is a jerk but I'll give him a little bit of a chance to sort of hang himself.
[laughter] So, what we do is we collect information and then you're trying to inject the fact you're being slightly to some degree being impartial.
Oh, let's collect a little bit more and see if this information substantiates our view.
So, what you're really doing is you are doing Bayesian statistics.
You've got this pre-assumption of what the person is and then you're collecting further information to validate your belief and that's what Sherlock Holmes did.
I don't think that we could say that we could be Sherlock Holmes to that degree because I also agree.
I don't think you can collect information about people to that level, but we certainly can do it for medical testing.
We can run a few tests; the test can be positive or negative, they can have false positives, they can have false negatives.
It's all probability, but we can say if you get an HIV test that if you do one test, there's quite a high chance it can be a false/positive.
If you run two or three tests, the chances it's going to be a false/positive go much, much further down.
So, all information is error prone but by questioning impartially and then by requestioning, we can validate the information and we can do that for any process.
We can do that for people, we can do that for crimes.
As we try to build up more complicated models, that's when it starts to fall down.
You can't look at all of you guys and characterize your all for 50,000 variables.
We could characterize you for height and weight, IQ or something like that.
You couldn't' characterize an individual for a lot of information because that's when the information starts to break down because
of the probabilities, but I think we all do it.
With training we can do it better and it is probability; it's not going to be absolute but it can help us to enrich our lives
to a more better chance of getting a positive outcome in our lives.
>> Dr. Sam Gosling: Are we done, Roxanne?
Do we have more time?
I don't want to keep people who have been very, very good and I don't want to keep them beyond the time.
It's up to you.
You're in charge.
>> Roxanne Bogucka: You have the room.
>> Dr. Sam Gosling: They're just being polite.
>> Dr. Jim Bryant: Bring it on.
>> I guess my question is so you can when you're snooping someone's room you can analyze all of that stuff, but if you never get the chance like see their room or whatever
but you just see them consistently can you make similar inferences from how they dress?
>> Dr. Sam Gosling: There are certain things you can learn.
This is one of the things, different domains afford information about different cues.
So, for example, yeah, you can learn so clothing tends to leave quite a lot of clues.
So, for example, we looked at a number of different traits.
One of the things we have looked at is narcissism as well and that reveals itself in much of the ways you might expect it to.
So narcissists as you may know they tend to be obsessed with ideas of beauty and their own superiority about others and how fantastic they are essentially, you know,
this is over the top kind of like a brittle sense of self but nonetheless that's sort of the view.
We find in our research the narcissists tend to wear more expensive clothes, they tend to be much less likely to wear glasses, they tend to put more makeup on
and in females they tend to show more cleavage; those sorts of things, yeah, but there are lots of clues.
Again, though, again, this is all just sort of probabilistic you think there are these signals because there are other reasons why you might not be wearing glasses like, you know,
you don't have bad vision, for example, you know.
>> So my question is in terms of when you're analyzing someone's personal space how many times would you say you have to revisit that place to see
and properly assess the characterization of the person?
For instance, Monday through Thursday my room would be a complete and utter mess.
Friday I clean, Friday I do my laundry, and Friday things are set.
So, for instance you show up on Wednesday and there's this outlier of like all my jackets are on the floor --
>> Dr. Jim Bryant: -- it might not technically be an outlier though.
[laughter] It might represent a norm.
>> Uh-oh. [laughter] There's something fundamentally wrong with me.
>> Dr. Jim Bryant: I didn't go that far.
I just said it wouldn't be an outlier.
>> Okay.
>> Dr. Sam Gosling: I think it depends, you're right, the more times you do it the better and it is true is that somebody could have come into my house the day
after I organized my CD collection and get it completely wrong.
So, the more, you know, as Jim was saying the more times you sample the better.
So, you know, sampling from the space too, but one thing to be careful about is I think people aren't very good at judging their especially for tidiness of their own spaces.
So, for example, so within your space you have a sort of amount spaces fluctuate and I'm sure it does,
but probably even though you have this fluctuation the mean is very different.
Probably at its tidiest is not as tidy as some places and at its messiest it's not as messy as some places and that's because, you know,
part of your personality is you can only tolerate a certain amount.
Not only tolerate it but also pull it off.
I was giving a small graduate class on some of this work once and I was co-teaching with somebody else and in the course of doing so I said, ah, well, my office is very,
very messy and a person [inaudible] to me said it's not messy, it's very tidy.
And so we said let's settle it.
We'll take the whole class up to my office and we'll look.
[laughter] We all went in and they all looked around and about a third of the class thought it was, you know, scarily organized, you know,
a third thought it was how can you possibly get any work done with all of the papers and a third thought it was about average.
So part of personality is not only the things you do but how you perceive the world and it's quite difficult to fake those differences.
Now, but getting back to your question so that's the general thing.
I think, you know, we perceive in ourselves more fluctuation than if you had gone to 20 different spaces you would see, but if you can go multiple times that's better.
>> So generally in your experience in the studies you've conducted how many times would you revisit those places?
>> Dr. Sam Gosling: We normally just send people in once and we do a pretty good job because there's a difference between a tidy space and a tidied space, you know.
So, for example, if I knew somebody was coming to assess my home or my office, there's only so much I could do, right, if I really wanted to give the impression I was tidy
because I don't have the means to suddenly file and organize everything and alphabetize the books and all these things.
So, I could tidy it up and get rid of some of the surface tidiness but the structure would be there and I bet even when your place is looking messy, the structure is there.
That you have hangers for the jackets, you know, some people wouldn't have done that.
Oh, man, I don't have hangers I'm going to put five of them on one wirey coat hanger and it's going to fall off, it always does this and oh,
I must get around to getting some hangers but I never do because, you know, I'm taking care of the parking ticket I got because, you know --
>> Dr. Jim Bryant: -- but if we came in --
>> -- my living room has no furniture.
So, I don't know what kind of judgment you'd make about me, but we have no time.
>> Dr. Sam Gosling: Normally we do once but more is better.
>> Dr. Jim Bryant: If we came in five times and we saw your room and four of the days it was untidy and one of the days it was tidy then we would be saying
on average it was slightly untidy and that would be a fair representation of the times that it was untidy.
So, we can sample and as long as that sample is representative of that reality, as long as that sample is representative of that reality, as long as it's impartial,
as long as we haven't chosen to tidy on the day we know it's going to be untidy, it should be representative of a trend and you're not saying, okay,
we can use statistics to understand an absolute.
You're saying we can understand a trend in you, a tendency in you, yeah.
>> Thank you.
>> Dr. Sam Gosling: We'll make this the last question because you guys have been very, very patient.
>> I'm wondering if Holmes isn't really the wrong case study.
I mean he is, we love him so much because he's an archetype; he's someone who can make these little fabulous distinctions between different things and organize them.
He's someone who is super human in a lot of ways.
>> Dr. Jim Bryant: Well, one thing is I think we have to say, yes, he's a caricature; however, he was also based on a true character.
Joseph Bell was a criminologist; Joseph Bell did help the police; Joseph Bell was also a proponent of impartial observation.
>> Oh, certainly.
>> Dr. Jim Bryant: So, these tendencies, these abilities can help us to understand our world.
They're not perfect, they're going to have a chance of being wrong, but by practice by knowing how to do some of these techniques we can improve our ability to make decisions
about the world and we saw by representation of the cancer data that once you question it and once you look at it, you can build up an understanding from relatively small pieces
of information as long as you validate that information.
I think I agree with you Holmes is a caricature but he's affected his character based on things that can be done.
Conan Doyle was a doctor, a medical doctor, and he had experience of Bell who was his archetype.
So I think that although we may never be able to achieve that level, fictitious level of Holmes, we can enrich our lives to be able to understand our world,
which can affect our lives in a positive direction to help us get what we want.
>> Dr. Sam Gosling: Sounded like you were going to suggest somebody maybe.
>> I was going to say that Holmes actually, you know, he's not a real person.
[laughter] And, therefore, you know where the situation we're looking at is one that is artificial and is, it's archetypical for a reason.
It's set up to show us something.
>> Dr. Sam Gosling: Absolutely.
>> Dr. Jim Bryant: And it shows us how we can improve understanding.
>> Dr. Sam Gosling: I must say when I consulted with professionals doing that what is kind of scary is how often they will use films
and TV shows as evidence for the validity of what they do.
[laughter] It's absolutely amazing.
I'm not joking.
Like one law enforcement agency and say, yeah, well, you know, it's really, did you see that film with Steve Martin?
He did this and he did that and see how great this is?
I'm thinking, yeah, but that's a movie.
[laughter] And I think this is one of the things that is quite dangerous because, and as a result of this, they are systematically collecting data
on when they get it right and when they get it wrong.
They're just looking at the positives like, oh, I got it right, you know, well done and here are the five times I got it right.
All right, well how many times did you get it wrong?
And they say we have no idea.
No idea.
>> Dr. Jim Bryant: The conception idea is very true and it works in the other direction as well.
So this idea, you know, are these people, criminologists, trained in statistics?
Are they making impartial observations?
That's questionable and you would have to determine whether they have that background.
The flip side is also becoming a problem as well.
So, jury selection now a lot of people on juries have seen NCI and now they start to believe that these observations from science are absolutely true and,
in which case this tends to lead to convictions that may not be of people who are actually guilty because they believe that the data is absolute
and is not just indicative and they don't question.
So, it can go both ways.
If you don't understand that all information by its nature has error in it and question whether your observation is due to that error, you can make false conclusions.
To understand the world, we have to understand that we only see that perception of the world, that perception has an error
and by appreciating the error we can then enrich the probability of making a correct decision and we don't get to [inaudible] ourselves that our decisions are absolute.
>> Dr. Sam Gosling: Cool.
Well, thank you very much.
[ Applause ]