Okay everybody, wrap up your conversations and let's get started. Even if your conversation is very important. Okay, so how was the test?
Mostly found it good. So the multiple choice grades are up now. The written question will actually be up sooner than I thought. I told the TA to have it done by the next executive review day but she's on track to have it up sooner than that. So today we're starting the chakra on intelligence, chapter 10. And one thing I'll recommend, this is not required but it may help you, if you did not take 104, I would recommend having a look at the chapter 10. at either the PowerPoint notes from 104 on chapter 4, especially the first half, or the lecture on the channel about behavior genetics, because we will be talking about the heritability of IQ.
queue and I won't be able to spend as much time on that as I would like, at least like the basics of how behavior genetics research is done. So that may be helpful to you, though it is not required. You can still get 100% with the information that I'll give you in lecture and it's in chapter 10. Another thing I'll say is I have added some terms to the key terms for chapter 10 and I've also added a slide. The only terms that will be relevant... for today that are added are ratio IQ and deviation IQ.
So those have been added. So after today, or after class, just download the newest version of the key terms list. But I won't add any more. I'm done the practice runs of the lecture, so I've added all the terms that I'll be adding.
Okay, so we will start with the basics. What is intelligence? And that might sound like a simple question, but it proves to be one of the most controversial questions in this area. Intelligence is one of the oldest topics in psychology. It goes back over a hundred years.
And right from the beginning, there were disagreements about how to define intelligence. Now in general, when we try to define terms, there are two different questions that we can ask. What do we mean by x? And what is a... good thing to me.
So, in the case of intelligence, the first question would be, would consist of asking what the term means in ordinary language. There's a philosopher who I added to the key terms, who's very important, named Wittgenstein. And Wittgenstein argued that the meaning of a term is found in its customary use.
This is often put as meaning is use or meaning is usage. Words are typically not, or words typically cannot be captured by a single one-sentence definition. They have various uses in different contexts.
And these different uses are related by what they're used for. called family resemblance. The uses have some things in common, maybe some things not. So in the case of intelligence, if we ask question one, what do we mean by X, the answer's going to be a little vague, right?
There's going to be some meanings in some contexts, in other contexts it might mean something else, maybe a little subjective, because that's the way ordinary language works. It's somewhat vague and context-dependent. However, when we try to advance and become more scientific, we ask question two. What is a good thing to mean by a term?
It's common in science to take terms from ordinary language and kind of sharpen them a bit, kind of give them a more precise... and more useful meaning. So take the concept temperature. Temperature is a word that existed before thermodynamics emerged as a science and it has sort of different uses, right? You may say, you know, that conversation was really heated or temperatures really ran high.
In that case, what you mean is people were pissed off. It has kind of a more metaphorical usage. And lots of words are like that. But when thermodynamics as a science developed, scientists tried to give a more precise and more useful definition. And what they have converged on is the idea that temperature is molecular motion.
In physics, what temperature essentially is, is how fast are the molecules vibrating, basically. That's a more theoretically useful and more precise definition of temperature. And in the PowerPoint notes, I go through various virtues that a technical definition should have.
It should be precise, right? If we're defining something in a science, we want to define it precisely, or at least as precisely as possible. We want the term to be useful.
We want it to be useful in terms of generating predictions and explaining data that we already have. Things like that. The definition should be useful. So in the case of temperature, defining temperature as molecular motion is very useful. It's informed a whole body of work that generates predictions and explains a wide range of data.
Lastly, and this one maybe is a little bit controversial, but the technical definition should conserve some of what we mean in ordinary language. It won't be exactly what we mean in ordinary language, but it should have some features in common, if nothing else, just for clarity's sake. So in the case of intelligence...
There are different attempts to define intelligence in a more precise and useful way, but it should have something to do with what we mean by intelligence in ordinary language. The silly example I gave in the PowerPoint notes is, if you defined intelligence as how many cartoons you watch per week, that would be precise, it would be well-defined, easy to measure, and it would probably allow you to generate predictions and explain data, but it's sort of a weird definition, right? Yes.
has nothing to do with what we ordinarily mean by a concept, so maybe it's not the best way to go about defining it. However. It is up to us how we define the term, right? As the great philosopher Thor from the Avengers movies says, all words are made up. All definitions are made up.
We make up the words, we choose how to define them. I'm saying in this context is there's different virtues of different definitions we can give, and we should pay attention to making them precise, useful, and making them have something in common with the ordinary language definition. So how do psychologists define intelligence?
Well, as I said, that is a matter of ongoing debate. I just finished saying that one virtue of a definition in a science is a definition should be precise, and definitions in psychology are typically not as precise as those in the hard sciences. But broadly speaking, the different definitions of intelligence that have been put forward by psychologists seem to have the following features in common.
They define intelligence as an ability, and more specifically the ability to acquire knowledge, to be able to learn stuff, and number two, to reason effectively. and to deal adaptively with the environment. So, broadly speaking, we think of intelligence as capturing the...
these abilities. Many research papers in psychology use a different term that maybe captures what we mean better, which is cognitive abilities. You will sometimes see intelligence, intelligence and cognitive ability used interchangeably, and that is basically what we mean. We mean how good are you, how able are you, at solving cognitive problems, at solving cognitively difficult mental tasks. So in terms of the theories we have, that's broadly speaking what we mean by intelligence.
Well, as we see, even around this, there's ongoing debate. Any questions about what I've said so far? So we'll start by going through some of the history.
And the history of research into intelligence usually begins with Francis Galton. He's widely thought to be the pioneer of this field. Galton is an interesting character, a little bit unsavory in several ways. So, Galton was Darwin's cousin, and was very much influenced by his cousin's ideas.
So, he emphasized the role of evolution and genetics in intelligence. He's actually the person who coined the term nature versus nurture debate, and he was very much on the nature side. He wrote a book called Hereditary Genius, in which he argued for the nature of Sodom.
Now, as I say, he's a bit of an unsavory character, because he is widely thought to be the founder of eugenics, which was the movement... to kind of improve the human genome by encouraging some people to reproduce and discouraging others to reproduce. This of course led to the eugenics movement in the Western world which included forced sterilization. so forcing certain people to be unable to have kids based on psychological tasks and other things, so a pretty dark part of our history.
That was done right here in Alberta until 1972, so not that long ago. Now, Galton, even though he's an unsavory character, he was not a dummy. He is actually the one who invented the correlation coefficient, and that may make some of you hate him even more, but he is the person who invented the correlation coefficient.
Another early pioneer of this work, who's a... Not so much of an unsavory character, is an outward. That shows how well I slept. Alfred Binet.
Alfred Binet was a Frenchman. And he developed some of the early actual intelligence tests. Dalton, to my knowledge, did not develop formal intelligence tests, but Binet did. And Binet developed them for actually a very good and benign reason. He wanted to be able to identify which kids needed help.
I believe it's the case that France was the first country to have universal education of all their children. And what Binet was tasked to do is give us a simple test that can identify which kids will need more help when they start school. So this is actually a very egalitarian and positive motivation for developing those tests.
So they would give various mental tests having to do with mathematics, reading, writing, etc. And he developed the idea of a mental age, which is the age at which you score. Or, to make that concrete, if you are doing as well as the average seven-year-old, that means you have a mental age of seven. Whatever your real age is, your mental age is, you know, where are you scoring?
You're scoring the same as the average five-year-old, the average seven-year-old, or whatever. And this idea of mental age formed the basis of the intelligence quotient. IQ is a very famous term.
You may wonder where does the Q come from? Where's the quotient part? And that was done by someone named Stern who took Sorry, there's a lot of names in this part, but he took Binet's idea of mental age and turned it into an equation from which you could get IQ. Here's the equation.
It's mental age divided by chronological age and multiplied by 100. Now to be more specific, this is ratio IQ. Later in the chapter there's a modern, more commonly used method of getting IQ, but this is ratio IQ. So let's say there's someone named Susan.
And Susan is 10. That's how old she is. That would mean her chronological age is 10. Your chronological age is how old are you, like, you know, for real. How many years have passed since you were born?
But if she is doing as well on the tests as the average 15-year-old, she would have a mental age of 15. So her IQ would be 150. 15 divided by 10 multiplied by 100. So 150. And yes, you should know how that works. There may be questions on the test with this. This was further developed into, for a long time, one of the most popular IQ tests, the Stanford-Binet test. So, Stanford-Binet, at least the early versions, used this.
It's called the Stanford-Binet because they took some of Binet's tests, but developed it at Stanford University. So for a long time this was a very commonly used IQ test. And it was mainly used for kids. As we'll see later. One reason this is not used as much anymore is it only really makes sense to do things this way with young people.
Like suppose you're 20 and you're scoring as well as the average 40-year-old. Does that mean you have an IQ of 200? Like that seems kind of strange, right?
Because actually performance on many IQ tests actually goes down the page. So it would be very strange to do it like that. So doing it this way only really makes sense for kids.
but widely used for a long time. Okay, to sum up some of the points I've made, I'm going to show a video that summarizes some of the history. And the reason I show this is to... kind of educate you on why this chapter is very controversial. This is arguably the most controversial chapter in the book, and one of the reasons for that is intelligence tests have a very dark history.
There's a long history of misuse, and it's good to be aware of that history. And this leads to a bit of a paradox, because on the one hand, IQ tests are widely criticized and widely looked down upon because of this history, but on the other hand, But despite that, it is in fact the case that there is serious work done in this area. IQ tests are among the best tests we have in psychology.
But they do have a history of misuse, and it's important to know about it. Okay, so a good summary of the history. Many of those terms we'll be coming back to, like the Flynn effect, etc.
One thing I forgot to mention that may be obvious, but given the way this works, the average... IQ is by definition 100. And that has stayed the same. Current versions of IQ tests do not use this equation, but they do make it such that the average is 100. Right, because if your mental age and your chronological age are the same, that means you're scoring the same as the average for people your age, which means you're average by definition.
Okay, next we have to talk about factor analysis. Any questions before we move on to that? So factor analysis is a technique for seeing how different types of mental tasks correlate with each other, how they tend to hang together. And those of you who are in my 104 class will know this.
I have a very simple example. So this technique was invented by someone named Charles Spearman. We're going to be saying more about the second.
Spearman is very important. This technique of factor analysis is used widely. It's used outside of psychology.
So let's say I gave everyone in this room four questions. How much alcohol have you had to drink in the last month? How much marijuana have you smoked in the last month?
How many... I'm hearing laughters of admission. How many parties have you been to in the last month?
And how many black playing cards, how many sets of playing cards do you owe? So, we've got three questions. Alcohol, red, marijuana, you can probably guess which color I'll make it. Let's say parties, I'll just make blue. And how many sets of cards do you own?
I'll make black. Okay, here's what I would predict. I don't know if this is true, but I would suspect the first three would kind of go together. People who say I've had more to drink in the last month, on average, have probably smoked more marijuana in the last month.
And probably... been to more parties. I don't know for sure if that's true, but to say it's that. Probably this one would not be very strongly correlated with these three. Maybe, but I would predict this one would not be.
Let's suppose that's true. Let's suppose that these three throw together, but this one doesn't. What you can do is, you can use factor analysis to see if you can extract one factor that explains variation in some of the items. And what that essentially means is, these three are varying together.
And when you represent it, it looks something like this. And this is not a scatterplot. I forget actually the term for this, but you can represent how they vary on a given factor, and these three would probably go together and this one would probably be off in statistical space.
It would not be the same. So this is the cards one. Now, this is not super precise, this is not good enough to pass a statistics class, but that should give you an intuitive sense of what factor analysis is. You're seeing whether items go together.
So what Spearman did was he took this technique that he had developed and applied it to mental tests. So you give people... Lots of different mental tasks. You give them vocabulary tasks, you give them math tasks, you give them spatial reasoning tests, you know, all sorts of stuff.
And what you're looking at is, do they vary together? If you do well on the verb... ones are you likely to also do well on the math ones and so on. And this could have gone either way, right?
You can imagine them all kind of going together or you can imagine them being very separate. I mean people often say, oh I'm really good at this. English classes, but I'm really bad at math classes.
Some people are like that, but what you actually find, and this is extremely well-applicated, there's really no controversy that this is the case, you find that mental tests are in one big club. You find that all mental tasks are strongly positively correlated. Meaning, if you do well on one, you are more likely to do well on the others. And again, it didn't have to be that way.
In the case of personality, the consensus nowadays is that it's the best summary of the data is that there's five clips, five factors. We'll see in personality, the leading personality theory today is the five factor model, where there's five. different main personality variables. But in the case of intelligence, to a first approximation there's one, is what Spearman found. And the term he used for this is the G factor.
Which you saw in the video. Sometimes called general intelligence or general cognitive ability. So this is what Spearman is most well known for, is the discovery that all mental tasks are positively correlated.
The g-factor explains something like 50% of the variation in the subtests. So it explains quite a bit of the variation. Now Spearman is not as well known for this, but if you zoom in to the big clump, you can find subclumps. Like yes, mathematical items do also tend to cluster together kind of within this. And his term for that was S-factors, or specific abilities.
And here's a representation of that. So you can imagine giving a bunch of verbal tasks, a bunch of math tasks, and they would kind of go together. Same with spatial tests.
So that would be the S factors, but the G factor is kind of underlying cognitive ability. It's the it's what all the tasks tap into. One thing we'll see later when we talk about neural correlates of intelligence IQ tests do not correlate very strongly with brain size only about 0.35 or 0.45. So not especially strong The g factor correlates with brain size at around 0.6, which is getting pretty good.
So you can think of the g as just how good your brain is in general at solving problems. Not just math, not just verbal, but in general, how good is your brain at solving problems. Here's another analogy that may help. 0.6.
So total IQ score correlates with brain size at around 0.3, 0.4, but G factor specifically is 0.6, which is pretty high, at least for psychology. Here's an analogy that may help you think about this. I don't know if anyone's tried this, but I strongly suspect that you can do the same thing for athleticism.
If you gave, let's say you go to a high school, and you give everyone a bunch of tests about athleticism. You look at how much they can curl, you look at... how far they can throw a ball, how fast they can run, how high they can jump, whatever, a bunch of tests.
And you would find that some people were better at one test than they were at another. But you would probably also find that people... who were good at one tended to be good at the others. You've all been in high school PE classes. There are some people who do better kind of at everything, and some people who do not so well kind of at everything.
And you can probably extract a G for athleticism. That's not testable, that's just not an analogy. Blasey's experiment is most famous for the g-factor, but some of the people we'll look at next kind of zoomed in on these. What's the nature of the s-factors?
What are the sub-abilities that are measured by IQ tells? The first example of that is Thurnstone, who argued that there are seven sub-abilities. They're listed in the PowerPoint notes, so they are space, spatial ability, how good you are at reasoning about spatial relationships, verbal comprehension, How good you are at understanding language, word fluency, how good you are at producing language, number of facilities of math, perceptual speed. So you give someone a pattern, how quickly can they identify the patterns?
Memory and reasoning, which is how good you are at dealing with novel, new cognitive problems. One more simple, and I think more widely agreed upon, way of dividing up G. When I say why they agreed upon, not to say that people think this is all that's going on, but I think people, even from different theoretical perspectives on intelligence, would probably agree on, at least on this, is the distinction between Sometimes called GC and GF. Crystallized intelligence is essentially how much stuff do you know? How many facts and techniques, etc., have you learned?
Whereas G, F, fluid intelligence, is how good are you at solving new problems? How quick and nimble and agile is your brain? So one example of a measure of crystallized intelligence would be something like vocabulary that is used on IQ tests. That's a measure of how many words have you learned.
One example of fluid intelligence, which is one of the best measures of IQ, is the Raven's Progressive Matrices. You don't need to know that word, but it's a helpful example. And what you have to do is you have to identify a pattern.
So there's a pattern in the top row. This one's pretty easy. There's a pattern in the top row, then a pattern in the second row.
And based on that, you have to infer what will be in that empty space. So if you're really getting a headache, that's a little concerning. So what's happening obviously is they are rotating clockwise. So what we're going to do in the last one is the same shape, like the diamond shape, but the black dot would be one turn over. I won't embarrass anyone or ask for guesses, but this would be a harder one.
You'd have to look at it a while to try to piece together what the pattern is. And this is an example of fluid intelligence because most of you have never seen anything like this, probably. Maybe you took an IQ test at some point, but if you haven't, you've probably never seen this. And your brain has to kind of figure out what to do and how to solve it.
So that's an example of a measure of fluid intelligence. And I show the examples I did, vocabulary and the Raven's progressive matrices, because these are two examples of tests that are highly G-loaded. So tests vary in terms of how well they tap into G.
I'll look at it later and tell you guys if you made one. So tests vary in terms of how well they tap into G. These tests are highly gene loaded, which means they correlate very strongly with gene. Which is really another way of saying they correlate very well with all the other IQ tests. If you're only going to give people one or two IQ tests in order to make a better guess of how they do on the others, these would be two really good ones.
They seem to be very strongly correlated with how well you do on other tests. By contrast, something like the digit span, which is in IQ-10, the digit span... You read someone a series of numbers and you see how many they can correctly read back to you.
You start with like 3, like 2, 7, 1. And then you slowly make the number longer until they can't do it. So you have 4, 8, 9, 2. And most people can't do more than five to nine. That's like the typical range. That is correlated with G, but not very strongly.
So if you're only going to give people one IQ test to get an estimate of how well they did on the others, these would be examples of better ones to go. And even though these are conceptually distinct, these two are also highly correlated. There are different explanations you could give for that. One explanation you could give is, people who have just kind of naturally more fluid intelligence, over time they tend to acquire more crystallized intelligence. If your brain is good at solving problems, and picking out patterns in the world, you tend to absorb more information from the environment.
Think about vocabulary. We're all exposed to thousands and thousands of words, and you don't necessarily learn them all through rote memorization. A lot of them you just learn by absorbing them from the environment.
And people will differ in terms of how readily they absorb information from the environment. Okay, the book goes through a number of other ways of dividing up intelligence. As always, focus on the ones that are in the key terms. I'm not going to go through every single division.
I'm going to focus on the two that are probably the most famous, which is Sternberg's theory, which has three types of intelligence, and Gardner's theory, which has eight. Sternberg, and Gardner's is, I would say, the most famous, the one with eight. But very quickly, Sternberg distinguishes between analytical intelligence, which is what IQ tests measure, so how good you are at solving cognitive problems, Practical intelligence, which is sometimes called street smarts.
So how good you are at dealing with problems in the everyday environment. And creative intelligence, which is how good you are at generating new ideas or new solutions. So that's Sternberg, but as I say, the most famous account that has multiple intelligences is Howard Gardner. And before I say a little bit about how that model works, I actually have a video of him describing the eight.
Practical intelligence is not defined in terms of whether you know a lot of information or whether you can solve problems quickly. It's about things like okay how good are you going to do in a job interview? Do you know how to use tools? Do you know how to navigate the streets or whatever? It's more of those kind of everyday skills.
So graduate students would probably be high in analytical intelligence, maybe not so high in practical intelligence. So as I say though, I have a video of Howard Gardner. So we'll watch him define the eight types of intelligence. Apparently he has some new ideas about some bonus ones, but I'm gonna stop the video before we get to that. We'll just focus on the eight that are most famous.
So, he went through them a little fast. So, one thing I want to correct though, he said that typical IQ tests only measure the first two, and that's not correct. Spatial intelligence would also be measured by standard IQ tests, as we'll see later.
when we look at the breakdown of the most popular IQ tests today. Just in case you didn't hear, so just very quickly, so linguistic means skill with language. Logical and mathematical is kind of self-evident. Logic and mathematics.
Spatial intelligence we've already talked about. Musical should be pretty obvious. Bodily kinesthetic is essentially hand-eye coordination.
So if you're good at sports or that kind of stuff, I think you would define that as a kind of intelligence. These are two, so inter-personal intelligence means skill with other people, and intra-personal means skill with yourself, so you're good at understanding your own... emotions, good at managing your own emotions, that kind of thing. And then naturalist intelligence is skill with the natural world.
You naturally pick up on patterns. You'll be good at bird seeing or being a biologist or a geologist or something like that. Okay, so those are the eight.
Any questions about the eight? They were widely influential in education. The controversy is really more of a semantic one.
He kind of alluded to it, which is, does it make sense to call these intelligences? And that's really a semantic question. Clearly there are different...
kinds of abilities and talents. Whether you want to call them intelligences or not, doesn't really interest me. Oh actually, I'm glad you asked, I forgot. I have one cartoon which I think reflects the motivation that Gardner had for developing these eight. I always found this kind of funny.
For a fair selection, everyone has to pass the same exam. Please climb that tree. You can see the monkey is very happy, and the other animals are not so happy. So clearly there are different abilities that people have.
Whether you want to call them intelligences or not is controversial, but maybe not all of them. Okay, the next one is emotional intelligence. The most widely validated emotional intelligence test has four sub-components.
I'm not going to go through all of them, they're in the book, but there's one that I want to look at, which is perceiving emotions. So, I couldn't find any items from the... The MSCEIT, which is the most commonly used emotional intelligence test, but there's one that is very similar to one of the four sub-components.
So one of the four sub-components is, as I say, perceiving emotions. There's a test called the Reading the Mind in Eyes test, which measures how good you are at perceiving emotions in other people's faces. This is a widely used test.
It's very well validated. Part of what it's used for is screening people who might be at risk for autism or who may... need further testing to determine whether they have autism.
Autism is correlated with, or associated with, difficulty perceiving emotions in people's faces. Interestingly, this is a test with a sex difference. Women are, on average, better at this test. And, probably not a coincidence, men are much more likely to have autism.
This test was developed by Simon Baron-Cohen. How many of you know Sacha Baron Cohen? It's a Borat dictator.
Those are his movies. So his cousin is a famous psychologist who studies autism. And as you can see there's a quote from Simon Baron Cohen, which is to be male is to have a mild form of autism, which I always thought was kind of funny. Anyway, so I want to quickly look at this. This is not the test that's mentioned in the book, but it's very similar.
So, what you do is you have to look at this and decide which emotion this is. Anyone want to say? Which emotion is this gentleman?
Playful. Playful? I think that's right. I don't remember for sure. How about this one?
Someone says upset? Okay, I'll go with that. I'm pretty sure that's right too.
I did look at the answer key. How about this one? Anyone know what this is? The bottom one is convinced?
Desire, good. Okay, so that is correct. I will note, I gave this test to my mom and dad.
I was curious how well they would do. My mom did better than my dad, but the ones like this my dad was very good at. He always got these ones right and my mom didn't, so. Make of that what you will.
Now, one issue that comes up as a criticism of emotional intelligence is some critics have alleged that emotional intelligence is basically just normal intelligence plus extroversion. intelligence, like you're good at detecting patterns in the world and so on, but you're also extroverted so you're interested in people, you'll kind of naturally learn different facial expressions. And there have been studies on that, so what you can do is you can test for what's called incremental validity.
I will not test you on this, but... I'll mention it for those of you who are interested in this. Incremental validity means, does a test have validity independent of other tests?
Is it kind of incrementally improving our understanding? What you can do is you can give a measure of emotional intelligence, and also give them an IQ test and a measure of extroversion, and control for how intelligent and how extroverted they are. And you do find that EQ, measures of emotional intelligence, do have validity that's independent of IQ and extroversion. Introversion helps you read people better? I would guess that's not true, but that would be an empirical question.
You have to see if they're quite right. Okay, now we'll move on to the types of IQ tests that are most widely used today. So whereas the early IQ tests used ratio IQ, today most tests use deviation IQ.
And what it boils down to is how many standard deviations away from the mean are there? That's what's used to calculate your IQ score. So I'll show you what the equation is for that. Now to show you how the equation works, I have to make some stuff up.
So let's say that the average, like you get an IQ test and people have a total score. Let's say the test is out of 68, and let's say the average is 42. Just making that up, I have no idea if that's right, but let's suppose it's that. And let's further suppose... This is the mean. And let's further suppose that the standard deviation is 4. So people on average get 42 out of 68, and the standard deviation is 4. Here's the equation for IQ, deviation IQ.
IQ equals 100 plus 15. multiplied by your score minus the mean divided by the standard deviation. So let's make that concrete. Let's say there's someone named Bob.
And let's say Bob got 50. out of 68. And again, I won't test you on these numbers because I'm making them up, but you should know this equation. This equation is good to know. So let's say Bob got 50 out of 68. So his score is 50. Actually, I would leave the equation. So Bob got 50 minus 42. And then you divide that by 4, which is the standard deviation.
You take all that, multiply it by 15, and that's going to be added to 100. So 15 minus 42 is 8. So he's 8, so that means he's 2 standard deviations above the mean. Right? 2 times 8 is 3. So it's going to be 8 divided by 4, which is 2. 2 times 15 is 30. So Bob is going to have 1 third. And so I should say... In the way that, what this results in, is the average will always be 100 and the standard deviation will always be 50. This is a method of transforming scores so that this is always the mean and this is always the standard.
These numbers you should know. These I just made up. So the way that IQ scores are transformed, 100 is always the average and 15 is always the standard deviation.
But this is how you calculate. I see some nods and some looks of confusion. Any questions about what I put up here? Yeah, so if I was going to ask you to use this equation, I would have to give you the mean and the standard deviation for the raw scores.
These would be the raw scores, but they're transformed such that it looks like this. And actually, just to show you what this looks like. So this. Once you transform the scores, this is what the distribution of IQ roughly looks like.
It's a little bit idealized, because it's a smooth curve. It wouldn't quite be that good, but approximately it looks like this. Most people will be close to the average.
Most people will be near 100. To be more precise, 68% of people are between 85 and 150. By the way, the book has... A distribution that breaks it down by 100, 110, and 120. Just ignore that. This is in the PowerPoint notes. Use this.
You should roughly remember these numbers. I'm not going to say, like, is it, you know, 14 or 13, but you should know roughly how many people fall in various ranges. Yes? How is IQ? Say again?
Oh, there's different ways of doing it, but there's no consensus on how to do that, and I certainly will not test you on that. She asked how is IQ measured for an analyst, and don't worry about that. Is the equation for variable standard deviation, is the standard deviation of the test right?
It's this standard. It's the standard deviation of the raw scores. The raw scores.
Yeah. So in this case it would be 4. But... Yeah, the way this works is, and the reason 100 is average, and let's say Bob got 42. Well, then 42 minus 42 is 0. 0 divided by 4 is 0. 15 times 0 is 0. So he's going to be 100. That's the reason it works out like that. Yes? Yes, it does.
So this is, so he asked, does this measure fluid and crystallize? And the answer is yes. In fact, I'll move to that right now.
So this is what you do with the total score. So let's say it's out of 68. But this is the actual elements of the IQ test. The most commonly used deviation IQ test today, in fact, it's the most commonly used IQ test in general, is the WAIS, the Wechsler Adult Intelligence Scale.
I think it's in its fourth edition. That's much. So what it is, is it's four scales, each of which consists of subscales, that are then totaled into a raw score, and then transformed with the equation that I showed you.
So the four basic scales, so verbal comprehension, basically how good you are with words, that includes vocabulary, information, and similarities, which I'll show you an example of that in a second. Working memory is how much information can you hold in your head and do stuff with it. So we talked about working memory in 1.04, and it's basically that, right?
How much can you hold in your head and manipulate and solve problems with? Perceptual reasoning is more of a spatial set of questions. So the Raven's progressive matrices we saw earlier, that's similar to some of the items in here.
And then processing speed is how quick your brain is. identifying patterns and solving problems. So I'll show you one or two examples of each. So here's some sample items from the similarities uh so i'm still these are fairly easy ones but what to an hour and a year have in common the answer would be unit of time if they are units of time what do a fish and a crow have in common and what do a poem and a statue have in common that was slightly more abstract so there you would say they're both you know human artwork I mentioned the digit span earlier. That would be an example of something in the working memory index.
So the test giver would read you these numbers and you have to say them back correctly. The researcher will slowly increase how long the list of numbers is and you just see how far it is before you make a mistake. Some of them yes, some of them no. So he asks, does it matter how long it takes you?
And the answer is for some of them yes, for some of them no. I won't test you on which is which, but the answer is it depends on the test. Yeah, verbal comprehension is basically how good you are with words. How good you are with language. So it would be a measure of what Gardner calls linguistic intelligence.
Here's something similar to the Raven's progressive matrices, this is from the the WAIS. So there's a general pattern, you have to figure out what it is and figure out which of the five would go in the blank space. So that's part of the perceptual reasoning index.
I have two examples. There's another example of the perceptual reasoning index. This is block design.
So you're actually literally given nine blocks that look like this. Some of them are full red, some of them are half white, half red, and you're given this shape or this pattern on a piece of paper. You have to make that with the blocks. This one is timed.
Most people can get it eventually, but how quickly can you do it? This is from the processing speed scale. So there's two symbols on the right, left, or right. And you have to identify whether at least one of them is in the set on the right. And you just circle a yes or no.
There are 60 of them, and you have two minutes to do it. And the question is, how far do you get? How quickly can you go through it, and how many do you get right? So, actually, I just wanna go back to this. So that should give you a sense.
of what each of these is measuring. So, you add up your score on all of those, that's what gives you your raw score on the W and I. Yeah, so all the ones I showed you would definitely be fluid intelligence, yes.
I could imagine the questions with shapes that are... Like, if I asked you, what's the name for this shape? That would be crystallized intelligence, but the ones I showed you would definitely be fluid intelligence. Okay, any other questions about this? So another distinction that comes after this is the distinction between achievement tests and aptitude tests.
This distinction is somewhat an idealization. The line is kind of fuzzy. But in theory, the difference is achievement tests measure how much you have learned.
So most of the tests you take in your life are achievement tests. Test one that we did last week, that was an achievement test. I gave you guys a bunch of information.
There was some in the book and PowerPoint notes. How much of it did you learn? That's an achievement test.
Aptitude tests measure your ability. Good are you at learning stuff. It's sort of similar to the distinction between crystallized and fluid intelligence.
It's not exactly right, but similar to that distinction. IQ tests, in theory, are supposed to be aptitude tests, though that's a little bit fuzzy, actually. But in theory, that's the difference between these two.
We have about 10 minutes left, and I'll see you before I get into the next section. The next section essentially answers this question. Why should we care about IQ tests?
Are these tests any good? And you'll remember from Chapter 2 that there are two different ways of getting at whether a test is good or not. Reliability and validity.
So I'll start with reliability. Reliability is about how consistent a test is. So there are three subtypes that you should know.
Test, retest, and this one's called internal consistency. Yes, internal consistency. And, uh, inter, sometimes it's called interjudge, sometimes it's called inter-raider, I'll, I'll, the book calls it interjudge, uh, interjudge. So the first one is the simplest one, it's the one that people usually talk about.
You give a group of people the test at one time, you give them the same test later, a week later, a year later, whatever, and you look at how closely correlated people's scores are. If there's a strong correlation between how well people do at time one and how well they do at time two, that would mean the test has good test-retest reliability. Internal consistency means within the test, how consistently do people do.
So IQ tests do quite well on this score, and that's why we can extract the g-factor. Because items within the test are all strongly positively correlated with each other. If the items within the test are kind of unrelated to how well you do on the other tests, or the other subtests, then the test would have low internal consistency.
And inter-judge reliability is how consistent is the test when different people administer it. So an IQ test, a good one, has to be administered one-on-one. The clinical psychologist and the person taking the test.
If you get totally different scores, whether Bob the psychologist is doing the test, or Susan the psychologist is doing the test, like administering it to someone, that would mean as low inter-judge reliability. But you don't want that. You want it to be the case that it doesn't matter who gives the test, the scores are roughly the same.
So how does IQ fare on these measures? So I already mentioned it does very well on internal consistency. That's why we can extract the g-factor. It also does very well on test-retest reliability.
And when I say very well, I mean better than almost anything in psychology. So the book gives some numbers. If you give someone a test at one time and give them the same test in two to twelve weeks, the correlation is around 0.95, which is basically one.
Fundamentally, that's one. So almost exactly the same. If you do it over the course of several years, the correlation goes down, but it's still quite high.
So if you give it to someone at age 9, and then again at age 40, so that's quite a long time, the correlation is between 0.7 and 0.8. So still quite high. And then the book also says if you go from age 11 to 80, so you give someone a test at age 11 and then again at age 80, the correlation is 0.66. So not 0.95, but still quite good. Any questions on reliability?
Okay, validity takes more time and only five minutes left. We'll stop there. Next class we will start with validity.