![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
potato_headThis came up in class today so I've been thinking about it; so let's discuss my experiences with aptitude tests, in particular the ones marketed as IQ tests.
Most moderately savvy people will tell you that IQ tests are essentially worthless. This isn't exactly true; they do measure something, and if used for their original purpose, can actually be very useful. The problem is that in modern culture, especially American culture, IQ scores have come to be conflated with a lot of things that they just don't actually measure; additionally, there's no real regulatory forces out there, so you get organizations like Mensa and paysites that make their own IQ tests. This probably goes without saying, but all IQ tests are not created equal. It's very difficult to make an aptitude test that actually measures what you want to measure, and you have to test the shit out of it to make sure it's working right. And then, because of the nature of aptitude tests, you should continue testing it and making adjustments every 5-10 years!
There are some aptitude tests that do this. In particular, the SAT and ACT tests are adjusted regularly, and there is a known correlation between scores on these tests and GPA in college...for those who get into college. Unfortunately, we only have about half the data, since people with low scores on the tests are much less likely to get accepted into college, and so we don't have a GPA for them. So there are some definite problems here. It would be very expensive to get that missing data, so nobody's done it yet. But as far as use of aptitude tests in the US goes, the SAT and ACT are actually examples of good usage...until people start conflating their results with other things.
The SAT and ACT are NOT IQ tests. They predict your ability to perform in a standard college or university in the US. That's it. And even then they are not the sole predictor; for example, I performed incredibly high on the SAT and ACT, but my college performance is only medium-high, maybe 1.5-2 standard deviations above the mean. This is at least partly because I am a lazy-ass student and don't study for anything. Ever. And because I often skip more classes than is necessary or really advisable. This is a variable that the SAT and ACT don't really measure; work ethic. They can, to some extent, as people who would normally perform lower on the tests can raise their scores by studying etc., but there's a sort of flattening-out at the top; you can get a high score by studying your ass off, or by having a certain level of retention and test-taking ability.
Alright, now that I've talked about aptitude tests in general, let's talk about IQ tests in particular. And the issues with IQ tests in the US, and conflating all aptitude tests with IQ tests. The first ever aptitude test, developed by a French psychologist named Alfred Binet in 1904, is often called 'the first IQ test'. It was not a fucking IQ test. What Binet's test measured was the ability of students to perform in French public schools. This was a problem because there were a lot of students in the schools who spoke French as a second language or were otherwise at a disadvantage. The schools commissioned Binet to design this test specifically so they could locate students that needed extra help because of their disadvantages - not, as you will hear from various sources such as this website, because they were 'inferior' in any way.
Okay, you might ask, what exactly is the difference between measuring IQ and measuring ability to perform in school? Isn't that what IQ is?
In fact, your IQ score is supposed to measure your CAPACITY to learn. Your ability to perform in school depends upon what you ALREADY KNOW. Take, for example, two students that Binet's test would have been administered to: one is an upper-class boy whose family has lived in France for generation upon generation. He speaks French as a first language and is familiar with the culture, as well as having completed the standard number of years of school for his age in France. He is not especially skilled in a particular academic activity - let's take math as an example here. He struggles with mathematics. The second student is a young lower-class girl whose family has recently immigrated from another country in Europe. She speaks French as a second language rather than her first, and perhaps has only just begun to learn it. She has not completed the standard number of school years; perhaps because in the country her family came from, it was not standard for girls to attend school, or perhaps because the country did not offer public schools and her family couldn't pay for her to attend a private school, or because her help was required to earn income for her family. However, she does have natural skill in math.
Under ideal conditions, the boy would score higher than the girl on Binet's test, but lower than her on the math portion of an IQ test.
Unfortunately, these ideal IQ tests have not really been developed. Most of the IQ tests you encounter in the US today are based on a modified version of Binet's test called the Stanford-Binet. There are also completely unqualified tests, like the ones you usually see online, that are generally just pulled out of the test-maker's ass, or at most are designed as a general learnings test - in other words, the score is based on how much you've retained of what you've learned in school plus your ability to apply it, which is not a bad thing to test, but not exactly what they're advertising, either. However, I don't exactly fault them, as measuring retention is really the closest we have to measuring IQ in adults; the Stanford-Binet test only actually works for children still in grade school.
Now, there are very legitimate uses of the Stanford-Binet test, and it is legitimately used in public schools. Mainly to do what Binet originally designed it to - that is, to pick out students that require extra help. However, the tests are not really designed to pick out students that perform ABOVE average in school...and yet they are often used to do just that. In fact, most versions of the Stanford-Binet that you will find in public schools don't accurately measure above-average intelligence at all. It's not uncommon to hear from early readers that the Stanford-Binet pegged them as geniuses or even super-geniuses at a young age, when in fact they were only average to above average. This is because the Stanford-Binet is, again, intended to measure deficits. Most of its depth and accuracy is therefor in the below-average range. In the above-average range, it is comparatively shallow, and it is easy to shoot into very high scores on the test if one is above-average in one or two skills; particularly, reading comprehension, articulation and memory.
Now, many of the companies that design these modified Stanford-Binet tests for school systems are aware of this, and the schools are aware of it, too. The problem is when you get down to the teacher and parent level. It is very, very common for teachers and/or parents to read much more into high scores on these tests than they should. There are other companies lurking in the wings, ready to take advantage of this; places where you can re-test your child over and over (for a fee), conferences where you can bring your genius child to mingle with others of their kind (for a fee), and so on. I, personally, find taking advantage of people's ignorance to take their money reprehensible.
Alright, so that was a lot of words to get to talking about my own personal experience with 'IQ tests'. When I was in elementary school, I was given a LOT of test, all varieties of Stanford-Binet. This was because I kept testing out of them. As in they couldn't get an accurate score, because Stanford-Binet tests rely on getting answers wrong (either in getting a certain number wrong out of the total, or they continue testing you until you get a certain number wrong in a row; the second is closer to Binet's original design). The Stanford-Binet tests are basically designed to compare your 'intellectual age'/grade to your actual age/grade. So when I tested beyond the elementary grades, they gave me the middle school test. When I tested beyond that, they gave me the high school test. When I tested beyond that, they gave me more versions of the high school test, in hopes I would eventually get something wrong and they could categorize me. That didn't happen.
So what happened there, exactly? Well, I was reading at a very high level for my stage. In fact, I was actually reading beyond high-school level; in second grade, I was reading The Hobbit. Not to say an average high-schooler can't handle The Hobbit; keep in mind the Stanford-Binet is only designed to evaluate your ability to handle schoolwork, which in a public high school in America is not going to reach that level of difficulty; the closest you come is Shakespeare, usually with a 'translated' version. As I said before, it is very easy to get very high scores if you excel at reading; in fact, a lot of the Stanford-Binet score is based on reading. For good reason, as reading (in English) is an essential skill in American public schools; and it does have a positive correlation with other academic skills as well (that is, if you read well, it is likely that you perform well in other academic tasks, too). But I am certainly no super genius, I'm sure we all can agree, let alone even higher than that, which is what my scores indicated if one were to take a chance at interpreting 'off the charts'; my point is, simply, if the score is above-average, it is unlikely to really have any meaning.
A few final thoughts: I mentioned before that most IQ tests today are based on the Stanford-Binet, and others measure retention, and regardless, neither measure intelligence as we would like to measure it. So how can we measure intelligence? Well, frankly, to measure somebody's capacity to learn, you have to observe them learning and measure it. Now, there are laboratory tests that could be constructed for this, but it would be just as useful on an individual level for teachers and parents to simply keep an eye on any student that learns quickly, either in a particular subject or in general. You don't need a test to quantify that student's ability, so you can place them in accelerated courses based on their score; show them how to seek knowledge for themselves and they'll likely do it at their own pace, and keep letting them take on accelerated coursework until they say they're not bored with school. Because of the many different ways in which a person can be intelligent, and the many different ways people learn effectively, teaching and facilitating learning is really still more an art than a science, even if companies looking to take your money to quantify you would like you to believe otherwise.
Most moderately savvy people will tell you that IQ tests are essentially worthless. This isn't exactly true; they do measure something, and if used for their original purpose, can actually be very useful. The problem is that in modern culture, especially American culture, IQ scores have come to be conflated with a lot of things that they just don't actually measure; additionally, there's no real regulatory forces out there, so you get organizations like Mensa and paysites that make their own IQ tests. This probably goes without saying, but all IQ tests are not created equal. It's very difficult to make an aptitude test that actually measures what you want to measure, and you have to test the shit out of it to make sure it's working right. And then, because of the nature of aptitude tests, you should continue testing it and making adjustments every 5-10 years!
There are some aptitude tests that do this. In particular, the SAT and ACT tests are adjusted regularly, and there is a known correlation between scores on these tests and GPA in college...for those who get into college. Unfortunately, we only have about half the data, since people with low scores on the tests are much less likely to get accepted into college, and so we don't have a GPA for them. So there are some definite problems here. It would be very expensive to get that missing data, so nobody's done it yet. But as far as use of aptitude tests in the US goes, the SAT and ACT are actually examples of good usage...until people start conflating their results with other things.
The SAT and ACT are NOT IQ tests. They predict your ability to perform in a standard college or university in the US. That's it. And even then they are not the sole predictor; for example, I performed incredibly high on the SAT and ACT, but my college performance is only medium-high, maybe 1.5-2 standard deviations above the mean. This is at least partly because I am a lazy-ass student and don't study for anything. Ever. And because I often skip more classes than is necessary or really advisable. This is a variable that the SAT and ACT don't really measure; work ethic. They can, to some extent, as people who would normally perform lower on the tests can raise their scores by studying etc., but there's a sort of flattening-out at the top; you can get a high score by studying your ass off, or by having a certain level of retention and test-taking ability.
Alright, now that I've talked about aptitude tests in general, let's talk about IQ tests in particular. And the issues with IQ tests in the US, and conflating all aptitude tests with IQ tests. The first ever aptitude test, developed by a French psychologist named Alfred Binet in 1904, is often called 'the first IQ test'. It was not a fucking IQ test. What Binet's test measured was the ability of students to perform in French public schools. This was a problem because there were a lot of students in the schools who spoke French as a second language or were otherwise at a disadvantage. The schools commissioned Binet to design this test specifically so they could locate students that needed extra help because of their disadvantages - not, as you will hear from various sources such as this website, because they were 'inferior' in any way.
Okay, you might ask, what exactly is the difference between measuring IQ and measuring ability to perform in school? Isn't that what IQ is?
In fact, your IQ score is supposed to measure your CAPACITY to learn. Your ability to perform in school depends upon what you ALREADY KNOW. Take, for example, two students that Binet's test would have been administered to: one is an upper-class boy whose family has lived in France for generation upon generation. He speaks French as a first language and is familiar with the culture, as well as having completed the standard number of years of school for his age in France. He is not especially skilled in a particular academic activity - let's take math as an example here. He struggles with mathematics. The second student is a young lower-class girl whose family has recently immigrated from another country in Europe. She speaks French as a second language rather than her first, and perhaps has only just begun to learn it. She has not completed the standard number of school years; perhaps because in the country her family came from, it was not standard for girls to attend school, or perhaps because the country did not offer public schools and her family couldn't pay for her to attend a private school, or because her help was required to earn income for her family. However, she does have natural skill in math.
Under ideal conditions, the boy would score higher than the girl on Binet's test, but lower than her on the math portion of an IQ test.
Unfortunately, these ideal IQ tests have not really been developed. Most of the IQ tests you encounter in the US today are based on a modified version of Binet's test called the Stanford-Binet. There are also completely unqualified tests, like the ones you usually see online, that are generally just pulled out of the test-maker's ass, or at most are designed as a general learnings test - in other words, the score is based on how much you've retained of what you've learned in school plus your ability to apply it, which is not a bad thing to test, but not exactly what they're advertising, either. However, I don't exactly fault them, as measuring retention is really the closest we have to measuring IQ in adults; the Stanford-Binet test only actually works for children still in grade school.
Now, there are very legitimate uses of the Stanford-Binet test, and it is legitimately used in public schools. Mainly to do what Binet originally designed it to - that is, to pick out students that require extra help. However, the tests are not really designed to pick out students that perform ABOVE average in school...and yet they are often used to do just that. In fact, most versions of the Stanford-Binet that you will find in public schools don't accurately measure above-average intelligence at all. It's not uncommon to hear from early readers that the Stanford-Binet pegged them as geniuses or even super-geniuses at a young age, when in fact they were only average to above average. This is because the Stanford-Binet is, again, intended to measure deficits. Most of its depth and accuracy is therefor in the below-average range. In the above-average range, it is comparatively shallow, and it is easy to shoot into very high scores on the test if one is above-average in one or two skills; particularly, reading comprehension, articulation and memory.
Now, many of the companies that design these modified Stanford-Binet tests for school systems are aware of this, and the schools are aware of it, too. The problem is when you get down to the teacher and parent level. It is very, very common for teachers and/or parents to read much more into high scores on these tests than they should. There are other companies lurking in the wings, ready to take advantage of this; places where you can re-test your child over and over (for a fee), conferences where you can bring your genius child to mingle with others of their kind (for a fee), and so on. I, personally, find taking advantage of people's ignorance to take their money reprehensible.
Alright, so that was a lot of words to get to talking about my own personal experience with 'IQ tests'. When I was in elementary school, I was given a LOT of test, all varieties of Stanford-Binet. This was because I kept testing out of them. As in they couldn't get an accurate score, because Stanford-Binet tests rely on getting answers wrong (either in getting a certain number wrong out of the total, or they continue testing you until you get a certain number wrong in a row; the second is closer to Binet's original design). The Stanford-Binet tests are basically designed to compare your 'intellectual age'/grade to your actual age/grade. So when I tested beyond the elementary grades, they gave me the middle school test. When I tested beyond that, they gave me the high school test. When I tested beyond that, they gave me more versions of the high school test, in hopes I would eventually get something wrong and they could categorize me. That didn't happen.
So what happened there, exactly? Well, I was reading at a very high level for my stage. In fact, I was actually reading beyond high-school level; in second grade, I was reading The Hobbit. Not to say an average high-schooler can't handle The Hobbit; keep in mind the Stanford-Binet is only designed to evaluate your ability to handle schoolwork, which in a public high school in America is not going to reach that level of difficulty; the closest you come is Shakespeare, usually with a 'translated' version. As I said before, it is very easy to get very high scores if you excel at reading; in fact, a lot of the Stanford-Binet score is based on reading. For good reason, as reading (in English) is an essential skill in American public schools; and it does have a positive correlation with other academic skills as well (that is, if you read well, it is likely that you perform well in other academic tasks, too). But I am certainly no super genius, I'm sure we all can agree, let alone even higher than that, which is what my scores indicated if one were to take a chance at interpreting 'off the charts'; my point is, simply, if the score is above-average, it is unlikely to really have any meaning.
A few final thoughts: I mentioned before that most IQ tests today are based on the Stanford-Binet, and others measure retention, and regardless, neither measure intelligence as we would like to measure it. So how can we measure intelligence? Well, frankly, to measure somebody's capacity to learn, you have to observe them learning and measure it. Now, there are laboratory tests that could be constructed for this, but it would be just as useful on an individual level for teachers and parents to simply keep an eye on any student that learns quickly, either in a particular subject or in general. You don't need a test to quantify that student's ability, so you can place them in accelerated courses based on their score; show them how to seek knowledge for themselves and they'll likely do it at their own pace, and keep letting them take on accelerated coursework until they say they're not bored with school. Because of the many different ways in which a person can be intelligent, and the many different ways people learn effectively, teaching and facilitating learning is really still more an art than a science, even if companies looking to take your money to quantify you would like you to believe otherwise.
