Searching "comments":
| # | Comment | Survey |
|---|---|---|
| 11 | The danger of saying it isn't important lies in what you are talking about. Learning to master these concepts (and no one is born with concept mastery of anything, so it is all learned, yes), is important to IQ scores. IQ scores are important to life success.. Mastering these concepts is important to life success. Just admitting that will do good things for society as we can develop ways to guide people (especially young children) along the path of concept mastery, and increase their IQ. The "Head-Start" program takes for granted that IQ is measuring something meaningful and sets out with the goal of increasing it's students IQ's by preventing intellectual neglect from hurting them. I don't know if you've seen the studies on smart-bred vs. dumb-bred rats in enriched vs. unenriched environments, but this is kind of the idea. For rats there are mazes to assess abillities, and we take for granted that the mazes are measuring some meaningful generalizable characteristic about the rats. In people we just use these tests that are like mental mazes, and we take for granted that we are measuring some meaningful generalizable characteristic of the people. IQ right now is the best tool we have for this job. Sure, there are other things like "musical abillities" and social skills which are more difficult to quantify and which it is also important to nurture early on if the child is going to be good at them, but that's just plain out-side of the scope of the IQ test... Saying the IQ test doesn't measure anything meaningful about intelligence would be like saying an aural-skills test doesn't measure anything meaningful about musical abillity. And saying that the particular slice of "intelligence" that is measured by the common IQ tests doesn't predict anything meaningful about the life outcome of an individual is an idea that needs to do something impressive in order to stand up to the considerable research correlating the two variables and the reasoning I presented earlier. | Do you think IQ tests are a true measure of intelligence? |
| 12 | Zang, you and MSG are suffering from the same problem. You want to attribute the relationship to some tangential factor instead of going with the obvious one. If the correlation doesn't back up my obvious causal schematic, it sure as hell doesn't back up your extraneous one. :). It's fine to do what you are doing, but you are going to be wrong more often than not. What we need is we need a correlation between two variables, and then generally there is a fairly obvious relationship in terms of which is affecting which or what a third variable might be. There are always going to be some other factors slipping in there, but generally this isn't going to have a significant effect on the correlation coefficient. The way you do this, incidentally, is that you do a multiple regression with any other factors you want to blame. It is a simple statistical test and they are done all the time because people like you often try to argue that two variables which are obviously in a causal relationship are not in that causal relationship. Doing the multiple regression "weeds out" or controls for other factors and then we can find out if you are right or not. What has to be done is that there has to be a study, and there have been studies addressing specifically what you are talking about. "This goes back to my initial statement regarding testing in general. People who do well on tests tend to do well in other areas related to approval by authorities. It is an indicator of a certain personality type." You've presented here a testable hypothesis as though it were true. Your testable hypothesis is either so obviously wrong that nobody has ever bothered to test it, or such an unbelievable show of genius that nobody has ever thought of it before, because there is no research on the specific subject of whether SAT scores correlate significantly with any of the tests of obedience to authority that are out. I can tell you from seeing similar studies what it's going to tell you, but I'm sure you won't be convinced until you do the study yourself. They aren't going to show a correlation of practical significance. You will eventually find a correlation of statistical significance if you include enough subjects, but the "r^2" or how much of the variance in SAT scores is attributable to this particular variable is going to be rather low, possible along the lines of .01 or .02. My null hypothesis is that there is no way that the one variable you have supplied is having a significant effect on the scores. If we bundled that together with every other personality factor imaginable that might influence scores on the test, we'd still see a fairly low correlation because the tests are designed to measure intellect instead of personality factors. They did research as the tests were being designed to try to eliminate any effects of the type of personality factors you are describing, and the research was well-controlled and the result is a test that does dodge these concerns as well as any test could be expected to. If desire to be liked by authorities is the predominate factor in the score people are getting on these tests, it should be easy enough to prove. If you are right, your article will be published in the Journal of the American Psychological Association, without any doubt. Feel free to do the research Zang, but, don't be surprised when it doesn't pan out. | Do you think IQ tests are a true measure of intelligence? |
| 13 | No, no, I don't think we are having a major disagreement. I think you're one of the few people on SC who is reasonable and combined with my being reasonable it would be kind of tough for two reasonable people to have a profound disagreement on something because reason dictates certain things and will lead any two people to the same conclusion. Your statement about correlating SAT and High School attendance struck me as though you were arguing that the predominate factor I'd described (intellect) wasn't the predominate factor and that instead you were suggesting the majority of the difference was a result of the alternative factor you supplied. I guess I misunderstood you. There are standard methodologies for the type of controls on the tests I am talking about. The agency that runs SAT doesn't release the research related to their tests (for various and oftentimes obvious reasons) so I don't know specifically what they have done, but they employ countless Ph.D.'s in psychological testing and related subjects, so it's pretty clear that they are doing the research. It wouldn't have been any trouble for them to design good tests surrounding what I'm describing, if that's the issue. The standard ways to do the tests on the tests would be multiple regression charts with people whom you have loads of data on (and if you need a sample group you just put the survey questions on the front page of the SAT TEST that year and people will answer them... (they do do this)), and then you try to figure out what factors are playing into the scores kids are getting... whether they studied, where they went to school, etc. There's a lot of information you give them when you register, and then they can get other data from your university once you start due to a form you never-knew-you-signed. The data is available to them. I'm sure they do experiments as well, I can picture hundreds of variations on how to do such experiments myself. Just trust me, there are ways. It's amazing how hard it can be to pack years of psychological learning into a post like this Zang, I'm sorry. There are just methods to do the research with, many of them are extremely clever but they were only invented once and then other people re-use them. | Do you think IQ tests are a true measure of intelligence? |
| 14 | http://www.mugu.com/cgi-bin/Upstream/Issues/psychology/IQ/apa.html Here, read this, it'll get you started. | Do you think IQ tests are a true measure of intelligence? |
| 15 | "As you point out yourself, determining causality could only be done with a tightly controlled study that is effectively impossible in human society. So all the rest of your comments are simply speculation, however plausible they may sound." You are overly skeptical, more skeptical than would be taken seriously in any of the major journals. You can never know anything... you can always bring up things like this but all you are doing is obscuring what's being discussed instead of saying something meaningful about what the evidence is suggesting. You aren't telling me anything I didn't already know about the process, MSG, you are stonewalling. The evidence is suggesting certain things. You can be "certain" of very little, but in as much as a 95% confidence interval is good enough to correlate two variables, so too is a good research base and a sound logical foundation enough to determine causation. Intelligence, beyond any personality factors, is the primary predictor of life success. If you have to choose one aspect of a person which seperates the "successfull" from the unsuccesfull, this is it. If you understood more of the science behind the conclusions you'd have more confidence in them on a personal level. I'm done arguing this particular subject with you because I don't feel like either of us are going to gain anything meaningful from it until you get a background in the research. The link I posted for Zang is good for you too probably. | Do you think IQ tests are a true measure of intelligence? |
| 16 | A correlation coefficient is always reversable. If you reverse the dependent and independent variables the relationship will remain unchanged. Any other questions that would have been addressed in a psych 101 class or high school statistics, had you bothered to take either before arguing with me about a statistical/psychological issue? | Do you think IQ tests are a true measure of intelligence? |
| 17 | Could is a dangerous word Kristal. It could be that Zang is the only real human being in the world and that the rest of us are computers. I'd thought that's what he meant originally, too, Kristal, but he clarified. It did kind of seem like that was what he was saying. | Do you think IQ tests are a true measure of intelligence? |
| 18 | (Addendum)-- For the population sampled. That doesn't apply in this case because IQ tests sample the general population as opposed to some sub-set, but I wouldn't want you to get confused in other cases where specific populations have been sampled. For instance, you couldn't sample a bunch of Alzeihmers patients and discover that 90% of them have gray hair, and then turn around and say, "90% of people with gray hair have Alzeihmers." But if you find a correlation between TWO NUMERICAL VARIABLES such as %gray hair and %Alzeihmers symptoms, then in that group the correlation does go both ways. You are confusing inter-group comparisons with comparisons of two numerical variables. In the general population, if two numerical variables correlate (such as "theoretical IQ" and "Tested IQ") the correlation is the same whether it's coming or going. Before MSG points it out, the causal relationship only goes one way (the high theoretical IQ was causing the high tested IQ), and that can't be reversed, but if there is a correlation of say .1 between gray hair % and number of alzeihmers symptoms in a sample of the general population there is also going to be a correlation of equal value between numbr of alzeihmers symptoms and gray hair % in those samples. | Do you think IQ tests are a true measure of intelligence? |
| 19 | Your skepticism is a result of not experiencing the science first hand. Scientists are always very skeptical MSG, but some conclusions are more certain than others. It's like Christian Apologists who try to argue for the reasonableness of the religion on the basis that "science isn't sure of anything." Some things are simply concluded, unchangable, realities of the sciences-- even the social ones. There are basic ideas that have been researched to death and exist in a state of very high certainty. There is no doubt where the truth lies in these situations. Hearing "science has determined this" is something I'd be skeptical of. But when you've read the articles and actively questioned the logic, as I do, you eventually realize that there comes a point where you can't question the conclusions anymore and expect to be taken seriously. The scientific method is extremely exact, and the major journals are very good about ensuring that it is followed through their open forum policies where all realistic view-points are discussed. And if you think you can, you do a study or publish an article. I wish science had more flaws-- the reality is that finding them is tough. | Do you think IQ tests are a true measure of intelligence? |
| 20 | You haven't experienced THIS science, MSG. What is your background in psychology, specifically? Unless there's some bomb that you've been waiting to drop, it's clear to me from the way you are forming your arguments that you don't have any knowledge that is directly relevant to this discussion. Today, I raised these questions with two people. The first has a Ph.D. in Psychology, emphasis statistics, and the second has a Ph.D. in Behavioral Genetics. I said, "Given the fact that you can only establish cause with 100% certainty with a controlled study, do you think it is reasonable to establish cause in a situation based on a sound argument and a good research base of correlations?" The statistician nodded in generalized agreement. He said, "You'll be wrong sometimes, but, often that's the best you have to go on." I asked him specifically if there were people who drag their feet intentionally on these kinds of things and are anal about the lack of specific controlled studies. He thought for a minute and said essentially what I'm saying, which was, that there comes a point where you just have to cut through the crap and say something is established. He cited a few examples. 1) I can't remember the name he said but there was a statistician who refined the multiple regression test and therefore had the "F" statistic that you see named after him. He argued his whole life that correlation doesn't suggest a causal relationship. One of his big arguments was against tobacco and cancer correlations. He said they were merely correlations. The irony of course is that he was a smoker and eventually died of lung cancer. 2) The first time Polio vaccine was administered nationally, 1/2 of the school-children received a placebo because someone "wanted more data". He said this was absurd and he never likes to hear people say, "We need more data." when there is clearly enough for a reasonable person to be convinced of something. He said that there comes a point where consensus are formed on things and nobody is making an argument against them that is being taken seriously. I asked him specifically about IQ tests in general, and their validity. He said, "Their validity for what?" (Paraphrasing). And I said, "Do they measure something meaningful about an individual?" He said of course they measure something meaningful, they measure intelligence. He looked at me like it was an extremely stupid question and I decided to end the conversation there. The behavioral genetics Ph.D. said, and I quote, "IQ is the most solidly based research area in all of behavioral genetics, due to the # of studies that have been done on it." (From here I am paraphrasing.) "There is no reasonable case to be made that the figure does not reflect something meaningful about an individual. If it didn't reflect something meaningful, people wouldn't be studying it. (He tangentially added.) There is no reasonable case to be made that variation in IQ in the US population isn't primarily accountable to genetic factors." I asked if he would agree that it was the single best variable for predicting life success. He pulled a Zang and said, "Define life success." I said "Income and level of educational attainment." He thought about it for a few minutes and said, "I can't imagine anything that would correlate more with educational level than IQ..." And he kind of suggested it was obvious. He went on about the personality factors, I asked him if they were secondary, he said, "I don't like to think of anything as "secondary", but, certainly they aren't playing the role that intelligence does. I mean, look at the situation, we are talking about level of educational attainment..." And went on to explain how intuitive it was that intellect was going to be tied to that. This was before class, so during class, he took about 15 minutes to say a lot more about intelligence. These are appeals to authority but I encourage you to read into them as far as you wish. | Do you think IQ tests are a true measure of intelligence? |
