While browsing through this thread the topic of IQ tests come up and their apparent validity (or not). Of course that thread has long since degraded into mindless drivel, but still, I am interested in the topic in general. In general I've noticed that people tend not believe in the results of an IQ test, or other tests that attempt to measure intelligence in general. I myself am curious to know just how valid they. While the opponents of IQ tests tend to blurr the definitions of "intelligence", there is in reality no debate among psychologists about their ability to measure certain aspects of intelligence, in particular general intelligence g. I've even done basic statistical analysis (in particular the use of confidence intervals )of the results and most of the time they check out. That is, they are not completely bogus from the standpoint of statistics. The content in the tests are reasonable, because most of the time they don't test for knowledge specific to a given culture. And any results that come out of it are just that, results; they don't really say why or how it is possible for a person to be smarter than another. Although we have considerable data on what effects a person's IQ... So, what exactly do IQ tests miss, that would make their results suspect? And please, try to be civil here, I am genuinely interesting in the reasoning why many distrust them.... Because, I'm not entirely convinced that intelligence is "not defined" or "can't be defined". I think after over 100 years of research we have a very good idea of what constitutes intelligence, if not the actual "core/substance" of it, so to speak (or whether or not there are different types of intelligence...)
There are two assumptions in IQ tests. 1. There is no definition of intelligence 2. Correlation is not causation. I would say that IQ tests are indicative of how well people do in IQ tests. They do not measure, for example, creativity, success or aptitude. They are good measures for how well you will do in school, since schools are pretty regimented in what they allow you to think. However, they are correlations, ie they measure incidence of the ability to take the test and relate it to other variables and should not be confused with causation.
1) Why not? 2) True, but IQ tests don't do that. They strictly measure, or try to measure, the intelligence of a person. The bell curve on which it is placed is just a consequence of the data gathered. Granted, it was imperfect back then, but by now they have improved greatly... They don't measure it, but IQ's are very strongly correlated with higher levels of creativity, success, and aptitude. But it doesn't say anything about it's cause.... How so? Standarized testing, which is what all schools pretty much train you to do, can only test for how well you've memorized things or follow directions. IQ tests, though, are not necessarily restricted to content in school, but rather test your ability to do certain things (e.g. rotate an object in your head, recall any number of things, etc.) --------------------------------------------------- I should say this that the correlation does not equal causation is quite abused or misused quite often, because it normally applies (and indeed was developed for the purpose of) interpreting the results of a hypothesis test, e.g. whether or not one decides to reject the null hypothesis. Since it is entirely possible that one can find patterns and correlations that have nothing to do with each other, that's why we conduct hypothesis tests and, more importantly, construct confidence intervals at a certain level of significance. So that we can be sure that the correlation is indeed real, and not just a fluke. They all help to explain how likely it is to find the true mean.... So, I'm still not seeing what the real problems are with IQ tests, other than objections to how intelligence is defined. In other words, a matter of semantics. What exactly is wrong with current definitions of intelligence that invalidate IQ tests? Why isn't intelligence defined?
Emotional stress of a person taking the IQ tests at the time...for example will change the IQ test results.
Can you show me where IQ levels are indicative of creativity or success? Feynmans was 120 something while this bouncer is reportedly 200 or more http://www.sciforums.com/showthread.php?t=62839
and all IQ tests have to be the same, otherwise we cannot trust the results or correlate between them.
For the success part: I was actually mistaken, I was just citing something from hearsay (not rigorous I know). Looking online only academic performance is correlated with higher IQ. At least from what I've found... For creativity, you can read this: http://inst.sfcc.edu/~mwehr/StudyGHB/11Ovr_Int.htm And http://www.hno.harvard.edu/gazette/2003/10.23/01-creativity.html Both of which cite studies that show correlation between higher IQ (or intelligence) and higher levels of creativity. Again, I'm going to have to find more detailed studies later (I'll read a book on this topic later...) As IQ's are formed from statistical data, it is not surprising that there will be a few outliers. Of course, I think it's been noted that IQ doesn't necessarily test for how much you know (or how much you think you know). All this is moot anyway, because what I'm interested in is not whether or not they correlate with success (which is actually a strawman on the part of the critics), I want to know if it can adequately measure intelligence. Contrary to popular belief, IQ tests do not test for, nor are intended to, knowledge or cultural adherence or academic performance. While how we define intelligence is certainly debatable, there is no question that lower IQ does certainly mean lower cognitive abilities. That is, those with lower IQ's learn much slower in general, have a harder time developing self-help skills, and if it is sufficiently low it can certainly put a limitation on your ability to understand things in general (i.e. can't recognize patterns/relationships, can't learn basic motor skills, can't think as clearly in general, can't reason as well, etc.). There is no denying it, this is very well documented. (Also, people who are clever in one area tend to be clever in many others in general too). So clearly IQ tests do measure something, but if it is not intelligence, then what exactly is it? How valid are they for quantifying how smart people are, or what predicting what their capabilities will be in terms of mental ability? If they are not valid, then what exactly are they doing wrong? Specifically, what should they measure instead, and why isn't the criteria they test you on adequate? It's pretty clear cut that lower IQ = lower cognition in general. Because, as it stands for me, there doesn't seem to be anything wrong with using them for that purpose, they do indeed measure intelligence of some form. If intelligence isn't defined, then what does it mean when a person with an IQ of 55 can't live on their own, or learn how to, and have to be supervised in daily life skills, for the rest of their lives...
Marilyn vos Savant is now on Youtube. She makes quite a few good remarks on the concept of intelligence in general.
