Question by Terra
Many people have different opinions on what numbers are the range for the bell curve of IQ. Please tell me the correct answer.
Answer by Cliff-BRD
This depends on the kind of population you are looking at. Usually, IQ scores, randomly selected from the general population, follow the normal curve, AKA the bell curve. That is to say, tens of thousands of people randomly selected from the population take the test, and the pattern follows the bell curve.
But note that this is not an ideal way of considering IQ, because IQ scores are normed not only to the general population, but also to age. Children do better as they mature, and are given qualitatively different tasks; the nature of the construct of intelligence is literally thought to change. General IQ scores peak in young adulthood and slowly decline, and decline rapidly as one approaches the last five years of one’s life (often as a result of major and minor brain damage from events such as stroke). There are also qualitative differences in the tests given to children and those given to adults.
Taking these factors into consideration, IQ will, as mentioned, generally follow a bell curve (there is some evidence, however, that IQ test scores do not perfectly fit a bell curve…instead, it would seem that IQ scores are positively skewed). Using a bell curve as a rough model, an IQ of 100 (the mean IQ) is a higher IQ than half the population, 115 (one standard deviation above the mean) is higher than roughly 84% of the population. Plus-two standard deviations (IQ 130) is in the highest 2.3%, +3 SD (IQ 145) is in the highest 0.14%, and so on. The current population of the world is approaching about 7,000,000,000 people. You would need an IQ of about 6.3 standard deviations above the mean to reach the highest score out of 7 billion people, or IQ 195. Similarly, the lowest IQ would theoretically be around 5.
Even using a curve to talk about the whole range of intelligence is kind of silly, though, isn’t it? There are lots of people in comatose states, and the like. How would you ever give them an IQ test? Is it fair to say that they have “no intelligence” (IQ of zero)? How about someone who simply refuses to take the test? Furthermore, the key to an IQ test being valid is having a large population take the test to establish it’s norms; there would be no distinction between IQ 195 and IQ 160 on a test given to 30,000 people. You would need to IQ test the whole world with a test thousands of items long–no one would want to sit through that (well, a few people, perhaps!). :^)
There may be ways to develop a test to actually determine IQs above the limits of normed tests (using Item Response Theory), but you still would need a representative sample of people with IQs above a certain point. Good luck finding them, let alone getting them to all participate! :^) Then comes the nature of what kinds of high-difficulty items would be written. In other words, just as the nature of intelligence changed qualitatively in children, the nature of IQ would also presumably change at such high levels. This would mean, though, that it’s very unlikely that there would be a consensus about whether these high-difficulty questions represent the same kind of intelligence as those represented by traditional IQ tests. Additional validity evidence would be needed!
For practical purposes, it’s fair to say that most of the validity evidence for IQ (i.e., studies using IQ test scores) looks at a range of 40 at the lowest to 160 on the high end. Outside this range, it’s both hard to test intelligence, and it’s hard to make any firm conclusions about it. Hope this helps! :^)
Read Full Post »
