But the instrument for measuring doesn’t change. That’s what makes it dependable, predictable and something to measure against. You can’t measure against a moving target.
I don't see why this should prevent it from being considered "scientific". These things limit how precise it can be but that doesn't really matter much. A few points don't matter.
Do the test when you have whatever is a normal amount of caffeine for you and after a normal nights sleep.
Ouch... The psychology is not science, though. It is kind of shamanism that is trying to be science. Raven progressive matrices are measuring your statistical biases in the first hand. The better you solve the patterns in this test, the more often you'll jump to conclusions on a base of insufficient data. And that is what they call an IQ. I have passed that test with a very high score, and I was surprised this test was as easy, and my results were as high. Then I had to think about this test for a while. The problem that I fight in myself (jumping to conclusions with little statistical evidence and insufficient data) for decades is regarded as high IQ. Bullshit.Disclaimer: I probably jumping to conclusion right now based on my own test, which is not a statistical evidence.
I cannot understand you position. You seem to assume that exactly one of these is true:
1) IQ is a measure of "true" intelligence
or
2) IQ completely fails as a measure and is meaningless
and since 1) is too nebulous to be true then 2) must be true.
I am not an expert on IQ, but if it was highly correlated with being able to recognize and guess patterns from incomplete data it would not be a surprise to me.
Even just the ability to formulate possible solutions to ambiguous problems could credibly have high correlation with IQ.
I am not saying that these are true, I am just wondering why they sound so impossible to you.
My point is: IQ tests fail to provide a measure of intelligence. The linked video at the end discusses some reasons why, such as environmental, educational, and cultural differences. The primary problem with these tests is that they are created by groups of people who have their own specific view on what intelligence is.
Yes, these tests assess certain traits of an individual, but intelligence is likely a more complex topic. Consider the language part of some tests, for example. These tests include vocabulary that the test subjects must know in order not to be discriminated against due to cultural and environmental factors. Instead, vocabulary capacity is measured. Language tests should include logical tasks of varying complexity, which must be solved without providing a variety of answers. If answers are provided, this will test for pattern recognition traits.
This is only evidence that we should not (tw: strawman argument) create a classist society where your lifetime status is algorithmically decided by a single test taken once at 18.
If you assume that IQ measures $SOMETHING this just mean that both sleep quality and caffeine influence $SOMETHING