Hey Psychology Lovers. IQ Question

Post Reply New Topic

http://www.apa.org/news/press/releases/ ... ffect.aspx

This article reports that the re-norming of IQ scores causes problems with special education placements and other determinations based on IQ testing. I know that test creators aim for culturally unbiased IQ test ? so the Flynn effect should be obvious evidence that they are failing. However, it seems that the Flynn effect is interpreted as people just getting smarter, rather than as a sign that IQ tests become obsolete as the culture changes with time.

If we look at this as a result of outdated IQ, an obvious solution would be to create new tests. How much work would be required to create a test that is sensitive enough to contemporary culture to preserve the old score norms? Would that be practical? And how practical would it be to repeat the process however often is necessary to offset the Flynn effect?

We update the IQ tests regularly--there are new editions, just like textbooks change when new information is discovered.

We try to keep tasks on the IQ test relatively culture-independent; for example, putting blocks together to form an image on a card requires familiarity with manipulating blocks, something that almost everyone has done, whether with wooden blocks or bricks in a vacant lot or Tetris on a computer screen.

When an IQ test is updated, it's administered to many people as a test. The average score is arbitrarily set at 100, with a standard deviation at 85 and 115, and so on. The Flynn effect is real and measurable, but it doesn't invalidate IQ tests (at least as far as IQ tests measure intelligence at all) because we keep on adjusting for it. An average person today scores 100, just like an average person in 1950 would have scored 100.

For practical purposes, we're not really worrying about how people scored decades ago, because the test still serves its purpose despite the Flynn effect.

That makes the IQ test a useful tool to figure out who might need special education, because it doesn't really measure how smart you are--it measures whether you're significantly above or below average. Since the average classroom is geared to the average person, someone who's far above or below that--say, below 70 and above 130--might not be able to benefit from the average classroom, and therefore might need a slower or faster-paced education.

But IQ is not the only criterion for diagnosing intellectual disability. A person can score quite low on an IQ test and still not be diagnosed with intellectual disability, because ID also requires developmental problems--adaptive skills delays, problems in one's everyday life, with taking care of oneself and solving everyday problems. If a person is just bad at taking IQ tests, there's no need for a diagnosis of ID--being uneducated can produce a low IQ score in someone who is coping just fine, and solving everyday problems very handily. Poor environment can lower IQ scores in children who haven't had the chance to do the experimenting that lets them solve IQ-test style problems; they're spending their brainpower on survival. So there's more to ID than an intelligence test--but it does offer a useful clue, because the tasks on an IQ test use a lot of the same cognitive skills a person will use in school; and a low score, combined with adaptive-skills delays, can mean a need for special education.

The core issues (problems) are to establish reliability and validity measures that are truly robust enough. A related issue is that the general public is never informed on the problems to do with r and v measures on the tests that have been used (and any new ones will probably suffer from the same editing out of this vital information).

You should check out Flynn's book Are We Getting Smarter?, a really detailed discussion of the Flynn effect and what it means.

I think there are a couple different aspects to it. Firstly, it's possible we really are getting smarter, due to things like removing lead from our environment, higher rate of breastfeeding, so many parents trying to cognitively stimulate their kids, etc. Each of those things has only a tiny effect on IQ, but all put together, they could add up to drag the average IQ up a bit.

Secondly, despite efforts to make culture-fair IQ tests, there is still a significant cultural bias to IQ. The easy part is removing question content that depends on cultural knowledge, such as modifying verbal reasoning items so they don't refer to objects that aren't present in that person's culture. But there's also evidence that understanding how to take a test is a learnt behavior in itself, which is affected by prior experience with being tested. (Incidentally, the often-touted increase in IQ scores in autistic kids receiving ABA seems to be mostly because ABA trains kids in 'testability', rather than actually increasing their intelligence.) A person who has never gotten any formal testing before will do poorly on an IQ test, not because they're not intelligent, but because they don't understand that the person is asking questions to assess their knowledge (as opposed to genuinely seeking information) or are not used to answering questions designed to assess their knowledge. This can result in them giving 'practical' answers, rather than the expected logical reasoning answers. For example, when asked which two items 'go together' among screwdrivers, hammers and nails, the expected answer might be 'screwdrivers and hammers' because they're both tools, but someone unfamiliar with testing might reply 'hammers and nails' because you use a hammer to pound in a nail. Since each generation tends to be more and more highly educated, and educated people 'test' better, average IQ scores go up.