IQ tests

how accurate are IQ tests anyway?

They are roughly as accurate as this piece of string.
IMO IQ test are a good guide for general intelligence but are not all that reliable when you consider that some people paid attention at school in mathematics, while others didn’t.

The people who paid attention will do well on the mathematics questions, while the people who didn’t, won’t. Even if they are more intelligent.

I am hopeless at maths, yet [big head] I believe I am of above average intelligence[/big head]. Less intelligent people who know their mathematics might get a higher score than me on an IQ test.

Or, ‘Not Very’.

A real IQ test theoretically measures aptitude, not achievement. How much attention you paid in math class should not enter into it. There may be analogies and logic questions, for example, but should not be any algebra or other subject-specific knowledge. As to how accurate they are, that’s another issue. Absent mitigating factors, they should give a general idea of one’s relative intelligence, but a difference of a few points between one person and another are meaningless. Among the mitigating factors are such things as one’s general state of well-being, unusually high or low test anxiety, and one’s prior experience in taking tests. Absent one or more of such conditions, if one person scores a 90 and another scores a 130 on the same test, it is a reasonable assumption that the second person is more intelligent.

MLS I took a few ‘pretend’ tests online. There are some questions that indirectly rely on knowledge of the times-tables and other mathematics.

It was a long time ago, I cannot remember what the questions were, but I remember at the time thinking a mathematically knowledgable person would find certain questions easy.

that really depends on how you define intelligence.

standard IQ tests, for instance, measure general information, comprehension, arithmetic, similarities, picture completion, object assembly etc. that is, they have an analytic orientation.

there are other definitions, such as * fluid intelligence, which represents reasoning and memory, and crystallized intelligence which judges the information, skills and strategies people learn through experience and apply in problem-solving situations.

then whatsisname (someone fill me in here) came up with a seven component definition…interpersonal, intrapersonal, logic-mathematical, linguistic, musical, body kinesthetic and…hmm…oh yes…spatial intelligence. if you use this definition of intelligence, then your standard IQ test measures only 3 of the components.

and then there are others like the Triarchic theory of intelligence and whatnot.

further, the fact that you can improve your performance on standard IQ tests with coaching and practice, does call into question their reliability.

on the whole, though, from a practical point of view, and as a general and convenient measure of intelligence, researchers have concluded that a your standard IQ tests are acceptable. that is why most educational institutions use such methods.

IQ tests for school-age kids have fairly good predictive validity for their performance in school. the predictive validity of tests like SATs for college performance and life success is not very clear. some research says good, some says not so good, some says lousy.

so there you have it.

i would seriously not recommend those online pretend tests. i have had scores ranging from 140 to 178 on those.

(178, ya riiiiiiighht. :rolleyes: )

further, different IQ tests are developed for different age-groups, so if you scored well on something developed for 10-year olds, whoopee for you! :wink:

Just to protect my rep - The ones I did weren’t for 10 year olds.

(And I didn’t put much stock in the scores I got as they were ‘pretend’, I.e. not official Mensa IQ tests)

very well, lobsang, your rep is intact. nods graciously

[hijack]
the OPs name reminded me of a beautifully bad joke i created the other day. i must share it…someone please find it amusing!!

What did the guy atom say to the girl atom in the bar?
“Baby, you make me get a hadron!”
What did the girl atom reply to the guy atom in the bar?
“Muon, you loser!”

i am thoroughly pleased with myself.
[/hijack]

Howard Gardner, author of “Multiple Intelligences.”

Howard Gardner actually has recently revised his book “Frames of Mind” and reached a conclusion that there are eight intelligences

Logical / Mathematical
bodily / kinesthetic
Interpersonal
Intrapersonal
Musical
Visual / spatial
Linguistic

and his latest:

Naturalistic intelligence
He is also considering adding a spiritual intelligence to the list

Finally my University Education classes are paying off!
:stuck_out_tongue:

what would an IQ in the eighties suggest?

I used to work for psychologists who did IQ testing, mostly for people who were applying for disability as mentally retarded. If they scored below 70, they could be classified as retarded. For the reason of obtaining benefits, the tests were useful.

Your IQ score has a 20-point margin of error, that is it could actually be 10 points lower or 10 points higher than the actual number score. I’ve taken the Wechsler Adult Intelligence Scale, which was widely used at the time I worked for the psychologists. Some of the general knowledge questions I remember were, “How far is it from New York to Paris?” and “What is an altar rail?” This would suggest that the test is biased toward upper middle class Christians.

That you have bushy hair, and wear brown cordural trousers and diamond-pattern yellow jumpers?

that you wear corduroy pants? :stuck_out_tongue:

okay, seriously…

on a normal distribution, it would mean that an IQ of, say, 85, is one standard distribution below the average of 100. so, depending on the IQ test in used, roughly 16% of the population falls below you. about 68% of the population has IQ scores in the 85-115 range.

intellectual functioning is probably quite okay.

if your question was an oblique way of asking whether it qualifies as retardation, the answer is no.
IIRC, the table is

IQ-category
<25 = profound retardation
25-39 = severe
40-54 = moderate
55-69 = mild

For online IQ test, press CTRL-ALT-DEL twice in rapid succession.

-FK

I took two IQ tests several years apart. The test was the same but had been revised in between my taking it. In between the two I went to grad school. According to my scores I added 14 points to my IQ my going to school, or just in that time (from 27 years old to 34 years old). Seems to me that that is too much to add for a legit test. In addition, the fact that I studied psych in college makes me wonder if specific experiences have a more profound impact than the “innate” interpretation of the test implies. (I took the WAIS)

The test, however, does have some interesting features: among them the subtest scores. Certain patterns in the sub-test scores might indicate test anxiety or depression, or organic learning disability, etc.

the reason why I asked was that I had an IQ test done when I was three or four. I believe its results said that I had an IQ in the eighties. However, I imagine that things have changed.

That IQ test is particularly good at identifying computer illiterates.

Not all IQ tests use the same scoring and rating schemes.

A score of 140 on Stanford-Binet would correspond roughly to a score of 160 on the Cattell scale (Cattell is used by Mensa), according to this site

As to whether IQ tests are accurate - it all depends what you mean by accurate.

There are many differing views of exactly what the tests measure, but one view of accuracy would be whether the results of the tests are consistent.

Allowing for different schemes of scoring and grading, and allowing for the real life imperative (90% of everything is crap, aka Sturgeon’s law) which indicates that few free tests will be reputable - then the tests I have taken all result in broadly comparable ratings.

The test I took for Mensa (172 - many years ago) and the test run recently by the BBC (145 - Test the Nation) produced very different scores, but when run through the conversion site linked above the results were within a couple of percentage points (well inside the margin for error).