Int = IQ?

[Delta]

A site quoting "The Intelligence of Dogs" by Stanley Coren, published by Hodder Headline, London 1994, presents an IQ test for dogs. Results are read as follows:
http://www.abc.net.au/animals/dog_test/results.htm

Over 25 points... Your dog is a genius
15 - 25 points.... Your dog is smart, but won't go to Harvard
5 - 15 points...... Your dog is not too bright, but is most likely very cute
Below 5 points.. Your dog must be an Afghan

Again, that's more evidence that a working animal with average Int 2 in D&D converts nicely by the IQ = Int * 10 formula.

 

[S'mon]

I actually read (most) of his book, The Mismeasure of Man. S'mon is kind of right.

I didn't think it was controversial that Gould was a Marxist, I didn't intend it as some kind of insult. I'm an academic, here in the UK many academics would happily call themselves Marxist. Often staff seminars include statements like "As Marx teaches us..."

 

[S'mon]

I remember that IQ = 79 or less was defined as 'mildly retarded' and 59 or less 'severly retarded' (clinically and legally speaking).

It's not true though - a lower than population-median IQ might be due to organic retardation, but not necessarily. These kind of blanket declarations result in practices like classing perfectly normal healthy children as retarded, which can even result in them being taken away from their parents and institutionalised. It tends to happen most often when population A is dealing with population B and they have different measured median IQs, eg white Australians and Australian Aborigines, and the A group applies these kind of benchmark rules, developed for their group, to the B group. This kind of ethnocentricity has very harmful effects IMO.

 

[mara]

I didn't think it was controversial that Gould was a Marxist, I didn't intend it as some kind of insult. I'm an academic, here in the UK many academics would happily call themselves Marxist. Often staff seminars include statements like "As Marx teaches us..."

All right then. The way you mentioned it made me think it was knee-jerk. Thanks for clarifying.

 

[S'mon]

As much as I like dogs, I have to acknowledge that they are simply not that smart - there is no way you can get a dog to talk (using proxy signals), even though it often seems that way with our best friends.

Whereas chimpanzees and dolphins can be taught simple language, though not much in the way of grammar. Of course humans have evolved to have large chunks of our brains dedicated to language; human superiority in language (compared to other smart animals) is likely greater than in other areas of intelligence.

 

[Roman]

Correct, it's not a perfect (1:1) correlation by any manner of means.

So, have you come accross any conversion of IQ into more 'objective' (as in not relative) computational units?

Whereas chimpanzees and dolphins can be taught simple language, though not much in the way of grammar.

Yes, I could probably envision a chimpanzee or a dolphin having the intelligence of a three year old child. With dogs, it is much more problematic. The site that gives 'dog IQ' grades dog intelligence (or motivation), and assigns points in an ordinal manner, but then arbitrarily claims these are equivalent to IQ.

If similar tests (to those for dogs) were done with normal healthy 3-year old children they would get full scores (assuming proper motivation).

Nevertheless, thanks Delta, for the link - it is interesting to see how tests of animal intelligence might be conducted.

Of course humans have evolved to have large chunks of our brains dedicated to language; human superiority in language (compared to other smart animals) is likely greater than in other areas of intelligence.

This seems likely, but it could be argued that language underlies even other forms of intelligence, including logic, mathematics and spatial intelligence (for example, language helps us define discrete concepts of shapes to be rotated).

 

[S'mon]

So, have you come accross any conversion of IQ into more 'objective' (as in not relative) computational units?

Not sure what you mean by conversion - measurable intelligence appears to correlate positively with various objective measurements, like brain volume and amount of electrical brain activity, but the only thing IQ tests can measure entirely objectively is the ability to do IQ tests!

 

[Roman]

Not sure what you mean by conversion - measurable intelligence appears to correlate positively with various objective measurements, like brain volume and amount of electrical brain activity, but the only thing IQ tests can measure entirely objectively is the ability to do IQ tests!

My point is that if 'g' represents the computing power of the brain, it should be measurable in computational, absolute units. IQ, by contrast, is a relative measurement (relative to the rest of the population). It follows that if IQ measures 'g' and 'g' is the computational power of the brain than it should be possible to convert IQ, through the use of some conversion formula, into a measurement using absolute computational units (operations per second, variable throughput, or something along those lines).

To give you an example. If we look at the height of people, we could say that the mean height of adult men in the world is 100 and the standard deviation is 15 (pun intended with the choice of numbers). We could call the the HQ (height quotient) and it would give us a relative measure of height of adult men. An absolute (or objective measure) would be to state that the mean height of adult men is 180 centimeters and the standard deviation is 27 centimeters. This means that 1 point of HQ (a relative unit) would convert to 1.8 centimeters (an absolute units).

 

[S'mon]

My point is that if 'g' represents the computing power of the brain, it should be measurable in computational, absolute units. IQ, by contrast, is a relative measurement (relative to the rest of the population). It follows that if IQ measures 'g' and 'g' is the computational power of the brain than it should be possible to convert IQ, through the use of some conversion formula, into a measurement using absolute computational units (operations per second, variable throughput, or something along those lines).

My understanding is that measured IQ correlates with other indicators of brain computational power but it's not a perfect 1:1 correlation. AFAIK there is no accepted accurate measurement of absolute brain processing power; although I think it might be theoretically possible to measure neural activity in such a way as to reach a figure. An obvious problem is that large parts of brain activity do not go towards producing intelligence but towards other functions, such as keeping the body functioning. Intelligence seems related more to the brain mass/body mass ratio than to absolute brain mass.

 

[Roman]

Indeed, this lack of our ability to disentangle these kinds of issues is one of the reasons for my scepticism that intelligence can currently be measured accurately.