h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Hmmm, I think you ought to be careful when you state that EPs are "a way of directly measuring general intelligence". Mainly because I personally disagree with almost every part of that sentence!

Sensory EPs are a very blunt tool that measure no more and no less than how long it takes nerves to tramsit an electical message from the sensory organs to the brain. It doesn't say anything about what the brain does with that message!

There are 3 sorts of sensory EPs that I have heard of.

1) Visual evoked potentials, the stimulus here is usually an alternating checkerboard pattern. The result will be a wave of certain amplitude after a delay (latency).

This test is more usually done NOT to "measure general intelligence" but to test whether there is damage to the pathways that transmit visual information to the brain. For example, a person who has had one episode of optic neuritis (inflammation of the optic nerve at or just behind the retina) will have a delayed VEP. It doesn't follow that they are less intelligent! and in fact I have met someone who has had optic neuritis and is VERY intelligent by any scale you care to mention.

2) Auditory evoked potentials, as you say the stimulus is a click. I've only heard about these being used in physiology research.

3) Somatosensory evoked potentials, the stimulus is a little electric shock somewhere on the body (like arm or leg), this can be useful in testing to see if the nerves that take the message up to the brain are damaged.

Any type of nerve's conduction speed depends on its physical properties but won't differ much between individuals unless the nerve has been damaged. However, the speed of thought (as measured by reaction times, time taken to complete an IQ test etc) varies MUCH more between individuals and I would imagine would be MUCH better correlated to what is generally understood as intelligence. (But different sorts of intelligence, and also reflecting "training": as in David Beckham's reaction times in football, or a chess grandmaster's speed of working out possible outcomes of a move).(Another problem is that reaction times might be speeded up by other factors, such as being under the influence of amphetamines - but obviously being on amphetamines doesn't in itself make you more intelligent).

Furthermore it might be argued that there is NO such thing as "general" intelligence and that defining "general" intelligence is simply a way of stating which type of intelligence you value most highly. For example the old "11 plus" test was supposed to be a sort of intelligence test and did appear to be correlated with success in school tests generally - but did not correlate so well with emotional intelligence or physical David-Beckham type intelligence. But at that time, the education system valued success in school tests much more highly than success on the football field.

both the auditory and visual evoked potentials correlate very well with the general intelligence reading that ordinary people get on well calibrated and non-culturally biased intelligence tests. Obiously, if you were to do both sorts of tests, and they gave very different results, you would look much more closely at the affected system.

once you are getting outside the normal range of iq, it gets very difficult to callibrate the intelligence tests, ending up with things like the four sigma test for the very intelligent (top 1%) where it is more the number of questions you can answer rather than the time you take to answer them which becomes significant.

Quite a lot of people seem to imply that general intelligence doesn't exist, but everything I have read seems to suggest it does.

to use a computer analogy, general intelligence would equate to clock speed of the system, while other factors would significantly modify the expected usefullness of such a clock speed.

It does appear from research in the last couple of decades, that there is definately both a general intelligence parameter, and then a whole set of other parameters which are orthoganal to it. for any individual specific intelligence, it seems to be the vector which is the product of the general and specific intelligences which are significant.

while we do seem to have found quite a lot of specific intelligences, they all partially correlate with general intelligence, and current results don't seem to be throwing up evidence which contradicts this current belief.

general reaction times seem to be a bit of a red herring, as they measure the observe, notice, decide, react pathways, and as such depend greatly on the level of training the individual has had.

even now you have people like linford christie who can repeatedly react faster than the accepted normal range.

while I would agree that more attention needs to be given to looking at the different specific intelligences both seperately, and in combination, it doesn't really impact on general intelligence and evoked potentials.

similarly, people on different substances, or with damage don't really impact on this usagefor evoked potentials either.

both the auditory and visual evoked potentials give very good results, which as I have said correlate well with general intelligence.

unfortunately, I don't know much about the somatosensory evoked potentials as they apply to this sort of research, and thus can't really comment on them very much.

interesting.

what aspect of VEPs correlates with intelligence?
simple latency, or a more complicated measure?

I think it is how short the latency is, but I would have to look it up.

I will get back to you.

I have just looked it up too. It appears that greater intelligence is associated with decreased P300 latencies and increased P300 amplitudes, among other things.

And, I discovered that (as you say) scientists *do* use event-related and evoked potentials to make conclusions about intelligence. To me this seems bizarre - surely intelligence is a phenomenon that can be studied at many levels (event-related potentials / functional MRI imaging / psychophysical tests all included if you like) and the most complete assessment of a person's intelligence would require looking at as many of these levels as possible - no one test would give you all the information you need to know.

But that's partly because I have an (unprovable) conviction that a single scale of "general intelligence" does not exist in reality, which would take us back round to where we started...

Anyway thank you for taking the trouble to reply to me - it's an interesting topic, isn't it.

breakfastlunchandtea

it sure is an interesting subject.

you seem to be partially right about general intelligence. it definately does exist, but it isn't all of the answer

you are correct that one single scale doesn't exist, what you have instead is the general intelligence, and a large and partially unknown set of general intelligences that also seem to be important.

if you only use the general intelligence information, it doesn't tell you enough.

ps what do you think of the other entries in the intelligence project?

pretty ambitious and some good stuff there (apart from my objection to the evoked potentials bit)! it makes me wonder about some wider aspects, eg:

I think it'd be interesting to have some more info about the social implications of intelligence testing. eg. in some parts of the UK, entry to secondary school for *all* state pupils is governed by results of a verbal reasoning test. The american SATS exam seems to have a large element of being a test of general intelligence from what I've heard (is this true?). In the mid-twentieth century there was a vogue for compulsory sterilisation of those with low "intelligence" for their own good and those of the race (and not just in Nazi Germany).

What right have we to test people's "intelligence" (as opposed to what a child has actually *learnt*, or what a person can show they can *do*) in this way without their consent - and not only that but to use the results to make life choices on their behalf?

And how do we define the rights of those with severe learning difficulties in situations where they only have partial understanding?

And just for fun, what about phrenology - very important in its day you know!

breakfastlunchandtea

sorry for the delay in replying, but I had a funeral to partially organise.

phrenology is only relevent as a historical blind alley.

american S.A.T.'s and the U.K. 11+ are indeed measures of general intelligence (and to a certain extent of general knowledge as well).
unfortunately for the people who advocate them, the results change more between 12 and 14 than at just about any other time, thereby making their predictive value largely worthless.

as to the ethical aspects of abusing intelligence testing, this is part of a much larger can of worms dealing with abuse of testing and prejudice, and I don't intend to go there until I have most of the rest largely completed.