Friday, 10 May 2013

Intelligence declined one SD since Victorian times - why NOT?


Okay - so simple reaction time data indicates that average general intelligence has declined by about one standard deviation (15 IQ points) since the late 1800s. 

What next?

Should we believe it?


Well, here are a few considerations:

1. Simple reaction times are the most objective evidence we have concerning trends in general intelligence.

Therefore we should:

a. believe them, or

b. come-up with something better, or

c. show how these studies of reaction times are incompetent, or

d. assume that the studies of reaction times are dishonest, or

e. assume that this data is wrong on the basis that most data is wrong (and because we are under no obligation to believe any particular bit of data - quite the contrary, there must be good reasons for believing any specific proposition)


2. Any other data which can be brought to bear on the matter of declining intelligence is (I think) relatively 'soft', subjective and imprecise compared with reaction times - (things like rates of major innovation, rates of geniuses etc) so - although I see much of what seems to be consistent with a significant and relatively rapid decline in general intelligence - it is unlikely to convince someone who does not want to believe in the first place.

On the other hand, even if it does not support, does any of this long terms evidence of general intelligence contradict the thesis that it has declined?

I mean, is there any decent evidence that general intelligence has actually risen? (aside from the Flynn effect of increasing pen and paper IQ test scores - which is not relevant here since it is what is being explained, and therefore inadmissible as evidence).

Has, for instance, per capita performance improved since the late 1800s in any 'g' related and quantifiable human endeavour? I know this is tricky to answer, since there are shifts in the specialized activities of intellectuals (e.g. away from study of The Classics and towards some of the sciences, or medicine) - but is there is reasonably compelling evidence of this sort of improvement?


3. If is is decided that we should, after all, accept the best evidence as showing a significant decline in average intelligence since the late 1800s, then there is quite a lot of further work to be done on the mechanism - because, on the whole, present understanding of 'dysgenic' mechanisms in relation to intelligence is not adequate to explain the rapidity of this decline.

(I mean, that calculations based on differential reproduction and heredity of intelligence predict a much slower rate of decline than could lead to one SD reduction in 'g' in the space of just four or five generations.)

The accumulation of deleterious mutations with the relaxation of selection due to high rates of child mortality is my first best guess at the likely 'extra' mechanism (and Michael Woodley agrees)

but really this is just a best guess, and the detailed mechanisms of how this might work are unclear.

However, it could make interesting science in trying to find-out!



Matthew C. said...

You're hinging a lot on a robust relationship between IQ and reaction time.

As someone whose reaction time is right about at the modern average, and IQ quite significantly above it, I wonder what the quality of the IQ / reaction time relationship is.

Adam G. said...

My IQ is well above average and my reaction time is below average.

Anecdote, therefore QED.

Bruce Charlton said...

AG - You are probably on the best-fit line showing the inverse correlation between IQ and RT, MC - you are part of the (considerable) scatter around that line!

Wm Jas said...

I have no idea how my IQ and RT compare to the average, but I also wonder if you're assuming too much about the relationship between the two. Even if IQ are RT are strongly correlated, it does not necessarily follow that a change in one implies a change in the other.

There is a strong positive correlation between height and weight, for example. It is also true that the weight of the average American has increased considerably in recent years -- but it does not follow that there has been a corresponding increase in height.

Bruce Charlton said...

@WmJas - I personally am not assuming anything! The assumption is widespread in IQ research.

The correlation between IQ and RT has been known and generally accepted since Francis Galton - and there are plausible reasons why IQ and RT are both aspects of the same attribute (something like speed and efficiency of CNS processing) which probably underlies 'g'.

Woodley's new paper (linked above) gives the key references, especially Arthur Jensen's reviews.

Matthew C. said...

Actually I bet hat size is a better correlate to IQ than reaction time.

Of course, I say that as someone with a big gourd. . .

Ballomar said...

The Weston A Price people would say it is down to malnutrition - i.e. modern diets are nutrient poor, which lead to a number of problems including a decline in intelligence.

The Crow said...

I am by far the fastest human I have ever seen, in raw reaction times. And one of the highest IQs too.
My speed serves me far better, though, than my IQ.
Speed is almost always useful, whereas undisciplined IQ can get you into no end of trouble.
Uh-oh: gotta run; the IQ police just arrived in my driveway...