Monday, 24 March 2014

Researching the decline of intelligence measured by reaction times. Where next?


Now that the approximate magnitude of previously estimated simple reaction times slowing over the past century or so has been confirmed

the evidence of a significant decline in intelligence since Victorian times can no longer be dismissed or ignored.

So, the question arises what next?

1. Other researchers than myself and Michael A Woodley's group need to get involved. So long as all the results come from one group, uncertainty remains, Independent testing/ replication should be attempted.

2. While there is probably no more historical reaction time data to be had; the LPC-SO method of comparing longitudinal with cross sectional data could be applied to further samples of simple reaction times - and to other possible objective measures correlated with general intelligence and with plausible biological links to intelligence.

3. Further historical data on the effects or outcomes of intelligence may be discovered, to supplement Woodley's reanalysis of innovation rates and the incidence of creative geniuses. Any quantifiable human activity or achievement which depends strongly upon intelligence ought to show evidence of decline in-line with slowing simple reaction times.

Historical variability in heritable general intelligence: Its evolutionary origins and socio-cultural consequences. MA Woodley, AJ Figueredo - 2013 - University of Buckingham Press.

4. The effects of normal ageing on simple reaction times needs to be known with more precision - age of onset, shape of curve, sex differences and so on.

5. The quantitative relationship between simple reaction time and currently-measured IQ needs to be known - so as to make a valid conversion formula. The sRT IQ correlation coefficient is too low to make such a formula useful for individuals, but in terms of group averages it could be valuable. Such a conversion formula might turn out to be non-linear - and opens the possibility of measuring (group) intelligence on an interval or even a ratio scale.

6. At a deeper level, an understanding the relationship between general intelligence and reaction times needs development - in particular, can 'g' be coherently defined terms of the objectively measurable speed of processing? What is the minimum possible sRT? What is the effect of slowing sRT on intelligence in terms of interactions with other cognitive constraints?

7. Assuming it is agreed that intelligence has declined very substantially over the past 150 years or so; then the mechanism of this decline needs elucidation - since the rate of decline seems to be faster (maybe even twice as fast?) than the rate predicted by the differential reproductive success of people with different IQ. My preferred explanation of intelligence being damaged by the generation-upon-generation accumulation of novel deleterious mutations (mutations which would, through most of history have led to a high probability of early death during childhood - and thereby the filtering of such mutations from the gene pool).



pyrrhus said...

Have you looked into the decline in the US of SAT and ACT scores over the last 50 years or more? Given that the test has been dumbed down twice, with #3 coming up, it is quite substantial. I intend to do some work on it.

Bruce Charlton said...

@p - The only time I personally looked at SATs was in relation to sex differences in intelligence