Hydrus, I do have a level of sympathy with your stance.
When I was first introduced to psychology and other "social sciences" I really did not know what to make of them.
A cigar being related to a phallus is some great insight into human psychology?
I ran away screaming as fast as I could from this insanity.
I would have loved to have gone to a school where math/comp sci was the only language accepted.
Words are a great mystery to me.
I was also not entirely comfortable with psychometrics.
Intelligence is at least partly genetic, but we are not quite sure which part.
This was not overly convincing to me.
However, the latest wave of GWAS research over the last 10 years has me
firmly in the Amen corner.
There is now waves of mathematicians entering genetic research.
Typically when you start to see something math up you know that the real science has started.
As you said much of the till now published psychology research literature was a great waste of
trees. There is a certain inherent truth in numbers that words lack. Sure some people can fool others
and sometimes themselves with bad quantitative analysis, though given the caliber of quant that is now being
done, I no longer accept that we are still in a pre-science era for genetics and IQ research.
Intelligence has been given an operational definition.
Massive GWAS are done using this operational definition and SNPs of very small effects are found.
These SNPs are genome wide significant and they replicate in other samples.
The logic is now tightening up.
I do not see as much hand waving anymore.
It can no longer be plausibly argued that the operational definition does not correspond to the SNPs.
If any old invalid definition were to be used for intelligence, then significant SNP results would not emerge.
GIGO.
For example, if the intelligence test used a cognitive test with little if any g loading without time limits etc.
then SNPs would not be found. Intelligence GWAS research is measuring something;
clearly it is because significant results are being found and are being replicated.
This of course leaves the question: What is it that is being measured?
Does operationally defined intelligence relate to what we understand to be real world intelligence?
Spearman's original research from 1904 sheds some light on this. When they simply asked teachers
to rank their students according to their perception of the students' intelligence, the g factor emerged.
This is fairly startling. From the vantage point of a teacher it would be quite clear what the rank of their
students would be. If you want to have a very good psychometric assessment of your IQ, probably
one of the best places to turn to would be senior level high school teachers. Possibly not so much
at the university level because it is no longer as uniform an environment.
It is quite strange that so many people would instead turn to the internet or even
a professional psychologist for such an assessment. A valid IQ test really only relates to finding your relative
ranking within your peer group within your particular environment. The idea that IQ tests would be normalized at
a larger level (say a national level) is essentially nonsensical. Who knows what environmental variables might
differ at such a scale? I do not understand why the same description used in heritability (... for a given population
at a given time") was not ported to intelligence.
In some recent research from the Nordic nations, they found that 50 years ago some students were as much as 10 miles away from
their school. As this factor has changed through time the PGS for Educational Attainment has changed. Bascially,
those children who had the genes for walking 10 miles to school probably achieved higher EA. Yet, this is entirely unrelated
to actual cognitive ability.
Edited by mag1, 01 April 2018 - 04:44 PM.