Hi @
Lux de Veritas ,
But let me infer my point of view when in regards to these statistical tables. One issue affecting generalization concerns tests of statistical significance. Most tests of statistical significance are based on the data obtained from samples, but they are really testing a certain hypothesis on the population. That said, they are usually testing the null hypothesis and the research hypothesis. Statistical significance is part of a larger complex picture.
Secondly, the issue that affects generalization from the sample results to the population is attrition of the sample. Attrition is usually thought of as loss of cases over time in a longitudinal design, but multivariate analyses are subject to this concern as well. Most of the multivariate statistical procedures require participants to have valid values on all measures. With multiple measures taken on each case, it is more likely that some of the cases will have missing data on at least one of them, and this can affect analysis.
The issue as i mentioned earlier , on matters of statistical significance , depends on the sample size as well as the presence of
Random Sampling. We do not know if the samples taken from East Asian populations were from academic institutions or were from the general population. The former can distort the data analysis in favor.
Lastly, i would also caution the studying of the variable of national stability. The mere fact that just because some countries in East Asia (Japan, South Korea) have high IQ scores or low crime rates may be due to the fact that these countries are largely developed with police services that are effective. Comparing this to countries , say in Africa, which are largely underdeveloped and lack proper civil infrastructure of crime force may misconstrue data.
I am not in the view that human intellect is genetic, but rather, influenced multifactorially. Environment has a significant attribution to this -- dependent variable.
I remain,
@
Nihonjin1051