Statistical article about Stanford Binet iq scale. Evolution of intelligence with scores of Wechsler and Stanford Binet IQ scale.

Cover of the book the EDI Study. Dusk over the sea with clouds, Galicia.




Author: José Tiberius

q028 Statistical article about Stanford Binet iq scale.


The title of each graph of the statistical study indicates the parents' variables (R or M & F) to which the correlations relate. Each point of the colored lines represents the correlations with the observational C variables of the children.

Likewise, the variables of unknown order, formed by the different groups of 1 to 10 values from the 70 IQ values of each parent and children variables, appear on the left-hand side of the graph. The criteria order of the groups of 1 to 10 values located on the right-hand side is the variable mentioned at the bottom of the graph.

Indeed, there is an almost instantaneous perception of the exactitude of the particular specification of the statistical study; each graph shows sixty coefficients of determination (r²) highlighting the global and underlying relations of the involved data set.

See the methodology of the statistical abstract for more details



1. General statistical significance

The considerable increase of the correlation for the estimation of homogenous groups is not due to the reduction of 68 to 5 or 4 degrees of freedom, since the estimation with non-homogenous groups, without previous rearrangement, has the same degrees of freedom and the correlation even lowers concerning the sample without grouping.

In general, the model of the genetic evolution of intelligence (Mendelian geneticsConditional intelligenceGlobal Cognitive Theory) adjusts perfectly, showing an superior to 0.9 in several cases. Bearing in mind the tendency to increase the goodness of fit with the size of rearranged groups, we could assume it would be over 0,9 in almost all the cases for grander groups within a more significant sample.

2. Social Model with centered variables

As expected, the compensation of random deviations in the values of the centered variables makes the new statistical analysis fit significantly better than the model with original vectors.

It seems there is not much margin left to deny the hereditary nature of intelligence, not even to try to reduce it to less than 80%. Considering the model use groups with a maximum of ten elements and the observed tendency to increase with the number of elements, we would say the correlation should be almost 1 for groups of a hundred elements.

In other words, the general laws of innate intelligence proposed by the CEL are confirmed at a social level while maintaining deviations within individuals.

3. Significant comments on this particular graph

As you can clearly see by its form, the three dependent variables of the children, analyzed in the model, behave in a very similar way to the progenitors' explanatory variables M & F

The general multidimensional correlation index (GMCI) is 17,20 which is one of the biggest values within The EDI Study.

When using as arragne criterionvariable X6, which is the most centered of the IQ vectors of the data set, there is an important increase in the correlations.

The biggest determination coefficient of this graph is 0,88 which is high value within this type of statistical studies.