Statistical Attainment

No one reasonably doubts that inheritance plays a significant role in offspring traits. Does this extend to academic achievement? One would certainly think so.

Similarly, no one reasonably doubts that socioeconomic inequities contribute greatly to academic outcomes. Put simply, regardless of innate intelligence, poor kids have less opportunity to get a decent formal education.

What emerges is a complex picture, for which factor analysis fails in attributing causality in life's fortunes. Not-so-bright rich kids may be successful because they go to the best schools, and have the right social connections. Quite-bright poor kids are disadvantaged in both their educational opportunities and social connections. In socioeconomically-stratified societies, wealth and poverty tend to be self-perpetuating, with exceptions. So far, no controversy.

Now, let's talk about the most important fact of statistics: correlation is not causality. Just because two elements coincide does not mean that one causes the other.

Confusing correlation with causality is the greatest error a so-called scientist can commit. It is the primary source of "fake" science, which can be strongly self-perpetuating. And it is quite common.

A group of psychologists at King's College in London demonstrate their ignorance of statistics' most important rule with the following study results:

"Half of the variance in intergenerational educational attainment can be attributed to genetic differences. These results demonstrate that the educational outcomes of parents and their offspring are similar for genetic as well as environmental reasons." ~ psychologists Ziada Ayorech, Eva Krapohl, Robert Plomin, & Sophie von Stumm

What these foolish five are claiming is that genetics is half responsible (cause) for "educational attainment" (effect). In other words, half of the reason for getting a good education owes to genetics (as in, rich parents), while the other half owes to environmental factors (as in, money talks). That claim states causality, when all the researchers might have shown is correlation, except that a 50/50 variance is no correlation at all: it's nothing more than a dice roll.

What the published study results do unequivocably show is that educational attainment has a tenuous link with intelligence. But that is just a self-evident statement about the psychologists who ran the numbers.

(One more noteworthy mention: the editors at the journal Psychological Science published this rubbish.)


Ziada Ayorech et al, "Genetic influence on intergenerational educational attainment," Psychological Science (17 July 2017).

PrintView Printer Friendly Version

EmailEmail Article to Friend

« Concepts | Main | The Great Disrupter »