Yes, it turns out. I'd guess most Zoozers already take it for granted that education imparts something of value, and isn't just an expensive sorting mechanism. But it's nice to be able to put a number on it.
The researchers were able to disentangle correlation from causation by including three types of studies in their meta-analysis:
"The first type of study includes data collected from individuals over time, including intelligence measurements obtained before and after individuals complete their education. This allows researchers to adjust for participants' prior intelligence level when examining the association between years of school and later intelligence.
The second type of study takes advantage of "natural experiments" in the form of policy changes that result in individuals staying in school for different lengths of time. In one study, for example, researchers examined data from the 1960s when Norway gradually enacted a new policy that increased the basic education requirement by 2 years, testing whether IQ scores were higher for students who'd been given more compulsory schooling.
In the third study type, researchers use school-admission age cutoffs to compare children who are similar in age but who have different levels of schooling due to their specific birth dates...In each of the three types of studies, the researchers found that an additional year of education was associated with an increase in IQ that ranged from 1.197 IQ points to 5.229 IQ points. In combination, the studies indicated that an additional year of education correlated with an average increase of 3.394 IQ points."
Science Daily article here:
https://www.sciencedaily.com/releases/2018/06/180621112004.htm
--Frankie
I'm not massivley keen on IQ-related studies, but do have a soft spot for meta-analyses...will definitely be giving this one a look.
ReplyDeleteHow long, though, till some utter trumpwit starts to calculate the dollar cost of a 1-point gain, and / or find some way to blame proffies for it?
Along with the Flynn effect, this is all the more reason to think that IQ is a very clumsy way of measuring human intelligence. Yes, I know that IQ correlates with all kinds of positive "life outcomes," and that this among the most robust results in psychology: another way to look at this is that it speaks badly for the methods and scientific validity of psychology.
ReplyDeleteAnother problem here is that it is statistical nonsense to quote these (or any) IQ scores to three digits to the right of the decimal point. This is because the range they are describing is 3.213 +/- 2.016, since the average of 1.197 and 5.229 is 3.213, and 2.016 = (5.229 - 1.197)/2, assuming a linear relation. It should be 3.2 +/- 2.0, or in other words, 1.2 to 5.2.
ReplyDeleteMoreover, what relevance could a thousandth, or even a tenth, of an IQ point possibly make? IQ tests struggle to be consistent to one significant digit. It's much like what Carl Sagan said about horoscopes: "These predictions are NOT predictions, they tell you what to do, not what's going to happen, and they don't agree with each other."
Reminded me of this podcast:
ReplyDeletea16z Podcast: The Case Against Education, From Signaling to Rainbow’s End
with Bryan Caplan, Marc Andreessen, and Sonal Chokshi