Steve, are you familiar with the research showing (or claiming to show) that self-control correlates significantly more to academic and career performance, than IQ does?
(eg. the sources discussed in the following article pop-science article by The New Yorker: http://www.newyorker.com/reporting/2009/05/18/090518fa_fact_lehrer)
Take age 9 to 17. 5.7/15 = .38. So rho = sqrt(1-.38^2) = .923. Adjusted for reliability and the age 9 IQ should correlate perfectly with the age 17 IQ
For normalized n(0,1) data the slope of the regression line equals rho. So the mean change is greater than sqrt(2/pi), but it shouldn't be as large as it is.
I graduated with a 4.0Gpa in Accounting then went into Computer Science before dropping out after 2.5 years I had a 3.98Gpa. I got one 95 in a class and the way the school calculated it dropped my Gpa by .02 I plan on going back to college to finish my degree in Computer Science, although I may switch to Electrical Engineering once I complete coding of the current website I am working on. I find both to be insanely interesting.
I don't like that range of changes. Everyone gets smarter as they get older?
ReplyDeleteI believe they are giving you magnitudes of changes, not the sign.
ReplyDeleteYup, too bad the lower magnitude bound is a pretty useless piece of information...
ReplyDeleteSteve, are you familiar with the research showing (or claiming to show) that self-control correlates significantly more to academic and career performance, than IQ does?
ReplyDelete(eg. the sources discussed in the following article pop-science article by The New Yorker: http://www.newyorker.com/reporting/2009/05/18/090518fa_fact_lehrer)
Search under "marshmallow experiment" on this blog. You'll find links to the original papers.
ReplyDeleteI got that. And I double down on what Christopher followed up with.
ReplyDeleteNot credible.
ReplyDeleteTake age 9 to 17. 5.7/15 = .38. So rho = sqrt(1-.38^2) = .923. Adjusted for reliability and the age 9 IQ should correlate perfectly with the age 17 IQ
For normalized n(0,1) data the slope of the regression line equals rho. So the mean change is greater than sqrt(2/pi), but it shouldn't be as large as it is.
That's all rot MtMoru.
ReplyDelete"Because I care", here's what a little plug and chug give.
I. 2(1-r)(1-2/pi) = var(|test -retest|)
II. 2*sqrt((1-r)/pi)) = mean(|test - retest|)
So for age 9 estimated r is .8 using I and about .7 using II.
The two may not agree because the data are not actually sampling a biveariate normal distribution. I doubt it's sample bias or statistical +/-.
I graduated with a 4.0Gpa in Accounting then went into Computer Science before dropping out after 2.5 years I had a 3.98Gpa. I got one 95 in a class and the way the school calculated it dropped my Gpa by .02 I plan on going back to college to finish my degree in Computer Science, although I may switch to Electrical Engineering once I complete coding of the current website I am working on. I find both to be insanely interesting.
ReplyDeleteI am sure people who use their brain and think a lot manage to maintain their IQ for most of their adult life until older age just gets to you.
ReplyDelete