Actually, the Bell Curve used multiple regression (the appropriate method) and data from the National Longitudinal Survey of z Youth (the best available data). They included all of their data results. That looked good and fished me in. While I have done more number crunching than the average person, I am not a statistician.
Actually, the Bell Curve used multiple regression (the appropriate method) and data from the National Longitudinal Survey of z Youth (the best available data). They included all of their data results. That looked good and fished me in. While I have done more number crunching than the average person, I am not a statistician.
What did I miss, they actually told us but I failed to notice the significance. They wrote, "In the text, we do not refer to the usual measure of goodness of fit for multiple regression, R2 (R squared), but they are presented here (an appendix) for the cross sectional analysis." As Stephen Gould points out in "The Mismeasure of Man", putting it in an appendix that most readers won't study obfuscates the weakness of the associations. Squaring a number less than one makes it smaller, in this case so small that even non-statisticians would notice.
They did not lie and they were transparent in their methods. But this was a disingenuous way of making a mountain out of a mole hill case for the premises stated.
I find Gould's critique to be a proper one, better than just calling Murray a racist as a case against his book. I'm not a fan of name calling, one of the things that Steve talks about in this commentary.
Actually, the Bell Curve used multiple regression (the appropriate method) and data from the National Longitudinal Survey of z Youth (the best available data). They included all of their data results. That looked good and fished me in. While I have done more number crunching than the average person, I am not a statistician.
What did I miss, they actually told us but I failed to notice the significance. They wrote, "In the text, we do not refer to the usual measure of goodness of fit for multiple regression, R2 (R squared), but they are presented here (an appendix) for the cross sectional analysis." As Stephen Gould points out in "The Mismeasure of Man", putting it in an appendix that most readers won't study obfuscates the weakness of the associations. Squaring a number less than one makes it smaller, in this case so small that even non-statisticians would notice.
They did not lie and they were transparent in their methods. But this was a disingenuous way of making a mountain out of a mole hill case for the premises stated.
I find Gould's critique to be a proper one, better than just calling Murray a racist as a case against his book. I'm not a fan of name calling, one of the things that Steve talks about in this commentary.