The Greater Male Variability Hypothesis (GMVH) suggests that males exhibit greater variance than females in their cognitive abilities. I present some relevant empirical findings and make a suggestion for how the hypothesis should be conceptualized and evaluated.
I think you should not use the variance ratio, as this metric is not interpretable by intuition. I think people use it because there is a known statistical test for it. The ratio of the standard deviations is better for understanding. So in your results, VR's are 1.16, 1.22, 1.23, mean is 1.20. Sqrt of 1.20 is 1.10, i.e. 10% greater for males. So in terms of IQ, this would correspond to SD's of about 15.7 to 14.3.
But taking VR0 is an egalitarian fault. We know males have higher intelligence by 4-5 points in adulthood, d=0.2 to 0.4 so true VR is not VR0 but about VR0.3 so clearly higher than 1,2...
But taking VR0 is an egalitarian fault. We know males have higher intelligence by 4-5 points in adulthood, d=0.2 to 0.4 so true VR is not VR0 but about VR0.3 so clearly higher than 1,2...
That is true too, and men have a much higher spatial, and quantitative intelligence than women if i recall correctly, the only "aspect" of intelligence in which they (women) have advantage is in verbal IQ.
Could this be the result of decreased variation in X-linked gene effects due to female X-inactivation?
Male X-linked genes are always producing a single set of genetics products, but female X-linked genes experience an averaging process in tissues expressing X-linked genes (i.e. practically all of them) that plausibly reduces variation in the X-linked effect sizes.
My goal in this post wasn't to explain the causes of the phenomenon, but, yes, that is one potential hypothesis of the cause of the greater male variability. At this point, I'm not confident in what the true explanation is. That hypothesis would explain greater male variability, but I'm not sure if it would help explain the positive relationship between mean effect size and variance ratio.
yeah, if you assume a model where the female has 5% of her genome mixed due to inactivation, uniformly distributed effect sizes across the genome for a quantitative trait, and male variance = sigma ^ 2, I think you'd get
VR0 = sigma ^ 2 / ((((.95 ^ 2) * sigma ^ 2) + ((.05 ^ 2) * (sigma ^ 2 / 4))) = 1.11, which isn't too much different than you'd suggest.
I think you should not use the variance ratio, as this metric is not interpretable by intuition. I think people use it because there is a known statistical test for it. The ratio of the standard deviations is better for understanding. So in your results, VR's are 1.16, 1.22, 1.23, mean is 1.20. Sqrt of 1.20 is 1.10, i.e. 10% greater for males. So in terms of IQ, this would correspond to SD's of about 15.7 to 14.3.
But taking VR0 is an egalitarian fault. We know males have higher intelligence by 4-5 points in adulthood, d=0.2 to 0.4 so true VR is not VR0 but about VR0.3 so clearly higher than 1,2...
Very insightful
But taking VR0 is an egalitarian fault. We know males have higher intelligence by 4-5 points in adulthood, d=0.2 to 0.4 so true VR is not VR0 but about VR0.3 so clearly higher than 1,2...
True, I just picked the values he emphasized. Adult male g advantage is about 0.3 d, so the VR will be somewhat higher and thus also the SDR.
That is true too, and men have a much higher spatial, and quantitative intelligence than women if i recall correctly, the only "aspect" of intelligence in which they (women) have advantage is in verbal IQ.
Could this be the result of decreased variation in X-linked gene effects due to female X-inactivation?
Male X-linked genes are always producing a single set of genetics products, but female X-linked genes experience an averaging process in tissues expressing X-linked genes (i.e. practically all of them) that plausibly reduces variation in the X-linked effect sizes.
My goal in this post wasn't to explain the causes of the phenomenon, but, yes, that is one potential hypothesis of the cause of the greater male variability. At this point, I'm not confident in what the true explanation is. That hypothesis would explain greater male variability, but I'm not sure if it would help explain the positive relationship between mean effect size and variance ratio.
yeah, if you assume a model where the female has 5% of her genome mixed due to inactivation, uniformly distributed effect sizes across the genome for a quantitative trait, and male variance = sigma ^ 2, I think you'd get
VR0 = sigma ^ 2 / ((((.95 ^ 2) * sigma ^ 2) + ((.05 ^ 2) * (sigma ^ 2 / 4))) = 1.11, which isn't too much different than you'd suggest.
Thanks. I like to see means and standard deviations, and also a plot of the shapes of the distributions, in case they give us some insights.