02-18-2008, 06:47 PM
February 18, 2008
Dear Dave,
My training and experience in statistics are solid.
Your statistician apparently did not carefully read what I wrote and you apparently did not understand it.
What I clearly wrote concerning the distribution of your observations is "...their variance about the expected mean values were sometimes much larger than predicted from the ideal theory for random letters." "Dave Bauscher reasoned that the larger variance than expected was a Godly miracle." THIS IS APPARENTLY STILL YOUR POSITION.
What I clearly wrote and you quoted was "All I found looking at his data was a typical Gaussian distribution of observations having a larger than ideal variance."
However, as I clearly wrote, this is the natural result of "...the natural increase in variance caused by the non-random organization of letters in the text...." Don't you get it? The variance is not ideal but rather is several times larger than the variance for observations that would result from a population of letters that were perfectly random. Your assumed variance is wrong because you assumnptions are wrong. As long as you insist on using an unrealistic theoretical variance that does not apply to this real situation you will reach the same incorrect conclusion.
"Alas, all Dave observed was a typical Gaussian (so-called "normal") distribution of observations having a larger than ideal variance."
Later this week I will calculate the variance correction factor and demonstrate that the results you observed fit quite well into the resulting Gaussian bell-shaped curve with corrected variance.
Sincerely,
Otto
Dear Dave,
My training and experience in statistics are solid.
Your statistician apparently did not carefully read what I wrote and you apparently did not understand it.
What I clearly wrote concerning the distribution of your observations is "...their variance about the expected mean values were sometimes much larger than predicted from the ideal theory for random letters." "Dave Bauscher reasoned that the larger variance than expected was a Godly miracle." THIS IS APPARENTLY STILL YOUR POSITION.
What I clearly wrote and you quoted was "All I found looking at his data was a typical Gaussian distribution of observations having a larger than ideal variance."
However, as I clearly wrote, this is the natural result of "...the natural increase in variance caused by the non-random organization of letters in the text...." Don't you get it? The variance is not ideal but rather is several times larger than the variance for observations that would result from a population of letters that were perfectly random. Your assumed variance is wrong because you assumnptions are wrong. As long as you insist on using an unrealistic theoretical variance that does not apply to this real situation you will reach the same incorrect conclusion.
"Alas, all Dave observed was a typical Gaussian (so-called "normal") distribution of observations having a larger than ideal variance."
Later this week I will calculate the variance correction factor and demonstrate that the results you observed fit quite well into the resulting Gaussian bell-shaped curve with corrected variance.
Sincerely,
Otto