Changes

:::::I'm sorry you take it as an ad hom; I'm sure you appreciate that in such an environment a certain caution about people's claimed expertise is perhaps excusable. And I certainly accept your word as to your post and experience. But this is more than a pure, abstract, mathematical, dare I say professorial treatment of statistics. This is the application of statistics in the real world, down and dirty, relying also upon a knowledge of human nature, and in particular the nature of the Liberal-minded human, which is an additional layer of expertise which possibly you do not have - although, to be fair, working in a university (is it a public funded institution?) you would very likely be exposed daily to all kinds of Liberal nonsense. [[User:Bugler|Bugler]] 13:19, 16 December 2008 (EST)

:::::I'm sorry you take it as an ad hom; I'm sure you appreciate that in such an environment a certain caution about people's claimed expertise is perhaps excusable. And I certainly accept your word as to your post and experience. But this is more than a pure, abstract, mathematical, dare I say professorial treatment of statistics. This is the application of statistics in the real world, down and dirty, relying also upon a knowledge of human nature, and in particular the nature of the Liberal-minded human, which is an additional layer of expertise which possibly you do not have - although, to be fair, working in a university (is it a public funded institution?) you would very likely be exposed daily to all kinds of Liberal nonsense. [[User:Bugler|Bugler]] 13:19, 16 December 2008 (EST)

::::::"This is the application of statistics in the real world, down and dirty, relying also upon a knowledge of human nature..." The thing is, Bugler, that statistics are an applied branch of mathematics. Their express purpose is to allow us to extract certainty from the vapor of possibility or, less metaphorically, to allow us to derive useful knowledge from messy research situations. Indeed, multivariate methods like regression are often used when the purity of experimental control is unattainable. Statistics can only provide this service reliably, however, when there is a full disclosure of how they are obtained. If the data are valid, the procedures appropriate, and the necessary mathematical assumptions not grievously violated, then the conclusions may be trusted to a particular degree of certainty. That ASchlafly's statistics are being used to asses the "...nature of the Liberal-minded human..." is effectively irrelevant. I do not ask that ASchlafly produce perfect statistics, but unless and until he provides a basis for judging his numbers, they cannot be regarded as anything more than his personal opinion. I appreciate your accepting my claimed expertise on my word, however. I have worked at both public and private universities and you will understand that the various strains of "nonsense" that I have encountered from a variety of ideological camps underlies my interest in seeing that statistical methods are not abused. If statistics are to be used to adjudicate between rival claims, then we must emphasize the need for them to be performed correctly. More simply, the extent to which I agree or disagree with the outcome of an analysis has no bearing on whether or not it was done correctly. And, as a side note, I have enjoyed our chat, but will probably be away from my computer for the next day or so for travel. If I do not respond to further comments, please understand it is not out of rudeness. -Drek

::::::"This is the application of statistics in the real world, down and dirty, relying also upon a knowledge of human nature..." The thing is, Bugler, that statistics are an applied branch of mathematics. Their express purpose is to allow us to extract certainty from the vapor of possibility or, less metaphorically, to allow us to derive useful knowledge from messy research situations. Indeed, multivariate methods like regression are often used when the purity of experimental control is unattainable. Statistics can only provide this service reliably, however, when there is a full disclosure of how they are obtained. If the data are valid, the procedures appropriate, and the necessary mathematical assumptions not grievously violated, then the conclusions may be trusted to a particular degree of certainty. That ASchlafly's statistics are being used to asses the "...nature of the Liberal-minded human..." is effectively irrelevant. I do not ask that ASchlafly produce perfect statistics, but unless and until he provides a basis for judging his numbers, they cannot be regarded as anything more than his personal opinion. I appreciate your accepting my claimed expertise on my word, however. I have worked at both public and private universities and you will understand that the various strains of "nonsense" that I have encountered from a variety of ideological camps underlies my interest in seeing that statistical methods are not abused. If statistics are to be used to adjudicate between rival claims, then we must emphasize the need for them to be performed correctly. More simply, the extent to which I agree or disagree with the outcome of an analysis has no bearing on whether or not it was done correctly. And, as a side note, I have enjoyed our chat, but will probably be away from my computer for the next day or so for travel. If I do not respond to further comments, please understand it is not out of rudeness. -Drek

== My political alignment ==

== My political alignment ==