Runtu wrote:I assume the best of others, but I have learned never to take conclusions from statistics at face value. When I looked at the stuff you were citing, there were obvious problems apparent immediately; on further investigation, the problems were worse than I had expected.
You seem to believe that I have an agenda to discredit those who oppose same-sex marriage. That is not the case, but I will point out poor and dishonest pseudo-scholarship whenever I see it, and despite your quibbles with my assessment, I stand by my judgment: the FRC piece was undermined by the very statistics it was misusing. This Schumm guy, who was unknown to me until today, not only did not counter anything in the study he ostensibly responded to, but came up with some of the most bizarre speculations about the relationship between gay teens and binge drinking that I have ever heard.
It's not my fault that these sources have turned out to go nowhere fast.
What I find fascinating is that much of the back and forth today has been in regards to the Schumm critique of the Columbia study that you cited.
And, rather than you looking carefully at your own citation (the Columbia study), and jumping all over it for: 1) the data in figure 1 not matching what was claimed in the narrative of the study; 2) The N's used in the study not being reported; 3) and most important, that the data in figure 1 of the study was "reported as statistically insignificant," but was still used by the study to call for significant policy changes; 4) the data in figure one didn't provide the actual count, particularly specifying precisely for heterosexual youth; and finally, the lack of discussion of the advantages of using generalized estimating equations rather than hierarchical linear modeling, which would have been useful for readers less familiar with the former.
Instead, you fixated about the "re-created" data of the critique I cited, as if it made a difference to what I had said here or even what Schumm said in his critique, and then you got exercised over Schumms exploratory hypothesis.
In other words, in your quest to uncover the misuse of data wherever it raises its ugly head, you "conveniently" overlooked the documented misuse of data in the study you cited, and went to great lengths disparaging the critique that pointed out the misuse of data in your cited study, without demonstrating a single misuse of the data in the said critique. At best, you showed that there was no way to test and validate the "re-created" data (which was only used to make a single point that didn't conflict with what was reported in the study you cited), and that the author of the critique had the temerity to proffer an exploratory hypothesis that you found absurd.
not to put too fine a point on it, but much of our discussion was kicked off by your scathing rebuke: "Come on, Wade. If we're going to have a reasonable and reasoned discussion of these issues, you need to provide real data from reputable sources."
Do you see the irony?
Thanks, -Wade Englund-