View Single Post
Old January 14, 2013, 03:04 PM   #51
Evan Thomas
Join Date: July 7, 2008
Location: Upper midwest
Posts: 5,187
Originally Posted by tyme
With no knowledge of the MMPI (other than having taken the MMPI-A once), I'm assuming higher scores on the introversion scale mean more introverted.

If a highly introverted person is .37 times as likely to be an SSRI responder, that means SSRIs don't work as well for that group, right? If introversion were not correlated positively or negatively, they wouldn't be able to state a figure like 0.37 comparing SSRI effectiveness between groups with higher introversion and lower introversion, right?
Yeah... It is "as likely." Sorry about that. And I did get the MMPI-scoring thing backward...

That said, what makes this hard to interpret is that they report only the age results as a strong predictor, with younger people about 3 times as likely to respond positively.

I find it odd that they described their findings in such an apples-and-oranges way. Why wouldn't they report the introversion results the same way: people who scored low on the introversion scale were 2.7 times as likely to respond positively to the medications? If so, there's not much difference in the predictive value of the two measures.

Or they could have said that older people were 0.33 times as likely to respond to treatment, and that introverts were 0.37 times as likely to respond. (If in fact that's what they found.)

If that was the case, I don't see why their conclusion in the abstract wasn't that both were good predictors, but that age was a teeny bit better. Confusing. I was reading the numbers from a zero baseline, in which any positive number would indicate a positive correlation with response to the drug, rather than a baseline of 1; so, in effect, I took 0.37 as a percentage...

I guess I shouldn't try to analyze this stuff while watching a football game.
Evan Thomas is offline  
Page generated in 0.03822 seconds with 7 queries