Gene, I believe I addressed that issue with my very first post on the subject. I'm happy to re-post it.Gene wrote:Shall we also compare their public statements AFTER their various stints at grad. school, Bill? When we approximate, we need to use ALL the available information, no?
We went into the whole speech deflect thing, southern accents, stuttering, etc. We went into judging someone's intelligence and character by their speech patterns.Man, I'm glad I don't have press following me with a microphone 24/7...
George the silver-tongued he is not.
I can keep going with this, just for entertainment.
A pattern recognition exercise, whether it be cerebral or digital, involves a variable selection step so as to determine the drivers (principle components) in the system. All information is considered at once, and correlation amongst all the variables is determined. Those variables (such as speaking ability) that don't independently contribute to the variability of the dependent variable (in this case, intelligence) in a statistically significant manner are removed. A classic approach to such using modeling software is stepwise regression. There are other more sophisticated methods.
Or...you can use a little common sense and do the same thing.
OK?
Sometimes you just have to call a spade a spade. If you don't like Bush, just say so. There's no harm in that. He is who he is. One of his strengths (consistency) will mean that he will not be liked by some people all of the time. That's cool.
It reminds me a bit of people who say cats are stupid. Just because a cat won't fetch for you doesn't mean it's stupid. It could be that it's smarter than the person trying to make it play the stupid game, no?
Two and a half more months of this. Oye!
- Bill