Razib has an essay on the limits of certitude in science, which is saying something coming from a “three cheers for reductionism” type of fellow (which I am also). Meanwhile Andrew Gelman points out some heretics who dare to blaspheme The Great God Randomized Trial. Among them is James Heckman, whom you may have heard about recently.
Le Carrefour de la Sagesse casts some doubt on a theory I was groping toward in an old post. Kudos to him (or her) anyway.
The title of this post stolen in part from Jake Prescott.
January 14, 2009 at 8:46 pm
I once saw Rubin blaspheme the Great God of Randomized trial’s distant cousin: the intent to treat principle. It was a biostatistical riot. This is fun stuff and statisticians need a new source of tension since the old one (Bayesians vs. Frequentists) is boring everyone to tears.
January 14, 2009 at 11:35 pm
I had never heard of that principle before. I’ve never actually taken a statistics course either (just probability).
January 15, 2009 at 9:01 am
The intent to treat principles says that all experimental units that you intended to treat,i.e.,that were randomized, must be used in all calculations. For example, you do not get to ignore randomized patients in a clinical trial if they failed to receive any treatment since this could lead to bias.
January 15, 2009 at 4:48 pm
Thanks for the shout-out, TGGP. By the way, do you happen to know any primary literature on the cognitive profiles of Ashkenazi Jews and Eastern Asians? I’m looking for linkable papers on Google that explain why, or how, Ashkenazi Jews excel at verbal g-components and East Asians excel at visuo-spatial g-components.
January 15, 2009 at 7:29 pm
You should ask some of the folks at GNXP, they’re much more familiar with the literature than I am.
January 15, 2009 at 8:04 pm
OK, thanks!
March 18, 2010 at 9:06 pm
[…] experiments”. I mentioned some heretics who fail to pay homage to randomized trials here. Among the party-poopers is a recent commenter here, Thorfinn, in “When Numbers Fail“. […]