In article <Ko_m9.11882$[EMAIL PROTECTED]>, <[EMAIL PROTECTED]> wrote:
>Assumptions: > >1. Causes are uncorrelated and form a factorial manifold of effects. > >2. Causes are linear and fall between r= +/-.95 and +/-.30. > >3. There is more than one cause. > > CR works because (in the additive model) the causes correspondig to the >extremes of the effect are positively correlated. In the mid range of the >effect, the causes are negatively correlated. This is called polarization. I haven't been paying much attention to this thread, being turned off by the uncivility. (I'll leave it to each participant to determine for themselves whether this adjective applies to them or to their opponents.) However, seeing as you've nicely summarized things here, could you continue and explain under what circumstances you think a method that is based on these assumptions would be useful? Presumably, you are intending it to be used to analyse observational data - for a well-designed experiment, what causes what isn't an issue. So how do you arrange for the distribution (which you don't control), to be uniform? Similarly, how do you arrange for the causes to be uncorrelated in this distribution? And how do you know that there is more than one cause? Finally, if you did somehow manage to arrange all that, why would you need CR to determine which are the causes and which is the effect? It would seem that you just need to compute correlations. The uncorrelated variables would be causes, and the one that correlates with other variables would be the effect - if the assumptions are true. Radford Neal ---------------------------------------------------------------------------- Radford M. Neal [EMAIL PROTECTED] Dept. of Statistics and Dept. of Computer Science [EMAIL PROTECTED] University of Toronto http://www.cs.utoronto.ca/~radford ---------------------------------------------------------------------------- . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
