Tuesday, May 13, 2014

on ACA and the use of statistics...

The problems-- and more importantly, the inherent limitations-- of the ACA (even on paper) render it less interesting to discuss than the pundits imagine. (That's in terms of the economics. Oh yes, the politics of it are quite interesting!)

Aside from making an appropriate but narrow point, this article also points to broader measurement issues:

1a.) statistics are ALWAYS/MERELY a proxy for the state of the world they purport to measure (the questions are whether a stat is a good/bad proxy and what are its strength/weaknesses); and
1b.) thus, statistics are inherently limited;

2a.) the use of statistics is especially unimpressive when univariate and/or cherry-picked-- in trying to describe a complex social phenomenon; and
2b.) thus, people who use statistics in this manner are rubes or demagogues.


Post a Comment

Subscribe to Post Comments [Atom]

<< Home