Ecological Inference

[js_dukelow@ccmail.pnl.gov (James S Jr Dukelow)]

Professor Field, the BEIR VI committee, and Greenland and Morgenstern  
might profit from reading Gary King's book, _A Solution to the 
Ecological Inference Problem: Reconstructing Individual Behavior from 
Aggregate Data_, Princeton University Press, 1997.  King, a professor 
of government at Harvard University, received the American Political 
Science Association's Gosnell Award for the "best methodological work 
in political science in 1995-96" for the work described in his book.  

Actually, I am told that the bulk of the BEIR VI committee deferred to 
the three epidemiologists with regard to Cohen's study and that the 
report considers Cohen's study and others in an appendix, but does not 
contain any substantive refutation of the study, but rather just a 
restatement of the epidemiologist's assertion that it is flawed.  That 
characterization is hearsay, since I have not seen the full report (I 
can't decide whether to pop for the $45 or so the NAS is charging).

Ecological inference is an example of what mathematicians call an 
inverse problem.  Inverse problems are usually ill-posed and require 
additional regularizing assumptions to enable a unique solution.  
Although difficult to solve, inverse problems are important in a wide 
variety of applications, and are routinely solved in those areas: 
tomography, seismography, certain types of non-destructive examination 
techniques in materials science, and maximum likelihood estimation in 
statistics.  Rather than use the information in large ecological data 
sets, some epidemiologists prefer to chant "It's only an ecological 
study, therefore we don't have to ...".

I'll be happy to email anyone interested a preprint of the review I 
wrote of King's book, which appeared in the November 1997 issue of 
Health Physics.

Best regards.

Jim Dukelow
Pacific Northwest National Laboratory 
Richland, WA
js_dukelow@pnl.gov

These comments are mine and have not been reviewed and/or approved by my 
management or by the U.S. Department of Energy.