30037: Pershing (reply) Re: 30031: Morse (ask) Lancet Study (fwd)

[Bob Corbett <corbetre@webster.edu>]

From: tjpershing@aol.com

The Lancet study was done with a methodology that produces statistically 
surprising #'s, the same methodology that produced the highest #'s for Iragi 
civilian casualties and deaths during the US occupation. As social science it 
appears to be a solid system- but it still relies on the responses of the 
interviewed with no collaborating evidence- and this may be its fault line- as 
has been established in other methodologies it doesn't take long for a person 
to see what kind of answer the interviewer is looking for and follow the 
natural human tendency of exaggerating or fabricating to "please" the 
interviewer. I have not seen the questionnaire so I don't know if the questions 
on violence were masked by other general questions, such as opinions of 
governance, schooling, health etc. to throw off this tendency.

 In the end, the numbers people used in reports were statistical averages drawn 
from a wide range of possible levels- all based on a very few number of actual 
responses projected across the PaP region- at that level the effect described 
above could have played into the very high purely statistical projections 
produced. I say this because in talking with and looking at reports from all 
manner of other groups the Lancet study appeared high- it therefore needs to be 
scrutinized because if it is accurate its a bombshell-

 Unfortunately, it appears the authors missed a good chance of answering this 
question by running a parallel study of the two year time period before 
Aristides departure- which would have at least given us a percentage of change 
in these events regardless of the levels of statistical projections.

 As for the Authors duplicitous use of footnoting her own work under a 
different name and hiding her previous affiliations from readers of the report 
and the Lancet- this is a big social science no-no, and put in doubt so much of 
what still might turn out to be an important study.

 I suggest a social scientist with no previous association with Haiti reweiw 
the methodology and re-run the study with no outside interference-