29176: Leiderman re: Lancet study: good science? (fwd)

[Bob Corbett <corbetre@webster.edu>]

From: leiderman@mindspring.com


dear Readers:

the proof of good science is the ability to replicate experiments and get the same result.  for disbelievers, then, I think it'd be relatively easy to knock on different doors, ask the same (or more) questions, and then compare and contrast the conclusions with the original survey.  probably, this could be accomplished in less time than it's taking the factions to go at each other's throats.  so far, no one has suggested how the results could lead to new policies, implementation and enforcement.  does that sound too much like work?

further, instead of all the time and energy that's going into casting blame and bias, those who disbelieve the results, per se, should just state what additional evidence, method or quality control would persuade them that the results are fairly representative.  given adequate time and money, the survey team would probably like to do a follow-up.

as an aside, I don't quite see the fascination with using GPS for deciding which doors to knock.  good surveys from random samples have been done without electronics and earth-orbiting satellites for centuries.  in fact, there's no reason why the whole country couldn't/shouldn't be surveyed periodically.  Haitians might like to get their troubles off their chests; and others could get to know their country a little better.

thank you,

Stuart Leiderman
leiderman@mindspring.com