[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: County Data

On Wed, 19 Dec 2001, John Williams wrote:

> Dr. Cohen,


> You stated below, "Statistics is also an important issue as many

> counties have less than 5 cases per year."


> I think that was one of the points Field, Smith and Lynch made in

> their Health Physics Forum paper about your study.  If you only have

> 5 cases per county, the statistics are a problem for you in an

> ecologic study.How can you represent the whole county with such

> scant data per county.

	--I said 5 cases per year. My 1995 paper summed cases over a 10

year period, and my 2000 Update paper summed cases over a later 15 year


>  The mortality data is not quality data and

> you have very little data for each county.

	--Summing over 10 and 15 year time periods gives a fair amount of


  The major limitation of

> ecologic studies is the quality of data and you just admitted that

> the data is lacking.  That is why you should start by focusing on the

> states that have SEER quality data and lots of counties like Iowa.

	--The number of cases in the SEER data is much less than in my

data because they have only 5 years to sum over

> Just because you have a lot of counties (and therefore data), that

> does not mean you have more confidence in your data if your

> underlying county by county data is suspect.  These errors do not

> average out over the United States, but rather get propagated with

> the addition of each county.


	--Calling data "suspect" is misleading; you should say "has

statistical limitations". If you can explain your last sentence above.

please do. Every data point from any study has statistical limitations, so

you are saying that the more data points you have, the less reliable are

your results.


You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line.