On 19 Feb 2004 18:52:35 -0800, [EMAIL PROTECTED] (Michael
Eglinton) wrote:

> Hello all
> 
> I work for an organisation that receives counts of all 'notifable'
> diseases from around NZ.
> 
> I would like to compare this years figures with last years and in the
> past we have used the Mantel-Haenszel chi-square test to test if there
> has been a linear relationship between years (thus indicating a
> change).  However we have a problem with this test when cell counts
> are small.
> 
> But my question is 
> 
> Should I even be using this test at all - to my mind we are using
> population data so any differences are real and therefore significant.

The trick is this:  "Population differences"  are real, so long 
as it is *administrative* -- you only care about who wins the 
vote, or how many shoes are needed.  The minute that 
you start thinking about *meaning*, you have to bring in 
some imagined 'universe'  of comparisons.  

Is there a declining trend?  - Yes, if it measures that way.
Is the decline something that would happen by chance?
 - If you ask the question about 'chance', then you are going
to say something more meaningful than the previous  'Yes'. 
It might not be *enormously*  more meaningful, but the 
decline meets a minimal test for being 'interesting'  if it
*is*  larger than chance.

>  We do have non-sampling error in our estimates (e.g. people who do
> not go to a doctor, some diseases are not always recorded because of
> the number of cases received e.g. campy in Auckland) but I do not
> believe that we have a good handle on the size of this error and it
> may be similar from year to year.

Is there any change  to *explain*?  I hope that you don't
have a politician on hand who merely wants to claim that
there was a decline, and the Party deserves the credit, for
being there.  

Statistics provides a tool, a way to state more concretely
something that our intuitions can suggest.

By the way, I wonder how complete the record is, for
'notifiable diseases'?  - in NZ, or anywhere?
Do the doctors properly report all that they see, 
and do the doctors *see*  all, the cases, and do
the doctors successfully *diagnose*  all the cases
that they see?  

For some diseases, I'm pretty sure that the first 
success of a public health campaign  will  be a
big increase in the number of reported cases.  A 
measured decline may indicate that reporting is worse.


-- 
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
"Taxes are the price we pay for civilization." 
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to