normal value

[GoogleCodeExporter]

Peter Robinson says:
>6) Normal value.
>==>> Perhaps it would be better to write "Normal finding" so that
>one could include X-rays in this (it is awkward to say normal value
>for a chest X-ray)

Original issue reported on code.google.com by albertgo...@gmail.com on 27 Jul 2009 at 3:42

[GoogleCodeExporter]

Normal finding and normal value refer to different things.  I would keep normal 
value
as is and consider adding a definition for normal finding to OGMS.

Original comment by hoga...@gmail.com on 2 Aug 2009 at 11:24

[GoogleCodeExporter]

Issue from Melanie Courtot:
> 2.a. normal value
> - why is not asserted under ICE?
> - did you consider importing the IAO, and if yes, should normal value be
> asserted under measurement datum?

Original comment by albertgo...@gmail.com on 5 Jan 2010 at 7:16

[GoogleCodeExporter]

The definition says quality. But not all assays are of qualities, are they? 
E.g. clotting *time*.

Original comment by alanruttenberg@gmail.com on 9 Mar 2010 at 5:48

[GoogleCodeExporter]

'normal value' is currently defined as: "A value for a quality reported in a 
lab report and asserted by the testing lab or the kit manufacturer to be normal 
based on a statistical treatment of values from a reference population."

I agree with Melanie Courtot that normal value should be a child of 
'information content entity' instead of being a sister-class of ICE.

Regarding Alan Ruttenberg's suggestion that the term 'quality' should be 
omitted from the definition of 'normal value', please answer the following 
questions:

1. Does an (successfully completed) assay always have a measurement datum as a 
specified output?  While it is only asserted in OBI that an assay has an 
information content entity as a specified output, I think that this further 
claim is warrented and it even seems to be implied by the fact that 'measuring' 
is an alternative term for 'assay' in OBI.
2. Is every measurement datum the specified output of some assay?  This seems 
intuitively correct and, assuming that there cannot be an unintentionally 
created measurement datum, it seems to follow from the definition of assay: "A 
planned process with the objective to produce information about some evaluant".
3. Is a quality the only sort of thing that can be measured?  This, I think, is 
a difficult question to answer.
4. Is a normal value always about a (or a set of) measurement datum?  This 
seems implied by the current definition of 'normal value' but is not made 
explicit.

If you answer "yes" to all 4 questions, I think you are committed to accept the 
claim that a normal value is always about a value of a quality.  If, however, 
you answer "no" to at least 1 of these questions, then the definition may need 
to be revised.  As Alan's example suggests, question 3 is, I think, the best 
candidate to say "no" to.  If this approach is taken, the definition of 'normal 
value' should be revised as: "A value for a quality, entity1, or entity2 
reported... ".

The answer to question 3 is a matter for OBI, but, assuming that the answer to 
4 is "yes", OGMS can sidestep the issue by changing the definition of 'normal 
value' to: "A value for a measurement datum reported... ".

Original comment by Alexande...@gmail.com on 12 Jun 2012 at 6:22