Nested character states
una at LANL.GOV
Tue Sep 18 16:52:33 CEST 2001
- From: Jaan Palisaar
> If you code a species as living in acid grasslands but your end-user
> just remembers (or the herbarium lable only states this) that he found
> it in the open, he will be lost. But if you defined acid grassland as
> a subset of open lands the species will still be included. And
> entering data should be easier since once you defined the hierachy of
> character states you only have to enter a low level character state
> and have everything above automatically included.
Coding and using character taxon data bases are two separate issues.
First, coding. It may seem easy to code a dependent character and have
the inferred states of controlling characters set automatically. But
that introduces *false* data. The observation that your species occurs
in acid grasslands does not tell you whether or not the same species
occurs outside grasslands.
Next, using. Why should the user be lost? For identification, the user
should code all states that *might* apply.
Los Alamos National Laboratory, Mailstop K-710, Los Alamos, NM 87545
More information about the delta-l