Duplicate specimens

From systematics

Jump to: navigation, search

Excerpt from email to Sharon and Bil, 27 July 2009

(ahipp 20:10, 29 July 2009 (UTC))

Bil wrote: What mechanism will we put in place to deal with duplicate records? (as opposed to duplicate specimens, which the annotation tool would vet; I recall talking about this but don't recall the resolution)

AH wrote: [This] question we... haven't discussed yet, and this could be complicated, because different sources for the same record are likely to send the data in different forms, so that it may not be obvious whether the records are identical. Probably the safest idea is to nest data specimen to three levels:

  1. collection (this is the collection event, e.g. Alverson & Iltis 20,199);
  2. specimen (this is the physical specimen, e.g. WIS234999, F1348888, K5488827); and
  3. instance (this is the instance of the specimen record, at least one per specimen, and more if there are duplicate data sources)

Then, we don't worry about whether the five instances of Alverson & Iltis 20,199 have identical data; we let the researcher make sense of this his or her ownself. Annotations would most often apply only to level 1 or 2; a person may want to flag a data copy (at level 3) as corrupt, but the final data source for all instances of a particular specimen are a single specimen. Now, that doesn't rule out annotation at level 3, but if we allow it, we need to guard against allowing people to annotate an instance when they mean to annotate a specimen or a collection (i.e., all specimens within a collection).

Personal tools