Data Entry Query

[Panza, Robin]

>>>From: Roger Hyam [mailto:roger at HYAM.NET]
3) Databasing *opinion* is of limited value and is really only a note on
the 'main' database record for a specimen. It may be very important from
the point of view of the output of the taxonomic process but not for the
number crunching to facilitate the process - <<<

As long as the database is being used in-house, I agree.  However, the main
reasons for databasing specimen data is to be able to quickly and easily
analyze the data, and to share the data with others.  When someone wants to
know what we have from location X (or time Y, or collected by Z, or
whatever), I send him a file from the database.  Some people know that if
there's an odd record, they need to contact me to either double-check the
specimen or send it to them for examination.  However, more people are using
the *data* instead of the *specimens*, and are not using the data with
appropriate skepticism.

There may never be a final answer to which genus or family this taxon
belongs, or in which subspecies (or whether subspecies exist).  To that
extent, the name attached to the record will always be an opinion.  However,
if the names in the records are too far off, there can be some serious
problems when policy-makers blindly use the data we send them.  Decisions
about which pieces of land need to be preserved and which can be paved over;
about which species can be left to their own devices, which need extra
protection by listing as threatened/endangered and which need to be
"managed" are being decided based on the data we export.

Many museum people are afraid government policy will be set based on grossly
inaccurate data.  This is why the name put into the record can be so
important.  Yes, it's just a handle, when the database owner is also the
specimen owner.  But as soon as the data are shared, it becomes politically
important.  I think this is what paralyzes some curators who are considering
data entry.

On the other hand, data "entry" shouldn't be considered complete till the
data are verified.  This includes making sure the tag and the record agree;
making sure the data are standardized, so that records can be reliably
retrieved; and making sure the name in the record is a reasonable
approximation of the current beliefs about how life should be named.  That
last, we agree, is unlikely to ever be fully settled.  However, we can be
sure it's a sparrow, not a vulture.

My recommendation is to find your own compromise level in data entry.  Some
shortcuts can be taken to get the data in, but remember that the data can't
be reliably gotten out till it's been cleaned.  The sloppier the entry, the
longer the cleaning, and it's not a usable database till it's reasonably
clean.  However, averaging 3 hours per specimen (allowing for taxonomic
decisions, looking up the geography, reading the journals, whatever) means
the data will never get in.  It's not a usable database till it's got data
in it, either!

OK, I work with birds--only 9,000 species, and 100,000 specimens is a
world-class collection--and this biases my view.  For people working with
taxa that are poorly understood and highly diverse (e.g., any
invertebrates), I would think setting up the records in an electronic
database would still be worthwhile, even if the name is not yet known.  As
long as there is a means of associating the record to the specimen (easy
with birds--write a catalogue number on the data tag), you can go back and
add, or change to, the currently correct name and authority information.
Yes, prioritize your data entry scheme--catalogue as specimens leave or
return from loans, or catalogue as new material arrives, or whatever.  But
it's dismaying to see how many people are paralyzed by the magnitude and
complexity of the job.

Robin K Panza
Section of Birds, Carnegie Museum of Natural History
4400 Forbes Ave., Pittsburgh  PA  15213  USA
phone:  412-622-3255;  fax:  412-622-8837
panzar at carnegiemuseums.org