Hi All

At 12:32 1999-05-21 +0100, Norman wrote:

A lot that I would agree with, given the similarity of the task. I'll comment 
as each point is raised.

>Glasgow Museums are seven years into such a task.  The end data is, er,
em, well
>that is not firmly inked in yet.  (If it ever is finishable.  There is always
Amen to that. Here at the Ulster Museum we have been at this since early
1990, 
after the establishment of the Registry in 1989. We are on just over
460,000 records
We already had data in Modes records that had been done on a pilot basis.
These
were patchy and depended on which curatorial departments had bothered to
try and
document their collections. There had been some grief with an early version of
Modes, so it was not a trusted program. We found that the newer PC version
was much
more reliable so it was used from then on in Registry. 

I was able to write data output routines to transform Modes data to the
Batchin format
of Minisis (the Quixis DBMS). 

>Methods of data entry we used:
>1.      Salaried staff (on 3 year contracts), based in the Documentation
office
This has been our main data entry method. We have had various staff over
the years.
Of the initial three people, one remains with us, but her job has changed
due to
reorganisations, so she only does sporadic data entry for us nowadays. 
We currently have 3 data entry staff, but now they are also doing cleanup
of the 
database and also spot checks of the collections, based on data retrieved
from 
the database.
One is also currently doing reconciliation of data from a sample accession
register
with data on the database. Our staff are on 1 year renewable contracts at
the moment. 
This is not a satisfactory situation, but is the way it is.

>2.      Short term contracted in staff, based in another part of the building
We have not taken in people on contracts on that basis. However we have taken 
in volunteers. These have been variable, depending on the level of interest
in 
and knowledge of Museums the degree of aptitude for the work, and the length 
of their stay. Having the right sort of work for short stay fixed term
volunteers 
is a must. Something small and finite which has been broken off from the
overall job is the best for that sort of volunteer.

>3.      Contracted out staff (to an ex-member of staff in their own home)
We have not done this.

>4.      Data entry company
We experimented with an outside company for one of our photographic
collections 
which was well documented in a publication. We gave the firm a version of the 
publication to input along with a format on what format the records were to
follow. 
This was to be the Batchin format so the records could be directly entered
to Quixis.
We thought that this well-defined subset of the collection, with simple
well-structured data took too much time and effort on our part to get back
useful 
records. We have not repeated the experiment.

>There is no doubt in my mind that these are listed in order of
effectiveness. 
>Having staff in the same room makes life much easier.  Communication is
the most
>important factor.  The data manager is able to give information on data
>standards, solve unforeseen problems, etc.  The data entry staff benefit
because
>they feel part of the project.  You can also monitor their work better! 
Amen to all that.

>We did give a data entry company a lot of Natural History registers to
input. 
That was brave.

>They did do a lot of records over a number of months (25,000 in 4 months
>approximately) but on their return I had to then convert them into our
database.
> We also did have to do a lot of editing after, in order to tidy up the
data to
>our standards.  
Was this editing automatic and did this take away time from other work?

>by documentation staff).  The enthusiastic, PC literate curators were able
to be
>given more involvement, but you to be on your toes to keep your data
standards
>intact!  
Now I will bring this one up as our early Modes records were all done by
curators.
All that Norman says applies. I saw so much variation in some of them that
a lot of 
cleaning up was needed. 
Later, when we knew better what we were doing, we could lay down Modes
templates for
curators to follow. We foud that the Natural History types could cope with
this better
as their data is inherently better structured anyway. 
That having been said, we have had one curator who must have input about
15,000 coins
and other history records for us, mostly on a 286 PC, using Modes 2 and
later Modes 4.
When he was doing Coins the records were mostly very good. Oddly enough the
quality
went down when he had to document other types of object. 
So we have had some degree of success with curatorial record input, but it
*does*
need a tight hand on the reins.

One last method that we were lucky enough to benefit from, was data held by
the MDA
which was done on a pilot basis for our Fine Art and Mineral collections.
It was 
extracted into Modes tagfile format ready to be loaded into Modes. Mind you
anyone 
who had data held by the MDA would have recovered it long ago.

So there are some aspects of the Ulster Museum Documentation Project. 

Michael
-- 
Michael Comiskey, Systems Manager, Ulster Museum
[log in to unmask]



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%