[Cross-posted to "archives-nra" and "Archives & Archivists" lists]
 
I have recently been working on a feasibility study for documenting and
digitising the collections of John Ruskin, which are going to be set up
in a new purpose-built library at the University of Lancaster. The
collections include a few thousand each of books, manuscripts, letters,
drawings and watercolours, photographs and some three-dimensional
objects. I have completed my draft report on that project, but it made
me consider a wider issue about estimating times which might be an
interesting topic for wider discussion.
 
Part of this project required me to estimate the time required for
documenting these collections, and I have been seeking information on
how archivists make such estimates. It seems that most people do it very
intuitively, getting a general impression of the size and complexity of
a collection and comparing it with collections they have dealt with in
the past. "You should allow a couple of years for that job" is a typical
answer.
 
Many people are making bids for funding for archival projects, and I
wonder whether anyone has tried to develop a formula or rule of thumb to
decide how much staff time it is reasonable to ask for to catalogue a
given collection.
 
Is it possible, in other words, to elicit from experts the basis on
which they make these estimates? (This is a common problem in developing
computerised "knowledge-based systems", but I am not suggesting that
anything like that need be used here.)
 
I have been able to gather a small amount of raw data from a few
specific projects, for number of items catalogued and time taken.
Allowing 220 working days per year and 6 actively working hours per day,
gives a range of 10 to 30 minutes per item. (I'm surprised that the
range is not wider).
 
If we take this as a starting point, we can then consider the effect of
factors such as:
 
Complexity of material: "C"
From simple letters with clear handwriting and routine subjects, to
complex, illegible, philosophical or technical manuscripts.
 
Depth of cataloguing or listing: "D"
Information recorded about each item might range from the name of the
writer and the date to a digest of the content with many name and
subject index terms.
 
System used: "S"
Free text descriptions or structured records. Availability of authority
files for name, place and subject access points. Ease of use.
 
If we give each of these three factors a rating as follows
 
Easy            0 minutes
Moderate        5 minutes
Difficult       10 minutes
 
We might then estimate the overall time to be allowed per item as
 
                10 + C + D + S minutes
 
This gives a range from 10 minutes per item for the simplest material to
40 minutes per item for complex material, thoroughly catalogued in a
difficult system.
 
I assume that preparing higher level records, for groups and series,
will be included in these times, though separate allowance might have to
be made for these if there are a lot of them in relation to the number
of lowest level records.
 
Of course this is a simplistic approach, but I am just suggesting it as
something to consider and build on. Does anyone have ideas for refining
it, by adjusting the numbers, adding other factors, or combining them in
a different way? It would be particularly interesting if anyone else has
data from projects they have completed which they could plug into this
model to see how well it fits, and if not why not.
 
Any thoughts?
 
Leonard Will
 
--
Dr Leonard D Will                    Tel: +44 181 372 0092
Information Management Consultant    Fax: +44 181 372 0094
27 Calshot Way, ENFIELD, Middlesex   Email: [log in to unmask]
EN2 7BQ, United Kingdom