The purpose of the workshop is to develop a research agenda for
the future research on the foundations of spreadsheets. To
facilitate the exchange of ideas, the workshop is organized into
several interactive discussion sessions. Invited speakers will
each provide a perspective to get the discussion going, and each
participant will contribute from their views and expertise. Two
of the invited speakers are:
- Alan Blackwell, University of Cambridge, UK
- Margaret Burnett, Oregon State University, USA
The workshop discussion will be guided by a number of
frame questions shown below.
The workshop will start with a session in
which participants introduce themselves and their research and
can give a position statement.
The following collection of frame questions will help to guide
the workshop discussions. Position statements (see below) would
be most useful if they addressed one or more of these frame
What are the essential attributes or dimensions of spreadsheet
usefulness? What are possible ways of measuring effectiveness?
How can we employ "people first" design approaches for new
Spreadsheets produce numbers (and charts), but eventually
spreadsheets are often used to support business decisions.
What mechanisms would help people to extract more useful
information from a spreadsheet to support such decisions?
Is it enough to rely on tools for this purpose? Maybe we need
specialized design methodologies and strategies?
Programming (language) perspective:
Spreadsheets encourage copy/paste as a reuse mechanism.
Programmers have typically used abstraction. Is there something
in the middle? For example, copy/paste retaining provenance?
Spreadsheets encourage an element-wise scalar view of the world.
Functional languages, for example, encourage the use of so-called
"bulk operations", such as array formulae, but they make bigger
demands on end users. Are we just stuck with this dichotomy, or
is there something in the middle? Can the demands of "bulk operations"
be reduced or eliminated? Can they be characterized, as a first step?
One of the dominant uses of spreadsheets is as small databases.
The integration of spreadsheets with database systems is rather
loose, requiring lots of transformations and data redundancies.
What would a combined spreadsheet and database look like?
In traditional programming languages, concepts, such as testing
and type systems have been used very successfully to enhance the
quality and reliability of software. What does it take to
transfer these ideas to spreadsheets so that they are considered
useful (and not constraining) tools/concepts by end users?
How can the idea of quality be promoted to end-user programmers?
("Getting the spreadsheet to compute correct values deserves at
least as much attention as getting the formatting right.")
You can register for the workshop on the following web site.
(See the second pull-down menu in Section 3 of the form.)
All registered participants are encouraged to send a position
statement (1 paragraph to 1 page) by email to the
These position statements will be included in the workshop
proceedings that will be distributed during the workshop. The
position statements will provide an additional stimulus for
discussions. Position statements will be most helpful that
address one or more of the frame questions. The deadline for
submitting a position statement is September 3, 2004.