Summary: Putting It All Together - Foundations for Designing User-Centered Systems - page 396

designs based on a new analysis of the problems. Similar effects of transfer will
happen after computing Fitts' Law a few times or comparing mouse movement
times to keystroke times.
To facilitate learning, shared representations need to be made available within
projects, across the system development team and across the different stages of the
design process for a single project. We extend this, though, to re-using repre-
sentations across multiple projects. This may mean that the shareable represen-
tations have to be able to reflect different levels of abstraction (based on their
dependency on their context of use) so that they can be learned from and re-used.
The corollary of this is that those people developing the shared representations are
likely to need to document the shared representations with information about
context.
In addition to sharing representations within projects, it may be possible to re-
use these representations across projects. This would increase the value of these
representations, and achieve the re-use that is called for by Pew and Mavor ( 2007 ).
This would also broaden the audience for the shared representations to include
managers and developers on future projects who may be able to learn about
potential risks and which stakeholders to consult.
14.4.6 Summary
The RD-ICM model is just one way of creating systems that are useful and safe. It
involves identifying the stakeholders—including the designers—and coming up
with a system through satisficing their different requirements, finding the best
solution matching their constraints, including the cost to find better solutions as a
constraint. If stakeholders' needs are not adequately supported, they may not
participate in the system development process fully, and may either obstruct the
process or even refuse to accept the final system. An advantageous side effect of
using the RD-ICM is that it can provide new ways of summarizing HCI methods
and, more broadly, Human-System Integration (HSI) as an integral part of design
and development.
In the vast majority of systems, a failure to consider the users as stakeholders,
and a lack of understanding of the abilities, capabilities, and limitations of users
will constitute a risk to system development and use. The main exceptions are
where the system being developed is not novel, so the risks in these areas are likely
to have been considered and dealt with on other projects. Where there are risks,
however, work has to be done to manage them appropriately, given the available
resources, and to balance all of the identified risks (user related risks will need to
be balanced against technical risks, for example).
Managers and developers are usually familiar with the technical risks associ-
ated with software and hardware. To make them more aware of user-related risks,
they are likely to require some level of education and training about the capa-
bilities and limitations of users. This education and training should take them to a