Challenges of Evaluation in Biomedical Informatics - Evaluation Methods in Biomedical Informatics - page 16

FIGURE 1.5. Tower model. (Adapted from ref. 28 the Journal of the American
Medical Informatics Association , with permission.)
information science, cognitive science, decision science, statistics, and
linguistics. Other fields supporting informatics are applied more in their
orientation, including software and computer engineering, clinical epi-
demiology, and evaluation itself. One of the strengths of informatics has
been the degree to which individuals from these different disciplinary back-
grounds, but with complementary interests, have learned not only to coexist,
but also to collaborate productively.
However, this diverse intellectual heritage for informatics can make it
difficult to define creative or original work in the field. 28 The “tower” model,
shown in Figure 1.5, asserts that creative work in informatics occurs at four
levels that build on one another. Projects at every level of the tower can be
found on the agenda of professional meetings in informatics and published
in journals within the field. The topmost layer of the tower embraces empir-
ical studies of information resources (systems) that have been developed
using abstract models and perhaps installed in settings of ongoing health
care or education. Because informatics is so intimately concerned with the
improvement of health care, the value or worth of resources produced by
the field is a matter of significant ongoing interest. 29 Studies occupy the
topmost layer because they rely on the existence of models, systems, and
settings where the work of interest is underway. There must be something
to study. As will be seen later, studies of information resources usually do
not await the ultimate installation or deployment of these resources. Even
conceptual models may be studied empirically, and information resources
themselves can be studied through successive stages of development.
Studies occupying the topmost level of the tower model are the focus of
this topic. Empirical studies include measurement and observations of the
performance of information resources and the behavior of people who in