The Structure of Objectivist Studies - Evaluation Methods in Biomedical Informatics - page 107

2. A new clinical workstation is introduced into a network of medical
offices. Logs of one week of resource use by nurses are studied. The report
enumerates sessions on the workstation,

broken down into logical

categories.
3. A number of computer-based information resources have been
installed to support care on an inpatient service. The information resources
log the identity of the patients about whom inquiries are made. By chart
audit, the investigators identify a number of clinical characteristics of each
patient's clinical problems. The investigators then study which characteris-
tics of patients are predictive of the use of resources to obtain further infor-
mation about that patient.
4. A researcher compiles a database of single nucleotide polymorphisms
(SNPs), which are variations in an individual's genomic sequences. The
researcher then examines these SNPs in relation to diseases that these indi-
viduals develop, as reflected in a clinical data repository.
5. Students are given access to a database to help them solve problems
in a biomedical domain. By random assignment, half of the students use a
version of the database emphasizing hypertext browsing capabilities; half
use a version emphasizing Boolean queries for information. The proficiency
of these students at solving problems is assessed at the beginning and end
of the second year of medical school.

Meshing Aims, Approaches, and Designs in
Demonstration Studies

This chapter closes with a description of the steps required to develop a
demonstration study. This discussion is a preview of the much more com-
plete exploration of designs for demonstration studies in Chapters 7 and 8.
It is included to alert the reader to the complexities of demonstration
studies, particularly those carried out in ongoing patient care, research, or
educational settings. In Chapter 12, we will return to issues of study plan-
ning, as well as how study plans are formally expressed in proposals and
evaluation contracts.
Planning a demonstration study can be seen as a three-stage process, as
shown in Figure 4.4. First, one must carefully define the problem the study
is intended to address. This step requires eliciting the aim of the study and
the main questions to be answered. It is then useful to classify the study,
based on its aims and the setting in which it will be conducted, using the
distinctions offered in Chapter 3. A study may emphasize structure, func-
tion, or impact, and it may be seen as occurring primarily in a laboratory
or field setting. From this information, a general design for the study—
descriptive, comparative, correlational—can be selected. The needs of
studies that address resource structure and function often are satisfied by
an objectives-based approach (Is the resource performing up to the level