Information Technology Reference
In-Depth Information
representation. Whether it is optimal for decision makers to choose among forms and depictions
based on their assessment of problem novelty or to have this done without user input using within
subjects experiments is also worthy of investigation.
This research presents two treatments for problem novelty based on dividing the problems into
the first four and the next six problems. But this treatment is actually a combination of task-system
treatment, since both the task and the system are more novel in the first four problems than in the
next six. More complex experiments, where treatments are divided into four, with the task and/or
system systematically varied among the subjects, could be conducted. These would further com-
ment on the task-system aspect of decision making, and on how it affects decision-maker calibra-
tion, by determining how task and system combinations can affect problem novelty—which, in
turn, will alter the properties of the optimal DSS dialog design to achieve perfect calibration.
Larkin and Simon (1987), Bauer and Johnson-Laird (1993), and Speier and Morris (2003)
report that diagrams improved decision quality. The results reported here demonstrate that visi-
bility diagrams also improve user calibration. Collectively, these studies suggest that visibility
results in better decisions and that decision makers are better calibrated as to the quality of their
decisions, at least in the environment in which the experiments were carried out. Specifically,
when problems are new and novel, visibility improves user calibration.
The process of refutation is central to science, and this experiment is but one of many with
varying methods, designs, instantiations, and mixes of expressiveness, visibility, and inquirability
required to provide sufficient evidence of the theory's utility. It is in the collective that greater
understanding emerges. As a first study, the treatments used in this study are modest approxima-
tions of the theory. However, before great time and effort is expended, this first study seems pru-
dent. Researchers can now invest their time, effort, and resources with greater confidence.
Further research is needed to investigate the theory with much greater specificity, including
variations on the mix of expressiveness, visibility and inquirability. Follow-up studies will neces-
sarily be more detailed, incorporate more of the theory—especially in terms of an “effective mix”
of expressiveness, visibility, and inquirability—and will eventually need to move out of the labo-
ratory and into the real world. No matter how “realistic,” all laboratory studies are contrived
approximations of reality. Nevertheless, this first study is a necessary step in substantiating the
theory of DSS design for user calibration.
APPENDIX 5.1. PEOPLE AND PLACES QUESTIONS IN
EXPRESSIVENESS TREATMENT
1.
When the event is occurring:
Sally is in Ann Arbor or Jack is in Pittsburgh (or both)
Jack is not in Pittsburgh or Jason is in New York (or both)
The event is occurring. What follows?
a. Sally's in Ann Arbor/Jason's not in New York
b. Sally's in Ann Arbor/Jack's not in Pittsburgh
c. Sally's not in Ann Arbor/Jack's in Pittsburgh
d. Jason's in New York/Sally's not in Ann Arbor
2.
When the event is occurring:
Jack is in Pittsburgh or Carmen is in Chicago (but not both)
Jason is in New York or Carmen is in Chicago (but not both)
The event is occurring. What follows?
a. Jack's in Pittsburgh/Carmen's in Chicago
Search WWH ::
Custom Search