Information Technology Reference
In-Depth Information
This paper extends, revises, and reports a partial test of the theory of DSS design for user cal-
ibration originally posited by Kasper (1996). Using Walls et al.'s (1992) framework for building
IS design theory (ISDT), Kasper (1996) developed a DSS design theory for user calibration. In
a recent review of literature, Walls et al. (2004) reported that Kasper (1996) was one of the few
researchers who made extensive use of the ISDT concepts. However, two important components
of the framework were not adequately addressed by Kasper. In this paper we extend the original
theory of DSS design for user calibration to address all the components specified in Walls et al.'s
framework. Further we report a partial test of the theory by comparing the user calibration effi-
ciency of the visual computing paradigm with that of the conventional text paradigm over two lev-
els of problem novelty. A preliminary version of these results is reported elsewhere (Ashford and
Kasper, 2003). Here we clarify that reporting and provide more detail and discussion regarding
the study and further studies of user calibration.
To frame the discussion within the broader notion of DSS, we begin by reviewing the general
concepts and components of DSS. This properly frames the theory of DSS design for user cali-
bration as a dialog design theory. Next, we draw extensively upon Kasper (1996) to describe the
theory. We begin by describing the kernel theories. Next we present the properties needed to
achieve the goal of perfect calibration, including meta-design. The laboratory test and test results
follow. Finally, we summarize the paper, highlighting the contributions.
BACKGROUND
The works of Gorry and Scott Morton (1971), Keen and Scott Morton (1978), Bonczek, Holsapple,
and Whinston (1981), Ginzberg and Stohr (1982), and Sprague and Carlson (1982) define the con-
ceptual and theoretical foundations of DSS. Coined by Gorry and Scott Morton (1971), decision
support systems are motivated by decisions and decision making, as opposed to systems support-
ing problem or opportunity identification, intelligence gathering, performance monitoring, com-
munications, and other activities supporting organizational or individual performance.
If we use Simon's (1955) classic four phases of the decision-making process—intelligence,
design, choice, and implementation and control—DSS concentrate on the design and choice
phases (Fazlollahi and Vadihov, 2001). As a field of study, DSS argues that systems built to sup-
port the design and choice phases of decision making present unique challenges, both in theory
and in practice. The term “decision making” implies judgment, personal assessment, and evalua-
tion. In the DSS literature, the amount of decision-making judgment needed to solve a problem
reflects the problem's structure. Problem structure is the degree to which aspects of design or
choice are subject to computerization (Gorry and Scott Morton, 1971). If for any decision, all
aspects of design and choice are structured and therefore computerizable, then no situational
judgment is needed and the problem is said to be structured. DSS are not designed to play a role
in structured problem environments; other forms of information systems, such as transaction pro-
cessing systems, serve these situations. At the other extreme, if no structure can be applied to any
part of the design or choice phases, then what is to be computerized? If nothing can be specified,
what role can the computerized part of the DSS play in the overall human-machine system?
For DSS to be of any use, there must be some part of design or choice that can be computer-
ized while other parts are left to the decision maker's expertise, judgment, and cognition. This
partnership of human and machine dates to the work of Licklider (1960), who envisioned a
man-machine symbiosis in which the human-computer decision-making whole would produce
holistic results where the performance of the human and computer combination would exceed
Search WWH ::
Custom Search