Information Technology Reference
In-Depth Information
Problem novelty, according to the theory, is based on both the degree and type of abstraction
needed to generate an appropriate representation of the problem. The type of abstraction ranges
from analytic to holistic. Familiar to MIS/DSS researchers as individual cognitive styles, the ana-
lytic type of abstraction requires decomposition and detail whereas the holistic type of abstraction
requires aggregate and heuristic processing.
The degree of abstraction ranges from none to extensive. When no abstraction is needed, a
problem can be solved by directly applying the decision maker's memory-based knowledge. At
the other extreme, problems eliciting a high degree of abstraction require extensive reasoning and
inference for the decision maker to formulate a mental representation of a problem. The degree of
abstraction reflects the proportion of memory contained in one's representation of a problem. If
the degree of abstraction is low, a very high proportion of one's mental representation of a prob-
lem is based on memory. Conversely, if the degree of abstraction is high, a very low proportion of
one's mental representation of a problem is based on memory. The degree and type of abstraction
combine to define problem novelty.
Problem novelty also plays a role in decision-confidence and decision-maker calibration. The
current evidence suggests that decision-maker calibration is directly related to the degree of problem
novelty. As the degree of problem novelty increases (the degree and type of abstraction increase),
the likelihood of miscalibration increases (Juslin, 1993; Wagenaar, 1988).
The role of decision confidence in outcomes makes it clear that an effective DSS design must not
only improve the quality of decisions; it must also facilitate the user's interpretation of the quality of
decisions he or she makes using the aid. The effect of DSS design alternatives on the user's inter-
pretation of the quality of decisions has been ignored by the IS research community, despite warn-
ings of the dangers inherent in this neglect (Mason, 1969; Weizenbaum, 1966). To begin to consider
the effects of DSS design on decision-confidence and decision-maker calibration, a theory is needed
to guide the research. To provide this guidance, a DSS dialog design theory is needed.
A THEORY OF DSS DESIGN FOR USER CALIBRATION
The theory of DSS design for user calibration presented in this paper follows Walls et al.'s (1992)
framework. The goal, the conceptual properties, and the interaction among the properties in spe-
cific situations needed to achieve the goal are introduced and developed below.
Goal of the Theory
Because the focus of DSS is decision quality, the DSS design theory for user calibration is pri-
marily concerned with the realism needed for deciding on a decision. Perfect calibration requires
that the decision maker's belief in the quality of a decision equal the quality of the decision (Russo
and Schoemaker, 1992). Hence, the goal of the theory of DSS design for user calibration is to pre-
scribe the DSS properties needed for users to achieve perfect calibration.
Properties of the Theory
The conceptualizations of symbolic representations in problem solving/decision making as devel-
oped by Kaufmann encourage proposing a design theory of symbols and actions for DSS design
for user calibration consisting of linguistic expressiveness, visibility of imagery, and overt exploita-
tion of inquiry. Expressiveness is the manner or tone of the exchange conveyed by the dialog sym-
bols (e.g., condescending, matter-of-fact, supportive, directive, etc.). Expressiveness recognizes
Search WWH ::
Custom Search