Information Technology Reference
In-Depth Information
responses, while analytical processes lead to accurate responses. Hence a time requirement is
consistent with the perceptual processes used to support spatial problem-solving elements. With
a requirement for accuracy, however, decision makers may be induced to switch from perceptual
to analytical processes. If this were so, a performance effect for graphs over tables would be
unlikely, as processes would be uncontrolled. Similarly, decision makers using symbolic problem-
solving elements may be induced to change to perceptual processes when a time requirement is
imposed. Vessey (1991) found no evidence of strategy change occurring when cognitive fit
existed. Hence the original study on cognitive fit provided substantial evidence to support the
descriptive power and stability of the theory for information acquisition tasks.
Cognitive fit may also occur in tasks involving
well-defined evaluation
, that is, tasks that have
just a few sub-tasks that are readily identifiable (Vessey 1991). Two situations may arise: (1) the
processes required to support information acquisition and evaluation are similar; that is, both are
perceptual or both analytical in nature, and therefore best supported by graphs and tables, respec-
tively; or (2) the processes required to support information acquisition and evaluation do not
match, in which case it is unlikely that use of either graphs or tables will result in performance
advantages. Again, analysis of prior studies supported this assessment. Further, no support was
observed for strategy change, which, in this case, is really process change.
Applying Cost-Benefit Principles to Complex Tasks with Numerous Ill-Defined Sub-Tasks
As we have seen, for tasks involving information acquisition and simple information evaluation,
choosing a problem representation that matches the type of task being solved so as to achieve cogni-
tive fit results in effective control of decision-making processes. Strategy shift may occur, however, in
more complex tasks that require substantial evaluation. A number of different strategies might be used
to address such tasks and there is scope for applying perceptual or analytical processes in the sub-tasks.
When complex spatial tasks are supported by spatial representations, and no undue emphasis is
placed on accuracy, cognitive fit will exist and there will be no incentive to change strategy. Strategy
shift may occur, however, on symbolic tasks as the inherent complexity of the task increases; that is,
decision makers may prefer to use perceptual rather than analytical processes. At least two general
situations may help to induce strategy shift during solution of complex symbolic problems. First, the
high analytical demands of the task per se may induce the problem solver to expend less effort in
decision making, resulting in the conscious choice of a less effortful strategy (Einhorn and Hogarth,
1981; Russo and Dosher, 1983). Second, performance constraints may be placed on the decision-
making exercise, leading to strategy shift. Both of these factors are regarded as task variables and
can be expected to lead to strategy shift largely through their influence on effort. As a result, in deci-
sion making using graphs or tables, the appropriate problem representation might not be a table,
which supports analytical processes, but a graph, which supports the more parsimonious perceptual
processes. We state the following propositions.
Proposition 2
:
As the amount of analytical evaluation required increases, decision makers
choose between analytical processes resulting in high effort and accuracy
and perceptual processes resulting in considerably lower effort and some-
what lower accuracy.
There may also exist the limiting case for which the use of perceptual processes alone is feasi-
ble. Again, perceptual processes are best supported with graphs.
Proposition 3
:
As the amount of analytical evaluation required increases, the problem can
only be solved using perceptual processes.
Search WWH ::
Custom Search