THE THEORY OF COGNITIVE FIT - Human-Computer Interaction and Management Information Systems: Foundations - page 127

Characteristics of Relevant Tasks
Although there is no comprehensive theory of tasks (see, for example, Campbell, 1988; Fleishman,
1982; Wood 1986), we can examine tasks used in graph/table decision making to determine the
abstract characteristics that might be facilitated by each type of representation. According to Einhorn
and Hogarth (1981), decision making consists of three interrelated sub-tasks: information acquisition,
evaluation, and feedback/learning. The first two sub-tasks are relevant to the analysis of graph/table
decision making.
Information Acquisition Tasks. Information acquisition tasks are those whose solution is
achieved directly via information acquisition processes. Several researchers have identified two
basic types of information acquisition tasks used in graphs versus tables studies (Umanath and
Scammell, 1988; Umanath, Scammell, and Das, 1990; Washburne, 1927). One type of task, spa-
tial, assesses the problem area as a whole rather than as discrete data values and therefore requires
making associations or perceiving relationships in the data. The following example involves a
comparison of trends and is therefore a spatial task:
Between the years 1100 and 1438 whose earnings increased most rapidly, those of the wool,
silk, or Calimala merchants? (Washburne, 1927)
Such tasks are best accomplished using perceptual processes.
The second type of task, symbolic, involves extracting discrete, and therefore precise, data val-
ues (Umanath and Scammell, 1988; Umanath, Scammell, and Das, 1990; Washburne, 1927). The
following example requires a specific amount as the response and is therefore a symbolic task:
How much did the wool merchants earn in the year 1100? (Washburne, 1927)
Such tasks are best accomplished using analytical processes.
Information Evaluation Tasks. Certain evaluation tasks can be analyzed in a similar manner to
tasks that involve information acquisition alone. In tasks involving fairly simple evaluation, the eval-
uation sub-tasks will be either spatial or symbolic in nature. For example, a task may require the
extraction of given data values followed by a series of calculations. The critical point, here, is that we
can infer the nature of both the information acquisition and the evaluation sub-tasks to be completed.
With regard to the type of evaluation, we can again identify, as for information acquisition tasks:
(a) tasks that are basically spatial in nature; and
(b) tasks that are basically symbolic in nature.
Tasks involving information acquisition and well-defined information evaluation may therefore
be accomplished via one of the two processes, spatial or symbolic, depending on the nature of each
of the sub-tasks; that is, when the sub-tasks are well defined, so, too, are the processes that a prob-
lem solver will use to solve them. Hence, the distinction between process and strategy, highlighted
in the introduction to the theory, is critical here.
When a task is sufficiently complex, however, its solution can be achieved via a number of dif-
ferent sets of sub-tasks, that is, strategies. Different strategies may have significant implications