THE THEORY OF COGNITIVE FIT - Human-Computer Interaction and Management Information Systems: Foundations - page 135

studies based on domain did not produce a sufficiently rich characterization of the studies. As a result
of our experiences here, we can add a further criterion for a good classification system: A good clas-
sification system should result in balanced categories.
In the classification system presented here, we employed some aspects of the alternative bases
for classification presented above. However, because our objective is to determine how the theory
of cognitive fit has been used since its initial presentation, we classified studies based on the type
of contribution that they make to the literature. Our taxonomy is based on mutually exclusive cat-
egories, each of which is divided into two sub-categories. Our first major category is based on
studies that test the original formulations of cognitive fit (Vessey, 1991, 1994), with sub-categories
based on task complexity. Our second major category is based on studies that extend the theory to
new domains. Our subcategories are positioned in two specific domains, those of human judg-
ment, and multi-criteria decision making. Our third major category is based on studies that use
different dimensions of fit. In addition to studies that have investigated traditional notions of fit
using new dimensions, a number of studies have applied the notions of fit to the complexity of the
relationship between the task and the problem representation.
From the viewpoint of the process of classification, we first examined a study to determine
whether it tested one of the original formulations of the theory. If it did not, then it was further
examined to determine whether it addressed a new dimension of fit. If this category was not appro-
priate, either, then the study was examined to determine whether it displayed a new dimension of
fit. In this way, we achieved a classification in which the categories are both mutually exclusive and
have theoretical significance.
Table 8.2 presents the studies examined according to the classification system.
TESTING THE THEORY OF COGNITIVE FIT
In this section, we test the theory of cognitive fit as presented in Vessey's two conceptual papers
(1991, 1994); that is, we examined studies of cognitive fit that involved quite simple information
acquisition and information evaluation tasks, followed by those that investigated cognitive fit in
more complex tasks.
Cognitive Fit in Simple Information Acquisition and Information Evaluation Tasks
We identified three studies that addressed the match of problem representation to task on simple
information acquisition and information evaluation tasks. Table 8.3 presents the details. We first
present the studies and then evaluate the findings.
Studies of Cognitive Fit on Simple Information Acquisition and Information Evaluation Tasks
Vessey and Galletta (1991) conducted a study to evaluate the characteristics of the basic theory of
cognitive fit, several extensions to the basic theory, and several premises on which the theory is
based. Here we present just those aspects of the study that evaluated the theory of cognitive fit.
Because spatial and symbolic tasks cannot be compared in any meaningful way, the study used two
experimental designs, one for each type of task. Problem representation was a between-subject vari-
able, with five repetitions of the task type as a within-subject factor.
As expected, users made faster and more accurate decisions on symbolic tasks with tables than with
graphs. However, users with graphs, though faster than those with tables on spatial tasks, as expected,
were less accurate. Hence the authors found only partial support for cognitive fit on spatial tasks.