TASK-TECHNOLOGY FIT - Human-Computer Interaction and Management Information Systems: Foundations - page 175

Because the concept of task-technology fit is so dependent upon the task, determining how to
measure TTF requires at least some conceptualization of the task. A task may have several or
many steps, and a given technology may “fit” or provide the needed functionality for some of
those steps but not others. This suggests that before trying to measure task-technology fit, it is first
necessary to be more explicit about exactly what the task is, and what the several steps are in car-
rying out that task. We can then ask to what extent the technology supports the user in carrying
out each of those steps.
An Example of Developing a Task Model Before Measuring TTF
To give an example of how this might be done in practice, consider work by Goodhue (1995,
1998). He developed a task model for the task domain of “using quantitative information in man-
agerial tasks,” as described below.
The model includes three interacting steps of managerial use of information in their tasks: iden-
tifying information, accessing it, and interpreting it. Arrow (1974) and Ferrence (1970) have made
a distinction between identifying needed information and acquiring it. Others (Cowan, 1986;
O'Reilly, 1983; Saunders and Jones, 1990) have made a distinction between acquiring information
and using it. Therefore, a rough process model of information use could consist of three interact-
ing steps: identification of the data needed, acquisition of that data, and integration and interpreta-
tion of the data. These three steps are not necessarily sequential, and may be repeated.
To carry the example further, Goodhue was concerned with managers or knowledge workers
for whom using quantitative information is only part of a larger business task. As these users
attempt to identify, acquire, and interpret data, they will be frustrated in their efforts (and their
overall performance) when the technology does not support them. For each subtask, there are sev-
eral important dimensions of TTF on which systems and services might meet or fail to meet task
needs. Enumerating these dimensions leads to a more detailed model of the task, which can be the
basis for evaluating TTF at the detailed level.
Identifying Needed Data
There are at least five dimensions of TTF along which systems and services might be evaluated
when users are identifying needed data. Users need information systems to contain the right data
(Epstein and King, 1982; Bailey and Pearson, 1983), at the right level of detail (O'Reilly, 1982;
Swanson 1987). The cognitive difficulty increases if the organization of the files is confusing
(Bailey and Pearson, 1983), making it difficult to locate the needed data, if the meaning of the
data elements is unclear (Epstein and King, 1982).
Accessing Identified Data
There are at least seven dimensions of TTF along which information systems and services might
be evaluated for users accessing identified data. Users could be frustrated by poor accessibility of
the data (Bailey and Pearson, 1983; Culnan, 1984) or if authorization is difficult. For those who
use query languages or download data on their own, users will be frustrated if the hardware
and software is not easy to use (Bailey and Pearson, 1983), they have received insufficient train-
ing (Bailey and Pearson, 1983), or systems are plagued by poor reliability (i.e., unexpected down-
times) (Bailey and Pearson, 1983; Swanson, 1987). When enlisting the aid of user support