Filetype PDF | Posted on 13 Sep 2022 | 3 years ago
Authentication
digital resources
144x Filetype PDF File size 0.25 MB Source: directory.umm.ac.id
Information & Management 36 (1999) 9±21
Research
Extending the technology acceptance model
with task±technology ®t constructs
a,* 1,b
Mark T. Dishaw , Diane M. Strong
a
University of Wisconsin ± Oshkosh, College of Business Administration, 800 Algoma Blvd., Oshkosh WI 54901, USA
bWorcester Polytechnic Institute, Department of Management, 100 Institute Road, Worcester MA 01609, USA
Received 8 June 1998; accepted 12 October 1998
Abstract
During the past decade, two signi®cant models of information technology (IT) utilization behavior have emerged in the MIS
literature. These two models, the technology acceptance model (TAM) and the task±technology ®t model (TTF), provide a
much needed theoretical basis for exploring the factors that explain software utilization and its link with user performance.
These models offer different, though overlapping perspectives on utilization behavior. TAM focuses on attitudes toward using
a particular ITwhich users develop based on perceived usefulness and ease of use of the IT. TTF focuses on the match between
user task needs and the available functionality of the IT. While each of these models offers signi®cant explanatory power, a
model that integrates constructs from both may offer a signi®cant improvement over either model alone. We discuss the
theoretical foundation of both these models and present a theoretical rationale for an integrated model. The result is an
extension of TAM to include TTF constructs. We test our integrated IT utilization model using path analysis. Our integrated
model provides more explanatory power than either model alone. Research using the integrated model should lead to a better
understanding of choices about using IT. # 1999 Elsevier Science B.V. All rights reserved.
Keywords: Technology acceptance model; Task±technology ®t; TAM; TTF; Software utilization
1. Introduction in the MIS literature. The second model, the task±
technology ®t model (TTF) [22, 23], addresses utili-
While information technology (IT) utilization stu- zation from a different, although not entirely ortho-
dies are common in the MIS literature [13, 38], early gonal, perspective.
studies lacked a strong theoretical foundation. Two We believe that TAM and TTF overlap in a sig-
signi®cant models have emerged which provide a ni®cant way and, if integrated, could provide an even
strong theoretical base for studies of IT utilization stronger model than either standing alone. Both these
behavior. The ®rst model, the technology acceptance models were developed to understand users' choices
model(TAM)[10],iswellknownandwidelyaccepted and evaluations of IT. The outcome variable for both,
TAM and TTF, is the actual use of IT or a related
variable. Applications of TAM usually focus early in
*Corresponding author. E-mail: dishaw@uwosh.edu the outcome chain on intention to use or actual use,
1E-mail: dstrong@wpi.edu whereas TTF applications focus later in the outcome
0378-7206/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved.
PII: S-0378-7206(98)00101-3
10 M.T. Dishaw, D.M. Strong/Information & Management 36 (1999) 9±21
chain on actual use or individual performance attri- that a behavior is determined by intention to perform
butable to actual use. the behavior. Actual behavior and intention have been
The objective of this research is to develop and foundtobehighlycorrelated[10,20].Intention,itself,
evaluate an integrated TAM/TTF model. To accom- is determined by attitude toward the behavior.
plish this objective, we examine the theory underlying Davis' research, in essence, examines the external
TAMandTTFandassesstheirsimilarities and differ- variables whichdetermineorin¯uenceattitudetoward
ences, which provides the theoretical foundation for ITuse. The TAM identi®es perceived ease of use, and
ourintegratedTAM/TTFmodel.Weempiricallycom- perceived usefulness as key independent variables.
pare the three models, TAM, TTF, and TAM/TTF, Perceived usefulness is also in¯uenced by perceived
using the same set of data collected in several orga- ease of use. The TAM includes the very important
nizations. Our results indicated that adding TTF assumption that the behavior is volitional, which is to
constructs to TAM explains signi®cantly more of sayvoluntaryoratthediscretionoftheuser.TheTAM
the variance in utilization than either TAM or TTF has been tested in several studies of IT use [1, 12, 29,
alone. 32].
An integration of TAM and TTF will be useful in TAM differs from TRA in two key ways. First, it
understandingsoftwareutilization in a broader variety speci®esusefulnessandeaseofuseasthetwoexternal
of circumstances, which is extremely important for variables or beliefs that determine attitude toward an
software authors and managers of the software users. IT, intention to use, and actual use. Thus, TAM does
They need to understand how the customers and end- notneedtobetailoredtoeachbehavior,aslongasthat
users of software actually choose to use or not use behavior is use of IT. Second, TAM does not include
certain functions. The key to understanding software the subjective norms construct in TRA. Subjective
use decisions lies in understanding how the functions norms, along with attitude, account for intention to
provided by the software ®t the perceived needs of the perform a behavior in TRA. For TAM, the subjective
user. normsconstruct has not been signi®cant [12, 29]. One
possible explanation is the use of students in many of
the tests of TAM; subjective norms may be more
2. Modeling information technology utilization important in an organizational setting [36]. Since
we have data from an organizational setting in which
2.1. The technology acceptance model (TAM) users may feel some social pressure to use the IT, we
test the signi®cance of subjective norms.
The technology acceptance model [10, 12], in The theory of planned behavior (TPB) [2, 3, 5], an
Fig. 1, is a speci®c adaptation of the theory of rea- extension of TRA, includes behavioral control as a
soned action (TRA) model [4, 20] to the study of IT construct to measure and account explicitly for the
usage. The TRA and its successor, the theory of extent to which users have complete control over their
plannedbehavior(TPB)[2],arewellknown,andhave behaviors, that is, the extent to which the behavior is
been widely employed in the study of speci®c beha- truly at the discretion of the user. TAM does not
viors [4]. In general, these theories (TRA, TAM) state include the behavioral control construct. In TPB,
behavioralcontroldirectlyaffectsintentiontoperform
a behavior, and may directly affect behavior in situa-
tions where the user intends to perform the behavior,
butispreventedfromdoingso[2].Whetherbehavioral
control is signi®cant depends on the particular beha-
vior. For example,behavioralcontrolwasimportantin
accounting for whether students intended and actually
received an `A' in a course, but not for whether
students attended classes [5].
For IT usage behavior, behavioral control has had
Fig. 1. Technology acceptance model (TAM). limited importance. Comparisons between TAM and
M.T. Dishaw, D.M. Strong/Information & Management 36 (1999) 9±21 11
TPB have largely concluded that TAM's ability to
accountforvarianceinintentiontouseorinactualuse
is about the same as TPB's [29, 36]. Like subjective
norms, the reason for this lack of effect could be the
use of students in comparing the models. Since we
collected data from ®rms, we test whether the addition
of behavioral control improves TAM. We expect
behavioral control to have some, but minor, signi®- Fig. 2. A basic task±technology ®t (TTF) model.
cance as compared to the other variables in TAM. In
the organizations we studied, use is voluntary. The ities of the user (see Fig. 2). Rational, experienced
information technologies we studied are tools that users will choose those tools and methods that enable
software maintainers may use, but do not need to themtocompletethetaskwiththegreatestnetbene®t.
use, to complete their maintenance tasks. While man- Information technology that does not offer suf®cient
agement in these organizations invested heavily in advantage will not be used.
advanced support tools for maintainers, maintainers In contrast to TAM, TTF, as a theoretical and
generally complete their maintenance tasks in theway measurable MIS construct and as part of a model of
they deem most effective. Such voluntary use at the IT utilization and performance, is still evolving. The
individual level is common with CASE tools [9]. basic ideas of TTF and models built around it are
In summary,TAMrepresentsthetailoringofawell- showninFig. 2,whichisoneTTFmodelthathasbeen
developed social psychology theory, the TRA [19], to tested [23]. Since other versions of TTF-based models
the speci®c behavior of using IT, by de®ning and exist2, the model shown in Fig. 2 should not be
developing measures for two variables, usefulness interpreted as the `TTF model'. For example, utiliza-
and ease of use [10, 11, 12]. The choice of these tion is a recent addition to the technology to perfor-
two variables is consistent with previous empirical mance chain [23]. Earlier TTF models employed
research in several MIS-related disciplines [11]. TAM individual performance as the only outcome variable
hasalsobeentestedandcomparedtorevisionsofTRA because these models were developed from work
by several authors independent of the original devel- adjustment theory which does not include a construct
opers of TAM. for behaviors, such as utilization.
AweaknessofTAMforunderstandingITutilization Acommon addition to a TTF model is individual
is its lack of task focus. IT is a tool by which users abilities [21, 23]. The inclusion of individual abilities
accomplish organizational tasks. The lack of task is supported by both, work adjustment theory from
focus in evaluating IT and its acceptance, use, and which TTF was originally derived and recent MIS
performance contributes to the mixed results in IT studies in which experience with particular IT is
evaluations [23]. While TAM's usefulness concept generally associated with higher utilization of that
implicitly includes task, that is to say usefulness IT [24, 37]. In tests of TTF models, individual abil-
means useful for something, more explicit inclusion ities, operationalized as computer literacy, negatively
of task characteristics may provide a better model of affected perceived ®t between task and technology
IT utilization. The task±technology ®t perspective [22] and, operationalized as experience with the par-
addresses this problem. ticular IT, positively affected utilization [17].
Although TTF is relatively new in the MIS litera-
2.2. Task±technology fit model ture, the concept of ®t, also called correspondence or
matching, is common in organizational theories. For
The ability of IT to support a task is expressed by example, the theory of work adjustment, from which
the formal construct known as task±technology ®t TTF was originally developed, considers the corre-
(TTF), which implies matching of the capabilities spondence between the abilities of an individual and
of the technology to the demands of the task [23]. the ability requirements of a job in determining an
TTF posits that IT will be used if, and only if, the
2
functions available to the user support (®t) the activ- See Refs. [22, 23] for a review of TTF and related models.
12 M.T. Dishaw, D.M. Strong/Information & Management 36 (1999) 9±21
individual's satisfactoriness for the job [21, 23]. incorporating both attitudes toward IT and the ®t
Research on strategic ®t, which is the correspondence between IT functionality and the characteristics of
between an organization and its environment, has the tasks that IT users are accomplishing with IT.
in¯uenced the methods for computing TTF [18, 22,
40]. 2.3. Integration of the TAM and the task±technology
The general concept of ®t has appeared in the MIS fit model
literature. For example, research on data representa-
tion, such as tables and graphs, has concluded that the Ajusti®cation for elaborating the TAM to include
best representation dependsontaskrequirements[35]. explicit references to task and technology is provided
Systemsimplementationresearchnotestheneedfor®t by the arguments of Goodhue [21, 22]. Goodhue
between tasks, technologies, and users [30]. Data linked his TTF model with the technology usage
quality research emphasizes the need for data to ®t model of Bagozzi [7], which, like TAM, was devel-
the needs of user tasks [16, 33, 34]. Research on oped from attitude/behavior models to explain tech-
problem solving and problem representation has nology utilization. The general argument for
developed the concept of cognitive ®t, which means combining the models is that they capture two differ-
that problem solving works best when the problem ent aspects of users' choices to utilize IT. TAM, and
representation and any tools or aids all support the the attitude/behavior models on which it is based,
processes required to perform that task [42, 43]. assume that users' beliefs and attitudes toward a
Wetest the TTF model shown in Fig. 3. Since we particular IT largely determine whether users exhibit
are focusing on models of IT utilization, utilization is the behavior of utilizing the IT. Critics note that users
our only outcome variable. Like the basic TTF model regularly utilize IT that they do not like because it
shown in Fig. 2, task and technology characteristics improves their job performance. TTF models take a
are the antecedents of TTF. We also test for direct decidedly rational approach by assuming that users
effects of task and technology characteristics on uti- choose to use IT that provides bene®ts, such as
lization (the dotted lines in Fig. 3). Tool experience, improvedjobperformance,regardless of their attitude
representing individual abilities, is expected to towardtheIT[22].Bothaspects,attitudetowardtheIT
directly affect utilization. and rationally determined expected consequences
While TTF models explicitly include task charac- from using the IT, are likely to affect users' choices
teristics, which is a weakness of TAM, they do not to utilize IT. That is, combining the two models is
explicitly include attitudes toward IT, which is the likely to provide a better explanation of IT utilization
core of TAM. Rather than arguing for TTF as an than either an attitude or a ®t model could provide
alternative to TAM, we propose adding the strengths separately.
ofTTFmodelstoTAMtoproduceanintegratedmodel We posit that constructs in the TTF model deter-
mine, in part, three variables in the TAM. TTF con-
structs are expected to directly affect utilization, as
they do in TTF models. TTF constructs may also
determine,inpart,TAM'stwodeterminantsofattitude
toward IT, namely perceived usefulness and perceived
ease of use. User beliefs about usefulness and ease of
use are likely to be developed, in part, from rational
assessments of the characteristics of the IT and the
tasks for which it could be used. In addition, these two
TAMvariables indirectly include aspects of the tech-
nologyandthetaskforwhichthetechnologycouldbe
used. For example, the whole notion of usefulness
implies that the software is useful for something. The
proposed integrated TAM/TTF model is shown in
Fig. 3. TTF model. Fig. 4.
no reviews yet
Please Login to review.
