Determinants of Perceived Ease of Use: Integrating Control, Intrinsic Motivation, and Emotion into the Technology Acceptance Model

Abstract

Much previous research has established that perceived ease of use is an important factor influencing user acceptance and usage behavior of information technologies. However, very little research has been conducted to understand how that perception forms and changes over time. The current work presents and tests an anchoring and adjustment-based theoretical model of the determinants of system-specific perceived ease of use. The model proposes control (internal and external--conceptualized as computer self-efficacy and facilitating conditions, respectively), intrinsic motivation (conceptualized as computer playfulness), and emotion (conceptualized as computer anxiety) as anchors that determine early perceptions about the ease of use of a new system. With increasing experience, it is expected that system-specific perceived ease of use, while still anchored to the general beliefs regarding computers and computer use, will adjust to reflect objective usability, perceptions of external control specific to the new system environment, and system-specific perceived enjoyment. The proposed model was tested in three different organizations among 246 employees using three measurements taken over a three-month period. The proposed model was strongly supported at all points of measurement, and explained up to 60% of the variance in system-specific perceived ease of use, which is twice as much as our current understanding. Important theoretical and practical implications of these findings are discussed.

Full text

Please cite this article as:
Venkatesh, V. “Determinants of Perceived Ease of Use: Integrating Control, Intrinsic
Motivation, and Emotion into the Technology Acceptance Model,” Information Systems
Research (11:4), 2000, 342-365.
https://doi.org/10.1287/isre.11.4.342.11872
DETERMINANTS OF PERCEIVED EASE OF USE:
INTEGRATING CONTROL, INTRINSIC MOTIVATION, AND EMOTION INTO
THE TECHNOLOGY ACCEPTANCE MODEL
Viswanath Venkatesh
University of Maryland
vvenkatesh@vvenkatesh.us
This is a pre-publication version and was subject to copyediting and proofing prior to
publication.

2
DETERMINANTS OF PERCEIVED EASE OF USE: INTEGRATING CONTROL,
INTRINSIC MOTIVATION, AND EMOTION INTO THE TECHNOLOGY
ACCEPTANCE MODEL
ABSTRACT
Much previous research has established that perceived ease of use is an important factor
influencing user acceptance and usage behavior of information technologies. However, very
little research has been conducted to understand how that perception forms and changes over
time. The current work presents and tests an anchoring and adjustment-based theoretical model
of the determinants of system-specific perceived ease of use. The model proposes control
(internal and external—conceptualized as computer self-efficacy and facilitating conditions
respectively), intrinsic motivation (conceptualized as computer playfulness), and emotion
(conceptualized as computer anxiety) as anchors that determine early perceptions about the ease
of use of a new system. With increasing experience, it is expected that system-specific perceived
ease of use, while still anchored to the general beliefs regarding computers and computer use,
will adjust to reflect objective usability, perceptions of external control specific to the new
system environment, and system-specific perceived enjoyment. The proposed model was tested
in three different organizations among 246 employees using three measurements taken over a
three-month period. The proposed model was strongly supported at all points of measurement,
and explained up to 60% of the variance in system-specific perceived ease of use, which is twice
as much as our current understanding. Important theoretical and practical implications of these
findings are discussed.

3
INTRODUCTION
Information technology (IT) acceptance and use is an issue that has received the attention
of researchers and practitioners for over a decade. Successful investment in technology can lead
to enhanced productivity, while failed systems can lead to undesirable consequences such as
financial losses and dissatisfaction among employees. Despite significant technological
advances and increasing organizational investment in these technologies, the problem of
underutilized systems plagues businesses (Johansen and Swigart 1996; Moore 1991; Norman
1993; Wiener 1993). For example, the Internal Revenue Service (IRS) invested about $4B on a
system aimed at simplifying the processing of tax returns for 1996 by computerizing the process.
However, reports in early 1997 (e.g., Johnston 1997) indicated that the IRS was forced to revert
to the manual method of processing returns. In this and other cases, users have found the system
to be too difficult to use and have not been able to scale that hurdle to user acceptance and usage
of the new system (e.g., Venkatesh 1999). Such low usage of installed systems has been
suggested to be a possible key explanation for the “productivity paradox” (Landauer 1995;
Sichel 1997). Thus, understanding user acceptance, adoption, and usage of new systems is a
high priority item for researchers and practitioners alike.
A significant body of research in information systems (IS) (e.g., Davis, Bagozzi, and
Warshaw 1989; Venkatesh 1999) and human-computer interaction (HCI) (e.g., Gould and Lewis
1985) has accumulated supporting the importance of such perceived ease of use on initial user
acceptance and sustained usage of systems. Although there is a large body of research on the
perceived ease of use construct, very little work has been done to understand the determinants of
this important driver of technology acceptance and use. Understanding the determinant structure
of this key driver of user acceptance and usage is critical because it will provide leverage points

4
to create favorable perceptions and thus, foster user acceptance and usage. The importance of
such a line of inquiry has been highlighted by recent research (Taylor and Todd 1995) focused
on the determinant structure of key constructs in the Theory of Planned Behavior (see Ajzen
1985, 1991).
The current research attempts to further our understanding of the determinants of
perceived ease of use of a system by focusing on how these perceptions form and change over
time with increasing experience with the system. Typically, researchers and practitioners have
restricted their attention to system design characteristics (e.g., Davis et al. 1989) or training (e.g.,
Venkatesh 1999) when trying to enhance user perceptions of the ease of use of a system, thereby
overlooking other controllable variables such as individual difference variables and variables that
are a result of a system-user interaction. Based on an anchoring and adjustment framework, a
theoretical model proposes that in forming system-specific perceived ease of use, individuals
anchor on key individual and situational variables that relate to control, intrinsic motivation, and
emotion. With increasing direct experience with the target system, individuals adjust their
system-specific perceived ease of use to reflect their interaction with the system.
BACKGROUND
There have been several theoretical models employed to study user acceptance and usage
behavior of emerging information technologies. While many of the models incorporate
perceived ease of use as a determinant of acceptance, the Technology Acceptance Model (TAM)
(Davis 1989; Davis et al. 1989) is the most widely applied model of user acceptance and usage.
TAM was adapted from the Theory of Reasoned Action (TRA) (Ajzen and Fishbein 1980;
Fishbein and Ajzen 1975). TAM suggests that two specific beliefs, perceived ease of use and
perceived usefulness, determine one’s behavioral intention to use a technology, which has been

5
linked to subsequent behavior (Taylor and Todd 1995; see Sheppard, Hartwick, and Warshaw 1988
for a meta-analysis of the intention-behavior relationship). Attitude towards using a technology was
omitted by Davis et al. (1989) in their final model (pp. 995-996) because of partial mediation of the
impact of beliefs on intention by attitude, a weak direct link between perceived usefulness and
attitude, and a strong direct link between perceived usefulness and intention. This was explained as
originating from people intending to use a technology because it was useful even though they did
not have a positive affect (attitude) toward using. The omission of attitude helps better understand
the influence of perceived ease of use and perceived usefulness on the key dependent variable of
interest—intention.
FURTHER, TAM POSITS THAT PERCEIVED USEFULNESS WILL BE
INFLUENCED BY PERCEIVED EASE OF USE BECAUSE OTHER THINGS BEING
EQUAL, THE EASIER A TECHNOLOGY IS TO USE, THE MORE USEFUL IT CAN BE.
CONSISTENT WITH TRA, TAM SUGGESTS THAT THE EFFECT OF EXTERNAL
VARIABLES (E.G., SYSTEM DESIGN CHARACTERISTICS) ON INTENTION IS
MEDIATED BY THE KEY BELIEFS (I.E., PERCEIVED EASE OF USE AND
PERCEIVED USEFULNESS). TAM HAS RECEIVED EXTENSIVE EMPIRICAL
SUPPORT THROUGH VALIDATIONS, APPLICATIONS, AND REPLICATIONS
(ADAMS, NELSON, AND TODD 1992; CHIN AND GOPAL 1993; CHIN AND TODD
1995; DAVIS 1993; DAVIS AND VENKATESH 1996; GEFEN AND STRAUB 1997;
HENDRICKSON, MASSEY, AND CRONAN 1993; IGBARIA, ZINATELLI, CRAGG,
AND CAVAYE 1997; MATHIESON 1991; SEGARS AND GROVER 1993;
SUBRAMANIAN 1994; SZAJNA 1994, 1996; TAYLOR AND TODD 1995; VENKATESH
1999; VENKATESH AND DAVIS 1996) BY RESEARCHERS AND PRACTITIONERS
i
,

6
SUGGESTING THAT TAM IS ROBUST ACROSS TIME, SETTINGS, POPULATIONS,
AND TECHNOLOGIES.
PERCEIVED EASE OF USE IS THE EXTENT TO WHICH A PERSON BELIEVES
THAT USING A TECHNOLOGY WILL BE FREE OF EFFORT. PERCEIVED EASE OF
USE IS A CONSTRUCT TIED TO AN INDIVIDUAL’S ASSESSMENT OF THE
EFFORT INVOLVED IN THE PROCESS OF USING THE SYSTEM (SEE DAVIS 1989
FOR A DETAILED DISCUSSION OF THE THEORETICAL AND EMPIRICAL
DEVELOPMENT OF THE CONSTRUCT). ALTHOUGH THIS RESEARCH FOCUSES
ON PERCEIVED EASE OF USE IN THE CONTEXT OF TAM, IT IS WORTH NOTING
THAT OTHER THEORETICAL PERSPECTIVES STUDYING USER ACCEPTANCE
HAVE ALSO EMPLOYED SIMILAR CONSTRUCTS—THOMPSON, HIGGINS, AND
HOWELL (1991) USE A CONSTRUCT CALLED “COMPLEXITY,” AND MOORE AND
BENBASAT (1991) EMPLOY A CONSTRUCT CALLED “EASE OF USE.”
ALTHOUGH PERCEIVED EASE OF USE IS ASSOCIATED WITH INTENTION IN
TAM, THE UNDERLYING OBJECTIVE IS TO PREDICT USAGE BEHAVIOR. IN
THIS CONTEXT, IT IS IMPORTANT TO HIGHLIGHT THAT A VAST BODY OF
RESEARCH IN BEHAVIORAL DECISION MAKING (E.G., PAYNE, BETTMAN, AND
JOHNSON 1993) AND IS (E.G., TODD AND BENBASAT 1991, 1992, 1993, 1994)
DEMONSTRATING THAT INDIVIDUALS ATTEMPT TO MINIMIZE EFFORT IN
THEIR BEHAVIORS, THUS SUPPORTING A RELATIONSHIP BETWEEN
PERCEIVED EASE OF USE AND USAGE BEHAVIOR, ALBEIT THROUGH
INTENTION AS SUGGESTED BY TAM. IN CONTRAST, THE OTHER TAM BELIEF
(I.E., PERCEIVED USEFULNESS) IS DEFINED AS THE EXTENT TO WHICH A

7
PERSON BELIEVES THAT USING A TECHNOLOGY WILL ENHANCE HER/HIS
PRODUCTIVITY.
THE PARSIMONY OF TAM COMBINED WITH ITS PREDICTIVE POWER
MAKE IT EASY TO APPLY TO DIFFERENT SITUATIONS. HOWEVER, WHILE
PARSIMONY IS TAM’S STRENGTH, IT IS ALSO THE MODEL'S KEY LIMITATION.
TAM IS PREDICTIVE BUT ITS GENERALITY DOES NOT PROVIDE SUFFICIENT
UNDERSTANDING FROM THE STANDPOINT OF PROVIDING SYSTEM
DESIGNERS WITH THE INFORMATION NECESSARY TO CREATE USER
ACCEPTANCE FOR NEW SYSTEMS (MATHIESON 1991). SPECIFICALLY, IT IS
IMPORTANT TO EMPHASIZE THAT ALTHOUGH PERCEIVED EASE OF USE HAS
BEEN EMPLOYED EXTENSIVELY IN USER ACCEPTANCE RESEARCH IN
GENERAL AND TAM RESEARCH IN PARTICULAR, VERY LITTLE HAS BEEN
DONE TO UNDERSTAND THE DETERMINANTS OF PERCEIVED EASE OF USE.
DAVIS’ MORE RECENT WORK ACKNOWLEDGES THIS POTENTIAL
LIMITATION: “WHILE BEING VERY POWERFUL IN HELPING US PREDICT
ACCEPTANCE, ONE OF THE LIMITATIONS OF TAM IS THAT IT DOES NOT HELP
UNDERSTAND AND EXPLAIN ACCEPTANCE IN WAYS THAT GUIDE
DEVELOPMENT BEYOND SUGGESTING THAT SYSTEM CHARACTERISTICS
IMPACT EASE OF USE... THIS PLACES A DAMPER ON OUR ABILITY TO
MEANINGFULLY DESIGN INTERVENTIONS TO FOSTER ACCEPTANCE. IN
ORDER TO BE ABLE TO EXPLAIN USER ACCEPTANCE AND USE, IT IS
IMPORTANT TO UNDERSTAND THE ANTECEDENTS OF THE KEY TAM
CONSTRUCTS, PERCEIVED EASE OF USE AND USEFULNESS” (VENKATESH AND

8
DAVIS 1996, PP. 472-473). UNDERSTANDING THE DETERMINANTS OF
PERCEIVED EASE OF USE IS FURTHER UNDERSCORED BY THE TWO
MECHANISMS BY WHICH IT INFLUENCES INTENTION: (1) PERCEIVED EASE OF
USE HAS A DIRECT EFFECT ON INTENTION, AND AN INDIRECT EFFECT ON
INTENTION VIA PERCEIVED USEFULNESS, AND (2) IT IS AN INITIAL HURDLE
THAT USERS HAVE TO OVERCOME FOR ACCEPTANCE, ADOPTION, AND
USAGE OF A SYSTEM (SEE DAVIS ET AL. 1989).
THEORETICAL FRAMEWORK AND MODEL DEVELOPMENT
This paper proposes a theoretical framework that describes the determinants of system-
specific perceived ease of use as individuals evolve from the early stages of experience with the
target system to stages of significant experience. Prior research in IS and psychology has
established the importance of actual behavioral experience in shaping the evolution of beliefs
such as perceived ease of use (Doll and Ajzen 1992; Davis et al. 1989; Fazio and Zanna 1978a,
1978b, 1981; Venkatesh and Davis 1996). The framework presents an anchoring and adjustment
perspective on the formation and change of system-specific perceived ease of use over time with
increasing experience with a target system. Behavioral decision theory suggests that “anchoring
and adjustment” is an important general decision making heuristic that is often used by
individuals (Slovic and Lichtenstein 1971; Tversky and Kahneman 1974; see Northcraft and
Neale 1987 for an example). In the absence of specific knowledge, the heuristic suggests that
individuals rely on general information that serves as an “anchor” and in fact, individuals are
often unable to ignore such anchoring information in decision-making processes. If additional
information becomes available (typically following direct experience with the target behavior),
individuals tend to adjust their judgments to reflect the new information but still rely on the

9
initial anchoring criteria. Specifically, Helson (1964) suggests that a subject’s response to a
judgmental task is based on three aspects: (1) sum of the subject’s past experiences, (2) the
context or background, and (3) the stimulus (see also Streitfeld and Wilson 1986). To the extent
that minimal context (i.e., specific system information) is given, the subject will make system-
specific perceived ease of use evaluations based on prior experiences with systems. As more
contextual information (i.e., system-specific information) becomes available, the more the
judgment will be made within that context rather than based on previous experience.
Specifically, prior to direct experience with the target system, individuals are expected to
anchor their system-specific perceived ease of use of a new system to their general beliefs
regarding computers and computer use. With increasing experience with the system, individuals
are expected to adjust their system-specific perceived ease of use to reflect their interaction with
the system (Figure 1). The framework can also be explained in terms of the general-specific
distinction from psychology and abstract-concrete distinction from marketing (Bettman and
Sujan 1987; Mervis and Rosch 1981). In the absence of much knowledge about the target
system and limited direct behavioral experience with the system, individuals will base their
perceived ease of use of the target system on general, abstract criteria. With increasing learning
and direct experience with the target system, user judgments about the ease of use of the system
are expected to reflect specific, concrete attributes that are a result of an individual’s direct
experience with the system.
In addition to research in psychology, organizational behavior, and marketing that was
discussed earlier, the basic arguments of anchoring and adjustment can also be supported from
empirical evidence from prior user acceptance research. In the absence of much direct hands-on
experience with new systems, user perceptions of ease of use of systems are not distinct across

10
the different new systems, thus suggesting that in the early stages of user experience with new
systems, there are a set of “common” determinants for system-specific perceived ease of use
(Venkatesh and Davis 1996). Specifically, in the early stages of user experience, the initial
anchors for system-specific perceived ease of use of a new/target system are expected to be
individual difference variables and general beliefs regarding computers based on prior
experience with computers/software in general and other systems in the organization. There is
some evidence supporting this line of reasoning—general computer self-efficacy (Compeau and
Higgins 1995a) has been shown to be a strong determinant of perceived ease of use before
hands-on experience (Venkatesh and Davis 1996). As users gain experience with the target
system, their assessment of the ease of use of the system, while still being anchored to individual
difference variables and general beliefs, will adjust to reflect unique attributes of their interaction
with the system and the system environment. Recent empirical research demonstrated low
correlations between initial perceived ease of use and perceived ease of use after significant
direct experience providing support for the idea that adjustments based on direct experience can
be important in shaping perceived ease of use over time (Venkatesh and Davis 1996). Further,
the original conceptualization of TAM presents an expectancy model, consistent with social
cognitive theory (Bandura 1986) that dictates perceived ease of use (a process expectancy) and
perceived usefulness (an outcome expectancy) would be key predictors of intention/behavior.
Since the current research builds on TAM, there is an implicit assumption incorporating an
outcome-process perspective that dictates other constructs would be “external variables”
influencing key expectancies, thus lending support to the examination of other constructs as
possible determinants of perceived ease of use.

11
A theoretical model based on the framework is proposed. Figure 2 presents the proposed
model. Constructs related to control, intrinsic motivation, and emotion are proposed as general
anchors for the formation of perceived ease of use regarding a new system. Specifically, control
is divided into perceptions of internal control (computer self-efficacy) and perceptions of
external control (facilitating conditions), intrinsic motivation is conceptualized as computer
playfulness, and emotion is conceptualized as computer anxiety. Computer self-efficacy,
facilitating conditions, computer playfulness, and computer anxiety are system-independent,
anchoring constructs that play a critical role in shaping perceived ease of use about a new
system, particularly in the early stages of user experience with a system. With increasing
experience with the system, objective usability, perceptions of external control (facilitating
conditions) as it pertains to the specific system,
ii
and perceived enjoyment from system use are
adjustments (resulting from the user-system interaction) that will have an added influence on
system-specific perceived ease of use.
Anchors
Control: Computer Self-Efficacy and Facilitating Conditions
Control is a construct that reflects situational enablers or constraints to behavior (Ajzen
1985). In IS (Taylor and Todd 1995) and psychology (Ajzen 1991), control has been treated as a
perceptual construct since that is of greater interest (from a psychological perspective) than
actual control when understanding behavior (see Ajzen 1991). Specifically, control relates to an
individual’s perception of the availability of knowledge, resources, and opportunities required to
perform the specific behavior. Perception of control was the key addition to the Theory of
Reasoned Action (TRA) (Ajzen and Fishbein 1980; Fishbein and Ajzen 1975) to arrive at the
Theory of Planned Behavior (TPB) (Ajzen 1985). Given that TAM was developed from TRA,

12
the predecessor to TPB, the role of control was not explicitly incorporated in the theoretical
development of TAM. Subsequent research also has not fully detailed the role of control in the
context of TAM (cf. Venkatesh and Davis 1996).
Control has been shown to have an effect on key dependent variables such as intention
and behavior in a variety of domains (see Ajzen 1991 for a review). In IS research, Mathieson
(1991) applied TPB to a technology acceptance context and found that while control was a
significant determinant of intention, TPB explained about the same variance as TAM. In a more
recent study, Taylor and Todd (1995) found a similar pattern of results. However, the effect of
control on intention over and above what is explained by the TAM constructs of perceived ease
of use and perceived usefulness is not known. As mentioned earlier, the final model of TAM
excludes the attitude construct and helps understand the explanatory power of perceived ease of
use and perceived usefulness on intention. This final model of TAM was not tested in Mathieson
(1991) and Taylor and Todd (1995), thus not providing information about the effect of control on
intention over and above TAM beliefs. Another point related to control is worthy of note—in IS
research, perceived ease of use has been seen to be a determinant of attitude consistent with TPB
(see Davis et al. 1989; Taylor and Todd 1995), while internal and external control have been
related to perceived behavioral control in TPB. The current work relates control to perceived
ease of use, thus departing from the basic framework of TPB. However, such “crossover effects”
have been observed in prior research (Oliver and Bearden 1985; Warshaw 1980; Venkatesh and
Davis 1996)
Although IS research has typically viewed control as unidimensional with a control belief
structure that includes self-efficacy, technology facilitating conditions, and resource facilitating
conditions (Taylor and Todd 1995), the unidimensionality of the construct has been challenged

13
in psychology research. Azjen’s (1985, 1991) conceptualization of control refers to internal and
external constraining factors. Specifically, internal control relates to knowledge/self-efficacy
and external control relates to the environment (Terry 1993). Empirical evidence confirms this
bi-dimensionality—there has been evidence of low interitem correlations among measures of
control based on the original conceptualization of the construct (see Beale and Manstead 1991;
Chan and Fishbein 1993; Sparks 1994). Based on such results, it has been argued that
individuals perceive internal and external control differently (Chan and Fishbein 1993; Sparks
1994; Sparks, Guthrie, and Shepherd 1997; Terry 1991, 1994; Terry and O’Leary, 1995). The
bi-dimensional conceptualization allows the role of the two control dimensions to be studied,
understood, and managed separately (Terry 1993; Terry and O’Leary, 1995; White, Terry, and
Hogg 1994) at the conceptual, operational, and practical levels. Despite the controversy about
the conceptualization of control, both internal (e.g., de Vries, Dijkstra, and Kuhlman 1988;
McCaul, O’Neill, and Glasgow 1988; Ronis and Kaiser 1989; Terry 1993; Wurtele 1988) and
external control (e.g., Kimieck 1992; Schifter and Ajzen 1985) have an important role in shaping
intention and behavior in a variety of domains.
In an IT usage context, internal control is conceptualized as computer self-efficacy, an
individual difference variable that represents one’s belief about her/his ability to perform a
specific task/job using a computer (see Compeau and Higgins 1995a, 1995b). There is
experimental evidence supporting the causal flow from computer self-efficacy to system-specific
perceived ease of use (Venkatesh and Davis 1996). The link was justified on the basis that in the
absence of direct system experience, the confidence in one’s computer-related abilities and
knowledge can be expected to serve as the basis for an individual’s judgment about how easy or
difficult a new system will be to use.

14
While there has been some theoretical and empirical support for the influence of general
perceptions of internal control (i.e., computer self-efficacy) on system-specific perceived ease of
use, the role of general perceptions of external control in determining system-specific perceived
ease of use has been overlooked. As Mathieson (1991) pointed out, external control issues are
not explicitly included in TAM, or the perceived ease of use construct. Mathieson (1991) argued
that while perceived ease of use could potentially encompass control over resources, this was not
made explicit. For instance, an item such as "I would find <particular system> easy to use" (e.g.,
"I would find Word easy to use") could result in a response wherein the respondent has taken
into account constraints placed not only by system characteristics but also by availability of
knowledge, resources, and opportunities—i.e., the underlying elements of control. Given the
broad conceptualization of perceived ease of use, we expect that user judgments of the difficulty
of using a system will incorporate both internal and external dimensions of control. External
control is expected to exert its influence in the form of individual perception of technology and
resource facilitating conditions (see Taylor and Todd 1995). In the context of workplace
technology use, specific issues related to external control include the availability of support staff,
which is an organizational response to help users overcome barriers and hurdles to technology
use, especially during the early stages of learning and use (e.g., Bergeron, Rivard, and De Serre
1990). In fact, consultant support has been conceptually and empirically shown to influence
perceptions of control (e.g., Cragg and King 1993; Harrison, Mykytyn, and Riemenschneider
1997). Users in organizational settings will have general perceptions of external control based
on prior technology introductions in the organization. Prior to direct experience with the new
system environment, such general perceptions of external control are essentially system-
independent and serve as situational anchors in the formation of perceived ease of use of the new

15
system. Thus, the model proposes that internal and external control will be important anchors in
the formation of early system-specific perceived ease of use.
Intrinsic Motivation: Computer Playfulness
The next anchor proposed is related to intrinsic motivation. There are two main classes
of motivation: extrinsic and intrinsic (Vallerand 1997). Extrinsic motivation relates to the drive
to perform a behavior to achieve specific goals/rewards (Deci and Ryan 1987), while intrinsic
motivation relates to perceptions of pleasure and satisfaction from performing the behavior
(Vallerand 1997). In TAM, extrinsic motivation and the associated instrumentality are captured
by the perceived usefulness construct (see Davis et al. 1989; Davis, Bagozzi, and Warshaw 1992;
Morris and Venkatesh 2000; Venkatesh, Morris, and Ackerman 2000; Venkatesh and Speier
1999). TAM does not explicitly include intrinsic motivation. The current work proposes that the
role of intrinsic motivation will relate to the perceived ease of use construct. In relating intrinsic
motivation to general computer usage contexts, the construct of computer playfulness has been
successfully applied and operationalized in prior research (Webster and Martocchio 1992).
Computer playfulness is an individual difference variable defined as “the degree of cognitive
spontaneity in microcomputer interactions” (Webster and Martocchio 1992, p. 204). There is a
significant body of theoretical and empirical evidence regarding the importance of the role of
intrinsic motivation on technology use in the workplace (Davis et al. 1992; Malone 1981a,
1981b; Webster and Martocchio 1992; Venkatesh and Speier 1999, forthcoming). Webster,
Trevino, and Ryan (1993) called for research on important outcomes of computer playfulness as
it relates to human-computer interaction. Although TAM predicts technology acceptance based
on user perceptions following such interactions, research to date has not studied how computer
playfulness fits into the nomological network of TAM. The current research addresses this issue

16
by proposing computer playfulness as a system-independent, motivation-oriented anchor for
system-specific perceived ease of use.
At the outset, it is important to address the basic rationale for such a causal flow.
iii
Computer playfulness represents an abstraction of the openness to the process of using systems
and such an abstract criterion is expected to serve as an anchor for the perceived ease of use of a
specific new system. Computer playfulness is an individual difference variable that is system-
independent. Those who are more “playful” with computer technologies in general are expected
to indulge in using a new system just for the sake of using it, rather than just the specific positive
outcomes associated with use. Such playful individuals may tend to “underestimate” the
difficulty of the means or process of using a new system because they quite simply enjoy the
process and do not perceive it as being effortful compared to those who are less playful. This
implies that there is likely to be a positive relationship between general computer playfulness and
system-specific perceived ease of use. Although individuals may not expect systems in
organizational settings to necessarily prompt high levels of fun (on-task or off-task), computer
playfulness is still expected to be a relevant factor influencing user perceptions about a system
since the construct of computer playfulness not only includes the desire for fun but also involves
exploration and discovery. Computer playfulness may also include challenge and curiosity (see
Malone 1981a, 1981b). Thus, in general, more playful individuals are expected to rate any new
system as being easier to use compared to those who are less playful.
Higher levels of computer playfulness lead to an internal locus of causality (Deci 1975;
DeCharms 1968) that in turn lowers perceptions of effort. Gattiker (1992) suggested that
motivation in general will have an impact on substantive complexity, a construct similar to
perceived ease of use. More specifically, from a theoretical standpoint, research in psychology

17
suggests that higher levels of intrinsic motivation typically lead to willingness to spend more
time on the task (e.g., Deci 1975). We extend this argument to suggest that higher levels of
computer playfulness will lead to lowered perceptions of effort—i.e., for the same level of actual
effort/time invested, perceptions of effort/time will be lower in the case of a more “playful” user
when compared to a less “playful” user. In the absence of much direct experience with the
specific system, the user does not possess much information about the extent to which using the
specific system is enjoyable, but one’s desire to explore and play with a new system in general is
expected to influence her/his perceived ease of use of the new system.
Emotion: Computer Anxiety
The anchors related to control capture knowledge and resource aspects, and the intrinsic
motivation anchor captures computer playfulness. The emotional aspect of technology usage is
expected to be captured via a construct called computer anxiety. Computer anxiety is defined as
an individual’s apprehension, or even fear, when she/he is faced with the possibility of using
computers (Simonson, Mauer, Montag-Torardi, and Whitaker 1987). Computer anxiety, like
computer self-efficacy and computer playfulness, relates to users’ general perceptions about
computer use. While computer self-efficacy relates to judgments about ability and computer
playfulness relates to the spontaneity in an individual’s interaction with a computer, computer
anxiety is a negative affective reaction toward computer use. A significant body of research in
IS and psychology has highlighted the importance of computer anxiety by demonstrating its
influence on key dependent variables. For example, computer anxiety has been shown to have a
significant impact on attitudes (Howard and Smith 1986; Igbaria and Chakrabarti 1990; Igbaria
and Parasuraman 1989; Morrow, Prell, and McElroy 1986; Parasuraman and Igbaria 1990;
Popovich, Hyde, Zakrajsek, and Blumer 1987), intention (Elasmar and Carter 1996), behavior

18
(Compeau and Higgins 1995a; Scott and Rockwell 1997; Todman and Monaghan 1994),
learning (Liebert and Morris 1967; Martocchio 1994; Morris, Davis, and Hutchings 1984), and
performance (Anderson 1996; Heinssen, Glass, and Knight 1987).
Given its important role in influencing key dependent variables, prior research has
devoted much attention to the causes of anxiety in a variety of domains including computer use
(Anderson 1996; Cambre and Cook 1985; Chu and Spires 1991; Igbaria and Chakrabarti 1990),
and on prescriptions and potential interventions to reduce computer anxiety (Bohlin and Hunt
1995; Chu and Spires 1991; Crable, Brodzinski, Scherer, and Jones 1994; Emanuele, Dale, and
Klions 1997; Keeler and Anson 1995; Leso and Peck 1992; Reznich 1996; Schuh 1996). From a
pragmatic standpoint, with the increasing pervasiveness of computers in the workplace and
homes, there may be some question about whether the construct of computer anxiety, which was
of much significance over a decade ago when individuals in organizations exhibited such
emotion (e.g., Zoltan and Chapanis 1982; see Maurer 1994 for a review), is still relevant—in
fact, there is recent field evidence to indicate the existence of computer anxiety and high
variability across individuals (Bozionelos 1996; Marcoulides, Mayes, and Wiseman 1995).
Although computer anxiety has been researched upon extensively in IS and psychology, its role
in the nomological net of TAM has not been investigated.
Based on the general framework proposed, we hypothesize that general computer anxiety
will be an anchor exerting a negative influence on the perceived ease of use of a new system.
The theoretical underpinnings for such a link are drawn from classical theories of anxiety
(Epstein 1972; Philipi, Martin, and Meyers 1972) that suggest the consequences of anxiety
include a negative impact on cognitive responses, particularly process expectancies. In related
research, Morris et al. (1984) suggest that there are two key components of anxiety: cognitive

19
and emotional. The cognitive component leads to negative expectancies while the emotional
element leads to negative physiological reactions. Tobias (1979) argued that even though
anxiety is an affective state, its effects on behavior and performance are mediated by cognitive
processes.
Social cognitive theory suggests that anxiety and expectancies (e.g., efficacy, ease of use)
are reciprocal determinants (Bandura 1986). Specifically, depending on which of the two
variables serve as the stimulus, an effect on the other may be observed. Within the context of the
current work since the effects on perceived ease of use are being studied, anxiety is viewed as a
determinant of the process expectancy—i.e., perceived ease of use. Further evidence for the
impact of computer anxiety on perceived ease of use comes from prior research demonstrating an
anxiety-attitude link (e.g., Igbaria and Parasuraman 1989). On this basis, the current research
argues that the role of computer anxiety in the context of TAM will play out as an effect on
perceived ease of use, a belief shown to be closely related to attitude (e.g., Davis et al. 1989). In
this context, the other belief of TAM (i.e., perceived usefulness) may have outcome-oriented
anchors (see Venkatesh and Davis 2000), but perceived ease of use is expected to have a
process-oriented anchor (i.e., computer anxiety), with increasing levels of general computer
anxiety leading to lowering of system-specific perceived ease of use.
The impact of anxiety on individual attentional resource allocation strategies provides
additional causal bases for the negative influence of computer anxiety on perceived ease of use.
Anxiety, typically, has an adverse effect on the attention devoted to the task at hand (Eysenck
1979; Tobias 1979). From the perspective of resource allocation theory (e.g., Kanfer, Ackerman,
and Murtha 1994), it can be argued that some of the attentional resources will be directed to the
off-task activity of anxiety reduction, thus increasing the effort required to accomplish tasks.

20
Given that perceived ease of use is an individual judgment about the ease of behavioral
performance based on effort, higher levels of general computer anxiety are expected to cause
lowering of system-specific perceived ease of use.
Interrelationships Among Anchors
Prior research has established interrelationships among the different anchors, particularly
among internal control (computer self-efficacy), computer playfulness, and computer anxiety.
Computer self-efficacy and highly similar constructs (e.g., perceived knowledge, computer
confidence) have been related to computer playfulness (Webster and Martocchio 1992) and
computer anxiety (Compeau and Higgins 1995a; Hunt and Bohlin 1993; Igbaria and Ilvari 1995;
Loyd, Loyd, and Gressard 1987; Martocchio 1994; Heinssen et al. 1987). Although anxiety in a
variety of domains, including computer use, has been negatively related to self-efficacy, there is
empirical evidence to suggest that computer anxiety and computer self-efficacy are distinct
constructs (e.g., Compeau and Higgins 1995b). Also, the two constructs have been shown to
explain unique variance in key dependent variables such as behavior (e.g., Compeau and Higgins
1995b). Similarly, there is evidence to suggest that computer playfulness is related to anxiety
(Bozionelos 1997; Webster and Martocchio 1992). Due to the extensiveness and complexity of
the proposed model of determinants, the interrelationship among the different constructs is not
expounded upon.
iv
Thus, the current work is interested in the interrelationships among the
constructs from a rather limited perspective of discriminant validity across the constructs and the
unique variance in perceived ease of use explained by each construct.
Anchors and Adjustments Over Time: The Role of Experience
What will the role of anchors be over time? What will the adjustments be? Prior user
acceptance research has shown experience to have an important influence on several key

21
constructs and relationships (e.g., Davis et al. 1989; Taylor and Todd 1995; Szajna 1996).
Similarly, in the proposed model also, experience is expected to have direct and moderating
effects on constructs and relationships. In this regard, the impact of experience on the different
anchors and adjustments is expected to be different.
Prior research has shown that perceptions of internal control (computer self-efficacy) will
continue to be a determinant of perceived ease of use of a specific system even after significant
direct experience with the system (Venkatesh and Davis 1996). Building on this finding, the
current research expects that even though individuals may have acquired significant system
specific knowledge and experience, their perceived ease of use of the target system will continue
to draw from their general confidence in their computer related abilities. Similarly, even with
increasing experience, general computer anxiety is expected to continue to have an effect on
system-specific perceived ease of use. As user experience with the specific system increases, the
knowledge and anxiety related adjustment is expected to be objective usability. Prior TAM
research (Venkatesh and Davis 1996) has defined and operationalized objective usability
consistent with its conceptualization in human-computer interaction research (Card, Moran, and
Newell 1980). Objective usability is a construct that allows for a comparison of systems based
on the actual level (rather than perceptions) of effort required to complete specific tasks. The
role of direct behavioral experience and results of such experiences are expected to be important
in shaping system-specific perceived ease of use over time. This also follows from
attitude/intention theories (Doll and Ajzen 1992; Fazio and Zanna 1978a, 1978b, 1981), and
anxiety research (Cambre and Cook 1985) which suggests that actual behavioral experience
shapes beliefs such as perceived ease of use. For instance, even if an individual possesses low
computer self-efficacy and high computer anxiety which shaped initial perceived ease of use,

22
with increasing direct experience with the target system, s/he is expected to perceive the system
to be easy or hard to use partly depending on the extent to which the system is easy to use from
an objective standpoint.
The next adjustment with experience relates to external control. Before direct behavioral
experience with the actual system environment issues (e.g., login/access time, response time,
etc.) and organizational environment as it relates to the specific system (e.g., support staff), the
general perception of external control (facilitating conditions) was expected to serve as an
anchor. Since perceptions of external control are not directly related to the user interface of the
system, individuals will simply modify/change their original perceptions of external control to
reflect the organizational environment as it relates to the specific system and other aspects of the
system environment. In operational terms, the adjustment is expected to be quite simply a
change in the mean value of the perceptions of external control, thus suggesting that
experience will have a direct effect on perceptions of external control. Over the long term, the
relationship between external control and perceived ease of use is expected to continue as
individuals are expected to factor in external control when judging the ease of use of a system.
The role of intrinsic motivation as a determinant of system-specific perceived ease of use
is expected to change over time. Early perceptions of system-specific ease of use were said to be
anchored to general computer playfulness. With increasing experience, perceived ease of use is
expected to reflect the unique attributes of enjoyment as it relates to the user-system interaction.
A conceptualization of intrinsic motivation that is system-specific is perceived enjoyment.
Perceived enjoyment is adapted from Davis et al. (1992) and defined as the extent to which the
activity of using a specific system is perceived to be enjoyable in it's own right, aside from any
performance consequences resulting from system use. With increasing direct experience with

23
the target system, the role of general computer playfulness as a determinant of perceived ease of
use of the target system is expected to diminish, and system-specific perceived enjoyment is
expected to dominate. In many Windows-based systems, software manufacturers are attempting
to provide interfaces that are fun, “cute,” and tie into social functioning (e.g., the “animated
assistant icons” in Office 97). Such design features aim to create enjoyment albeit with the goal
of enhancing perceived ease of use of the specific system. We expect that with increasing
experience, system use may become more routinized, less challenging, and less discovery-
oriented. In such cases, the lack of enjoyment may cause system use to be perceived to be more
effortful.
Often, researchers have manipulated task difficulty and examined its effect on intrinsic
motivation (e.g., Hirst 1988) and concluded causal flow from perceptions of ease to intrinsic
motivation, consistent with Davis et al. (1992). In marked contrast to these prior findings, there
is some recent evidence that favors a causal flow from perceived enjoyment to perceived ease of
use (Venkatesh 1999). By manipulating the level of system-specific enjoyment through training,
not only was it found that perceived ease of use could be enhanced but also the salience of
perceived ease of use as a determinant of intention increased (Venkatesh 1999), thus suggesting
that perceived ease of use could certainly be influenced by system-specific perceived enjoyment.
The current research argues for a causal flow in keeping with this recent empirical evidence.
Thus, depending on the extent to which actual system use is perceived to be enjoyable or boring,
perceived ease of use of the target system may increase or decrease over time.
Summary
The current research proposed a theoretical framework (Figure 1), based on an anchoring
and adjustment perspective, to explain the determinants of perceived ease of use, a key driver of

24
user acceptance and usage of information technologies. The theoretical model (Figure 2) based
on the framework proposes control—internal (computer self-efficacy) and external (facilitating
conditions), intrinsic motivation (computer playfulness), and emotion (computer anxiety) as
general anchors that influence early perceptions of ease of use of a new system. With increasing
experience with the target system, an individual is expected to adjust her/his perceived ease of
use of the system. Specifically, the role of computer self-efficacy and computer anxiety are
expected to continue. The role of computer playfulness is expected to diminish over time, giving
way to system-specific perceived enjoyment. In addition, objective usability is expected to serve
as an adjustment for internal control and computer anxiety. Finally, facilitating conditions will
undergo a shift from being general perceptions and expectations of the system and organizational
environment to being system-specific.
METHOD
Three longitudinal field studies were conducted to test the model of determinants of
perceived ease of use. The studies were designed to sample for heterogeneity (see Cook and
Campbell 1979) in terms of industry and target system being introduced to end-users. In all
three studies, the use of the system was voluntary. Three measurements of user reactions were
made over a three-month period of time in each of the three studies. The first measurement was
following initial training (T1), the second measurement was after one month of use (T2), and the
third measurement was after three months of use (T3). All constructs were measured at T1, T2,
and T3 with one exception—objective usability was measured only at T1 because of the involved
nature of measurement that requires about 45 minutes of the subject’s time. Table 1 presents a
summary of the measurement.

25
Study 1
The subjects were seventy employees in a medium-sized retail electronic store. They
were being introduced to a new interactive online help desk system, which was to be used in
responding to customer queries received in-person and via telephone. Fifty-eight subjects
completed the study and provided usable responses at all three points of measurement. Prior to
the training, none of the subjects possessed specific knowledge about the system or how it
worked. The training was conducted by a group of three individuals unaware of the research or
its objectives.
Study 2
The subjects were one-hundred and sixty employees in a large real estate agency. They
were being introduced to a new multimedia system for property management, which was to be
used to manage all information related to new properties available for sale, properties sold in the
past, and to help customers. One-hundred and forty-five subjects completed the study and
provided usable responses at all three points of measurement. The subjects possessed no prior
(pre-study) knowledge about the system. Similar to the first study, the subjects participated in a
training program. The training was conducted by a group of three individuals who did not know
about the research or its objectives.
One point worth noting is that there were two separate groups of subjects in this
organization in that the two groups were introduced at different points in time. Forty-nine
subjects were introduced to the system first, and forty-one of them provided completed responses
at all points of measurement (Study 2a). One-hundred and eleven subjects in two different

26
branch offices were introduced to the system about a year after study 2a, and one-hundred and
four subjects in this group completed the study (Study 2b) at all three points of measurement.
Study 3
The subjects were fifty-two employees in a medium-sized financial services firm. The
payroll department of the organization was moving from a proprietary IBM-mainframe
environment to a PC-based (Windows95) environment for the company payroll application.
Forty-three subjects completed the study and provided usable responses at all three points of
measurement. As in the previous two studies, the subjects possessed no prior knowledge about
the system and were trained to use the system by three trainers unaware of the research or its
objectives.
Measurement
As mentioned earlier, all three studies in this research measured user reactions related to
a new system. User reactions were tracked over time as users progressed from being novices to
fairly experienced users of the system. The instrument primarily used validated items from prior
research (see Appendix 1 for list of items employed in this research). The TAM constructs of
perceived ease of use (EOU), perceived usefulness (U), and behavioral intention to use (BI) were
measured using scales adapted from Davis (1989) and Davis et al. (1989). Perceived
voluntariness of use (VOL) was measured using a scale adapted from Moore and Benbasat
(1991). This measure was treated as a check to ensure that the study contexts were perceived to
be voluntary by the users.
The anchors measured were perceptions of internal control (computer self-efficacy;
CSE), perceptions of external control (facilitating conditions; FC), intrinsic motivation
(computer playfulness; PLAY), and emotion (computer anxiety; CANX). Internal control

27
(computer self-efficacy) was measured by adapting the scale of Compeau and Higgins (1995a),
consistent with previous work on the determinants of perceived ease of use (Venkatesh and
Davis 1996). Perceptions of external control (facilitating conditions) were measured using the
scale adapted from Mathieson (1991) and Taylor and Todd (1995). Intrinsic motivation
(computer playfulness) was measured using the scale adapted from Webster and Martocchio
(1992). The only anchor for which a scale was not readily adaptable was emotion (computer
anxiety). Although there are several scales available to measure computer anxiety, the reliability
and validity of prior scales, including the widely used Computer Anxiety Rating Scale (CARS)
(Heinssen, et al. 1987), have been challenged (Compeau and Higgins 1995a; Ray and Minch
1990). Given some of the concerns regarding the multidimensionality of CARS, a new scale
(Brown and Vician 1997) is employed in this research. While their scale builds on CARS, it
addresses some of the problems of reliability and validity of the older scale.
The two adjustments measured were objective usability (OU) and perceived enjoyment
(ENJ). Objective usability was operationalized consistent with the keystroke model from
human-computer interaction research (Card et al. 1980) and prior user acceptance research
(Venkatesh and Davis 1996). The suggested guideline for operationalization of this construct is
to compute a novice to expert ratio of effort. Specifically, the time taken by an expert to perform
a set of tasks using the system in an error-free situation is compared with the time taken by a
novice (subject). In this research, following each training program, subjects were assigned a set
of tasks to be completed. The time taken by each individual subject to complete the tasks was
recorded by the system, which was then compared to the time taken by an expert to arrive at a
ratio that would serve as the measure of objective usability for the particular subject. The higher
the objective usability estimate (novice to expert ratio), the harder the system is to use from an

28
objective standpoint. Perceived enjoyment was measured using the scale adapted from Davis et
al. (1992); this scale was also used recently in organizational behavior research (Venkatesh and
Speier 1999).
A pre-test of the instrument was conducted to ensure that the items were adapted
appropriately to the current context. A group of thirty undergraduate students were chosen at
random to participate in the pre-test of the instrument. The reliability and validity of the scales
were consistent with prior research. Of particular interest to us was the reliability and validity of
the new computer anxiety scale. The Cronbach alpha estimate for reliability was 0.81.
Exploratory factor analysis using principal components analysis with direct oblimin rotation and
an extraction criterion of eigenvalue greater than one was conducted—the computer anxiety
items loaded on one factor with loadings greater than 0.70 and cross-loadings less than 0.25 on
factors related to the other anchors (i.e., computer self-efficacy, facilitating conditions, and
computer playfulness). In contrast, the scale of Heinssen et al. (1987) had a Cronbach alpha
estimate of 0.43 and loaded on three separate factors.
Next, a focus group of five business professionals evaluated the instrument. The reaction
to the instrument was largely positive. The key change made to the instrument following the two
pre-tests was the inclusion of titles for each of the constructs in Study 1, Study 2a, and Study 3.
While research on the topic of intermixing vs. grouping of items has suggested that grouping of
items may lead to inflated reliability and validity estimates (e.g., Budd 1987), there is some
work, including Budd (1987) that suggests that there is a possibility that intermixing of items in
the case of validated scales leads to measurement errors, confusion, and irritation among the
respondents (see Davis and Venkatesh 1996). Thus, in this research, items related to each
construct were grouped to avoid possible measurement errors, particularly since validity and

29
reliability of the different scales had been established by prior research and such a validation was
not a focus of this work. While there is empirical evidence in other contexts suggesting that
grouping and titling causes no negative consequences (Davis and Venkatesh 1996), this research
sought to eliminate the possibility of biases empirically as well. Therefore, construct titles were
not included in Study 2b that was conducted after the three aforementioned studies were
completed. The objective of intermixing items in Study 2b was to examine possible artifactual
inflation of reliability/validity, and path coefficients (because of grouping and titles) in Studies 1,
2a, and 3.
RESULTS
Prior to analyzing the data, we examined support for the assumption that the technology
introduction contexts were indeed voluntary, one of the boundary conditions of TAM. In all
studies at all points of measurement, the mean score of perceived voluntariness was greater than
6.0 on a 7-point scale with a standard deviation less than 0.5, supporting the idea that the users
indeed saw the usage contexts to be voluntary.
The structural equation modeling technique of Partial Least Squares (PLS) was used to
analyze the data. PLS analyzes measurement and structural models with multi-item constructs
that include direct, indirect, and interaction effects. There are several excellent examples of the
use of PLS in IS research (see Barclay, Higgins, and Thompson 1995; Chin, Marcolin, and
Newsted 1996; Compeau and Higgins 1995a, 1995b; Sambamurthy and Chin 1994). The
software package used to perform the analysis was PLS Graph, Version 2.91.03.04.
The measurement model was assessed separately for each of the studies (1, 2a, 2b, and 3)
at each of the three points of measurement, thus resulting in 12 models. All constructs in all
models satisfied the criteria of reliability and discriminant validity, therefore, no changes to the

30
constructs were required. This pattern was very consistent with expectations since all
measurement scales, with the exception of computer anxiety, had been tested and validated in
prior research. Since PLS-Graph does not produce loadings and cross-loadings, the DOS version
of PLS was used and the procedure LVPC was employed to generate the factor structure.
Appendix 2 reports the results of the measurement model for the data pooled across
organizations at T1 (the rationale for presenting pooled data is presented in the next paragraph).
The basic factor structure indicated all cross-loadings were lower than .35 in all studies at all
three points of measurement. This basic pattern was found in all studies at all points of
measurement. The reliability and discriminant validity coefficients were examined, and the
pattern of results in all studies at all points of measurement were supportive of high reliability
within constructs, and discriminant validity across constructs (square root of the shared variance
across items measuring a construct was higher than correlations across constructs). The results
pertaining to the measurement model are reported based on the data pooled across studies in the
next paragraph.
Given that the measurement models were found to be acceptable, we conducted tests to
examine whether the data could be pooled across studies. There were two key differences across
the studies that needed to be examined before pooling: (1) Was there any difference in the path
coefficients across each of the different studies given that different systems were being
introduced in each of the organizations? (2) Studies 1, 2a, and 3 employed an instrument with
titles whereas Study 2b had items intermixed across constructs—did the titles result in any
artifactual inflation of path coefficients? To address the first issue, the data were then pooled
across the four studies (1, 2a, 2b, and 3) and dummy variables (DUMMY1, DUMMY2, and
DUMMY3) were introduced and coded as (0,0,0), (1,0,0), (0,1,0), and (0,0,1) to represent each

31
of the studies. The models were analyzed including interaction terms of all constructs with the
dummy variables—e.g., CSE X DUMMY1, CSE X DUMMY2, etc. (see Chin 1996 for a
discussion). Non-significant interaction terms suggested that the models were statistically
equivalent across sites at each of the three points of measurement (Pindyck and Rubenfeld 1981).
To address the second issue, the data were pooled across sites, and one dummy variable,
STUDY_DUMMY, was introduced--0 to represent studies 1, 2a, and 3 (i.e., where the
instrument was administered with titles) and 1 to represent study 2b (i.e., where the instrument
was administered with items being intermixed). The models were analyzed including interaction
terms of all constructs with the variable STUDY_DUMMY—e.g., CSE X STUDY_DUMMY,
PEC X STUDY_DUMMY, etc. Non-significant interaction terms suggested statistical
equivalence across studies 1, 2a, 3, and 2b at each of the three points of measurement.
Therefore, the data were pooled across the different sites and the measurement model was re-
estimated. The reliability and discriminant validity coefficients are reported in Tables 2(a), 2(b),
and 2(c).
Once the measurement model corresponding to the pooled data set was found to be
acceptable, the structural model results were examined at all three points of measurement.
Figures 3(a), 3(b), and 3(c) summarize the results at each time period. TAM was strongly
supported at all three points of measurement, consistent with the vast body of prior research on
the model, with perceived ease of use and perceived usefulness explaining about 35% of the
variance in intention. In order to examine full mediation by perceived ease of use, additional
models were tested by including direct links from the proposed determinants to intention.
v
The
effects of all proposed determinants of perceived ease of use at all points of measurement were
fully mediated (by perceived ease of use) and no direct effects were observed on intention. Of

32
particular interest from the perspective of this research is that perceived ease of use was found to
have a direct effect and indirect effect (via perceived usefulness) on intention at all three points
of measurement. The results indicated that the proposed framework and model of determinants
of perceived ease of use were strongly supported.
vi
It is particularly worth noting that
perceptions of external control not having any direct effect on intention runs counter to TPB,
although it is consistent with the proposed model. As expected, at T1, the proposed anchors
were the only determinants of perceived ease of use, with the variance explained being 40%.
With increasing experience (i.e., T2 and T3), adjustments were found to be play a key role in
determining perceived ease of use, with the variance explained increasing to up to 60%. The
current work thus explains twice as much variance in perceived ease of use when compared to
Venkatesh and Davis (1996), the previous model of the determinants of perceived ease of use.
DISCUSSION
Several key findings emerged from the current work. There was significant support for
the model of the determinants of perceived ease of use with the hypothesized determinants
playing a role as expected over time with increasing experience with the target system. We
found that control (internal and external conceptualized as computer self-efficacy and facilitating
conditions respectively), intrinsic motivation (computer playfulness), and emotion (computer
anxiety) serve as anchors that users employ in forming perceived ease of use about a new
system. With increasing direct experience with the target system, the adjustments playing a role
were objective usability, perceptions of external control as it related to the specific system
environment, and perceived enjoyment from system use. Interestingly, with increasing
experience, although adjustments played an important role in determining system-specific
perceived ease of use, the general beliefs regarding computers and computer use continued to be

33
important factors driving system-specific perceived ease of use. In fact, certain general anchors
(computer self-efficacy and facilitating conditions) were stronger determinants than were
adjustments resulting from the user-system interaction.
Theoretical Contributions and Implications
From the perspective of user acceptance, the current research significantly expands our
understanding of factors influencing user acceptance. Research on the Technology Acceptance
Model has led to various applications and replications. However, with the exception of some of
Davis’ work (e.g., Venkatesh and Davis 1996), research has not focused on understanding the
determinants of TAM’s key constructs. In this research, we attempt to go beyond the
determinants of perceived ease of use identified by Venkatesh and Davis (1996). The
determinants of perceived ease of use were developed and justified from a theoretical standpoint
and validated empirically in three separate longitudinal field studies. By significantly increasing
the variance explained in perceived ease of use, a much clearer picture of the factors influencing
user judgments of system-specific ease of use has now emerged. Specifically, the proposed
model of determinants explained up to a total of 60% variance in perceived ease of use, thus
doubling our current understanding. The model presents an exposition of how the different
determinants influence perceived ease of use and how their influence is affected by increasing
user experience with a target system, thus providing researchers and practitioners with an in-
depth understanding of the dynamics underlying the formation and change of perceived ease of
use of a specific system.
The findings suggest that initial drivers of system-specific perceived ease of use are
largely individual difference variables and situational characteristics, whose effect becomes
stronger with experience. With increasing user experience with the target system, characteristics

34
of the user-system interaction play a role in driving perceived ease of use of the target system,
although their effect is still not as strong as the system-independent constructs. This is a very
powerful result because it suggests that long-term perceived ease of use of a specific system are
strongly anchored to general beliefs about computers that are system-independent and can be
measured without much experience with the target system. This pattern of findings runs
somewhat counter to what would be predicted by attitude/intention theories that suggest that
experience will play a very key role in shaping attitudinal beliefs. One potential explanation for
the current findings is that in the case of perceived ease of use about a specific system,
individuals are driven by their general beliefs, even after significant direct experience with the
system, as long as the specific system fits with the individual’s broad expectations and industry-
standard user interface conventions. It is, therefore, possible that adjustments will play a more
important role in influencing perceived ease of use of the new system if one’s continuing
experience is inconsistent with the anchors. When experiences are consistent with expectations,
there is no adjustment necessary (see also Szajna and Scamell 1993). It is also possible that the
users under-adjusted and did not fully taken their experiences into account, thus explaining the
relatively weak influence of adjustments. Such an under-adjustment could be attributed to the
short time-frame (three months) of the current research. In understanding the relative influence
of anchors and adjustments, given the nascent state of research on determinants of perceived ease
of use, we examined anchors and adjustments as main effects, something that is consistent with
prior work in this area (Venkatesh and Davis 1996). Future work should examine possible
moderation of anchors by adjustments.
The proposed theoretical framework and model of determinants of perceived ease of use
integrates important constructs from other user acceptance models/research into the nomological

35
net of TAM by positioning them as determinants of perceived ease of use. By demonstrating
that constructs used in prior user acceptance research are indeed determinants of perceived ease
of use, the current research presents a more complete, coherent and unified view of user
acceptance with TAM as the focal point. Also, by validating one of TAM’s fundamental
assumptions of mediation of external variables by perceived ease of use, the robustness of TAM
as a powerful model to understand and predict user acceptance is further established. This
represents an important theoretical contribution since there has been limited research (e.g., Davis
et al. 1992; Venkatesh and Davis, 1996) focused on testing the core assumption of mediation of
the effect of other constructs on intention by the TAM constructs of perceived ease of use and
perceived usefulness. The current work focused on the determinants of perceived ease of use,
one of the two key drivers of acceptance per TAM. In related work, the determinants of
perceived usefulness have been identified (Venkatesh and Davis 2000). It is important to test
these two models in one study in order to present an integrated view of TAM and its
determinants.
Limitations and Additional Future Research Directions
While this research possesses the advantage of field data from three different
organizations, one potential direction for future research is to test the model in experimental
settings as a way to provide internal validity for the model emerging from this work. The current
work presented a cross-sectional analysis of the data, thus relying to a great extent on the theory
to support causality. Future work should focus on a longitudinal analysis in order to strengthen
the direction of causality proposed by the model. The current work was conducted in voluntary
settings, which is quite suitable to the type of models/theories typically employed in user

36
acceptance research. However, future research should examine mandatory usage contexts to test
the boundary conditions of the proposed model.
There are some limitations related to measurement that should be noted. Although most
scales employed in this work had been validated in prior research, the computer anxiety scale
had not been previously used. The scale exhibited high reliability and validity in all studies at all
points of measurement, but additional work is certainly warranted to further validate the scale
and the role of the construct using this scale in more traditional anxiety research and other user
acceptance research. In the current work, objective usability was operationalized using only one
measurement during training, and the operationalization was consistent with prior human-
computer interaction and user acceptance research. Future research should devise methods of
measuring objective usability over time. However, even as it stands now, objective usability
measured in the very early stages of user interaction appears to serve as a very useful predictor of
long term perceptions of ease of use of a target system. Another limitation relates to the
measurement of intention. The scales do not specify a time frame for use, largely because
participating organizations were not specific at the outset about when subsequent administrations
of the survey could be conducted. Also, usage behavior was not measured. One of the
fundamental assumptions of research in the area of user acceptance is that the determinant
constructs being studied are good predictors of usage behavior. There has been some concern
about the predictive ability of TAM (see Straub, Limayem, and Karahanna-Evaristo 1995).
However, given that there is a significant body of research in IS (Taylor and Todd 1995),
organizational behavior (Venkatesh and Morris 2000; Venkatesh and Speier 1999; Venkatesh et
al. 2000), and psychology (see Sheppard et al. 1988 for a meta-analysis) supporting intention as a
predictor of actual behavior, the issue is somewhat less critical. Future research should

37
nevertheless examine the findings of the current work in a context where usage can be measured
in order to add additional credibility to the model.
Practical Implications
System designers typically attempt to build systems that are easy to use while providing
the functionality that the users need to accomplish their tasks. While user interface design is the
typical focal point to enhance user acceptance, this research shows that there are multiple factors
not directly related to the user-system interaction that are perhaps more important (e.g., computer
self-efficacy). While a large amount of time during system design and development is typically
spent on the user interface, this research suggests that practitioners should spend more time
creating a favorable impact on system-independent factors, which have clearly been shown to be
more important than user perceptions that relate to the user-system interaction in determining
perceived ease of use of a specific system. This is particularly important since at all stages of
user experience with a system, general, system-independent constructs play a stronger role than
constructs that are a result of the user-system interaction.
The next steps should focus on designing and testing interventions to enhance perceived
ease of use by targeting the identified determinants with an eye toward fostering increased
technology acceptance in the workplace. As discussed earlier, there exists extensive research in
IS and psychology that describes methods of enhancing self-efficacy and reducing anxiety–
practitioners should attempt to adapt such interventions to end-user training contexts. Similarly,
research should focus on designing managerial interventions that will provide facilitating
conditions that favor the creation of positive perceptions about the ease of use of a specific
system via perceptions of external control. Researchers and practitioners should attempt to better
leverage the individual difference variable of computer playfulness and system-specific

38
perceived enjoyment during the design and development phases of system building, and attempt
to incorporate it into end-user training situations. In general, practitioners should design
interventions directed at the various determinants of perceived ease of use that go beyond
traditional training methods, which typically aim to impart only conceptual and procedural
knowledge about a specific system. Organizations should consider putting in place general
computer training programs that target increasing computer awareness, enhancing computer self-
efficacy, and reducing computer anxiety among employees. Such training programs combined
with appropriate facilitating conditions should pave the path for acceptance and usage of new
systems. In fact, organizations will benefit particularly from system-specific training
interventions that enhance user perceptions about the specific system and their general beliefs
about new information technologies (see Compeau 1992).
One of the areas that has not been exploited in practice is the potential for intrinsic
motivation to enhance user acceptance and usage. Much prior research (Davis et al. 1992;
Malone 1981a, 1981b; Webster and Martocchio 1992; Venkatesh and Speier 1999) has found
intrinsic motivation to be an important factor influencing user acceptance and learning. This
research has further refined our understanding in this regard by suggesting that general computer
playfulness and perceived enjoyment are determinants of perceived ease of use. One example is
“fun icons” like the ones introduced in MS-Office 97. A similar example is the use of “warm
and fuzzy” screen savers (e.g., flashing cartoons on the screen, some action related to your
favorite basketball team, etc.) as a way to cause perceived ease of use of specific systems (used
by the individual) to be more favorable. Recent work in IS (Venkatesh 1999) and organizational
behavior (Venkatesh and Speier 1999) suggests that training environments can be tailored to
exploit intrinsic motivation with a view toward enhancing acceptance and usage of new systems.

39
This may be important from the standpoint of breaking the monotony in the extensive use of
software in today’s workplace.
CONCLUSIONS
This research investigated the determinants of perceived ease of use, a key driver of
technology acceptance, adoption, and usage behavior. Based on a field investigation in three
different organizations, strong support was found for the anchoring and adjustment model of
determinants. We found that an individual’s general beliefs regarding computers were the
strongest determinants of system-specific perceived ease of use, even after significant direct
experience with the target system. The findings of the current work point to the need for an
increased focus on individual difference variables in order to enhance user acceptance and usage,
rather than over-emphasizing system-related perceptions and design characteristics as has been
done in much prior information systems and human-computer interaction research. The current
work calls for practitioners to develop and implement general training programs on computer
skills as they will have a strong influence on the acceptance and sustained usage of new systems.

40
REFERENCES
Adams, D. A., Nelson, R. R., and Todd, P. A., “Perceived Usefulness, Ease of Use, and Usage of
Information Technology: A Replication,” MIS Quarterly, 16, 2 (1992), 227-250.
Ajzen, I., “From Intentions to Actions: A Theory of Planned Behavior,” in Kuhl, J., and
Beckmann J. (Eds.), Action Control: From Cognition to Behavior, Springer Verlag, New
York, 1985, 11-39.
Ajzen, I., and Fishbein, M., Understanding Attitudes and Predicting Social Behavior, Englewood
Cliffs, Prentice-Hall, New Jersey, 1980.
Ajzen, I., “The Theory of Planned Behavior,” Organizational Behavior and Human Decision
Processes, 50, 2 (1991), 179-211.
Anderson, A. A., “Predictors of Computer Anxiety and Performance in Information Systems,”
Computers in Human Behavior, 12, 1 (1996), 61-77.
Bandura, A., Social Foundations of Thought and Action: A Social Cognitive Theory, Prentice-
Hall, Englewood Cliffs, NJ, 1986.
Barclay, D., Higgins, C., and Thompson, R.. “The Partial Least Squares Approach to Causal
Modeling: Personal Computer Adoption and Use as an Illustration,” Technology Studies,
Special Issue on Research Methodology, 2,2 (1995), 285-324.
Beale, D. A., and Manstead, A. S. R., “Predicting Mothers’ Intentions to Limit Frequency of
Infants’ Sugar Intake: Testing the Theory of Planned Behavior,” Journal of Applied
Social Psychology, 21, 5 (1991), 409-431.
Bergeron, F., Rivard, S., and De Serre, L. “Investigating the Support Role of the Information
Center,” MIS Quarterly, 14, 3 (1990), 247-259.
Bettman, J. R., and Sujan, M., “Effects of Framing on Evaluations of Comparable and
Noncomparable Alternatives by Expert and Novice Consumers,” Journal of Consumer
Research, 14, 2 (1987), 141-154.
Bohlin, R. M., and Hunt N. P., “Course Structure Effects on Students’ Computer Anxiety,
Confidence and Attitudes,” Journal of Educational Computing Research, 13, 3 (1995),
263-270.
Bozionelos, N., “Psychology of Computer Use: XXXIX. Prevalence of Computer Anxiety in
British Managers and Professionals,” Psychological Reports, 78, 3 (1996), 995-1002.

41
Bozionelos, N., “Psychology of Computer Use: XLV. Cognitive Spontaneity as a Correlate of
Computer Anxiety and Attitudes toward Computer Use,” Psychological Reports, 80, 2
(1997), 395-402.
Brown, S. A., and Vician, M., “Understanding Computer Anxiety and Communication
Apprehension as Antecedents to Student Experiences with Technology-Supported
Learning Environments,” Working Paper, Indiana University, Bloomington, Indiana,
1997.
Budd, R. J., “Response Bias and the Theory of Reasoned Action,” Social Cognition, 5, 2 (1987),
95-107.
Cambre, M. A., and Cook D. L., “Computer Anxiety: Definition, Measurement, and Correlates,”
Journal of Educational Computing Research, 1, 1 (1985), 37-54.
Card, S. K., Moran, T. P., and Newell, A., “The Keystroke-Level Model for User Performance
Time with Interactive Systems,” Communications of the ACM, 23, (1980), 396-410.
Chan, D. K. S., and Fishbein, M., “Determinants of College Women’s Intention to Tell Their
Partners to Use Condoms,” Journal of Applied Social Psychology, 23, 18 (1993), 1455-
1470.
Chin, W. W.. “The Measurement and Meaning of IT Usage: Reconciling Recent Discrepancies
Between Self Reported and Computer Recorded Usage”, in B. A. Aubert and W. W. Chin
(Eds.), Administrative Sciences Association of Canada - 24th Conference, IS Division
Proceedings, (1996), 65-74.
Chin, W. W., and Gopal, A., “An Examination of the Relative Importance of Four Belief
Constructs on the GSS Adoption Decision: A Comparison of Four Methods,”
Proceedings of the 26th Hawaii International Conference on System Sciences, (1993),
548-557.
Chin, W. W., Marcolin, B. L., and Newsted, P. R. (1996). A Partial Least Squares Latent
Variable Modeling Approach for Measuring Interaction Effects: Results from a Monte
Carlo Simulation Study and Voice Mail Emotion/Adoption Study. Proceedings of the 17th
International Conference on Information Systems, Cleveland, Ohio. December, 1996.
Chin, W. W., and Todd, P. A., “On the Use, Usefulness, and Ease of Use of Structural Equation
Modeling in MIS Research: A Note of Caution,” MIS Quarterly, 19, 2 (1995), 237-246.
Chu, P. C., and Spires, E. E., “Validating the Computer Anxiety Rating Scale: Effects of
Cognitive Style and Computer Courses on Computer Anxiety,” Computers in Human
Behavior, 7, 1/2 (1991), 7-21.
Compeau, D. R. Individual Reactions to Computer Technology, Unpublished Doctoral
Dissertation, University of Western Ontario, 1992.

42
Compeau, D. R., and Higgins, C.A., “Application of Social Cognitive Theory to Training for
Computer Skills,” Information Systems Research, 6, 2 (1995a), 118-143.
Compeau, D. R., and Higgins, C. A., “Computer Self-Efficacy: Development of a Measure and
Initial Test,” MIS Quarterly, 19, 2 (1995b), 189-211.
Cook, T. D., and Campbell, D. T., Quasi-Experimentation: Design and Analysis Issues for Field
Settings, Houghton Mifflin Company, Boston, 1979.
Crable, E. A., Brodzinski, J. D., Scherer, R. F., and Jones, P. D., “The Impact of Cognitive
Appraisal, Locus of Control, and level of Exposure on the Computer Anxiety of Novice
Computer Users,” Journal of Educational Computing Research, 10, 4 (1994), 329-340.
Cragg, P. B., and King, M. “Small-firm Computing: Motivators and Inhibitors. MIS Quarterly,
17 (1993), 47-60.
Davis, F. D., “Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information
Technology,” MIS Quarterly, 13, 3 (1989), 319-339.
Davis, F. D., “User Acceptance of Information Technology: System Characteristics, User
Perceptions and Behavioral Impacts,” International Journal of Man-Machine Studies, 38,
3 (1993), 475-487.
Davis, F. D., Bagozzi, R. P., and Warshaw, P. R., “User Acceptance of Computer Technology: A
Comparison of Two Theoretical Models,” Management Science, 35, 8 (1989), 982-1002.
Davis, F. D., Bagozzi, R. P., and Warshaw, P. R., “Extrinsic and Intrinsic Motivation to Use
Computers in the Workplace,” Journal of Applied Social Psychology, 22, 14 (1992),
1111-1132.
Davis, F. D., and Venkatesh, V., “A Critical Assessment of Potential Measurement Biases in the
Technology Acceptance Model: Three Experiments,” International Journal of Human-
Computer Studies, 45, 1 (1996), 19-45.
DeCharms, R., Personal Causation: The Internal Affective Determinants of Behavior, Academic
Press, New York, 1968.
Deci, E. L., Intrinsic Motivation, Plenum Press, New York, 1975.
Deci, E. L., and Ryan, R. M. “The Support of Autonomy and the Control of Behavior,” Journal
of Personality and Social Psychology, 53, 6 (1987), 1024-1037.
de Vries, H., Dijkstra, M., and Kuhlman, P., “Self-efficacy: The Third Factor Besides Attitude
and Subjective Norm as a Predictor of Behavioural Intentions,” Health Education
Research, 3, 3 (1988), 273-282.

43
Doll J., and Ajzen I. “Accessibility and Stability of Predictors in the Theory of Planned
Behavior,” Journal of Personality and Social Psychology, 63, 5 (1992), 754-765.
Elasmar, M. G., and Carter M. E., “Use of E-mail by college Students and Implications for
Curriculum,” Journalism and Mass Communication Educator, 51, 2 (1996), 46-54.
Emanuele, S., Dale, A., and Klions H. L., “Psychology of Computer Use: XLII. Problem Solving
and Humor as a Function of Computer Anxiety,” Perceptual and Motor Skills, 84, 1
(1997), 147-156.
Epstein, “The Nature of Anxiety with Emphasis upon Its Relationship to Expectancy,” in C. D.
Spielberger (Ed.) Current Trends in Theory and Research, Vol. II, Academic Press, New
York, 1972.
Eysenck, M. W., “Anxiety: A Reconceptualization,” Journal of Research in Personality, 13, 4
(1979), 363-385.
Fazio, R. H., and Zanna, M., “Direct Experience and Attitude-behavior Consistency,” in L.
Berkowits (Ed.), Advances in Experimental Social Psychology, 14, Academic press, San
Diego, CA, 1981, 161-202.
Fazio, R. H., and Zanna, M., “Attitudinal Qualities Relating to the Strength of the Attitude-
behavior Relationship,” Journal of Experimental Social Psychology, 14, 4 (1978a), 398-
408.
Fazio, R. H., and Zanna, M., “On the Predictive Validity of Attitudes: The Role of direct
Experience and Confidence,” Journal of Personality, 46 (1978b), 228-243.
Fishbein, M., and Ajzen, I., Belief, Attitude, Intention and Behavior: An Introduction to Theory
and Research, Reading, Addision-Wesley, MA, 1975.
Gattiker, U. E., “Computer Skills Acquisition: A Review and Future Directions for Research,”
Journal of Management, 18, 3 (1992), 547-574.
Gefen, D., and Straub, D. W., “Gender Differences in the Perception and Use of E-mail: An
Extension to the Technology Acceptance Model,” MIS Quarterly, 21, 4 (1997), 389-400.
Gould, J. D., and Lewis, C. “Designing for Usability: Key Principles and What Designers
Think,” Communications of the ACM, 28, 3 (1985), 300-311.
Heinssen, R. K. Jr., Glass, C. R., and Knight, L. A., “Assessing Computer Anxiety: Development
and Validation of the Computer Anxiety Rating Scale,” Computers in Human Behavior,
3, 1 (1987), 49-59.
Helson, H. Adaption-Level Theory. Harper and Row, New York, 1964.

44
Hendrickson, A. R., Massey, P. D., and Cronan, T. P., “On the Test-Retest Reliability of
Perceived Usefulness and Perceived Ease of Use Scales,” MIS Quarterly, 17, 2 (1993),
227-230.
Hirst, M. K., “Intrinsic Motivation as Influenced by Task Interdependence and Goal Setting,”
Journal of Applied Psychology, 73, 1 (1988), 96-101.
Howard, G. S., and Smith, R. D., “Computer Anxiety in management: Myth or Reality?”
Communications of the ACM, 29, 7 (1986), 611-615.
Hunt, N. P., and Bohlin, R. M., “Teacher Education Students’ Attitudes toward Using
Computers,” Journal of Research on Computing in Education, 25, 4 (1993), 487-497.
Igbaria, M., and Parasuraman, S., “A Path Analytic Study of Individual Characteristics,
Computer Anxiety, and Attitudes Toward Microcomputers,” Journal of Management, 15,
3 (1989), 373-388.
Igbaria, M., and Chakrabarti, A., “Computer Anxiety and Attitudes towards Microcomputer
Use,” Behaviour and Information Technology, 9, 3 (1990), 229-241.
Igbaria, M., and Iivari, J., “The Effects of Self-Efficacy on Computer Usage,” OMEGA
International Journal of Management Science, 23, 6 (1995), 587-605.
Igbaria, M., Zinatelli, N., Cragg, P., and Cavaye, A. L. M., “Personal Computing Acceptance
Factors in Small firms: A Structural Equation Model,” MIS Quarterly, 21, 3 (1997), 279-
305.
Johansen, R. and Swigart, R., Upsizing the Individual in the Downsized Organization: Managing
in the Wake of Reengineering, Globalization, and Overwhelming Technological Change.
Reading, MA: Addison-Wesley (1996).
Johnston, D. C., “Computers Clogged, IRS Seeks to Hire Outside Processors,” New York Times,
Jan. 31, 1997, A1, A18.
Harrison, D. A., Mykytyn, P. P., and Riemenschneider, C. K. “Executive Decisions about
Adoption of Information Technology in Small Business: Theory and Empirical Tests,”
Information Systems Research, 8 (1997), 171-195.
Kanfer, R., Ackerman, P. L., and Murtha, T. C., “Goal Setting, Conditions of Practice, and Task
Performance: A Resource Allocation Perspective,” Journal of Applied Psychology, 79,
(1994), 826-835.
Keeler, C. M., and Anson, R., “An Assessment of Cooperative Learning Used for Basic
Computer Skills Instruction in the College Classroom,” Journal of Educational
Computing Research, 12, 4 (1995), 379-393.

45
Kimieck, J., “Predicting Vigorous Physical Activity of Corporate Employees: Comparing the
Theories of Reasoned Action and Planned Behavior,” Journal of Sport and Exercise
Psychology, 14, 2 (1992), 192-206.
Landauer, T. K., The Trouble With Computers: Usefulness, Usability, and Productivity.
Cambridge, MA: MIT Press, (1995).
Leso, T., and Peck, K. L., “Computer Anxiety and Different Types of Computer Courses,”
Journal of Educational Computing Research, 8, 4 (1992), 469-478.
Liebert, R. M., and Morris, L. W., “Cognitive and Emotional Components of Test Anxiety: A
Distinction and Some Initial Data,” Psychological Reports, 20, 3 (1967), 975-978.
Loyd, B. H., Loyd, D. E., and Gressard, C., “Gender and Computer Experience as Factors in the
Computer Attitudes of Middle School Students,” Journal of Early Adolescence, 7, 1
(1987), 13-19.
Malone, T. W., “Toward a Theory of Intrinsically Motivating Instruction,” Cognitive Sciences, 5,
4 (1981a), 333-369.
Malone, T. W., “What Makes Computer Games Fun?” Byte, 6, 12 (1981b), 258-278.
Marcoulides, G. A., Mayes, B. T., and Wiseman, R. L., “Measuring Computer Anxiety in the
Work Environment,” Educational Psychological Measurement, 55, 5 (1995), 804-810.
Martocchio, J. J., “Effects of Conceptions of Ability on Anxiety, Self-efficacy, and Learning in
Training,” Journal of Applied Psychology, 79, 6 (1994), 819-825.
Mathieson, K., “Predicting User Intentions: Comparing the Technology Acceptance Model with
the Theory of Planned Behavior,” Information Systems Research, 2, 3 (1991), 173-191.
Maurer, M. M., “Computer Anxiety Correlates and What They Tell Us: A Literature Review,”
Computers in human Behavior, 10, 3 (1994), 369-376.
McCaul, K. D., O’Neill, H. K., and Glasgow R. E., “Predicting the Performance of Dental
Hygiene Behaviors: An Examination of the Fishbein and Ajzen Model and Self-Efficacy
Expectations,” Journal of Applied Social Psychology, 18, 2 (1988), 114-128.
Mervis, C. B., and Rosch, E., “Categorization of Natural Objects,” Annual Review of
Psychology, 32 (1981), 89-115.
Moore, G. A., Crossing the Chasm: Marketing and Selling High-Tech Products to Mainstream
Customers. New York: Harper-Collins, (1991).
Moore, G. C., and Benbasat, I., “Development of an Instrument to Measure the Perceptions of
Adopting an Information Technology Innovation," Information Systems Research, 2, 3
(1991), 192-222.

46
Morris, L. W., Davis, M. A., and Hutchings, C. H., “Cognitive and Emotional Components of
Anxiety: Literature Review and a Revised Worry-Emotionality Scale,” Journal of
Educational Psychology, 73, 4 (1984), 541-555.
Morris, M. G. and Venkatesh, V.
Morrow, P. C., Prell, E. R., and McElroy, J. C., “Attitudinal and Behavioral Correlates of
Computer Anxiety,” Psychological Reports, 59, 3 (1986), 1199-1204.
Norman, D. A., Things That Make Us Smart: Defending Human Attributes in the Age of the
Machine. Reading, MA: Addison-Wesley, (1993).
Northcraft, G. B., and Neale, M. A., “Experts, Amateurs, and Real Estate: An Anchoring-and-
Adjustment Perspective on Property Pricing Decisions,” Organizational Behavior and
Human Decision Processes, 39, (1987), 84-97.
Parasuraman, S., and Igbaria, M., “An Examination of Gender Differences in the Determinants
of Computer Anxiety and Attitudes Toward Microcomputers Among Managers,”
International Journal of Man-machine Studies, 32, 3 (1990), 327-340.
Payne, J. W., Bettman, J., and Johnson, E. J., The Adaptive Decision-maker. Cambridge University
Press, New York, 1993.
Pindyck, R. S., and Rubenfeld, D. L., Econometric Models and Economic Forecasts, McGraw-
Hill Book Company, New York, 1981.
Popovich, P., Hyde, K., Zakrajesek, T., and Blumer, C., “The Development of the Attitudes
Towards Computer Usage Scale,” Educational and Psychological Measurement, 47, 1
(1987), 261-269.
Philipi, B. N., Martin, R. P., and Meyers, J., “Interventions in relation to Anxiety in School,” in
C. D. Spielberger (Ed.) Current Trends in Theory and Research, Vol. II, Academic Press,
New York, 1972.
Ray, N. M., and Minch, R. P., “Computer Anxiety and Alienation: Toward a Definitive and
Parsimonious Measure,” Human Factors, 32, 4 (1990), 477-491.
Reznich, C. B., “Applying Minimalist Design Principles to the Problem of Computer Anxiety,”
Computers in Human Behavior, 12, 2 (1996), 245-261.
Ronis, D. L., and Kaiser M. K., “Correlates of Breast Self-examination in a Sample of College
Women: Analyses of Linear Structural Relations,” Journal of Applied Social Psychology,
19, 13 (1989), 1068-1084.

47
Sambamurthy, V. and Chin, W., “The Effects of Group Attitudes Toward GDSS Designs on the
Decision-making Performance of Computer-supported Groups,” Decision Sciences, 25, 2,
(1994): 215-242.
Schifter, D. E., and Azjen, I., “Intention, Perceived Control, and Weight Loss: An Application of
the Theory of Planned Behavior,” Journal of Personality and social Psychology, 49, 3
(1985), 843-851.
Schuh, K. L., “The Lecture Classroom Environment and its Effects on Change in Computer
Anxiety of Students Taking Computer Proficiency Classes,” Journal of Educational
Computing Research, 15, 3 (1996), 241-259.
Scott, C. R., and Rockwell, S. C., “The Effect of Communication, Writing, and Technology
Apprehension on Likelihood to Use new Communication Technologies,” Communication
Education, 46, 1 (1997), 44-62.
Segars, A. H., and Grover, V., “Re-Examining Perceived Ease of Use and Usefulness: A
Confirmatory Factor Analysis,” MIS Quarterly, 17, 4 (1993), 517-525.
Sheppard, B. H., Hartwick, J., and Warshaw, P. R., “The Theory of Reasoned Action: A Meta-
Analysis of Past Research with Recommendations for Modifications and Future
Research,” Journal of Consumer Research, 15, (1988), 325-343.
Sichel, D. E., The Computer Revolution: An Economic Perspective. Washington, DC: The
Brookings Institution, (1997).
Simonson, M. R., Maurer, M., Montag-Torardi, M., and Whitaker, M., “Development of a
Standardized Test of Computer Literacy and a Computer Anxiety Index,” Journal of
Educational Computing Research, 3, 2 (1987), 231-247.
Slovic, P., and Lichtenstein, S., “Comparison of Bayesian and Regression Approaches to the
Study of Information Processing in Judgement,” Organizational Behavior and Human
Performance, 6, (1971), 641-744.
Sparks, P., “Attitudes towards Food: Applying, Assessing and Extending the Theory of Planned
Behaviour,” in D. R. Rutter and L. Quine (Eds.), The Social Psychology of Health and
Safety: European Perspectives, Avebury Press, Aldershot, England, 1994, 25-46.
Sparks, P., Guthrie, C. A., and Shepherd R., “The Dimensional Structure of the Perceived
Behavioral Control Construct,” Journal of Applied Social Psychology, 27, 5 (1997), 418-
438.
Straub, D., Limayem, M., and Karahanna-Evaristo, E., “Measuring System Usage: Implications
for IS Theory Testing,” Management Science, 41, 8 (1995), 1328-1342.
Streitfeld, B., and Wilson, M. “The ABCs of Categorical Perception,” Cognitive Psychology, 18
(1986), 432-451.

48
Subramanian, G. H., “A Replication of Perceived Usefulness and Perceived Ease of Use
Measurement,” Decision Sciences, 25, 5/6 (1994), 863-874.
Szajna, B., “Software Evaluation and Choice: Predictive Validation of the Technology
Acceptance Instrument,” MIS Quarterly, 18, 3 (1994), 319-324.
Szajna, B., “Empirical Evaluation of the Revised Technology Acceptance Model,” Management
Science, 42, 1 (1996), 85-92.
Szajna, B., and Scamell, R.W. “The Effects of Information System User Expectations on Their
Performance and Perceptions,” MIS Quarterly, 17 (1993), 493-516.
Taylor, S., and Todd, P.A., “Understanding Information Technology Usage: A Test of
Competing Models,” Information Systems Research, 6, 2 (1995), 144-176.
Terry, D. J., “Coping Resources and Situational Appraisals as Predictors of Coping Behaviour,”
Personality and Individual Differences, 12, 10 (1991), 1031-1047.
Terry, D. J., “Self-efficacy Expectancies and the Theory of Reasoned Action,” in D. J. Terry, C.
Gallois, and M. McCamish (Eds.), The Theory of Reasoned Action: Its Application to
AIDS-preventive Behaviour, Pergamon, Oxford, 1993.
Terry, D. J., “The Determinants of Coping: The Role of Stable and Situational Factors,” Journal
of Personality and Social Psychology, 66, 5 (1994), 895-910.
Terry, D. J., and O’Leary, J. E., “The Theory of Planned Behaviour: The Effects of Perceived
Behavioural Control and Self-efficacy,” British Journal of Social Psychology, 34, 2
(1995), 199-220.
Thompson, R. L., Higgins, C. A., and Howell, J. M., “Personal Computing: Toward a
Conceptual Model of Utilization,” MIS Quarterly, 15, 1 (1991), 124-143.
Tobias, S., “Anxiety Research in Educational Psychology,” Journal of Educational Psychology,
71, 5 (1979), 573-582.
Todd, P., and Benbasat, I. “An Experimental Investigation of the Impact of Computer Based
Decision Aids on the Decision Making Process,” Information Systems Research, 2, 2
(1991) 87-115.
Todd, P., and Benbasat, I., “An Experimental Investigation of the Impact of Computer Based DSS on
Processing Effort. MIS Quarterly, 16, 3 (1992), 373-393.
Todd, P., and Benbasat I., “Decision-makers, DSS and Decision Making Effort: An Experimental
Investigation, INFOR , 31, 2 (1993), 1-21.

49
Todd, P. and Benbasat, I. (1994). The Influence of DSS on Choice Strategies: An Experimental
Analysis of the Role of Cognitive Effort, Organizational Behavior and Human Decision
Processes, 60 (1994), 36-74.
Todman, J., and Monaghan, E., “Qualitative Differences in Computer Experience, computer
Anxiety, and Students’ Use of Computers: A Path Model,” Computers in Human
Behavior, 10, 4 (1994), 529-539.
Tversky, A., and Kahneman, D., “Judgement Under Uncertainty: Heuristics and Biases,”
Science, 185, (1974), 1124-1131.
Vallerand, R. J., “Toward a hierarchical Model of Intrinsic and Extrinsic Motivation,” Advances
in Experimental Social Psychology, 29 (1997), 271-360.
Venkatesh, V., “Creating Favorable User Perceptions: Exploring the Role of Intrinsic
Motivation,” MIS Quarterly, 23, 2 (1999), 239-260.
Venkatesh, V., and Davis, F. D., “A Model of the Antecedents of Perceived Ease of Use:
Development and Test,” Decision Sciences, 27, 3 (1996), 451-481.
Venkatesh, V., and Speier, C., “Computer Technology Training in the Workplace: A
Longitudinal Investigation of the Effect of the Mood,” Organizational Behavior and
Human Decision Processes, 79, 1 (1999), 1-28.
Webster, J., and Martocchio, J. J., “Microcomputer Playfulness: Development of a Measure with
Workplace Implications,” MIS Quarterly, 16, 2 (1992), 201-226.
Webster, J., Trevino, L. K., and Ryan, L., “The Dimensionality and Correlates of Flow in
Human-Computer Interactions, Computers in Human Behavior, 9, 4 (1993), 411-426.
Weiner, L. R., Digital Woes: Why We Should Not Depend on Software. Reading, MA: Addison-
Wesley, (1993).
White, K. M., Terry, D. J., and Hogg, M. A., “Safer Sex Behavior: The Role of Attitudes,
Norms, and Control Factors,” Journal of Applied Social Psychology, 24, 24 (1994), 2164-
2192.
Wurtele, S. K., “Increasing Women’s Calcium Intake: The Role of Health Beliefs, Intentions and
Health Value,” Journal of Applied Social Psychology, 18, 8 (1988), 627-639.
Zoltan, E., and Chapanis, A., “What do Professional Persons Think about Computers?”
Behaviour and Information Technology, 1, 1 (1982), 55-68.

50
APPENDIX 1. QUESTIONNAIRE ITEMS
Behavioral Intention to Use
Assuming I had access to the system, I intend to use it.
Given that I had access to the system, I predict that I would use it.
Perceived Usefulness
Using the system improves my performance in my job.
Using the system in my job increases my productivity.
Using the system enhances my effectiveness in my job.
I find the system to be useful in my job.
Perceived Ease of Use
My interaction with the system is clear and understandable.
Interacting with the system does not require a lot of my mental effort.
I find the system to be easy to use.
I find it easy to get the system to do what I want it to do.
Perceptions of Internal Control (Computer Self-Efficacy)
(Note: Additional instructions were provided per Compeau and Higgins 1995a, 1995b)
I could complete the job using a software package...
...if there was no one around to tell me what to do as I go.
...if I had never used a package like it before.
...if I had only the software manuals for reference.
...if I had seen someone else using it before trying it myself.
...if I could call someone for help if I got stuck.
...if someone else had helped me get started.
...if I had a lot of time to complete the job for which the software was provided.
...if I had just the built-in help facility for assistance.
...if someone showed me how to do it first.
...if I had used similar packages before this one to do the same job.
Perceptions of External Control (Facilitating Conditions)
I have control over using the system.
I have the resources necessary to use the system.
I have the knowledge necessary to use the system.
Given the resources, opportunities and knowledge it takes to use the system, it would be easy for
me to use the system.
The system is not compatible with other systems I use.
Computer Anxiety
Computers do not scare me at all.
Working with a computer makes me nervous.
I do not feel threatened when others talk about computers.

51
It wouldn’t bother me to take computer courses.
Computers make me feel uncomfortable.
I feel at ease in a computer class.
I get a sinking feeling when I think of trying to use a computer.
I feel comfortable working with a computer.
Computers make me feel uneasy.
Computer Playfulness
The following questions ask you how you would characterize yourself when you use computers:
…spontaneous
…unimaginative
…flexible
…creative
…playful
…unoriginal
…uninventive
Perceived Enjoyment
I find using the system to be enjoyable.
The actual process of using the system is pleasant.
I have fun using the system.
Objective Usability
No specific items were used. It was measured as a ratio of time spent by the subject to the time
spent by an expert on the same set of tasks.
Experience
Was not explicitly measured--was coded based on point of measurement.
Perceived Voluntariness of Use
My superiors expect me to use the system.
My use of the system is voluntary.
My supervisor does not require me to use the system.
Although it might be helpful, using the system is certainly not compulsory in my job.
Note: All items were measured on 7-point Likert scale, except computer self-efficacy which was
measured using a 10-point Guttman scale.

52
Appendix 2. Factor Structure Based on Pooled Data at T1
1 2 3 4 5 6 7 8
BI1 0.95 0.08 0.07 0.05 0.03 0.01 0.09 0.19
BI2 0.91 0.05 0.09 0.09 0.07 0.12 0.09 0.17
U1 0.33 0.88 0.10 0.14 0.10 0.12 0.01 0.08
U2 0.28 0.90 0.13 0.19 0.11 0.07 0.03 0.19
U3 0.21 0.91 0.21 0.21 0.09 0.03 0.12 0.10
U4 0.18 0.93 0.05 0.10 0.17 0.07 0.12 0.11
EOU1 0.10 0.02 0.94 0.15 0.09 0.09 0.14 0.10
EOU2 0.02 0.07 0.96 0.17 0.01 0.03 0.02 0.14
EOU3 0.06 0.02 0.91 0.20 0.13 0.17 0.14 0.17
EOU4 0.11 0.10 0.90 0.21 0.08 0.05 0.11 0.19
CSE1 0.10 0.12 0.21 0.99 0.00 0.02 0.19 0.12
CSE2 0.20 0.02 0.03 0.88 0.02 0.03 0.23 0.20
CSE3 0.14 0.14 0.09 0.87 0.08 0.05 0.10 0.13
CSE4 0.15 0.01 0.00 0.75 0.09 0.08 0.11 0.15
CSE5 0.12 0.03 0.03 0.71 0.01 0.04 0.03 0.08
CSE6 0.03 0.00 0.10 0.79 0.02 0.06 0.05 0.15
CSE7 0.08 0.03 0.07 0.85 0.03 0.10 0.08 0.13
CSE8 0.10 0.12 0.02 0.89 0.04 0.11 0.03 0.15
CSE9 0.11 0.02 0.04 0.90 0.05 0.14 0.09 0.12
CSE10 0.07 0.08 0.03 0.82 0.01 0.15 0.02 0.10
EC1 0.05 0.02 0.07 0.03 0.81 0.10 0.04 0.22
EC2 0.02 0.13 0.17 0.06 0.80 0.09 0.01 0.13
EC3 0.06 0.09 0.19 0.14 0.77 0.07 0.09 0.12
EC4 0.15 0.00 0.09 0.11 0.70 0.02 0.01 0.14
EC5 0.19 0.12 0.10 0.10 0.72 0.19 0.03 0.04
ANX1 0.04 0.04 0.19 0.14 0.09 0.71 0.02 0.07
ANX2 0.05 0.09 0.14 0.13 0.03 0.79 0.09 0.29
ANX3 0.03 0.12 0.23 0.11 0.09 0.80 0.07 0.01
ANX4 0.09 0.01 0.08 0.12 0.08 0.77 0.10 0.20
ANX5 0.09 0.15 0.09 0.17 0.02 0.70 0.08 0.14
ANX6 0.04 0.03 0.11 0.22 0.08 0.77 0.14 0.18
ANX7 0.17 0.12 0.22 0.28 0.10 0.70 0.02 0.14
ANX8 0.18 0.03 0.03 0.05 0.08 0.80 0.08 0.19
ANX9 0.01 0.05 0.02 0.19 0.18 0.81 0.03 0.11
PLAY1 0.14 0.12 0.20 0.11 0.02 0.08 0.90 0.30
PLAY2 0.02 0.05 0.06 0.08 0.09 0.10 0.73 0.21
PLAY3 0.02 0.07 0.19 0.23 0.10 0.11 0.79 0.22
PLAY4 0.11 0.09 0.10 0.11 0.03 0.12 0.71 0.11
PLAY5 0.10 0.09 0.11 0.20 0.00 0.19 0.73 0.17
PLAY6 0.11 0.14 0.04 0.06 0.20 0.11 0.80 0.10
PLAY7 0.13 0.16 0.16 0.19 0.09 0.12 0.70 0.07
ENJ1 0.03 0.07 0.10 0.13 0.03 0.14 0.07 0.80
ENJ2 0.13 0.09 0.05 0.14 0.19 0.17 0.25 0.79
ENJ3 0.12 0.02 0.25 0.13 0.03 0.03 0.20 0.71

53
TABLES AND FIGURES
Table 1. Measurement
Post-training One month Three months
(T1) post-implementation post-implementation
(T2) (T3)
BI, U, EOU BI, U, EOU BI, U, EOU
VOL VOL VOL
CSE, PEC, PLAY, PECb PECb
CANXa
OU, ENJ ENJ ENJ
a Although CSE, PLAY, CANX were measured at T2 and T3, measures taken at T1
were used to examine the extent to which EOU was anchored to those beliefs.
Interestingly, the pattern of results was identical even CSE, PLAY, and CANX
measures from T2 and T3 were used.
b Although PEC is an anchor, the process of anchoring and adjustment (change
in mean value over time) for the construct required the use of the different
measures over time.

54
Table 2(a). Reliability and Discriminant Validity Coefficients at T1
ICR 1 2 3 4 5 6 7 8 9
BI .92 .94
U .93 .52*** .91
EOU .92 .34*** .33** .88
CSE .84 .24** .08 .39** .82
PEC .82 .21* .19* .35** .30*** .84
PLAY .88 .26** .12 .32*** .35** .13 .88
CANX .91 .07 -.20* -.40*** -.30** -.17 -.31*** .85
ENJ .90 .10 -.16* .06 .20* .19* .39*** -.23** .89
OU N/A .10 .15* .17* .19 .19 .10 -.21* .10 1.0
Table 2(b). Reliability and Discriminant Validity Coefficients at T2
ICR 1 2 3 4 5 6 7 8 9
BI .90 .92
U .91 .55*** .89
EOU .93 .30*** .35*** .91
CSE .90 .29*** .13 .43*** .84
PEC .88 .20* .11 .40*** .27*** .87
PLAY .85 .12 .18 .22** .28*** .21** .86
CANX .88 .11 -.26** -.35** -.33*** .06 -.30** .89
ENJ .92 .19* -.19* .25** .15 .22** .40*** -.25** .89
OU N/A .14 .12 .22** .22* .20* .18* -.20* .17 1.0
Table 2(c). Reliability and Discriminant Validity Coefficients at T3
ICR 1 2 3 4 5 6 7 8 9
BI .90 .95
U .91 .56*** .92
EOU .96 .24** .39*** .87
CSE .80 .20* .19* .46*** .79
PEC .85 .22** .20* .49*** .29*** .74
PLAY .81 .20* .14 .15 .30*** .10 .82
CANX .83 -.19* -.12 -.33*** -.31*** .05 -.33*** .84
ENJ .93 .18* .08 .29*** .28** .12 .40*** .09 .85
OU N/A .20* .21* .39*** .22** .26* .14 -.23** .06 1.0
Note: Diagonal elements are the square root of the shared variance between
the constructs and their measures. Off-diagonal elements are the
corerlations between the different constructs.
ICR = Internal Consistency Reliability
* p<.05; ** p<.01; *** p<.001

55
Figure 1. Theoretical Framework for the Determinants of Perceived Ease of Use

56
Figure 2. Theoretical Model of the Determinants of Perceived Ease of Use

57
Figure 3(a). Results at T1

58
Figure 3(b). Results at T2

59
Figure 3(c). Results at T3

60
ENDNOTES
i
The development and testing of TAM was based on studies conducted to examine potential acceptance of products
of IBM, Canada.
ii
Of the different adjustments, the adjustment for external control behaves differently. Specifically, general
perceptions of external control serve as an anchor and the perceptions adjust in mean value with increasing user
experience with the target system.
iii
It is possible to argue that perceived ease of use should influence intrinsic motivation (computer playfulness),
rather than intrinsic motivation influencing perceived ease of use, as proposed. The causal flow from perceived ease
of use to intrinsic motivation would be consistent with a motivational model where extrinsic and intrinsic motivation
are the key predictors of intention/behavior, resulting in perceived ease of use being examined as a determinant of
intrinsic motivation. However, given the focus on TAM, an outcome and process expectancy model, intrinsic
motivation is expected to influence perceived ease of use.
iv
Please see the correlations reported in the results section.
v
The results of the additional models are not reported here due to space constraints.
vi
The effect of experience on perception of external control was observed via a significant mean shift over time.