Davis 1989

Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information
Technology
Author(s): Fred D. Davis
Source: MIS Quarterly, Vol. 13, No. 3 (Sep., 1989), pp. 319-340
Published by: Management Information Systems Research Center, University of Minnesota
Stable URL: http://www.jstor.org/stable/249008
Accessed: 11-12-2016 20:40 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
Management Information Systems Research Center, University of Minnesota is

collaborating with JSTOR to digitize, preserve and extend access to MIS Quarterly
This content downloaded from 202.43.95.117 on Sun, 11 Dec 2016 20:40:58 UTC
All use subject to http://about.jstor.org/terms
IT Usefulness and Ease of Use
dent to
to perceived
perceived usefulness,
usefulness,asasopposed
opposedtoto
Perceived Usefulness, a parallel,
parallel, direct
direct determinant
determinantof
ofsystem
systemusage.
usage.
Perceived Ease of Implications
Implications are
are drawn
drawnfor
forfuture
futureresearch
researchonon
user acceptance.
Use, and User Keywords: User acceptance, end user
Acceptance of computing, user measurement
Information ACM Categories: H.1.2, K.6.1, K.6.2, K.6.3
Technology
Introduction
By: Fred
Fred D.
D. Davis
Davis Information technology offers the potential for sub-
Computer
Computer and
and Information
Information Systems
Systems stantially improving white collar performance
Graduate School of Business (Curley, 1984; Edelman, 1981; Sharda, et al.,
1988). But performance gains are often ob-
Administration
structed by users' unwillingness to accept and
University of Michigan use available systems (Bowen, 1986; Young,
Ann Arbor, Michigan 48109 1984). Because of the persistence and impor-
tance of this problem, explaining user accep-
tance has been a long-standing issue in MIS
research (Swanson, 1974; Lucas, 1975; Schultz
and Slevin, 1975; Robey, 1979; Ginzberg, 1981;
Swanson, 1987). Although numerous individual,
organizational, and technological variables have
Abstract
been investigated (Benbasat and Dexter, 1986;
Valid measurement scales for predicting user Franz and Robey, 1986; Markus and Bjorn-
acceptance of computers are in short supply. Anderson, 1987; Robey and Farrow, 1982), re-
Most subjective measures used in practice are search has been constrained by the shortage
unvalidated, and their relationship to system of high-quality measures for key determinants
usage is unknown. The present research de- of user acceptance. Past research indicates that
velops and validates new scales for two spe- many measures do not correlate highly with
cific variables, perceived usefulness and per- system use (DeSanctis, 1983; Ginzberg, 1981;
ceived ease of use, which are hypothesized to Schewe, 1976; Srinivasan, 1985), and the size
be fundamental determinants of user accep- of the usage correlation varies greatly from one
tance. Definitions for these two variables were study to the next depending on the particular
used to develop scale items that were pretested measures used (Baroudi, et al., 1986; Barki and
Huff, 1985; Robey, 1979; Swanson, 1982, 1987).
for content validity and then tested for reliability
and construct validity in two studies involving The development of improved measures for key
a total of 152 users and four application pro- theoretical constructs is a research priority for
grams. The measures were refined and stream- the information systems field.
lined, resulting in two six-item scales with reli-Aside from their theoretical value, better meas-
abilities of .98 for usefulness and .94 for ease
ures for predicting and explaining system use
of use. The scales exhibited high convergent, would have great practical value, both for ven-
discriminant, and factorial validity. Perceived use-
dors who would like to assess user demand for
fulness was significantly correlated with both self-
new design ideas, and for information systems
reported current usage (r=.63, Study 1) and managers within user organizations who would
self-predicted future usage (r= .85, Study 2). Per-
like to evaluate these vendor offerings.
ceived ease of use was also significantly corre-
lated with current usage (r=.45, Study 1) and Unvalidated measures are routinely used in prac-
future usage (r=.59, Study 2). In both studies, tice today throughout the entire spectrum of
usefulness had a significantly greater correla-design, selection, implementation and evaluation
tion with usage behavior than did ease of use. activities. For example: designers within vendor
Regression analyses suggest that perceived organizations such as IBM (Gould, et al., 1983),
ease of use may actually be a causal antece- Xerox (Brewley, et al., 1983), and Digital Equip-
MIS Quarterly/September 1989 319
ment
mentCorporation
Corporation(Good,(Good,
et al., 1986)
et al.,
measure
1986) measure
believe
believethat
thatthethe
systems
systemsis too
is too
hardhard
to use
to and
use a
user
userperceptions
perceptionsto guide
to guide
the development
the development
of ofthe
that
that theperformance
performance benefits
benefits
of usage
of usage
are out-
are ou
new
newinformation
informationtechnologies
technologies
and products;
and products;
inin- by
weighed
weighed bythe
the
effort
effortof of
using
using
the the
application.
applicatio
dustry
dustry publications
publications
often often
report user
report
surveys
user surveys
That
That is,
is,ininaddition
addition to to
usefulness,
usefulness,
usageusage
is theo-
is the
(e.g.,
(e.g.,Greenberg,
Greenberg,1984;1984;
Rushinek
Rushinek
and Rushinek,
and Rushinek,
rized
rized to
tobebeinfluenced
influenced by by
perceived
perceived
ease ease
of use.
of us
1986);
1986);several
several
methodologies
methodologiesfor software
for software
se- se-
Perceived usefulness is defined here as "the
lection
lection call
call
forfor
subjective
subjective
user inputs
user(e.g.,
inputs (e.g.,
Goslar,
Goslar, 1986;
1986;
Klein
Klein
and Beck,
and 1987);
Beck,and con- anddegree
1987); con- to which a person believes that using
temporary
temporary design
design
principles
principles
emphasize emphasize
meas- a particular system would enhance his or her
meas-
uring
uringuseruser
reactions
reactions
throughout
throughout
the entirethe
design job performance." This follows from the defini-
entire design
process
process (Anderson
(Anderson and Olson
and 1985;
Olson Gould
1985;
andGouldtion
andof the word useful: "capable of being used
Lewis,
Lewis, 1985;
1985;
Johansen
Johansen
and Baker,
and 1984;
Baker,
Mantei advantageously." Within an organizational con-
1984; Mantei
and
andTeorey,
Teorey, 1988;1988;
Norman,
Norman,
1983; Shneiderman, text, people are generally reinforced for good
1983; Shneiderman,
1987).
1987).Despite
Despitethe the
widespread
widespread
use of subjec- performance by raises, promotions, bonuses,
use of subjec-
tive
tivemeasures
measures in practice,
in practice,
little attention
little attention
is paid and other rewards (Pfeffer, 1982; Schein, 1980;
is paid
to
tothethe
quality
quality
of the
of measures
the measures
used or how
used or how well 1964). A system high in perceived use-
well Vroom,
they
theycorrelate
correlatewithwith
usage usage
behavior.behavior.
Given the Givenfulness,
the in turn, is one for which a user believes
low
lowusage
usagecorrelations
correlations
often observed
often observed
in rein
inthe
re-existence of a positive use-performance
search
search studies,
studies,
thosethose
who base
who important
base important
busi- relationship.
busi-
ness
nessdecisions
decisions on unvalidated
on unvalidated
measuresmeasures
may may ease of use, in contrast, refers to "the
Perceived
be
begetting
getting misinformed
misinformedabout aabout
system's
a system's
accept- accept-
degree to which a person believes that using
ability
ability to to
users.
users.
a particular system would be free of effort." This
The
Thepurpose
purpose of this
of this
research
research
is to pursue
is tobetter follows from the definition of "ease": "freedom
pursue better
measures
measures forfor
predicting
predicting
and explaining
and explaining
use. The from
use. Thedifficulty or great effort." Effort is a finite
investigation
investigation focuses
focuses
on twoon theoretical
two theoretical
con- resource
con- that a person may allocate to the vari-
structs,
structs, perceived
perceivedusefulness
usefulness
and perceived ous activities for which he or she is responsible
and perceived
ease of use, which are theorized to be funda- (Radner and Rothschild, 1975). All else being
mental determinants of system use. Definitions equal, we claim, an application perceived to be
for these constructs are formulated and the theo- easier to use than another is more likely to be
retical rationale for their hypothesized influence accepted by users.
on system use is reviewed. New, multi-item meas-
urement scales for perceived usefulness and per-
ceived ease of use are developed, pretested,
and then validated in two separate empirical stud- Theoretical Foundations
ies. Correlation and regression analyses exam- The theoretical importance of perceived useful-
ine the empirical relationship between the new ness and perceived ease of use as determinants
measures and self-reported indicants of system of user behavior is indicated by several diverse
use. The discussion concludes by drawing im- lines of research. The impact of perceived use-
plications for future research.
fulness on system utilization was suggested by
the work of Schultz and Slevin (1975) and Robey
(1979). Schultz and Slevin (1975) conducted an
exploratory factor analysis of 67 questionnaire
Perceived Usefulness and items, which yielded seven dimensions. Of
these, the "performance" dimension, interpreted
Perceived Ease of Use
by the authors as the perceived "effect of the
What causes people to accept or reject informa-
model on the manager's job performance," was
tion technology? Among the many variables mostthat
highly correlated with self-predicted use of
may influence system use, previous research sug-
a decision model (r=.61). Using the Schultz and
gests two determinants that are especially im-questionnaire, Robey (1979) finds the per-
Slevin
portant. First, people tend to use or not formance
use an dimension to be most correlated with
application to the extent they believe it will
two help
objective measures of system usage (r=.79
them perform their job better. We refer to andthis
.76). Building on Vertinsky, et al.'s (1975)
first variable as perceived usefulness. Second,
expectancy model, Robey (1979) theorizes that:
even if potential users believe that a given
"A ap-
system that does not help people perform
plication is useful, they may, at the sametheir
time, jobs is not likely to be received favorably
320 MIS Quarterly/September 1989
in spite
spite of
of careful
careful implementation
implementationefforts"
efforts"(p.(p.to learn a computer language. The self efficacy
537). Although
Although the the perceived use-performance paradigm does not offer a general measure ap-
perceiveduse-performance
contingency,
contingency, as as presented
presented in inRobey's (1979) plicable to our purposes since efficacy beliefs
Robey's(1979)
model,
model, parallels
parallels our
our definition
definitionofofperceived use-are theorized to be situationally-specific, with
perceiveduse-
fulness,
fulness, the
the use
use ofof Schultz
Schultz and
andSlevin's (1975) measures tailored to the domain under study
Slevin's(1975)
performance
performance factor
factor to to operationalize perform- (Bandura, 1982). Self efficacy research does,
operationalizeperform-
ance expectancies
expectancies is is problematic
problematicfor
forseveral rea-however, provide one of several theoretical per-
severalrea-
sons: the
the instrument
instrument is is empirically
empiricallyderived
derivedviaviapectives suggesting that perceived ease of use
exploratory
exploratory factor
factor analysis;
analysis; aasomewhat
somewhatlow ratioand perceived usefulness function as basic de-
lowratio
of sample
sample size
size to
to items
items is
is used
used(2:1);
(2:1);four ofterminants
fourof of user behavior.
thirteen
thirteen items
items have
have loadings
loadingsbelow
below.5,
.5,and
andsev-
sev-
eral of
of the
the items
items clearly
clearly fall
falloutside
outsidethe
thedefini-
defini-
tion of
of expected
expected performance
performanceimprovements
improvements
(e.g., "My
"My job
job will
will be
be more
more satisfying," "Others Cost-benefit paradigm
satisfying,""Others
will be
be more
more aware
aware of
of what
what I Iam
amdoing,"
doing,"etc.).
etc.). The cost-benefit paradigm from behavioral deci-
sion theory (Beach and Mitchell, 1978; Johnson
An alternative
alternative expectancy-theoretic
expectancy-theoreticmodel,
model,de-de-
and Payne, 1985; Payne, 1982) is also relevant
rived from
from Vroom
Vroom (1964),
(1964), was
wasintroduced
introducedand
and
to perceived usefulness and ease of use. This
tested by
by DeSanctis
DeSanctis (1983).
(1983). The
Theuse-perform-
use-perform-
research explains people's choice among vari-
ance expectancy
expectancy was
was not
not analyzed
analyzedseparately
separately
ous decision-making strategies (such as linear
from performance-reward
performance-reward instrumentalities
instrumentalitiesand
and
compensatory, conjunctive, disjunctive and elmi-
reward
reward valences.
valences. Instead,
Instead, aa matrix-oriented
matrix-orientedmeas-
meas-
nation-by-aspects) in terms of a cognitive trade-
urement
urement procedure
procedure was
was used
used totoproduce
producean
anover-
over-
off between the effort required to employ the strat-
all index of "motivational force" that combined
these three constructs. "Force" had small but
egy and the quality (accuracy) of the resulting
decision. This approach has been effective for
significant correlations with usage of a DSS
explaining why decision makers alter their choice
within a business simulation experiment (corre-
strategies in response to changes in task com-
lations ranged from .04 to .26). The contrast be-
tween DeSanctis's correlations and the ones ob-
plexity. Although the cost-benefit approach has
mainly concerned itself with unaided decision
served by Robey underscore the importance of
making, recent work has begun to apply the
measurement in predicting and explaining use.
same form of analysis to the effectiveness of
information display formats (Jarvenpaa, 1989;
Kleinmuntz and Schkade, 1988).
Self-efficacy theory
The importance of perceived ease of use is sup- Cost-benefit research has primarily used objec-
ported by Bandura's (1982) extensive research tive measures of accuracy and effort in research
on self-efficacy, defined as "judgments of how studies, downplaying the distinction between ob-
well one can execute courses of action required jective and subjective accuracy and effort. In-
creased emphasis on subjective constructs is war-
to deal with prospective situations" (p. 122). Self-
efficacy is similar to perceived ease of use as ranted, however, since (1) a decision maker's
choice
defined above. Self-efficacy beliefs are theorized of strategy is theorized to be based on
to function as proximal determinants of behav- subjective as opposed to objective accuracy and
ior. Bandura's theory distinguishes self-efficacyeffort (Beach and Mitchell, 1978), and (2) other
judgments from outcome judgments, the latterresearch suggests that subjective measures are
being concerned with the extent to which a be- often in disagreement with their ojbective coun-
havior, once successfully executed, is believed terparts (Abelson and Levi, 1985; Adelbratt and
to be linked to valued outcomes. Bandura's "out- Montgomery, 1980; Wright, 1975). Introducing
measures of the decision maker's own perceived
come judgment" variable is similar to perceived
usefulness. Bandura argues that self-efficacy costs and benefits, independent of the decision
and outcome beliefs have differing antecedents actually made, has been suggested as a way
and that, "In any given instance, behavior would of mitigating criticisms that the cost/benefit frame-
be best predicted by considering both self- work is tautological (Abelson and Levi, 1985).
efficacy and outcome beliefs" (p. 140). The distinction made herein between perceived
usefulness and perceived ease of use is similar
Hill, et al. (1987) find that both self-efficacy and to the distinction between subjective decision-
outcome beliefs exert an influence on decisions making performance and effort.
Adoption of innovations
innovations two
two components:
components:attributed
attributed information
informationquality
quality
and
and attributed
attributedaccess
accessquality.
quality.
Potential
Potential
users
users
areare
Research on thethe adoption
adoption of
of innovations
innovations also
alsohypothesized
hypothesizedto toselect
selectand
anduse
use
information
information re-re-
suggests a prominent
prominent role
role for
for perceived
perceived ease
easeports
ports based
basedon
onananimplicit
implicitpsychological
psychological
trade-
trade-
of use. In their
their meta-analysis
meta-analysis of
of the
the relationship
relationship
off
off between
betweeninformation
informationquality
quality
and
and
associated
associated
between the characteristics of an innovation and
costs
costs of
of access.
access.Swanson
Swanson(1987)
(1987)
performed
performed
an an
its adoption, Tornatzky and Klein (1982) find that
exploratory
exploratoryfactor
factoranalysis
analysisininorder
order to to
measure
measure
compatibility, relative advantage, and complex-
information
informationquality
qualityandandaccess
accessquality.
quality.A five-
A five-
ity have the most consistent significant relation- factor
factor solution
solutionwas
wasobtained,
obtained,with
withoneone
factor
factor
cor-
cor-
ships across a broad range of innovation types.
responding
respondingto toinformation
information quality
quality (Factor
(Factor
#3,#3,
Complexity, defined by Rogers and Shoemaker
"value"),
"value"), and
andoneonetotoaccess
accessquality
quality(Factor
(Factor#2,#2,
(1971) as "the degree to which an innovation "accessibility").
"accessibility").Inspecting
Inspectingthetheitems
items that
that
load
load
on on
is perceived as relatively difficult to understand
these
these factors
factorssuggests
suggestsa aclose
close
correspondence
correspondence
and use" (p. 154), parallels perceived ease of to perceived
perceivedusefulness
usefulnessand
andease
ease
ofof
use.
use.
Items
Items
use quite closely. As Tornatzky and Klein (1982)
such
such as
as "important,"
"important,""relevant,"
"relevant,"
"useful,"
"useful,"andand
point out, however, compatibility and relative ad-
"valuable"
"valuable" load
loadstrongly
stronglyononthe
the
value
value
dimension.
dimension.
vantage have both been dealt with so broadly
Thus,
Thus, value
valueparallels
parallelsperceived
perceivedusefulness.
usefulness.TheThe
and inconsistently in the literature as to be diffi- fact that relevance and usefulness load on the
cult to interpret.
same factor agrees with information scientists,
who emphasize the conceptual similarity be-
tween the usefulness and relevance notions
(Saracevic, 1975). Several of Swanson's "acces-
Evaluation of information reports sibility" items, such as "convenient," "controlla-
Past research within MIS on the evaluation of ble," "easy," and "unburdensome," correspond
information reports echoes the distinction be- to perceived ease of use as defined above. Al-
tween usefulness and ease of use made herein. though the study was more exploratory than con-
Larcker and Lessig (1980) factor analyzed six firmatory, with no attempts at construct valida-
tion, it does agree with the conceptual distinction
items used to rate four information reports. Three
items load on each of two distinct factors: (1)between usefulness and ease of use. Self-
perceived importance, which Larcker and Lessig reported information channel use correlated .20
define as "the quality that causes a particularwith the value dimension and .13 with the ac-
information set to acquire relevance to a deci- cessibility dimension.
sion maker," and the extent to which the infor-
mation elements are "a necessary input for task
accomplishment," and (2) perceived usable-
ness, which is defined as the degree to whichNon-MIS studies
"the information format is unambiguous, clear
or readable" (p. 123). These two dimensions are Outside the MIS domain, a marketing study by
similar to perceived usefulness and perceived Hauser and Simmie (1981) concerning user per-
ease of use as defined above, repsectively, al- ceptions of alternative communication technolo-
though Larcker and Lessig refer to the two di- gies similarly derived two underlying dimensions:
mensions collectively as "perceived usefulness." ease of use and effectiveness, the latter being
Reliabilities for the two dimensions fall in the similar to the perceived usefulness construct de-
range of .64-.77, short of the .80 minimal level fined above. Both ease of use and effectiveness
recommended for basic research. Correlations were influential in the formation of user prefer-
with actual use of information reports were not ences regarding a set of alternative communi-
addressed in their study. cation technologies. The human-computer inter-
action (HCI) research community has heavily
emphasized ease of use in design (Branscomb
and Thomas, 1984; Card, et al., 1983; Gould
and Lewis, 1985). For the most part, however,
Channel disposition model these studies have focused on objective meas-
Swanson (1982, 1987) introduced and tested aures of ease of use, such as task completion
model of "channel disposition" for explaining thetime and error rates. In many vendor organiza-
choice and use of information reports. The con- tions, usability testing has become a standard
cept of channel disposition is defined as having phase in the product development cycle, with
large
large investments
investments in
intest
testfacilities
facilitiesand
andinstrumen-
instrumen- retained,
retained,yielding
yielding1010
items
itemsforfor
eacheach
construct.
construct.
tation.
tation. Although
Although objective
objectiveease
easeofofuse
useisisclearly
clearly Next,
Next, aa field
fieldstudy
study(Study
(Study 1) 1)
of of
112112
users
users
con-con-
relevant
relevant toto user
user performance
performancegiven
giventhe
thesystem
system cerning
cerningtwotwodifferent
differentinteractive
interactivecomputer
computersys-sys-
is used,
used, subjective
subjective ease
easeof
ofuse
useisismore
morerelevant
relevant tems was conducted in order to assess the reli-
to the users' decision whether or not to use the ability and construct validity of the resulting
system and may not agree with the objective scales. The scales were further refined and
measures (Carroll and Thomas, 1988). streamlined to six items per construct. A lab
study (Study 2) involving 40 participants and two
graphics systems was then conducted. Data
from the two studies were then used to assess
Convergence of findings the relationship between usefulness, ease of
There is a striking convergence among the wide use, and self-reported usage.
range of theoretical perspectives and research
Psychometricians emphasize that the validity of
studies discussed above. Although Hill, et al. a measurement scale is built in from the outset.
(1987) examined learning a computer language,
As Nunnally (1978) points out, "Rather than test
Larcker and Lessig (1980) and Swanson (1982,
the validity of measures after they have been
1987) dealt with evaluating information reports,
constructed, one should ensure the validity by
and Hauser and Simmie (1981) studied com-
the plan and procedures for construction" (p.
munication technologies, all are supportive of the
258). Careful selection of the initial scale items
conceptual and empirical distinction between use-
helps to assure the scales will possess "content
fulness and ease of use. The accumulated body
validity," defined as "the degree to which the
of knowledge regarding self-efficacy, contingent
score or scale being used represents the con-
decision behavior and adoption of innovations
cept about which generalizations are to be
provides theoretical support for perceived use-
made" (Bohrnstedt, 1970, p. 91). In discussing
fulness and ease of use as key determinants
of behavior. content validity, psychometricians often appeal
to the "domain sampling model," (Bohrnstedt,
From multiple disciplinary vantage points, per- 1970; Nunnally, 1978) which assumes there is
ceived usefulness and perceived ease of use a domain of content corresponding to each vari-
are indicated as fundamental and distinct con- able one is interested in measuring. Candidate
structs that are influential in decisions to use in- items representative of the domain of content
formation technology. Although certainly not the should be selected. Researchers are advised to
only variables of interest in explaining user be- begin by formulating conceptual definitions of
havior (for other variables, see Cheney, et al., what is to be measured and preparing items to
1986; Davis, et al., 1989; Swanson, 1988), they fit the construct definitions (Anastasi, 1986).
do appear likely to play a central role. Improved
Following these recommendations, candidate
measures are needed to gain further insight into
items for perceived usefulness and perceived
the nature of perceived usefulness and per-
ease of use were generated based on their con-
ceived ease of use, and their roles as determi-
ceptual definitions, stated above, and then pre-
nants of computer use. tested in order to select those items that best
fit the content domains. The Spearman-Brown
Prophecy formula was used to choose the
number of items to generate for each scale. This
formula estimates the number of items needed
Scale Development and to achieve a given reliability based on the
Pretest number of items and reliability of comparable
existing scales. Extrapolating from past studies,
A step-by-step process was used to develop the formula suggests that 10 items would be
new multi-item scales having high reliability and needed for each perceptual variable to achieve
validity. The conceptual definitions of perceived reliability of at least .80 (Davis, 1986). Adding
usefulness and perceived ease of use, stated four additional items for each construct to allow
above, were used to generate 14 candidate for item elimination, it was decided to generate
items for each construct from past literature. Pre- 14 items for each construct.
test interviews were then conducted to assess
the semantic content of the items. Those items The initial item pools for perceived usefulness
that best fit the definitions of the constructs were and perceived ease of use are given in Tables
1 and
and 2,
2, respectively.
respectively.InIn preparing
preparing candidate
candidateother items in order to yield a more pure indi-
items,
items, 37
37 published
publishedresearch
researchpapers
papers
dealing
dealing withcant of the conceptual variable.
with
user
user reactions
reactionsto tointeractive
interactivesystems
systemswerewere
re- re-
viewed
viewed inin other
othertotoidentify
identifyvarious
various
facets
facets
of the
of thePretest interviews were performed to further en-
constructs that should be measured (Davis, hance content validity by assessing the corre-
1986). The items are worded in reference to "the spondence between candidate items and the defi-
electronic mail system," which is one of the two nitions of the variables they are intended to
test applications investigated in Study 1, reported measure. Items that don't represent a construct's
below. The items within each pool tend to have content very well can be screened out by asking
a lot of overlap in their meaning, which is con- individuals to rank the degree to which each item
sistent with the fact that they are intended as matches the variable's definition, and eliminat-
measures of the same underlying construct. ing items receiving low rankings. In eliminating
Though different individuals may attribute slightly items, we want to make sure not to reduce the
different meaning to particular item statements, representativeness of the item pools. Our item
the goal of the multi-item approach is to reduce pools may have excess coverage of some areas
any extranneous effects of individual items, al- of meaning (or substrata; see Bohrnstedt, 1970)
lowing idiosyncrasies to be cancelled out by within the content domain and not enough of
Table 1. Initial Scale Items for Perceived Usefulness
1.
1.My
My jobjob
would
would
be difficult
be difficult
to perform towithout
perform electronic
without
mail. electronic mail.
2.
2.Using
Using electronic
electronic
mail gives
mail megives
greatermecontrol
greater
over my
control
work. over my work.
3.
3.Using
Using electronic
electronic
mail improves
mail improves
my job performance.
my job performance.
4.
4.The
The electronic
electronic
mail system
mail system
addresses my
addresses
job-relatedmyneeds.
job-related needs.
5.
5.Using
Using electronic
electronic
mail saves
mailmesaves
time. me time.
6.
6.Electronic
Electronic mail mail
enablesenables
me to accomplish
me to accomplish
tasks more quickly.
tasks more quickly.
7.
7.Electronic
Electronic mail mail
supports
supports
critical aspects
critical
of my
aspects
job. of my job.
8.
8.Using
Using electronic
electronic
mail allows
mailme allows
to accomplish
me to moreaccomplish
work thanmorewould otherwise
work than be would otherwise be
possible.
9. Using electronic mail reduces the time I spend on unproductive activities.
10. Using electronic mail enhances my effectiveness on the job.
11. Using electronic mail improves the quality of the work I do.
12. Using electronic mail increases my productivity.
13. Using electronic mail makes it easier to do my job.
14. Overall, I find the electronic mail system useful in my job.
Table 2. Initial Scale Items for Perceived Ease of Use
1. I often become confused when I use the electronic mail system.

2. I make errors frequently when using electronic mail.
3. Interacting with the electronic mail system is often frustrating.
4. I need to consult the user manual often when using electronic mail.
5. Interacting with the electronic mail system requires a lot of my mental effort.
6. I find it easy to recover from errors encountered while using electronic mail.
7. The electronic mail system is rigid and inflexible to interact with.
8. I find it easy to get the electronic mail system to do what I want it to do.
9. The electronic mail system often behaves in unexpected ways.
10. I find it cumbersome,to use the electronic mail system.
11. My interaction with the electronic mail system is easy for me to understand.
12. It is easy for me to remember how to perform tasks using the electronic mail system.
13. The electronic mail system provides helpful guidance in performing tasks.
14. Overall, I find the electronic mail system easy to use.
others.
others. ByBy asking
askingindividuals
individualstoto
rate
ratethe
the similar-over work" was retained since, although it was
similar-
ity
ity of
of items
itemsto toone
oneanother,
another,we wecancan
perform
perform a aranked fairly low, it fell in the top 9 and may
cluster
cluster analysis
analysistotodetermine
determinethethe
structure
structure thethe tap an important aspect of usefulness.
of of
substrata,
substrata, remove
removeitems
itemswhere
where excess
excesscoverage
coverage
is suggested,
suggested,and
andadd
additems
itemswhere
whereinadequate
inadequateLooking now at perceived ease of use (Table
coverage
coverage is
is indicated.
indicated.
4), we again find three main clusters. The first
Pretest
Pretest participants
participantsconsisted
consistedofof
a sample
a sample 15 15 relates to physical effort, while the second re-
of of
experienced
experienced computer
computerusers
usersfrom
from thethe Sloan lates to mental effort. Selecting the six highest-
Sloan
School
School of
of Management,
Management,MIT,MIT,including
includingfive
five sec- priority items and eliminating the seventh pro-
sec-
retaries,
retaries, five
fivegraduate
graduatestudents
studentsand
andfive
five mem-vides good coverage of these two clusters. Item
mem-
bers
bers of
of the
the professional
professionalstaff.
staff.
InIn
face-to-face in-in- 11 ("understandable") was reworded to read
face-to-face
terviews,
terviews, participants
participantswere
wereasked
asked
toto
perform
performtwotwo "clear and understandable" in an effort to pick
tasks,
tasks, prioritization
prioritizationand
andcategorization,
categorization, which up some of the content of item 1 ("confusing"),
which
which has been eliminated. The third cluster is
were
were done
done separately
separatelyfor
forusefulness
usefulnessand
and
ease
ease
of use.
use. For
For prioritization,
prioritization,theythey were
werefirst
first givensomewhat more difficult to interpret but appears
given
a card
card containing
containingthethedefinition
definition ofofthe
the
target
target con- to be tapping perceptions of how easy a system
con-
struct
struct and
and asked
askedto
toread
readit.it.Next,
Next, they
they
were
were given is to learn. Remembering how to perform tasks,
given
13 index
index cards
cardseach
eachhaving
havingone oneofof
thethe items using the manual, and relying on system guid-
items
for
for that
that construct
constructwritten
writtenonon it.it.
The The
14th
14th
or or "over- ance are all phenomena associated with the proc-
"over-
all" item for each construct was omitted since ess of learning to use a new system (Nickerson,
its wording was almost identical to the label on 1981; Roberts and Moran, 1983). Further review
the definition card (see Tables 1 and 2). Partici- of the literature suggests that ease of use and
pants were asked to rank the 13 cards accord- ease of learning are strongly related. Roberts
ing to how well the meaning of each statement and Moran (1983) find a correlation of .79 be-
matched the given definition of ease of use or tween objective measures of ease of use and
usefulness. ease of learning. Whiteside, et al. (1985) find
that ease of use and ease of learning are
For the categorization task, participants were strongly related and conclude that they are con-
asked to put the 13 cards into three to five cate- gruent. Studies of how people learn new sys-
gories so that the statements within a category tems suggest that learning and using are not
were most similar in meaning to each other and separate, disjoint activities, but instead that
dissimilar in meaning from those in other cate- people are motivated to begin performing actual
gories. This was an adaptation of the "own cate- work directly and try to "learn by doing" as op-
gories" procedure of Sherif and Sherif (1967). posed to going through user manuals or online
Categorization provides a simple indicant of simi- tutorials (Carroll and Carrithers, 1984; Carroll,
larity that requires less time and effort to obtain et al., 1985; Carroll and McKendree, 1987).
than other similarity measurement procedures
such as paid comparisons. The similarity data
In this study, therefore, ease of learning is re-
was cluster analyzed by assigning to the same
cluster items that seven or more subjects placed
garded as one substratum of the ease of use
construct, as opposed to a distinct construct.
in the same category. The clusters are consid-
ered to be a reflection of the domain substrata Since items 4 and 13 provide a rather indirect
for each construct and serve as a basis of as- assessment of ease of learning, they were re-
placed with two items that more directly get at
sessing coverage, or representativeness, of the
ease of learning: "Learning to operate the elec-
item pools.
tronic mail system is easy for me," and "I find
The resulting rank and cluster data are summa- it takes a lot of effort to become skillful at using
rized in Tables 3 (usefulness) and 4 (ease of electronic mail." Items 6, 9 and 2 were elimi-
use). For perceived usefulness, notice that items nated because they did not cluster with other
fall into three main clusters. The first cluster re- items, and they received low priority rankings,
lates to job effectiveness, the second to produc- which suggests that they do not fit well within
tivity and time savings, and the third to the im- the content domain for ease of use. Together
portance of the system to one's job. If we with the "overall" items for each construct, this
eliminate the lowest-ranked items (items 1, 4, procedure yielded a 10-item scale for each con-
5 and 9), we see that the three major clusters struct to be empirically tested for reliability and
each have at least two items. Item 2, "control construct validity.
Table 3. Pretest Results: Perceived Usefulness
Old New
Item # Item Rank Item # Cluster
1 Job Difficult Without 13 C
2 Control Over Work 9 2
3 Job Performance 2 6 A
4 Addresses My Needs 12 C
5 Saves Me Time 11 B
6 Work More Quickly 7 3 B
7 Critical to My Job 5 4 C
8 Accomplish More Work 6 7 B
9 Cut Unproductive Time 10 B
10 Effectiveness 1 8 A
11 Quality of Work 3 1 A
12 Increase Productivity 4 5 B
13 Makes Job Easier 8 9 C
14 Useful NA 10 NA
Table 4. Pretest Results: Perceived Ease of Use
Old New
Item # Item Rank Item # Cluster
1 Confusing 7 B
2 Error Prone 13
3 Frustrating 3 3 B
4 Dependence on Manual 9 (replace) C
5 Mental Effort 5 7 B
6 Error Recovery 10
7 Rigid & Inflexible 6 5 A
8 Controllable 1 4 A
9 Unexpected Behavior 11
10 Cumbersome 2 1 A
11 Understandable 4 8 B
12 Ease of Remembering 8 6 C
13 Provides Guidance 12 (replace) C
14 Easy to Use NA 10 NA
NA Ease of Learning NA 2 NA
NA Effort to Become Skillful NA 9 NA
able on IBM
IBM systems
systems and
and offers
offers both
bothfull-screen
full-screen
Study 1 and command-driven
command-driven editing
editing capabilities.
capabilities.The
The
A field study was conducted to assess the reli- questionnaire
questionnaire asked
asked participants
participants to
torate
ratethe
the
ability, convergent validity, discriminant validity,
extent to
to which
which theythey agree
agree with
with each
eachstatement
statement
and factorial validity of the 10-item scales re- by circling
circling aa number
number fromfrom one
one to
toseven
sevenarranged
arranged
sulting from the pretest. A sample of 120 users horizontally
horizontally beneath
beneath anchor
anchor point
pointdescriptions
descriptions
within IBM Canada's Toronto Development Labo- "Strongly
"Strongly Agree,"
Agree," "Neutral,"
"Neutral," and
and "Strongly
"StronglyDis-
Dis-
ratory were given a questionnaire asking them agree." In
In order
order to to ensure
ensure subject
subject familiarity
familiaritywith
with
to rate the usefulness and ease of use of two the systems
systems being
being rated,
rated, instructions
instructionsasked
askedthe
the
systems available there: PROFS electronic mail participants
participants to to skip
skip over
over the
the section
sectionpertaining
pertaining
and the XEDIT file editor. The computing envi- to a given
given system
system if if they
they never
never use
useit.
it.Responses
Responses
ronment consisted of IBM mainframes accessi- were obtained
obtained from
from 112112 participants,
participants,for foraare-
re-
ble through 327X terminals. The PROFS elec- sponse rate of 93%. Of these 112, 109 were
tronic mail system is a simple but limited users of electronic mail and 75 were users of
messaging facility for brief messages. (See XEDIT. Subjects had an average of six months'
Panko, 1988.) The XEDIT editor is widely avail-
experience with the two systems studied. Among
the
the sample,
sample,1010percent
percentwerewere
managers,
managers, 35 per- different
35 per- different trait
trait or
orwith
withdifferent
differentitems
itemsused
used
toto
meas-
meas-
cent
cent were
wereadministrative
administrative staff,
staff,
and and
55 percent
55 percent ure aa different
different trait
trait(Campbell
(Campbelland andFiske,
Fiske,
1959).
1959).
were
were professional
professionalstaff
staff
(which
(which
included
included a broad For perceived
a broad perceived usefulness,
usefulness,1,800
1,800such
suchcompari-
compari-
mix
mix of
ofmarket
market analysts,
analysts,
product
product
development ana- ana- sons
development sons were
were confirmed
confirmedwithout
withoutexception.
exception. OfOfthe
the
lysts,
lysts,programmers,
programmers, financial
financial
analysts
analysts and re- 1,800
and re- 1,800 comparisons
comparisonsfor forease
easeof
ofuse
usethere
therewere
were
search
searchscientists).
scientists). 58 exceptions
exceptions (3%).
(3%).This
Thisrepresents
representsananunusu-
unusu-
ally
ally high
high level
level of
ofdiscriminant
discriminantvalidity
validity(Campbell
(Campbell
and Fiske,
Fiske, 1959;
1959; Silk,
Silk,1971)
1971)and
andimplies
impliesthat
that
the usefulness
usefulness and
andease
easeof
ofuse
usescales
scalespossess
possess
Reliability
Reliabilityand
andvalidity
validity
a high
high concentration
concentrationof
oftrait
traitvariance
varianceand
and
are
are
The
The perceived
perceivedusefulness
usefulnessscale
scale
attained
attained
Cron-Cron- not strongly
strongly influenced
influencedby bymethodological
methodological
bach
bach alpha
alphareliability
reliabilityof of
.97 .97
for for
bothboth
the elec-
the elec- artifacts.
artifacts.
tronic
tronicmail
mailandand
XEDIT
XEDITsystems,
systems,while
while
perceived
perceived
ease
ease of
ofuse
useachieved
achieved
a reliability
a reliability
of .86
offor
.86 elec-
for elec- Table
Table 55 gives
gives aa summary
summaryfrequency
frequencytable
table
ofofthe
the
tronic mail and .93 for XEDIT. When observa- correlations
correlations comprising
comprisingthe
theMTMM
MTMMmatrices
matrices for
for
tions were pooled for the two systems, alpha usefulness and ease of use. From this table it
was .97 for usefulness and .91 for ease of use.
is possible to see the separation in magnitude
between monotrait and heterotrait correlations.
Convergent and discriminant validity were tested
The frequency table also shows that the hetero-
using multitrait-multimethod (MTMM) analysis
trait-heteromethod correlations do not appear to
(Campbell and Fiske, 1959). The MTMM matrix
contains the intercorrelations of items (methods) be substantially elevated above the heterotrait-
monomethod correlations. This is an additional
applied to the two different test systems (traits),
electronic mail and XEDIT. Convergent validity diagnostic suggested by Campbell and Fiske
refers to whether the items comprising a scale
(1959) to detect the presence of method
variance.
behave as if they are measuring a common un-
derlying construct. In order to demonstrate con-
The few exceptions to the convergent and dis-
vergent validity, items that measure the same
criminant validity that did occur, although not ex-
trait should correlate highly with one another
tensive enough to invalidate the ease of use
(Campbell and Fiske, 1959). That is, the ele-
scale, all involved negatively phrased ease of
ments in the monotrait triangles (the submatrix
use items. These "reversed" items tended to cor-
of intercorrelations between items intended to
relate more with the same item used to meas-
measure the same construct for the same
ure a different trait than they did with other items
system) within the MTMM matrices should be
of the same trait, suggesting the presence of
large. For perceived usefulness, the 90 monotrait-
common method variance. This is ironic, since
heteromethod correlations were all significant at
the .05 level. For ease of use, 86 out of 90, reversed scales are typically used in an effort
to reduce common method variance. Silk (1971)
or 95.6%, of the monotrait-heteromethod corre-
similarly observed minor departures from con-
lations were significant. Thus, our data supports
vergent and discriminant validity for reversed
the convergent validity of the two scales.
items. The five positively worded ease of use
Discriminant validity is concerned with the abil- items had a reliability of .92 compared to .83
ity of a measurement item to differentiate be- for the five negative items. This suggests an im-
tween objects being measured. For instance, provement in the ease of use scale may be pos-
within the MTMM matrix, a perceived usefulness sible with the elimination or reversal of nega-
item applied to electronic mail should not corre- tively phrased items. Nevertheless, the MTMM
late too highly with the same item applied to analysis supported the ability of the 10-item
XEDIT. Failure to discriminate may suggest the scales for each construct to differentiate between
presence of "common method variance," which systems.
means that an item is measuring methodological
artifacts unrelated to the target construct (such Factorial validity is concerned with whether the
as individual differences in the style of respond- usefulness and ease of use items form distinct
ing to questions (see Campbell, et al., 1967; Silk, constructs. A principal components analysis
1971) ). The test for discriminant validity is that using oblique rotation was performed on the
an item should correlate more highly with other twenty usefulness and ease of use items. Data
items intended to measure the same trait than were pooled across the two systems, for a total
with either the same item used to measure a of 184 observations. The results show that the
Table
Table5. Summary
5. ofSummary
Multitrait-Multimethod
ofAnalyses
Multitrait-M
Construct
Perceived Usefulness Perceived Ease of Use

Same Trait/ Different Same Trait/ Different
Diff. Method Trait Diff. Method Trait
Correlation Elec. Same Diff. Elec. Same Diff.
Size Mail XEDIT Meth. Meth. Mail XEDIT Meth. Meth.
-.20 to -.11 1
-.10 to -.01 6 1 5
.00 to .09 3 25 2 1 32
.10 to .19 2 27 2 5 40
.20 to .29 5 25 9 1 11
.30 to .39 7 14 2 2 1
.40 to .49 9 9
.50 to .59 4 3 11
.60 to .69 14 4 3 13
.70 to .79 20 11 3 8
.80to .89 7 26 2
.90 to .99 4
# Correlations 45 45 10 90 45 45 10 90
usefulness and ease of use items load on dis- how

howto
toperform
performtasks"),
tasks"),
which
which
the pretest
the pretest
indi- indi-
cated
catedwas
tinct factors (Table 6). The multitrait-multimethod wasconcerned
concerned
with
with
ease ease
of learning,
of learning,
was was
replaced by a reversal of item 9 ("easy to
analysis and factor analysis both support the con-
struct validity of the 10-item scales. become skillful"), which was specifically de-
signed to more directly tap ease of learning.
These items include two from cluster C, one
Scale refinement
each from clusters A and B, and the overall item.
In applied testing situations, it is important to (See Table 4.) In order to improve representa-
keep scales as brief as possible, particularly tive coverage of the content domain, an addi-
when multiple systems are going to be evalu- tional A item was added. Of the two remaining
ated. The usefulness and ease of use scales A items (#1, Cumbersome, and #5, Rigid and
were refined and streamlined based on results Inflexible), item 5 is readily reversed to form "flex-
from Study 1 and then subjected to a secondible to interact with." This item was added to
round of empirical validation in Study 2, reportedform the sixth item, and the order of items 5
below. Applying the Spearman-Brown prophecy and 8 was permuted in order to prevent items
formula to the .97 reliability obtained for per- from the same cluster (items 4 and 5) from ap-
ceived usefulness indicates that a six-item scale pearing next to one another.
composed of items having comparable reliabil-
In order to select six items to be used for the
ity would yield a scale reliability of .94. The five
positive ease of use items had a reliability of usefulness scale, an item analysis was per-
formed. Corrected item-total correlations were
.92. Taken together, these findings from Study
1 suggest that six items would be adequate to computed for each item, separately for each
achieve reliability levels above .9 while main- system studied. Average Z-scores of these cor-
taining adequate validity levels. Based on the relations were used to rank the items. Items 3,
results of the field study, six of the 10 items for 5, 6, 8, 9 and 10 were top-ranked items. Refer-
each construct were selected to form modified ring to the cluster analysis (Table 3), we see
scales. that this set is well-representative of the content
domain, including two items from cluster A, two
For the ease of use scale, the five negatively from cluster B and one from cluster C, as well
worded items were eliminated due to their ap- as the overall item (#10). The items were per-
parent common method variance, leaving items muted to prevent items from the same cluster
2, 4, 6, 8 and 10. Item 6 ("easy to remember from appearing next to one another. The result-
Table
Table
6. Factor Analysis
6. of Perceived
Factor Usefulness andAnalysis
Ease
Ease
of Use Questions:
of StudyUse
1 Question
Factor 1 Factor 1
Scale Items (Usefulness) (Ease of Use)
Usefulness
1 Quality of Work .80 .10
2 Control over Work .86 -.03
3 Work More Quickly .79 .17
4 Critical to My Job .87 -.11
5 Increase Productivity .87 .10
6 Job Performance .93 -.07
7 Accomplish More Work .91 -.02
8 Effectiveness .96 -.03
9 Makes Job Easier .80 .16
10 Useful .74 .23
Ease of Use
1 Cubersome .00 .73
2 Ease of Learning .08 .60
3 Frustrating .02 .65
4 Controllable .13 .74
5 Rigid & Inflexible .09 .54
6 Ease of Remembering .17 .62
7 Mental Effort -.07 .76
8 Understandable .29 .64
9 Effort to Be Skillful -.25 .88
10 Easy to Use .23 .72
ing
ing
six-item usefulness
six-item and ease of use scales usefulness
(.69), and overall
overall (.64).
(.64). All
All correlationsand
correlations were
weresig-
sig- ea
are
are
shown in the
shown
Appendix. in the nificant at the .001 level.
Appendix.
Regression analyses were performed to assess

the joint effects of usefulness and ease of use
Relationship
Relationshipto use to
on usage. The effect ofuse
usefulness on usage,
Participants
Participants
were asked to self-report their were asked
controlling for ease to
of use, was significant at the sel
degree
degree
of current usage ofof
electroniccurrent
mail and .001 level forusage
electronic mail (b=.55),
of XEDITelec
XEDIT
XEDITon six-position categorical
on scales six-position
with (b=.69), and pooled (b=.57). Incategor
contrast, the
boxes
boxes
labeled "Don't uselabeled
at all," "Use less than "Don't use
effect of ease of use at all,"
on usage, controlling for
once
once
each week," "Use
each
about once eachweek,"
week," usefulness,"Use about
was non-significant across the board onc
"Use several times a week," "Use about once (b=.01 for electronic mail; b=.02 for XEDIT;
each day," and "Use several times each day." and b=.07 pooled). In other words, the signifi-
Usage was significantly correlated with both per- cant pairwise correlation between ease of use
ceived usefulness and perceived ease of use and usage vanishes when usefulness is con-
for both PROFS mail and XEDIT. PROFS mail trolled for. The regression coefficients obtained
usage correlated .56 with perceived usefulness for each individual system within each study
and .32 with perceived ease of use. XEDIT were not significantly different (F3, 178= 1.95,
usage correlated .68 with usefulness and .48 n.s.). As the relationship between independent
with ease of use. When data were pooled across variables in a regression approach perfect linear
systems, usage correlated .63 with usefulness dependence, multicollinearity can degrade the
and .45 with ease of use. The overall usefulness- parameter estimates obtained. Although the cor-
use correlation was significantly greater than the relations between usefulness and ease of use
ease of use-use correlation as indicated by a are significant, according to tests for multi-
test of dependent correlations (t181=3.69, collinearity they are not large enough to com-
p<.001) (Cohen and Cohen, 1975). Usefulness promise the accuracy of the estimated regres-
and ease of use were significantly correlated sion coefficients since the standard errors of the
with each other for electronic mail (.56), XEDIT estimates are low (.08 for both usefulness and
ease of
of use),
use), and
and the
the covariances
covariancesbetween
betweenthe theThe
The study
study involved
involvedevaluating
evaluatingtwo
twoIBM
IBMPC-
PC-
parameter
parameter estimates
estimates arearenegligible
negligible(-.004)
(-.004) based
based graphics
graphics systems:
systems:Chart-Master
Chart-Master (by
(by
De-
De-
(Johnston,
(Johnston, 1972;
1972; Mansfield
MansfieldandandHelms,
Helms,1982).
1982).cision
cision Resources,
Resources,Inc.
Inc.of
ofWestport,
Westport,CN)
CN)and
and
Pen-
Pen-
Based
Based on
on partial
partial correlation
correlationanalyses,
analyses,the vari- draw
thevari- draw (by
(by Pencept,
Pencept,Inc.
Inc.of
ofWaltham,
Waltham,MA).
MA).Chart-
Chart-
ance in
in usage
usage explained
explainedby
byease
easeof
ofuse drops Master
usedrops Master is
is aa menu-driven
menu-drivenpackage
packagethat
thatcreates
creates
by 98%
98% when
when usefulness
usefulnessisiscontrolled
controlledfor. The numerical
for.The numerical business
businessgraphs,
graphs,such
suchasasbar
barcharts,
charts,
regression
regression and
and partial
partialcorrelation
correlationresults suggestline
resultssuggest line charts,
charts, and
andpie
piecharts
chartsbased
basedononparameters
parameters
that usefulness mediates the effect of ease of defined
defined byby the
the user.
user.Through
Throughthethekeyboard
keyboard and
and
use on usage, i.e., that ease of use influencesmenus, menus, the
the user
userinputs
inputsthe
thedata
datafor,
for,and
anddefines
defines
usage indirectly through its effect on usefulnessthe desired
desired characteristics
characteristicsof,of,the
thechart
chart toto
bebe
(J.A. Davis, 1985). made.
made. The
The user
user can
canspecify
specifya awide
widevariety
varietyofof
options
options relating
relatingto
totitle
titlefonts,
fonts,colors,
colors,plot
plot
orienta-
orienta-
tion,
tion, cross-hatching
cross-hatchingpattern,
pattern,chart
chartformat,
format,
and
and
so on.
on. The
The chart
chartcan
canthen
thenbebepreviewed
previewed onon
the
the
screen,
screen, saved,
saved, and
andprinted.
printed.Chart-Master
Chart-Masteris is
a a
successful
successful commercial
commercialproduct
productthat
thattypifies
typifies
the
the
Study 2 category
category of
of numeric
numericbusiness
businesscharting
chartingprograms.
programs.
A lab study was performed to evaluate the six-
item usefulness and ease of use scales result-
Pendraw
Pendraw is
is quite
quitedifferent
differentfrom
fromthe
thetypical
typical
busi-
busi-
ing from scale refinement in Study 1. Study ness2
ness charting
charting program.
program.ItItuses
usesbit-mapped
bit-mapped graph-
graph-
was designed to approximate applied prototype ics and
and aa "direct
"directmanipulation"
manipulation"interface
interfacewhere
where
testing or system selection situations, an impor-
users
users draw
draw desired
desiredshapes
shapesusing
usinga adigitizer
digitizer
tant class of situations where measures of this tablet
tablet and
and an
an electronic
electronic"pen"
"pen"asasa astylus.
stylus.TheThe
kind are likely to be used in practice. In proto- digitizer
digitizer tablet
tablet supplants
supplantsthe
thekeyboard
keyboard as as
the
the
type testing and system selection contexts, pro- input
input medium.
medium. By Bydrawing
drawingonona atablet,
tablet,the
the user
user
spective users are typically given a brief hands- manipulates
manipulates the
theimage,
image,which
whichisisvisible
visibleonon the
the
on demonstration involving less than an hour of screen
screen as
as it
it is
is being
beingcreated.
created.Pendraw
Pendraw offers
offers
actually interacting with the candidate system. capabilities
capabilities typical
typicalof ofPC-based,
PC-based,bit-mapped
bit-mapped
Thus, representative users are asked to rate the "paint"
"paint" programs
programs(see (seePanko,
Panko,1988),
1988),allowing
allowing
future usefulness and ease of use they would the user user to
to perform
performfreehand
freehanddrawing
drawing and
andselect
select
expect based on relatively little experience with from
from among
among geometric
geometricshapes,
shapes,such
suchasas
boxes,
boxes,
the systems being rated. We are especially in- lines,
lines, and
and circles.
circles.AAvariety
varietyofofline
linewidths,
widths,color
color
terested in the properties of the usefulness and selections and title fonts are available. The
ease of use scales when they are worded in digitizer is also capable of performing character
a prospective sense and are based on limited recognition, converting hand-printer characters
experience with the target systems. Favorable into various fonts (Ward and Blesser, 1985).
psychometric properties under these circum- Pencept had positioned the Pendraw product to
stances would be encouraging relative to their complete with business charting programs. The
use as early warning indicants of user accep- manual introduces Pendraw by guiding the user
tance (Ginzberg, 1981). through the process of creating a numeric bar
chart. Thus, a key marketing issue was the
The lab study involved 40 voluntary participants
extent to which the new product would compete
who were evening MBA students at Boston Uni-
favorably with established brands, such as Chart-
versity. They were paid $25 for participating in Master.
the study. They had an average of five years'
work experience and were employed full-time in
several industries, including education (10 per-Participants were given one hour of hands-on
experience with Chart-Master and Pendraw,
cent), government (10 percent), financial (28 per-
cent), health (18 percent), and manufacturing (8using workbooks that were designed to follow
percent). They had a range of prior experience the same instructional sequence as the user
with computers in general (35 percent none or manuals for the two products, while equalizing
limited; 48 percent moderate; and 17 percent the style of writing and eliminating value state-
extensive) and personal computers in particular ments (e.g., "See how easy that was to do?").
(35 percent none or limited; 48 percent moder- Half of the participants tried Chart-Master first
ate; and 15 percent extensive) but were unfa- and half tried Pendraw first. After using each
miliar with the two systems used in the study. package, a questionnaire was completed.
Reliability
Reliability and
andvalidity
validity questions
questions were
wereworded
wordedasasfollows:
follows:"Assuming
"Assuming
Pendraw
Pendraw would
wouldbe
beavailable
availableononmy
myjob,
job,
I predict
I predict
Cronbach
Cronbach alpha
alphawas
was.98
.98for
forperceived
perceiveduseful-
useful- that
that II will
will use
useititon
ona aregular
regularbasis
basisinin
the
thefuture,"
future,"
ness
ness and
and .94
.94 for
forperceived
perceivedease
easeofof
use.
use.
Con-
Con- followed
followed by by two
twoseven-point
seven-pointscales,
scales,
oneone
with
with
vergent
vergent validity
validitywas
wassupported,
supported,with
withonly
only
two
two
of of
likely-unlikely
likely-unlikelyend-point
end-pointadjectives,
adjectives,the
the
other,
other,
re-re-
72 monotrait-heteromethod
monotrait-heteromethodcorrelations
correlations
falling
falling versed
versed in
in polarity,
polarity,with
withimprobable-probable
improbable-probableend-
end-
below
below significance.
significance.Ease
Easeofofuse
useitem
item
4 (flexibil-
4 (flexibil- point
point adjectives.
adjectives.Such
Suchself-predictions,
self-predictions,oror
"be-
"be-
ity),
ity), applied
applied to
toChart-Master,
Chart-Master,was
wasnot
not
significantly
significantly havioral
havioral expectations,"
expectations,"are
areamong
amongthe
the
most
most
ac-ac-
correlated
correlated with
witheither
eitheritems
items3 3(clear
(clear
and
and
under-
under- curate
curate predictors
predictorsavailable
availablefor
foranan
individual's
individual's
standable)
standable) or
or 55(easy
(easytotobecome
becomeskillful).
skillful).
This
This future
future behavior
behavior(Sheppard,
(Sheppard,etetal.,
al.,
1988;
1988;
War-
War-
suggests
suggests that,
that,contrary
contrarytotoconventional
conventionalwisdom,
wisdom, shaw
shaw and
and Davis,
Davis,1985).
1985).For
ForChart-Master,
Chart-Master, use-
use-
flexibility
flexibility is
is not
notalways
alwaysassociated
associatedwith
withease
ease
of of fulness
fulness was
was significantly
significantlycorrelated
correlatedwith
withself-
self-
use.
use. As
As Goodwin
Goodwin(1987)
(1987)points
pointsout,
out,
flexibility
flexibility
cancan
predicted
predicted usage
usage(r=.71,
(r=.71,p<.001),
p<.001),but
but
ease
ease
ofof
actually
actually impair
impairease
easeof
ofuse,
use,particularly
particularlyfor
for use was not (r=.25, n.s.) (Table 8). Chart-
novice users. With item 4 omitted, Cronbach
Master had a non-significant correlation between
alpha for ease of use would increase from .94 ease of use and usefulness (r=.25, n.s.). For
to .95. Despite the two departures to conver- Pendraw, usage was significantly correlated with
gent validity related to ease of use item 4, no both usefulness (r=.59, p<.001) and ease of
exceptions to the discriminant validity criteria oc-
use (r=.47, p<.001). The ease of use-useful-
curred across a total of 720 comparisons (360 ness correlation was significiant for Pendraw
for each scale). (r=.38, p<.001). When data were pooled across
Factorial validity was assessed by factor ana- systems, usage correlated .85 (p<.001) with use-
lyzing the 12 scale items using principal compo- fulness and .59 (p<.001) with ease of use (see
nents extraction and oblique rotation. The re- Table 8). Ease of use correlated with usefulness
sulting two-factor solution is very consistent with .56 (p<.001). The overall usefulness-use corre-
distinct, unidimensional usefulness and each of lation was significantly greater than the ease of
use scales (Table 7). Thus, as in Study 1, Study use-use correlation, as indicated by a test of de-
2 reflects favorably on the convergent, discrimi- pendent correlations (t77 = 4.78, p<.001) (Cohen
nant, and factorial validity of the usefulness and and Cohen, 1975).
ease of use scales.
Regression analyses (Table 9) indicate that the
effect of usefulness on usage, controlling for
ease of use, was significant at the .001 level
Relationship to use for Chart-Master (b = .69), Pendraw (b = .76) and
overall (b=.75). In contrast, the effect of ease
Participants were asked to self-predict their
future use of Chart-Master and Pendraw. The of use on usage, controlling for usefulness, was
Table
Table
7. Factor Analysis7.
of Perceived
FactorUsefulness Analys
and
and
Ease of Use Items:
Ease Study 2 of Use Item
Factor 1 Factor 2
Scale Items (Usefulness) (Ease of Use)
Usefulness
1 Work More Quickly .91 .01
2 Job Performance .98 -.03
3 Increase Productivity .98 -.03
4 Effectiveness .94 .04
5 Makes Job Easier .95 -.01
6 Useful .88 .11
Ease of Use
1 Easy to Learn -.20 .97
2 Controllable .19 .83
3 Clear & Understandable -.04 .89
4 Flexible .13 .63
5 Easy to Become Skillful .07 .91
6 Easy to Use .09 .91
Table
Table8. Correlations8.
Between
Correlations
Perceived Usefulness, Betwee
Perceived
Perceived
Ease of Use, and Self-Reported
Ease of Use, and
System
System Usage Usage
Correlation
Usefulness Ease of Use Ease of Use
& Usage & Usage & Usefulness
Study 1
Electronic Mail (n- 109) .56*** .32*** .56***
XEDIT (n=75) .68*** .48*** .69***
Pooled (n =184) .63*** .45*** .64***
Study 2
Chart-Master (n = 40) .71*** .25 .25
Pendraw (n = 40) .59*** .47*** .38**
Pooled (n = 80) .85*** .59*** .56***
Davis, et al. (1989) (n= 107)
Wave 1 .65*** .27** .10
Wave 2 .70*** .12 .23**
*** p<.001 ** p<.01 * p<.05
Table 9. Regression Analyses of the Effect of Perceived

Usefulness and Perceived Ease of Use on
Self-Reported Usage
Independent Variables
Usefulness Ease of Use R2
Study 1
Electronic Mail (n = 109) .55*** .01 .31
XEDIT (n = 75) .69*** .02 .46
Pooled (n =184) .57*** .07 .38
Study 2
Chart-Master (n = 40) .69*** .08 .51
Pendraw (n= 40) .76*** .17 .71
Pooled (n = 80) .75*** .17* .74
Davis, et al. (1989) (n= 107)
After 1 Hour .62*** .20*** .45
After 14 Weeks .71** -.06 .49
*** p<.001 ** p<.01 * p<.05
non-significant for both Chart-Master (b=.08, through through usefulness.

usefulness.Partial
Partialcorrelation
correlation
analysis
analysis
n.s.) and Pendraw (b=.17, n.s.) when analyzed indicatesindicates that
thatthe
thevariance
variancein in
usage
usage
explained
explained
separately and borderline significant when ob- by by ease
ease of
ofuse
usedrops
dropsby by91%
91%when
whenusefulness
usefulness
servations were pooled (b= .17, p<.05). The re- is is controlled
controlledfor.
for.Consistent
Consistent with
with
Study
Study1, these
1, these
gression coefficients obtained for Pendraw and regression
regressionandandpartial
partialcorrelation
correlationresults
results
suggest
suggest
Chart-Master were not significantly different (F3, that usefulness mediates the effect of ease of
74 = .014, n.s.). Multicollinearity is ruled out since use on usage. The implications of this are ad-
the standard errors of the estimates are low (.07 dressed in the following discussion.
for both usefulness and ease of use) and the
covariances between the parameter estimates
are negligible (-.004). Discussion
Hence, as in Study 1, the significant pairwise The purpose of this investigation was to develop
correlations between ease of use and usage and validate new measurement scales for per-
drop dramatically when usefulness is controlled ceived usefulness and perceived ease of use,
for, suggesting that ease of use operates two distinct variables hypothesized to be deter-
minants
minantsofof computer
computer usage.
usage.
This effort
This effort
was suc-was suc-fulness
fulness was
wascorrelated
correlated.63.63with
with
self-reported
self-reportedcur-cur-
cessful
cessfulininseveral
several respects.
respects.
The new
The scales
new were
scales wererent
rent use
use in
inStudy
Study1 1and
and .85.85
with
withself-predicted
self-predicted
found
foundtotohavehave strong
strongpsychometric
psychometric
properties
propertiesuse
use in
in Study
Study2.2.Perceived
Perceived ease
ease
of of
useuse
waswas
cor-
cor-
and
andto toexhibit
exhibit significant
significant
empirical
empirical
relationships
relationshipsrelated
related .45
.45with
withuse
useinin
Study
Study 1 and
1 and.69.69
in Study
in Study
with
withself-reported
self-reported measures
measures
of usage
of usage
behavior.
behavior.2. The
The same
samepattern
patternofof correlations
correlations is found
is found
Also,
Also,several
severalnew new
insights
insights
were were
generated
generated
about aboutwhen
when correlations
correlationsarearecalculated
calculated separately
separatelyforfor
the
the nature
natureofof perceived
perceivedusefulness
usefulness
and ease
andof ease of each
each ofof the
thetwo
twosystems
systems inin
each
each
study
study
(Table
(Table
use, and their roles as determinants of user 8).
8). These
These correlations,
correlations,especially
especially thethe
usefulness-
usefulness-
acceptance. use
use link,
link, compare
comparefavorably
favorably with
withother
other
correla-
correla-
tions
tions between
betweensubjective
subjective measures
measures andand
self-
self-
The new scales were developed, refined, and reported
reported use
usefound
foundininthe
the
MIS
MIS
literature.
literature.
Swan-
Swan-
streamlined in a several-step process. Explicit son's
son's (1987)
(1987)"value"
"value"dimension
dimension correlated
correlated.20 .20
definitions were stated, followed by a theoretical
with
with use,
use,while
whilehishis"accessibility"
"accessibility" dimension
dimension cor-cor-
analysis from a variety of perspectives, includ-
related
related .13
.13with
withself-reported
self-reported use.
use.
Correlations
Correlations
ing: expectancy theory; self-efficacy theory; be- between
between "user
"userinformation
information satisfaction"
satisfaction" andand
self-
self-
havioral decision theory; diffusion of innovations;
reported
reported useuseofof.39
.39(Barki
(Barki and
and Huff,
Huff,1985)
1985)
andand
marketing; and human-computer interaction, re-
.28
.28 (Baroudi,
(Baroudi,etetal.,
al.,1986)
1986)have
have been
been
reported.
reported.
garding why usefulness and ease of use are hy-
"Realism
"Realism of ofexpectations"
expectations" hashasbeen
been
found
found
to be
to be
pothesized as important determinants of system
correlated
correlated.22 .22with
withobjectively
objectively measured
measured use use
use. Based on the stated definitions, initial scale
(Ginzberg,
(Ginzberg,1981)
1981)andand.43.43
with
with self-reported
self-reported use use
items were generated. To enhance content va-
(Barki
(Barki and
andHuff,
Huff,1985).
1985)."Motiviational
"Motiviational force"
force"
waswas
lidity, these were pretested in a small pilot study,
correlated
correlated.25.25with
withsystem
system use,
use,objectively
objectively
meas-
meas-
and several items were eliminated. The remain-
ured
ured (DeSanctis,
(DeSanctis,1983).
1983).Among
Amongthethe
usage
usage
cor-
cor-
ing items, 10 for each of the two constructs, were
relations
relations reported
reportedininthe
the
literature,
literature,
thethe
.79 .79
corre-
corre-
tested for validity and reliability in Study 1, a
lation
lation between
between"performance"
"performance" and
anduseuse
reported
reported
field study of 112 users and two systems (the
by
by Robey
Robey(1979)
(1979)stands
standsout.
out.
Recall
Recall
that
that
Robey's
Robey's
PROFS electronic mail system and the XEDIT
expectancy
expectancymodel
modelwas
wasa key
a key
underpinning
underpinning forfor
file editor). Item analysis was performed to elimi-
the
the definition
definitionof
ofperceived
perceivedusefulness
usefulnessstated
stated
in in
nate more items and refine others, further stream- this
this article.
article.
lining and purifying the scales. The resulting six-
item scales were subjected to further construct
validation in Study 2, a lab study of 40 users
One
One of
of the
themost
mostsignificant
significant
findings
findings
is the
is the
rela-
rela-
and two systems: Chart-Master (a menu-driven
tive
tive strength
strengthofofthe
theusefulness-usage
usefulness-usagerelation-
relation-
business charting program) and Pendraw (a bit-
ship
ship compared
comparedtotothetheease
ease
ofof
use-usage
use-usage rela-
rela-
mapped paint program with a digitizer tablet as tionship. In both studies, usefulness was
its input device).
significantly more strongly linked to usage than
The new scales exhibited excellent psychomet- was ease of use. Examining the joint direct effect
ric characteristics. Convergent and discriminant of the two variables on use in regression analy-
validity were strongly supported by multitrait- ses, this difference was even more pronounced:
multimethod analyses in both validation studies. the usefulness-usage relationship remained
These two data sets also provided strong sup- large, while the ease of use-usage relationship
port for factorial validity: the pattern of factor load- was diminished substantially (Table 8). Multi-
ings confirmed that a priori structure of the two collinearity has been ruled out as an explana-
instruments, with usefulness items loading highly tion for the results using specific tests for the
on one factor, ease of use items loading highly presence of multicollinearity. In hindsight, the
on the other factor, and small cross-factor load- prominence of perceived usefulness makes
ings. Cronbach alpha reliability for perceived use- sense conceptually: users are driven to adopt
fulness was .97 in Study 1 and .98 in Study 2. an application primarily because of the functions
Reliability for ease of use was .91 in Study 1 it performs for them, and secondarily for how
and .94 in Study 2. These findings mutually con- easy or hard it is to get the system to perform
firm the psychometric strength of the new meas- those functions. For instance, users are often
urement scales. willing to cope with some difficulty of use in a
system that provides critically needed function-
As theorized, both perceived usefulness andality. Although difficulty of use can discourage
ease of use were significantly correlated with self-adoption of an otherwise useful system, no
reported indicants of system use. Perceived use- amount of ease of use can compensate for a
system
system that
that does
does not
notperform
performa auseful
usefulfunction.
function. tern of results in a two-wave study (Tables 8
The prominence
prominence of ofusefulness
usefulnessover
overease
easeofofuse
use and 9). In that study, MBA student subjects were
has important
important implications
implicationsfor
fordesigners,
designers,particu-
particu- asked to fill out a questionnaire after a one-hour
larly
larly in
in the
the human
humanfactors
factorstradition,
tradition,whowhohave
have introduction to a word processing program, and
tended
tended toto overemphasize
overemphasizeeaseeaseof
ofuse
useand
andover-
over- again 14 weeks later. Usage intentions were
look usefulness
usefulness (e.g.,
(e.g.,Branscomb
BranscombandandThomas,
Thomas, measured at both time periods, and self-
1984;
1984; Chin,
Chin, et
et al.,
al., 1988;
1988;Shneiderman,
Shneiderman,1987).
1987). reported usage was measured at the later time
Thus,
Thus, aa major
major conclusion
conclusionof
ofthis
thisstudy
studyisisthat
that period. Intentions were significantly correlated
perceived
perceived usefulness
usefulnessisisaastrong
strongcorrelate
correlateofof with usage (.35 and .63 for the two points in
user acceptance
acceptance and
andshould
shouldnot
notbebeignored
ignoredbyby time, respectively). Unlike the results of Studies
those
those attempting
attempting totodesign
designor
orimplement
implementsuc- suc- 1 and 2, Davis, et al. (1989) found a significant
cessful
cessful systems.
systems. direct effect of ease of use on usage, controlling
for usefulness, after the one-hour training ses-
From
From aa causal
causal perspective,
perspective,the
theregression
regressionre-
re- sion (Table 9), although this evolved into a non-
sults
sults suggest
suggest that
that ease
easeof
ofuse
usemay
maybebeananante-
ante- significant effect as of 14 weeks later. In gen-
cedent
cedent to
to usefulness,
usefulness,rather
ratherthan
thana aparallel,
parallel, eral, though, Davis, et al. (1989) found useful-
direct
direct determinant
determinantof ofusage.
usage.The
Thesignificant
significant ness to be more influential than ease of use in
pairwise
pairwise correlation
correlationbetween
betweenease
easeofofuse
useand
and driving usage behavior, consistent with the find-
usage
usage all
all but
but vanishes
vanisheswhen
whenusefulness
usefulnessisiscon-
con- ings reported above.
trolled
trolled for.
for. This,
This, coupled
coupledwith
witha asignificant
significantease
ease
of use-usefulness
use-usefulness correlation
correlationisisexactly
exactlythe
thepat-
pat- Further research will shed more light on the gen-
tern
tern one
one would
would expect
expectififusefulness
usefulnessmediated
mediated erality of these findings. Another limitation is that
between ease of use and usage (e.g., J.A. the usage measures employed were self-
Davis, 1985). That is, the results are consistent reported as opposed to objectively measured.
with an ease of use --> usefulness --> usage Not enough is currently known about how accu-
chain of causality. These results held both for rately self-reports reflect actual behavior. Also,
pooled observations and for each individual since usage was reported on the same ques-
system (Table 8). The causal influence of ease tionnaire used to measure usefulness and ease
of use on usefulness makes sense conceptu- of use, the possibility of a halo effect should not
ally, too. All else being equal, the easier a be overlooked. Future research addressing the
system is to interact with, the less effort needed relationship between these constructs and ob-
to operate it, and the more effort one can allo- jectively measured use is needed before claims
cate to other activities (Radner and Rothschild, about the behavioral predictiveness can be
1975), contributing to overall job performance. made conclusively. These limitations notwithstand-
Goodwin (1987) also argues for this flow of cau- ing, the results represent a promising step
sality, concluding from her analysis that: "There toward the establishment of improved measures
is increasing evidence that the effective func- for two important variables.
tionality of a system depends on its usability"
(p. 229). This intriguing interpretation is prelimi-
nary and should be subjected to further experi-
mentation. If true, however, it underscores the Research implications
theoretical importance of perceived usefulness. Future research is needed to address how other
variables relate to usefulness, ease of use, and
This investigation has limitations that should be acceptance. Intrinsic motivation, for example,
pointed out. The generality of the findings re- has received inadequate attention in MIS theo-
mains to be shown by future research. The fact ries. Whereas perceived usefulness is con-
that similar findings were observed, with respect cerned with performance as a consequence use,
to both the psychometric properties of the meas- intrinsic motivation is concerned with the rein-
ures and the pattern of empirical associations, forcement and enjoyment related to the process
across two different user populations, two differ- of performing a behavior per se, irrespective of
ent systems, and two different research settings whatever external outcomes are generated by
(lab and field), provides some evidence favoring such behavior (Deci, 1975). Although intrinsic
external validity. motivation has been studied in the design of com-
puter games (e.g., Malone, 1981), it is just be-
In addition, a follow-up to this study, reported ginning to be recognized as a potential mecha-
by Davis, et al. (1989) found a very similar pat- nism underlying user acceptance of end-user
systems
systems (Carroll
(Carroll and
and Thomas,
Thomas,1988).
1988).Currently,
Currently, through
throughconcept
concept screening
screening
andand
prototype
prototype
test- test-
the role
role of
of affective
affective attitudes
attitudesisisalso
alsoan
anopen
open ing
ing to
topost-implementation
post-implementation assessment.
assessment.
The fact
The fact
issue. While
While some
some theorists
theorists argue
arguethat
thatbeliefs
beliefs that
that the
themeasures
measures performed
performedwellwell
psychometri-
psychometri-
influence
influence behavior
behavior only
only via
via their
theirindirect
indirectinflu-
influ- cally
cally both
bothafter
afterbrief
brief
introductions
introductions
to the
to target
the target
ence on
on attitudes
attitudes (e.g.,
(e.g., Fishbein
Fishbeinand
andAjzen,
Ajzen, system
system(Study
(Study2,2,
and
and
Davis,
Davis,
et al.,
et 1989)
al., 1989)
and and
1975), others
others view
view beliefs
beliefs and
andattitudes
attitudesasasco-
co- after
after substantial
substantialuser
user
experience
experience
withwith
the system
the system
determinants
determinants ofof behavioral
behavioral intentions
intentions(e.g.,
(e.g.,Tri-
Tri- (Study
(Study1,1,and
andDavis,
Davis,et et
al.,al.,
1989)
1989)
is promising
is promising
andis, 1977),
1977), and
and still
still others
others view
viewattitudes
attitudesasas concerning
concerningtheir
theirappropriateness
appropriateness
at various
at various
antecedents
antecedents of
of beliefs
beliefs (e.g.,
(e.g., Weiner,
Weiner,1986).
1986). points
pointsin inthe
thelife
lifecycle.
cycle.
Practitioners
Practitioners
generally
generally
Counter
Counter toto Fishbein
Fishbein and
and Ajzen's
Ajzen's(1975)
(1975)position,
position, evaluate
evaluatesystems
systems not
not
onlyonly
to predict
to predict
acceptabil-
acceptabil-
both Davis
Davis (1986)
(1986) and
and Davis,
Davis, et
etal.
al.(1989)
(1989)found
found ity
ity but
butalso
alsototo
diagnose
diagnose thethe
reasons
reasons
underlying
underlying
that attitudes
attitudes do
do not
not fully
fully mediate
mediatethe theeffect
effectofof lack
lack of
ofacceptance
acceptance and
andto to
formulate
formulateinterven-
interven-
perceived
perceived usefulness
usefulness and
and perceived
perceivedeaseeaseof
ofuse
use tions
tions to
toimprove
improve user
user
acceptance.
acceptance.
In this
In this
sense,sense,
on behavior. research on how usefulness and ease of use
can be influenced by various externally control-
It should be emphasized that perceived useful- lable factors, such as the functional and inter-
ness and ease of use are people's subjective face characteristics of the system (Benbasat and
appraisal of performance and effort, respectively, Dexter, 1986; Bewley, et al., 1983; Dickson, et
and do not necessarily reflect objective reality. al., 1986), development methodologies (Alavi,
In this study, beliefs are seen as meaningful vari- 1984), training and education (Nelson and
ables in their own right, which function as be- Cheney, 1987), and user involvement in design
havioral determinants, and are not regarded as (Baroudi, et al. 1986; Franz and Robey, 1986)
surrogate measures of objective phenomena (as is important. The new measures introduced here
is often done in MIS research, e.g., Ives, et al., can be used by researchers investigating these
1983; Srinivasan, 1985). Several MIS studies issues.
have observed discrepancies between perceived
and actual performance (Cats-Baril and Huber, Although there has been a growing pessimism
1987; Dickson, et al., 1986; Gallupe and De- in the field about the ability to identify measures
Sanctis, 1988; Mcintyre, 1982; Sharda, et al., that are robustly linked to user acceptance, the
1988). Thus, even if an application would objec- view taken here is much more optimistic. User
tively improve performance, if users don't per- reactions to computers are complex and multi-
ceive it as useful, they're unlikely to use it (Alavi faceted. But if the field continues to systemati-
and Henderson, 1981). Conversely, people may cally investigate fundamental mechanisms driv-
overrate the performance gains a system has ing user behavior, cultivating better and better
to offer and adopt systems that are dysfunc- measures and critically examining alternative theo-
tional. Given that this study indicates that people retical models, sustainable progress is within
reach.
act according to their beliefs about performance,
future research is needed to understand why per-
formance beliefs are often in disagreement with
objective reality. The possibility of dysfunctional
impacts generated by information technology Acknowledgements
(e.g., Kottemann and Remus, 1987) emphasizes This research was supported by grants from the
that user acceptance is not a universal goal and MIT Sloan School of Management, IBM Canada
is actually undesireable in cases where systems Ltd., and The University of Michigan Business
fail to provide true performance gains. School. The author is indebted to the anony-
mous associate editor and reviewers for their
More research is needed to understand how
many helpful suggestions.
measures such as those introduced here per-
form in applied design and evaluation settings.
The growing literature on design principles (An-
References
derson and Olson, 1985; Gould and Lewis,
1985; Johansen and Baker, 1984; Mantei and Abelson, R.P. and Levi, A. "Decision Making and
Teorey, 1988; Shneiderman, 1987) calls for the Decision Theory," in The Handbook of Social
use of subjective measures at various points Psychology, third edition, G. Lindsay and E.
throughout the development and implementation Aronson (eds.), Knopf, New York, NY, 1985,
process, from the earliest needs assessment pp. 231-309.
Adelbratt,
Adelbratt, T.
T. and
andMontgomery,
Montgomery,H.H."Attractiveness
"Attractiveness Multitmethod
Multitmethod Matrix,"
Matrix,"Psychological
Psychological
Bulletin
Bulletin
of Decision
Decision Rules,"
Rules,"Acta
ActaPsychologica
Psychologica(45),
(45), (56:9),
(56:9),March
March1959,
1959,
pp.pp.
81-105.
81-105.
1980,
1980, pp.
pp. 177-185.
177-185. Campbell,
Campbell,D.T.,D.T.,Siegman,
Siegman, C.R.C.R.
and and
Rees,Rees,
M.B. M.B.
Alavi,
Alavi, M.
M. "An
"An Analysis
Analysisof
ofthe
thePrototyping
Prototyping Ap-
Ap- "Direction-of-Wording
"Direction-of-Wording Effects
Effects
in the
in Relation-
the Relation-
proach
proach toto Information
InformationSystems
SystemsDevelopment,"
Development," ships
ships Between
Between Scales,"
Scales,"
Psychological
PsychologicalBulle-Bulle-
Communications
Communicationsof ofthe
theACM
ACM(27:6),
(27:6),
June
June tin
tin (68:5),
(68:5),November
November 1967,
1967,
pp. pp.
293-303.
293-303.
1984,
1984, pp.
pp. 556-563.
556-563. Card,
Card,S.K.,
S.K.,Moran,
Moran, T.P.
T.P.
andand
Newell,
Newell,
A. The
A. Psy-
The Psy-
Alavi,
Alavi, M.
M. and
and Henderson,
Henderson,J.C.
J.C."An
"AnEvolutionary
Evolutionary chology
chologyof ofHuman-Computer
Human-Computer Interaction,
Interaction,
Strategy
Strategy forfor Implementing
Implementinga aDecision
DecisionSupport
Support Erlbaum,
Erlbaum,Hillsdale,
Hillsdale,
NJ, NJ,
1984.
1984.
System,"
System," Management
ManagementScience
Science(27:11),
(27:11), No- Carroll,
No- Carroll,J.M.
J.M.andand
Carrithers,
Carrithers,
C. "Training
C. "Training
WheelsWheels
vember
vember 1981,
1981, pp.
pp.1309-1323.
1309-1323. in
in aa User
UserInterface,"
Interface," Communications
Communications of the
of the
Anastasi,
Anastasi, A.
A. "Evolving
"EvolvingConcepts
Conceptsofof
Test
Test
Valida-
Valida- ACM
ACM (27:8),
(27:8),August
August1984,
1984,
pp. pp.
800-806.
800-806.
tion,"
tion," Annual
Annual Review
Reviewof
ofPsychology
Psychology(37),
(37), Carroll,
Carroll,J.M.
J.M.and
and
McKendree,
McKendree,J. "Interface
J. "Interface
1986,
1986, pp.
pp. 1-15.
1-15. Design
DesignIssues
Issuesfor
forAdvice-Giving
Advice-GivingExpert
Expert
Sys- Sys-
Anderson,
Anderson, N.S.
N.S.and
andOlson,
Olson,J.R.
J.R.(eds.)
(eds.)
Methods
Methods tems,"
tems,"Communications
Communications of the
of the
ACMACM(30:1,(30:1,
for
for Designing
Designing Software
SoftwaretotoFit
FitHuman
Human Needs
Needs January
January1987,
1987,pp.pp.
14-31.
14-31.
and
and Capabilities:
Capabilities:Proceedings
Proceedingsofofthe
the Work- Carroll,
Work- Carroll,J.M.,
J.M., Mack,
Mack,R.L.,
R.L.,
Lewis,
Lewis,
C.H.,C.H.,
Grishkow-
Grishkow-
shop
shop on
on Software
SoftwareHuman
HumanFactors,
Factors,National
National sky,
sky, N.L.
N.L.and
andRobertson,
Robertson, S.R.S.R.
"Exploring
"Exploring
Ex- Ex-
Academy
Academy Press,
Press,Washington,
Washington,D.C.,
D.C.,1985.
1985. ploring
ploringa aWord
WordProcessor,"
Processor,"Human-Computer
Human-Computer
Bandura,
Bandura, A. A. "Self-Efficacy
"Self-EfficacyMechanism
Mechanism ininHuman
Human Interaction
Interaction(1), (1),
1985,
1985, pp.pp.
283-307.
283-307.
Agency,"
Agency," American
AmericanPsychologist
Psychologist(37:2),
(37:2), Feb- Carroll,
Feb- Carroll,J.M.
J.M.and andThomas,
Thomas, J.C.J.C.
"Fun,"
"Fun,"
SIGCHI SIGCHI
ruary
ruary 1982,
1982, pp.
pp.122-147.
122-147. Bulletin
Bulletin (19:3),
(19:3),January
January 1988,
1988,
pp. 21-24.
pp. 21-24.
Barki,
Barki, H.H. and
and Huff,
Huff,S.S."Change,
"Change,Attitude
Attitude to to Cats-Baril,
Cats-Baril,W.L.W.L. and
andHuber,
Huber,G.P.G.P.
"Decision
"DecisionSup- Sup-
Change,
Change, and and Decision
DecisionSupport
SupportSystem
System Suc-
Suc- port
port Systems
Systems for
forIll-Structured
Ill-Structured Problems:
Problems:An An
cess,"
cess," Information
Informationand andManagement
Management (9:5),
(9:5),De-De- Empirical
EmpiricalStudy,"
Study," Decision
Decision Sciences
Sciences (18:3),
(18:3),
cember
cember 1985,
1985, pp.
pp.261-268.
261-268. Summer
Summer1987, 1987, pp.
pp.352-372.
352-372.
Baroudi,
Baroudi, J.J.,
J.J., Olson,
Olson,M.H.
M.H.andandIves,
Ives,
B.B."An"An Em-Em- Cheney,
Cheney,P.H.,
P.H.,Mann,
Mann, R.I.R.I.
andand
Amoroso,
Amoroso,D.L. "Or-
D.L. "Or-
pirical
pirical Study
Study ofofthe
theImpact
ImpactofofUserUser Involve-
Involve- ganizational
ganizationalFactors
Factors Affecting
Affecting the the
SuccessSuccess
of of
ment
ment on on System
SystemUsage
Usageand
andInformation
Information Sat-
Sat- End-User
End-UserComputing,"
Computing," Journal
Journalof Manage-
of Manage-
isfaction,"
isfaction," Communications
Communicationsofofthe theACM
ACM (29:3),
(29:3), ment
mentInformation
Information Systems
Systems (3:1),
(3:1),
Summer Summer
March
March 1986,
1986, pp.
pp.232-238.
232-238. 1986,
1986,pp.
pp.65-80.
65-80.
Beach,
Beach, L.R.
L.R. and
andMitchell,
Mitchell,T.R. Contingency Chin,
T.R."A"AContingency Chin,J.P.,
J.P.,Diehl,
Diehl, V.A.
V.A.andandNorman,
Norman, K.L. "De-
K.L. "De-
Model for the Selection of Decision Strate- velopment
velopmentofof ananInstrument
Instrument for for
Measuring
Measuring
gies," Academy of Management Review (3:3), User
User Satisfaction
Satisfaction of ofthetheHuman-Computer
Human-Computer In- In-
July 1978, pp. 439-449. terface,"
terface,"CHI'88
CHI'88 Human
Human Factors
Factors
in Comput-
in Comput-
Benbasat, I. and Dexter, A.S. "An Investigation ing
ing Systems,
Systems,Washington,
Washington, D.C.,D.C.,
MayMay 15-19, 15-19,
of the Effectiveness of Color and Graphical 1988,
1988,ACM,
ACM,New New York,
York, NY,NY,
pp. pp.
213-218.
213-218.
Presentation Under Varying Time Constraints, Cohen,
Cohen,J. J.and
andCohen,
Cohen, P. Applied
P. Applied
Multiple
Multiple
Re- Re-
MIS Quarterly (10:1), March 1986, pp. 59-84. gression/
gression/Correlation
Correlation Analysis
Analysis
for for
the Behav-
the Behav-
Bewley, W.L., Roberts, T.L., Schoit, D. and Ver- ioral
ioral Sciences,
Sciences, Erlbaum,
Erlbaum, Hillsdale,
Hillsdale,
NJ, NJ,1975.1975.
plank, W.L., "Human Factors Testing in the Curley,
Curley,K.F.
K.F."Are
"AreThere
There anyany
RealReal
Benefits
Benefits
from from
Design of Xerox's 8010 'Star' Office Worksta- Office
OfficeAutomation?"
Automation?" Business
BusinessHorizons
Horizons (4), (4),
tion," CHI '83 Human Factors in Computing July-August
July-August1984, 1984, pp.pp.37-42.
37-42.
Systems, Boston, December 12-15, 1983, Davis,
Davis,F.D.
F.D."A
"ATechnology
Technology
Acceptance
Acceptance
ModelModel
ACM, New York, NY, pp. 72-77. for
for Empirically
EmpiricallyTesting
Testing
New
New
End-User
End-User
Infor-
Infor-
Bohrnstedt, G.W. "Reliability and Validity Assess- mation
mationSystems:
Systems: Theory
Theory
andand
Results,"
Results,"
doc- doc-
ment in Attitude Measurement," in Attitude toral
toral dissertation,
dissertation,MIT
MIT
Sloan
Sloan
School
School
of Man-
of Man-
Measurement, G.F. Summers (ed.), Rand- agement,
agement,Cambridge,
Cambridge, MA,MA,
1986.
1986.
McNally, Chicago, IL, 1970, pp. 80-99. Davis,
Davis,F.D.,
F.D.,Bagozzi,
Bagozzi,R.P.
R.P.
andand
Warshaw,
Warshaw,
P.R. P.R.
Bowen, W. "The Puny Payoff from Office Com- User
User Acceptance
Acceptance ofof
Computer
Computer Technology:
Technology:
A A
puters," Fortune, May 26, 1986, pp. 20-24. Comparison
ComparisonofofTwo
Two
Theoretical
Theoretical
Models,"
Models,"
Man- Man-
Branscomb, L.M. and Thomas, J.C. "Ease of agement
agementScience
Science(35:8),
(35:8),
August
August
1989,
1989,
pp. pp.
Use: A System Design Challenge," IBM Sys- 982-1003.
tems Journal (23), 1984, pp. 224-235. Davis, J.A. The Logic of Causal Order, Sage,
Campbell, D.T. and Fiske, D.W. "Convergent Beverly Hills, CA, 1985.
and Discriminant Validation by the Multitrait- Deci, E.L. Intrinsic Motivation, Plenum, New
York,
York, NY,
NY,1975.
1975. Management Science
Science (27:1),
(27:1), January
January 1981,1981,
DeSanctis,
DeSanctis,G.G."Expectancy
"Expectancy Theory
Theoryas an an Expla- pp. 33-56.
as Expla-
nation
nation of
ofVoluntary
Voluntary Use
Useof of
a Decision
a Decision SupportHill, T., Smith, N.D.,
Support N.D., and
and Mann,
Mann, M.F.
M.F. "Role
"Role ofof
System,"
System,"Psychological
Psychological Reports
Reports(52),
(52), 1983, Efficacy Expectations
1983, Expectations inin Predicting
Predicting thethe Deci-
Deci-
pp.
pp. 247-260.
247-260. sion to Use Advanced
Advanced Technologies:
Technologies: The The
Dickson,
Dickson,G.W.,
G.W.,DeSanctis,
DeSanctis, G. G.
andand
McBride,
McBride,D.J. D.J. Case of Computers,"
Computers," Journal
Journal of of Applied
Applied Psy-
Psy-
"Understanding
"Understandingthe the
Effectiveness
Effectiveness of Computer chology, (72:2), May
of Computer May 1987,
1987, pp.
pp. 307-313.
307-313.
Graphics
GraphicsforforDecision
DecisionSupport:
Support: A CumulativeIves, B., Olson, M.H.
A Cumulative M.H. and
and Baroudi,
Baroudi, J.J.
J.J. "The
"The meas-
meas-
Experimental
ExperimentalApproach,"
Approach," Communications
Communications of of urement of User Information
Information Satisfaction,"
Satisfaction," Com-
Com-
the
the ACM
ACM(29:1),
(29:1),January
January 1986,
1986,
pp.pp.
40-47.
40-47. munications of the ACM ACM (26:10),
(26:10), October
October
Edelmann,
Edelmann,F.F."Managers,
"Managers, Computer
Computer Systems, 1983, pp. 785-793.
Systems,
and
and Productivity,"
Productivity," MIS
MISQuarterly
Quarterly(5:3), Sep-Sep- Jarvenpaa, S.L. "The
(5:3), "The Effect
Effect of of Task
Task Demands
Demands
tember
tember1981,
1981,pp.
pp.1-19.
1-19. and Graphical Format
Format on on Information
Information Process-
Process-
Fishbein,
Fishbein,M.M.and
andAjzen,
Ajzen,
I. "Belief,
I. "Belief,
Attitude, In- In- ing Strategies," Management
Attitude, Management Science
Science (35:3),
(35:3),
tention and Behavior: An Introduction to March 1989, pp. 285-303.
285-303.
Theory and Research," Addison-Wesley, Read- Johansen, R. & Baker
Baker E.,
E., "User
"User Needs
Needs Work-
Work-
ing, MA 1975. shops: A New Approach
Approach to to Anticipating
Anticipating UserUser
Franz, C.R. and Robey, D. "Organizational Con- Needs for Advanced
Advanced Office
Office Systems,"
Systems," Office
Office
text, User Involvement, and the Usefulness of Technology and People
People (2),
(2), 1984,
1984, pp.pp. 103-
103-
Information Systems," Decision Sciences 119.
(17:3), Summer 1986, pp. 329-356. Johnson, E.J. and Payne,
Payne, J.W.
J.W. "Effort
"Effort and
and Ac-
Ac-
Gallupe, R.B., DeSanctis, G. and Dickson, G.W.curacy in Choice," Management Science
"Computer-Based Support for Group Problem(31:4), April 1985, pp. 395-414.
Johnston, J. Econometric Methods, McGraw-
Finding: An Empirical Investigation," MIS Quar-
terly (12:2), June 1988, pp. 277-296. Hill, New York, NY, 1972.
Klein, G. and Beck, P.O. "A Decision Aid for
Ginzberg, M.J. "Early Diagnosis of MIS Implemen-
tation Failure: Promising Results and Unan- Selecting Among Information Systems Alter-
swered Questions," Management Science natives," MIS Quarterly (11:2), June 1987, pp.
(27:4), April 1981, pp. 459-478. 177-186.
Good, M., Spine, T.M., Whiteside, J. and George Kleinmuntz, D.N. and Schkade, D.A. "The Cog-
P. "User-Derived Impact Analysis as a Tool nitive Implications of Information Displays in
for Usability Engineering," CHI'86 Human Fac- Computer-Supported Decision-Making," Uni-
tors in Computing Systems, Boston, April 13- versity of Texas at Austin, Graduate School
17, 1986, ACM, New York, New York pp. 241- of Business, Department of Management Work-
246. ing Paper 87/88-4-8, 1988.
Goodwin, N.C. "Functionality and Usability," Com- Kottemann, J.E. and Remus, W.E. "Evidence
munications of the ACM (30:3), March 1987, and Principles of Functional and Dysfunctional
pp. 229-233. DSS," OMEGA (15:2), March 1987, pp. 135-
Goslar, M.D. "Capability Criteria for Marketing 143.
Decision Support Systems," Journal of Man- Larcker, D.F. and Lessig, V.P. "Perceived Use-
agement Information Systems (3:1), Summer fulness of Information: A Psychometric Exami-
1986, pp. 81-95. nation," Decision Sciences (11:1), January
Gould, J., Conti, J. and Hovanyecz, T. "Com- 1980, pp. 121-134.
posing letters with a Simulated Listening Type- Lucas, H.C. "Performance and the Use of an
writer," Communications of the ACM (26:4), Information System," Management Science
April 1983, pp. 295-308. (21:8), April 1975, pp. 908-919.
Malone, T.W. "Toward a Theory of Intrinsically
Gould, J.D. and Lewis C. "Designing for Usabil-
Motivating Instruction," Cognitive Science (4),
ity: Key Principles and What Designers Think,"
Communications of the ACM (28:3), March 1981, pp. 333-369.
Mansfield, E.R. and Helms, B.P. "Detecting Mul-
1985, pp. 300-311.
ticollinearity," The American Statistician (36:3),
Greenberg, K. "Executives Rate Their PCs," PC August 1982, pp. 158-160.
World, September 1984, pp. 286-292. Mantei, M.M. and Teorey, T.J. "Cost/Benefit
Hauser, J.R. and Simmie, P. "Profit Maximizing Analysis for Incorporating Human Factors in
Perceptual Positions: An Integrated Theory for the Software Lifecycle," Communications of
the Selection of Product Features and Price," the ACM (31:4), April 1988, pp. 428-439.
Markus,
Markus, M.L.
M.L.and
andBjorn-Anderson,
Bjorn-Anderson, N. N.
"Power
"Power ber-December
ber-December 1975,
1975, pp.
pp. 321-343.
321-343.
Over
Over Users:
Users:It's
It'sExercise
Exercisebyby
System
SystemProfes-
Profes- Schein,
Schein, E.H.
E.H. Organizational
OrganizationalPsychology,
Psychology,third
thirdedi-
edi-
sionals,"
sionals," Communications
Communications ofof
the
the
ACM
ACM(30:6),
(30:6), tion, Prentice-Hall,
Prentice-Hall, Englewood
EnglewoodCliffs,
Cliffs,NJ,
NJ,1980.
1980.
June
June 1987,
1987,pp.
pp.498-504.
498-504. Schewe,
Schewe, C.D.
C.D. "The
"The Management
ManagementInformation
Information
Mcintyre,
Mcintyre, S.S."An
"AnExperimental
Experimental Study
Studyof of
thethe System
System User:
User: An
An Exploratory
ExploratoryBehavioral
BehavioralAnaly-
Analy-
Impact of Judgement-Based Marketing sis," Academy
Academy of of Management
ManagementJournal
Journal(19:4),
(19:4),
Models," Management Science (28:1), Janu- December
December 1976,
1976, pp.
pp. 577-590.
577-590.
ary 1982, pp. 17-23. Schultz,
Schultz, R.L.
R.L. and
and Slevin,
Slevin,D.P.
D.P."Implementation
"Implementation
Nelson, R.R. and Cheney, P.H. "Training End and Organizational
Organizational Validity:
Validity:AnAnEmpirical
EmpiricalIn-
In-
Users: An Exploratory Study," MIS Quarterly vestigation,"
vestigation," in
in Implementing
ImplementingOperations
OperationsRe-
Re-
(11:4), December 1987, pp. 547-559. search/Management
search/Management Science,
Science,R.L.
R.L.Schultz
Schultz
Nickerson, R.S. "Why Interactive Computer Sys- and D.P. Slevin (eds.), American Elsevier,
tems Are Sometimes Not Used by People New York, NY, 1975, pp. 153-182.
Who Might Benefit from Them," International Sharda, R., Barr, S.H., and McDonnell, J.C. "De-
Journal of Man-Machine Studies (15), 1981, cision Support System Effectiveness: A
pp. 469-483. Review and Empirical Test," Management Sci-
Norman, D.A. "Design Principles for Human- ence (34:2), February 1988, pp. 139-159.
Computer Interfaces," CHI '83 Human Fac- Sheppard, B.H., Hartwick, J. and Warshaw, P.R.
tors in Computing Systems, Boston, Decem- "The Theory of Reasoned Action: A Meta-
ber 12-15, 1983, ACM, New York, NY, pp. 1- Analysis of Past Research with Recommen-
10. dations for Modifications and Future Re-
Nunnally, J. Psychometric Theory, McGraw-Hill, search," Journal of Consumer Research
New York, NY, 1978. (15:3), December 1988, pp. 325-343.
Panko, R.R. End-User Computing: Manage- Sherif, M. and Sherif, C.W. "The Own Catego-
ment, Applications, and Technology, Wiley, ries Approach in Attitude Research," in Read-
New York, NY, 1988. ings in Attitude Theory and Measurement, M.
Payne, J. W. "Contingent Decision Behavior," Fishbein (ed.), Wiley, New York, NY, 1967,
Psychological Bulletin, (92:2), 1982, pp. 382- pp. 190-198.
402. Shneiderman, B. Designing the User Interface,
Pfeffer, J. Organizations and Organization Addison-Wesley, Reading, MA, 1987.
Theory, Pitman, Boston, MA, 1982. Silk, A.J. "Response Set and Measurement of
Radner, R. and Rothschild, M. "On the Alloca- Self-Designated Opinion Leadership," Public
tion of Effort," Journal of Economic Theory Opinion Quarterly (35), 1971, pp. 383-397.
(10), 1975, pp. 358-376. Srinivasan, A. "Alternative Measures of System
Roberts, T.L. and Moran, T.P. "The Evaluation Effectiveness: Associations and Implications,"
of Text Editors: Methodology and Empirical Re- MIS Quarterly (9:3), September 1985, pp. 243-
sults," Communications of the ACM (26:4), 253.
April 1983, pp. 265-283. Swanson, E.B. "Management Information Sys-
Robey, D. "User Attitudes and Management In- tems: Appreciation and Involvement," Manage-
formation System Use," Academy of Manage- ment Science (21:2), October 1974, pp. 178-
ment Journal (22:3), September 1979, pp. 527- 188.
538. Swanson, E.B. "Measuring User Attitudes in MIS
Robey, D. and Farrow, D. "User Involvement in Research: A Review," OMEGA (10:2), March
Information System Development: A Conflict 1982, pp. 157-165.
Model and Empirical Test," Management Sci- Swanson, E.B. "Information Channel Disposition
ence (28:1), January 1982, pp. 73-85. and Use," Decision Sciences (18:1), Winter
Rogers, E.M. and Shoemaker, F.F. Communi- 1987, pp. 131-145.
cation of Innovations: A Cross-Cultural Ap- Swanson, E.B. Information System Implemen-
proach, Free Press, New York, NY, 1971. tation: Bridging the Gap Between Design and
Rushinek, A. and Rushinek, S.F. "What Makes Utilization, Irwin, Homewood, IL, 1988.
Users Happy?" Communications of the ACM Tornatzky, L.G. and Klein, K.J. "Innovation Char-
(29:7), July 1986, pp. 594-598. acteristics and Innovation Adoption-Implemen-
Saracevic, T. "Relevance: A Review of and a tation: A Meta-Analysis of Findings," IEEE
Framework for the Thinking on the Notion in Transactions on Engineering Management
Information Science," Journal of the Ameri- (EM-29:1), February 1982, pp. 28-45.
can Society for Information Science, Novem- Triandis, H.C. Interpersonal Behavior, Brooks/
Cole,
Cole, Monterey,
Monterey, CA,CA,1977.
1977. and
and Iconic
IconicInterfaces,"
Interfaces,"
CHI CHI
'85 Proceedings,
'85 Proceedings,
Vertinsky,
Vertinsky, I.,
I., Barth,
Barth,R.T.
R.T.and
andMitchell,
Mitchell,V.F.
V.F."A"A San
San Francisco,
Francisco,April
April
14-18,
14-18,
1985,1985,
ACM, ACM,
New New
Study
Study of
of OR/MS
OR/MS Implementation
Implementationasasa aSocial
Social York,
York,NY,
NY, pp.
pp.
185-191.
185-191.
Change
Change Process,"
Process," in
inImplementing
ImplementingOpera-
Opera- Wright,
Wright,P.P.
"Consumer
"ConsumerChoice
Choice
Strategies:
Strategies:
Sim- Sim-
tions
tions Research/Management
Research/ManagementScience,
Science,R.L.
R.L. plifying
plifyingvs.
vs.
Optimizing,"
Optimizing,"
Journal
Journal
of Marketing
of Marketing
Schultz and D.P. Slevin (eds.), American Research
Research(14:1),
(14:1),
February
February
1975,1975,
pp. 429-
pp. 429-
Elsevier, New York, NY, 1975, pp. 253-272. 433.
Vroom, V.H. Work and Motivation, Wiley, New Young, T.R. "The Lonely Micro," Datamation
York, NY, 1964. (30:4), April 1984, pp. 100-114.
Ward, J.R. and Blesser, B. "Interactive Recog-
nition of Handprinter Characters for Computer
About the Author
Input," IEEE Computer Graphics and Appli-
cations, September 1985, pp. 24-37. Fred D. Davis is assistant professor at the Uni-
Warshaw, P.R. and Davis, F.D. "Disentangling versity of Michigan School of Business Admini-
Behavioral Intention and Behavioral Expecta- stration. His doctoral research at the Sloan
tion," Journal of Experimental Social Psychol- School of Management, MIT, dealt with predict-
ogy (21), May 1985, pp. 213-228. ing and explaining user acceptance of computer
Weiner, B. "Attribution, Emotion, and Action," in technology. His current research interests in-
Handbook of Motivation and Cognition, R.M. clude computer support for decision making, mo-
Sorrentino and E.T. Higgins (eds.), Guilford, tivational determinants of computer acceptance,
New York, NY, 1986, pp. 281-312. intentions and expectations in human behavior,
Whiteside, J., Jones, S., Levy, P.S. and Wixon, and biased attributions of the performance im-
D. "User Performance With Command, Menu, pacts of information technology.
Appendix
Appendix
Final Measurement Scales for Perceived Usefulness and
Perceived Usefulness
Using CHART-MASTER in my job would enable me to accomplish tasks more quic
likely I I I I I I I unlikely
extremely quite slightly neither slightly quite extremely
Using CHART-MASTER would improve my job performance.
likely 1- I I I I I unlikely
Using CHART-MASTER in my job would increase my productivity.
likely I I I I I unlikely
Using CHART-MASTER would enhance my effectiveness on the job.
likely I I I unlikely
Using CHART-MASTER would make it easier to do my job.
likely II I I I I I I unlikely
I would find CHART-MASTER useful in my job.

Learning to operate CHART-MASTER would be easy for me.
likely I I I I I I I I unlikely
I would find it easy to get CHART-MASTER to do what I want it to do.
My interaction with CHART-MASTER would be clear and understandable.
I would find CHART-MASTER to be flexible to interact with.
It would be easy for me to become skillful at using CHART-MASTER.
I would find CHART-MASTER easy to use.

Davis 1989

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Davis 1989

Uploaded by

Copyright:

Available Formats

Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information

Management Information Systems Research Center, University of Minnesota is

Information ACM Categories: H.1.2, K.6.1, K.6.2, K.6.3

MIS Quarterly/September 1989 319

320 MIS Quarterly/September 1989

MIS Quarterly/September 1989 321

322 MIS Quarterly/September 1989

MIS Quarterly/September 1989 323

Table 2. Initial Scale Items for Perceived Ease of Use

1. I often become confused when I use the electronic mail system.

324 MIS Quarterly/September 1989

MIS Quarterly/September 1989 325

Table 3. Pretest Results: Perceived Usefulness

Table 4. Pretest Results: Perceived Ease of Use

326 MIS Quarterly/September 1989

MIS Quarterly/September 1989 327

Perceived Usefulness Perceived Ease of Use

usefulness and ease of use items load on dis- how

328 MIS Quarterly/September 1989

Regression analyses were performed to assess

MIS Quarterly/September 1989 329

330 MIS Quarterly/September 1989

MIS Quarterly/September 1989 331

Table 9. Regression Analyses of the Effect of Perceived

non-significant for both Chart-Master (b=.08, through through usefulness.

332 MIS Quarterly/September 1989

MIS Quarterly/September 1989 333

334 MIS Quarterly/September 1989

MIS Quarterly/September 1989 335

336 MIS Quarterly/September 1989

MIS Quarterly/September 1989 337

338 MIS Quarterly/September 1989

MIS Quarterly/September 1989 339

Perceived Ease of Use

340 MIS Quarterly/September 1989

You might also like