Extending the Theory of Planned Behavior to Predict Electronic Commerce Adoption

Understanding and Predicting Electronic Commerce Adoption: An Extension of the Theory of
Planned Behavior
Author(s): Paul A. Pavlou and Mendel Fygenson
Reviewed work(s):
Source: MIS Quarterly, Vol. 30, No. 1 (Mar., 2006), pp. 115-143
Published by: Management Information Systems Research Center, University of Minnesota
Stable URL: http://www.jstor.org/stable/25148720 .
Accessed: 14/01/2013 17:03
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
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.
Management Information Systems Research Center, University of Minnesota is collaborating with JSTOR to
digitize, preserve and extend access to MIS Quarterly.
http://www.jstor.org
This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

All use subject to JSTOR Terms and Conditions
Pavlou & Fygenson/ExtendingtheTPB
Qjarteny
Understanding and Predicting Electronic

Commerce Adoption: An Extension of
the Theory of Planned Behavior1
By: Paul A. Pavlou elicit and test a comprehensive set of salient beliefs for each
Anderson Graduate School of behavior.
Management
University of California, Riverside

A longitudinal study with online consumers supports the proposed
e-commerce model, the
Riverside, CA 92521 adoption validating predictive power ofTPB
and theproposed conceptualization a higher-order factor
U.S.A. ofPBCas
formed by self-efficacy and controllability. Our findings stress the
paul.pavlou@ucr.edu
importance of trust and technology adoption variables (perceived
usefulness and ease of use) as salient beliefs for predicting e
Mendel Fygenson
commerce adoption, justifying the integration of trust and tech
Marshall School of Business
nology adoption variables within the TPB framework. In addition,
University of Southern California technological characteristics (downloaddelay, Website navigability,
Los Angeles, CA 90089 and consumer time and monetary
information protection), skills,
U.S.A. resources, and product characteristics and
(product diagnosticity
mfygenson@marshall.usc.edu product value) add to the explanatory and predictive power of our
model. Implications for Information Systems, e-commerce, TPB, and
the study of trust are discussed.
Abstract of planned behavioral

Keywords: Theory behavior, perceived
control, self-efficacy, controllability, technology adoption, tech
This paper extends Ajzen 's (1991) theory of planned behavior (TPB)
nology acceptance model, trust, electronic commerce, consumer
to explain and predict the process of e-commerce adoption by behavior
consumers. The process is captured through two online consumer
behaviors: (1) getting information and (2) purchasing a product
a Web vendor. First, we model the association
from simultaneously
between these two contingent online behaviors and their respective
intentions by appealing to consumer behavior theories and the Introduction^ ^ i
theory of implementation intentions, respectively. Second,
Business-to-consumer (B2C) e-commerce is the activity in which
following TPB, we derive for each behavior its intention, attitude,
consumers get information and purchase products using Internet
subjective norm, and perceived behavioral control (PBC). Third, we
technology (Olson and Olson 2000). The potentialbenefitsof e
commerce have been widely touted (e.g., Gefen et al. 2003).
However, for these information technology-enabled benefits to
materialize, consumers must first adopt online activities, such as

!Ritu Agarwal was the accepting senior editor for this paper. Elena
Karahanna was the associate editor. D. Harrison McKnight, Jonathan getting information and purchasing products from commercial
Palmer, and David Gefen served as reviewers. websites. B2C e-commerce adoption?the consumer's engagement
MIS Quarterly Vol. 30 No. 1, pp. 115-143/March 2006 115

theTPB
Pavlou & Fygenson/Extending
1. of B2C e-commerce is not viewed as a monolithic

in online relationships with Web
exchange vendors?goes beyond Adoption
the realm of traditional marketing, and itmust thus be understood andgetting
behavior,but is ratherproposed ofhoXh.purchasing
from the viewpoint that online consumers are simultaneously IT information. Since TPB has not been used to simultaneously
users (Koufaris2002). According toTaylor and Todd (1995b), IT predict related behaviors, by modeling these two online
usage encompasses not only use of hardware and software, but also behaviors, we theoretically extend TPB.
the services that surround the IT and the people and procedures that
support its use. B2C e-commerce thus presents a unique opportunity 2. PBC is a key determinant of both focal e-commerce behaviors.
to examine a user's interaction with a complex IT system. To the best of our knowledge, most e-commerce studies do not
account for PBC 2002), nor has a set of
(e.g., George
E-commerce adoption is an instance of IT acceptance and use within antecedents of PBC ever been advanced or
theoretically
a setting that combines technology adoption with marketing examined.
empirically
elements, and it thus requires distinct theorization within the infor
mation systems literature. However, despite an emerging interest 3. PBC is viewed as a two-dimensional construct formed by two
underlying dimensions {self-efficacyand controllability),

among IS researchers toward the B2C e-commerce phenomenon,
there is only a limitedand fragmentedunderstandingof online allowing a more detailed examination of external control
consumer behavior. The
purpose of this study is to theoretically beliefs.
propose and test a set of factors that integrate
empirically
with marketing and economic variables to 4. Trust is viewed as an antecedent of both attitude and PBC, and
technology adoption
enhance our understanding of online consumer behavior. within theproposed TPB model.
therebyintegrated
B2C e-commerce has some notable differences compared to 5. Most factors are shown to be IT-related
empirically (e.g.,
traditional consumer behavior. First, the spatial and temporal ease of use, information protection), or within the
usefulness,
between consumers and Web vendors increases fears of IS domain (e.g., trust,navigability),highlightingthekey role
separation
uncertainty(Ba and
selleropportunismdue toproduct and identity of IT in online consumer behavior.
Pavlou 2002). Second, information can be easily collected,
personal
processed, and exploited bymultiple parties not directly linked to The paper proceeds as follows: the next section discusses the two
the transaction. Third, consumers must actively engage in extensive e-commerce describes the TPB framework and the nature
behaviors,
IT use when interacting with a vendor's which has become
website, and role of PBC, and linksTPB perceptionswith intentionsand
the store itself (Koufaris 2002). Fourth, there are concerns about the
behaviors. The section proposes and describes the elicited
following
Internet infrastructure that Web vendors
reliability of the open externalbeliefs and justifieshow they linktoTPB. The next two
to interface with consumers (Rose et al. 1999). These The final
employ sections present the research methodology and results.
differences stress the uncertainty of the online environment and and
section discusses the study's findings, contribution,
the of consumer trust and the significance of
emphasize importance
implications.
IT adoption. More importantly, they reduce consumers' perception
of control, confidence, and effortlessness over online activities,
a barrier to e-commerce Therefore, compared to
creating adoption.
traditional consumer behavioral control (PBC),
behavior, perceived
as described in thetheoryofplanned behavior (TPB) (Ajzen 1991), Electronic Commerce Adoption
is likely to play a critical role in B2C e-commerce.
Description of Online
TPB is a well-researched model that has been shown to predict Consumer Behaviors
behavior across a variety of settings. As a general model, it is
it Electronic commerce is broadly described as the
designed to explain most human behaviors (Ajzen 1991). Hence, adoption
is reasonable to expect that a TPB-based model could effectively consumer's engagement in online exchange relationships with Web
online consumer behavior. We thus create an extended vendors. From a consumer behavior standpoint, getting product
explain
version of TPB to predict two prevalent online behaviors: getting
information and purchasing products are generally viewed (among
Web vendors. This other activities) as the two key online consumer behaviors (Gefen
information and purchasing productsfrom study
aims to predict these two behaviors the major and Straub 2000). While most e-commerce studies have
largely
by examining
constructsof TPB (attitudeand PBC) and theirmost important focused on product purchasing, online consumer behavior is not
antecedents. This results in a comprehensive, monolithic since consumers must first engage in getting product
yet parsimonious
model thatatteststo the influentialrole of PBC, while identifying beforepurchasing. Choudhury et al. (2001) argue that
information
are consistent with the TPB consumers do not make a single, inclusive decision, but they rather
and validating important factors that
structure. Moreover, the derived model a consider two distinct stages: getting product information and then
nomological explains
substantial portion of the variance in e-commerce adoption. In purchasing theproduct. Gefen and Straub (2000) also distinguish
this study provides and empirical between the two behaviors that getting information is an
summary, conceptual clarity by arguing
validation on the following issues: activity intrinsicto the IT since theWeb system itselfpresents the
116 2006
MIS QuarterlyVol. 30 No. 1/March

theTPB
product information. Product purchasing, on the other hand, is a The Theory of Planned Behavior
Web systemprimarilyprovides the
taskextrinsicto the IT since the
means to achieve the purchase. TPB 1) is an extension of the theory of reasoned action
(Figure
(TRA) (Ajzen and Fishbein 1980). TPB has been one of themost
Getting informationinvolves the transferof informationfromthe influentialtheoriesinexplainingand predictingbehavior,and ithas
Web vendor to the consumer the vendor's website. to predict a wide et al.
through browsing been shown range of behaviors (Sheppard
information has been referred to as browsing or window
Getting 1988).
shopping(Gefen2002). The value of online information searchhas
beenwidely acknowledged (Bellman et al. 1999) since it is critical to TRA, the proximal determinant of a behavior is a
According
for learning about product specifications and potential alternatives, behavioral intention, which, in turn, is determined by attitude
(A)
determining requirements, and gaining sufficient knowledge tomake norm a person's
and subjective (SN). Attitude captures overall
well-informed decisions (Choudhury et al. 2001). Product pur evaluation of performing the behavior; SN refers to the person's
chasing refers to the procurement of a product by providing
perception of the expectations of important others about the specific
monetary information in exchange for the focal good. In addition to
behavior. Finally, the antecedents of attitude and SN are a set of
monetary information, purchasing usually involves
underlyingattitudinal(b{) and normativebeliefs (n{), respectively.
product
consumer information (e.g., address information,
providing product Attitudinal beliefs are assessments about the likelihood of the
preferences).2 behavior's normative beliefs are assessments about
consequences;
othersmight thinkof thebehavior. Attitudeand SN
what important
These two behaviors, getting information and product purchasing, are described via an expectancy-value formula:
constitute the major part of long-held consumer behavior models.
et al. (1973) describe a five-stage buyer decision-making
Engel
process that includes problem recognition, information search, Aoclb.-e, (1)
evaluation of alternatives, purchase decision, and post-purchase
behavior. Information search corresponds to getting information and SN rrii (2)
oc^n;
purchase decision to product purchasing. Ives and Learmonth
(1984) propose thecustomerresource lifecycle (CRLF) with three Where: ej is the person's subjective evaluation of the
key stages: prepurchase, during purchase, and post-purchase. of the outcome, and
desirability
Getting information is a prepurchase activity, while product
purchasing corresponds to during purchase activities. Similarly, is the person's motivation to comply with
mj important
Kalakota and Whinston (1997) introduce the consumer mercantile others.
model (CMM) that consists of three phases: prepurchase interaction,
purchase, and post-purchase interactions. Prepurchase interaction are to obstacles,

Recognizing that most human behaviors subject
consists of product search, while comparison-shopping corresponds
Ajzen (1991) introducedTPB, which generalizesTRA by adding a
to getting information.Choudhury et al. (2001) describe four third perception: behavioral control A set of
perceived (PBC).
transaction vendor
stages:
requirements determination, selection, controlbeliefs (c;) and theirperceived power (p;) (to facilitateor
and after-sales service. information corresponds
purchase, Getting inhibittheperformanceof a behavior) determinePBC throughan
to requirements determination, and product purchasing to purchase.
expectancy-value formula:
In sum, we focus on two behaviors?getting information and
product purchasing?that largely determine e-commerce adoption.3

PBC oc?c,. p. (3)
Behavioral Intentions and

Actual Behavior
The purchasing process may be supplemented by automatic information
extraction through cookies and data mining tools. However, it is beyond the Behavioral intentions are motivational factors that capture how hard
scope of this study to account for this type of information sharing, which is people arewilling to trytoperforma behavior (Ajzen 1991). TPB
not related to consumer behavior.
suggeststhatbehavioral intentionis themost influential
predictorof
behavior; after all, a person does what she intends to do. In a meta
We recognize the existence of other e-commerce activities, such as
analysis of 87 studies, an average correlation of .53 was
fulfillment and repeat buying. Yet, fulfillment is a vendor's behavior reported
(Kalakota andWhinston 1997). Even ifpost-purchase experience influences between intentionsand behavior (Sheppard et al. 1988). Following
future behaviors, for predicting a specific behavior, the proposed TPB we a for our two focal
TPB, expect positive relationship
variables are supposed to take into account all previous experiences (Ajzen
behaviors?getting information and purchasing?and their respec
1991). Most important, consumer post-purchase behavior is contingent upon
tive intentions.
fulfillment,which cannot be predicted before purchase.

2006 117

Pavlou & Fygenson/Extending the TPB
kr Attitudinal v
Attitude
_^
/r Beliefs \ _| \.
_ Subjective Intention Behavior

I Normative __j^ _^
In Beliefs I I_!Î___I Jh-J^T-7^-1
Control Perceived
Vl - I I Behavioral ^~~ ^~~~~~~*
\
f Beliefs 1-' j
Control
I
Figure 1. The Theory of Planned Behavior (Adapted from Ajzen 1991)
Connecting Getting Information

and Orbell 1999). We view purchasing a specific product from a
vendor as while
and Product Purchasing particular Web the goal behavior, getting
information about the product from theWeb vendor is viewed as a
means to achieve the goal behavior (implementation intention).
In the social psychology literature, many researchers model related
Therefore, a goal intention to purchase a product from aWeb vendor
are always
behaviors using the TPB framework, but these behaviors
activates an intention to get information about that product from the
modeled independently, without any attempt to capture the extent of
vendor's website.4 For a student that intends to buy a
their relationship et al. 2000). This raises example,
(e.g., Povey important
textbook from Amazon ismost likely to visit Amazon to get price
questions: Can two related behaviors be modeled simultaneously
information about the textbook. In terms of the temporal order,
within theTPB framework? If so, how? Through which TPB
consumers first form the intention to purchase a product to fulfill a
constructs should two related behaviors be connected? In principle,
particular need, and they then form the implementation intentions to
TPB at one level of specificity. How to relate one
only applies
facilitate fulfilling the need. Therefore, the product purchasing
behavior to another remains a crucial open question (personal
(goal) intention precedes and drives the getting product information
communication with I. Ajzen 2003).
(implementation) intentions. Salisbury et al. (2001) show that
intentions to purchase relate to intention to get information. The
To the relationship between the two focal behaviors, we
explain
draw upon three different of consumer behavior. preceding arguments suggest
aspects First,
product purchasing is contingent upon getting information. This

H2: Intentions to purchase a product from a Web vendor
notion is captured in the buyer's decision making model (Engel et
influence intentions to get information about
al. 1973), theCRLF (Ives and Learmonth 1984), and theCMM positively
a the product from the vendor's website.
(Kalakota and Whinston 1997), which assume sequential
relationship between getting information and purchasing. Second,
getting information facilitates purchases. For example, Kim and

Benbasat that consumers in getting information
(2003) argue engage Attitude
to reduce the uncertainty of product purchasing. Third, getting
information influences purchasing. This is captured in the theory of
Attitude has long been shown to influence behavioral intentions
mere exposure 1968), which holds that the frequency of
(Ajzen and Fishbein 1980). This relationshiphas received sub
(Zajonc
exposure facilitates a behavior. Empirical studies (Choudhury et al.
stantial empirical support. With regard to the focal behaviors, atti
2001; Gefen 2002) report a positive correlation between getting tude toward getting information and product purchasing is defined
information and purchasing. Therefore, we suggest as the consumer's evaluation of the desirability of using a website
to get information and purchase products from a Web vendor,
HI: Getting product information from a vendor's website
respectively. Using a deductive logic, favorable attitude is likely to
positively influences purchasing a product from that
Web vendor.
To link behavioral intentions between

getting information and Gollwitzer's (1999) theory suggests that a goal behavior can trigger several
product purchasing,
we to Gollwitzer's
refer (1999) theory of implementation intentions. Intention to purchase a product from a specific
Web vendor triggers intentions to get product information, not only from the
intentions, which are self-regulatory strategies that
implementation
to the theory, a specific vendor, but also from other sources. Both implementation intentions
aim to drive a goal-oriented behavior. According are potential consequences of the goal behavioral intention, but the intention
goal-driven behavior automatically activates a set of goal-enabling to get information about a specific product from a specific Web vendor is
(implementation) intentions that help realize the behavior (Sheeran more likely to occur, and it is thus examined.
118 MIS Quarterly Vol. 30 No. 1/March 2006

theTPB
encourage consumers to get information and purchase products from behavior toward product purchasing from theWeb
a vendor. vendor.
Underlying Dimensions of PBC

Subjective Norm
Since the early days of TPB, there has been someambiguity
SN suggests is instigated by one's desire to act as
that behavior
surrounding the nature of PBC. Recently, questions regarding its
important referent others act or think one should act. Applied to the
nature and measurement a
have been attracting lot of attention (e.g.,
two focal behaviors, SN reflects consumer perceptions of whether
these two behaviors are accepted, and implemented
Ajzen 2002b; Trafimow et al. 2002). In particular, empirical
encouraged, by
the consumer's circle of influence. The literature suggests a positive
findingshave cast doubt on Ajzen's (1991) original assertion that
PBC is a unitary construct, suggesting instead that PBC has two
relationshipbetween SN and intendedbehavior,and empiricalwork distinctdimensions: self-efficacy(SE) and controllability.5
While
has shown that SN influences behavioral intentions toward system
the conceptualization of SE and controllability is still controversial,
use (Karahanna et al. 1999). A positive relationship between SN
there is an emerging consensus that the two are the underlying
and intentions to get information and purchase products from aWeb
dimensions of PBC. We offerthefollowingdefinitions:
vendor is thus expected.
Self-Efficacy: Following Bandura (1986), we define SE as

individual judgments of a person's capabilities to perform a
Perceived Behavioral Control behavior. Applied to e-commerce, SE describes consumers'
judgments of their own capabilities to get product information
PBC is a topic thathas been debated in the social psychology and purchase products online.
literature (for a review, see Trafimow et al. 2002). This paper sheds
lighton thenatureand role ofPBC by (1) clarifyingitsrole inTPB, Controllability: We follow Ajzen (2002b) to define
(2) describing its underlying dimensions, and (3) proposing a controllability as individual judgments about the availability of
model that integrates its underlying dimensions and resources and opportunities toperform the behavior.
parsimonious Applied
their antecedents into a coherent model. to e-commerce, controllability describes consumers' percep
tions of whether getting information and purchasing products
online is completely up to them because of the availability of
resources and opportunities.
The Role of PBC inTPB
PBC is defined as a person's perception of how easy or difficult it

would be to carryout a behavior (Ajzen 1991). To differentiate The Nature of Perceived Behavioral Control
PBC fromattitude,Ajzen (2002b) emphasized thatPBC denotes a
of control over the performance of a behavior and Despite empirical evidence that SE and controllability can be
subjective degree
not the perceived likelihood thatperformingthe behavior will manipulated differentlyand can be reliably distinguishedacross
produce a given outcome. suggested
Ajzen that PBC "should be behaviors (e.g., Cheng and Chan 2000), Ajzen (2002b, p. 696)
read as perceived control over
the performance of a behavior"
maintains that"the fact that it is possible to reliably distinguish
(2002b, p. 668). Therefore, PBC is the consumer's perceived ease between twodifferenttypesofPBC?SE and controllability?does
or difficulty of getting product information from a vendor's website not invalidate the unitary nature of the [PBC] construct." To bridge
and purchasing a product from aWeb this inconsistency, he proposes a two-level hierarchical model to
vendor, respectively.
describe PBC as an "overarching, superordinate construct" (p. 697).
In general, PBC plays a dual role in TPB. First, along with attitude
and SN, it is a co-determinant of intention. Second, Hierarchical or models are used to explain the
together with higher-order
intention, it is a co-determinant of behavior. for the role of interrelations among lower-order factors that constitute an inte
Support
PBC on intentionand behavior isprovidedbyMathieson (1991) and grative latent construct. Higher-order models provide a more
Taylor and Todd (1995b). We thussuggest coherent description of multiple facets of a complex phenomenon
thatcould be described by a unitaryfactor(Law et al. 1998). The
over a Web relationships between lower and higher order constructs can be
H3a: PBC getting information from vendor
oxformative. While reflective structures assume that the
positively influences (1) intention and (2) actual reflective
behavior toward getting product information from

that Web vendor.
While the SE and controllability differ in their predictive validity (e.g.,
Conner and Armitage 1998), there is no evidence to support the common
H3b: PBC over product from a Web vendor
purchasing view that SE reflects internal factors whereas controllability reflects beliefs
positively influences (1) intention and (2) actual about external factors (Ajzen 2002b).

2006 119

Jfc
Attitudinal
_J Attjtude I
fW\ Beliefs_| H_l^^^-^-^^^
Normative -^ Subjective -^^^fc Intentions Behavior

/[. Beliefs -j^
\f. Norm '_'
| ^_' --'"''"
-
I I r? Controllability j j_j-1 v.
|--r- ^--~~" I
\ VW Beliefs ! r Controllability I Perceived ^ j-~~~
\ j Behavioral f"*" j _
V[_ _ j4j '
Self-EfficacyU-?| Self-Efficacy K*~ - - -?'- - I 1 1 First-Order
Construct
\jR Beliefs ! I-" .
^ Second-Order Construct
y\_I i_I
CONTROL BELIEFS PERCEIVED BEHAVIORAL CONTROL
Figure 2. The Proposed Extension of the Theory of Planned Behavior
latent second order construct causes the first order factors, formative correspond to specific open-ended questions (Table 1). Normative
structures assume that the second order construct is caused by the beliefs were not elicited since prior studies showed that SN has a
first order factors (for a review, see Edwards 2001). weak role inonline behaviors (George 2002). We solicited thekey
drivers for each behavior from a convenience sample of 56
Figure 2 depicts our proposed extension of TPB with PBC viewed participants, which included faculty, staff, and students of a major
as a second-order factor formed by the first-order dimensions of SE university in the United States. Their responses are sorted based on
and controllability. the frequency mentioned (Tables 2 and 3). We then chose the
beliefs that exceeded a 20 percent frequency cutoff, as prescribed by
The rationale for a formative model is based on the notion that SE Ajzen and Fishbein (1980, p. 68) (presented inbold in the tables).
and controllability are dynamic concepts (Bandura 1986), not stable
traits. As dynamic concepts, they are likely to change over time and The resulting set of beliefs span a wide range of characteristics,
be manipulated differently by other factors (Trafimow et al. 2002). which we grouped into six categories for better exposition: (1) trust
Hence, PBC cannot equally cause SE and controllability, thus inWeb vendor, (2) technology acceptance, (3) consumer resources,
rendering a reflective model unlikely. Moreover, since a change in (4) technological characteristics, (5) product characteristics, and
one of the lower-order factors does not necessarily imply an equal (6) consumer skills. These categories were derived based on
change in the other, a formative model is deemed more likely. literature grounding and practical empiricism. For getting infor
mation : (1) the attitudinal beliefs are trust, perceived usefulness and
In our endeavor to comprehensively predict the two key e-commerce ease of use; (2) the controllability beliefs are trust, ease of use, time
behaviors, the proposed TPB extension allows for a thorough resources, download delay, and website navigability; and (3) the SE
prediction of PBC through its underlying dimensions and their beliefs are ease of use and skills. For purchasing: (1) the attitudinal
antecedents, while maintaining a parsimonious view of beliefs are trust, usefulness, ease
of purchasing, and product value;
respective
PBC. The following section elicits the antecedents of PBC through (2) the controllability are trust, ease of purchasing, monetary
beliefs
its two underlying dimensions, in addition to eliciting the antecedent resources, product diagnosticity, and information protection; and
beliefs of attitude. (3) the SE beliefs are ease of use and skills. Figure 3 depicts our
proposed model.
TPB can aggregate beliefs to create measures of attitude, SN, and
Eliciting External Beliefs HHBHI^H PBC (Ajzen and Fishbein 1980). This aggregation has been
criticized for not identifying specific factors that might predict a
TPB includes three of external beliefs: attitudinal, behavior (e.g., Taylor and Todd 1995a) and for the biases itmay
categories
and control. These beliefs are scenario and a create (e.g., Karahanna et al. 1999). The idea that TPB beliefs can
normative, specific
cannot
be generalized. for each new behavior, one be decomposed intomultidimensional constructs has been credited
priori Hence,
must that are toTaylor and Todd (1995b), who introducedthedecomposed TPB
identify five to nine salient beliefs for each behavior
context and population specific (Ajzen and Fishbein 1980). (DTPB). While we stayfaithfultoTPB, we decompose thederived
beliefs following DTPB to provide a better understanding of each
We conducted a belief elicitation study using an

open-ended
behavior. In doing so, we aim not only to assure high explanatory
The aim was and predictive validity, but also to select managerially amenable
questionnaire, following Ajzen's (2002a) procedure.
to freely elicit themost salient attitudinal and control beliefs, which factors. We also use another variation of TPB to permit cross-over
120 MIS QuarterlyVol. 30 No. 1/March2006

Table 1. Questionnaire for Eliciting External Salient Beliefs
1. Getting informationabout this particular product from thisvendor's website inthe next 30 days:
Attitudinal 1a What do you beNeve are the advantages of doing this?
Beliefs
^k what do you believe are the disadvantages?
-
(Getting
2' Anvtnin9 e,se y?u associate with your getting informationabout this product fromthisvendor's
Information)
website?
3. What factors or circumstances would enable you toget informationabout this product fromthis
Control vendor's website?
Beliefs 4. What factors or circumstances would make itdifficultforyou toger informationabout this product
(Getting from this vendor's website?
Information) 5 Are tnere any 0tner jssues (barriers or facilitatingconditions) thatcome tomind when you think
about getting informationabout this product from thisvendor's website?
6. Purchasing the particular product from thisWeb vendor inthe next 30 days:
Attitudinal 6a_ What do you believe are the advantages of doing this?
Beliefs 6b. What do you believe are the disadvantages?
7. Anything else you associate with your purchasing this product from thisWeb vendor?
8. What factors or circumstances would enable you topurchase this product from thisWeb vendor?
Control 9. What factors or circumstances would make itdifficultforyou topurchase this product from thisWeb
Beliefs vendor?
(Purchasing) ^q Are tnere any 0tner issues (barriers or facilitatingconditions) thatcome tomind when you think
this product from thisWeb
_about your purchasing vendor?_
Table 2. Frequency of Elicited Beliefs (Getting Information)

Attitudinal Beliefs Frequency (%) Control Beliefs Frequency (%)
Trust - Getting Information 37 (66%) Getting Information Skills 31(55%)
Perceived Ease of Getting Info 33 (59%) Perceived Ease of Getting Info 30 (54%)
Perceived Usefulness of Getting 25 (45%) Trust - Getting Information 24 (43%)
Info
Download Delay 14 (25%) Download Delay 21 (38%)
Perceived Risk ofGetting Information 6 (11%) Time Resources
18(32%)
Perceived Enjoyment 5 (8%) Website Navigability 12 (21%)
Product Variety 5 (8%) Website Features (e.g., search engine, 7 (13%)
_FAQ)_
Instant 2 3 (5%)
Gratification_ (4%)_Website Personalization_

2006 121

Table 3. Frequency of Elicited Beliefs (Purchasing)

Attitudinal Beliefs Frequency (%) Control Beliefs Frequency (%)
Perceived Usefulness of 33 (59%) Monetary Resources 41 (73%)
Purchasing
Perceived Ease of Purchasing 32 (57%) Product Diagnosticity 33 (59%)
| j
i Trust - Purchasing 17 (30%) Perceived Ease of Purchasing 28 (57%)
|
ProductValue 15(27%) ProductValue
25(45%)
| | -
|
P" Monetary Resources 14(25%) [
Trust Purchasing 22(39%)
I Information Protection
_ProductDiagnosticity 13(23%) 18(32%)
Product
Quality 8(14%) Skills12(21%)
Purchasing
[
Perceived Risk of Purchasing 7(13%) Delayed Gratification3(5%)
Product Variety 2 (4%) Quick Pay Availability (e.g., one- 3 (5%)
click pay)
~
Trust -Getting Information Attitudetoward
~~Z^+ ?"
*""""~^^ Getting Info
\___
PU ofGetting Info \T^ >v
^^^/^-'
- Intentionto
PEOU ofGetting Info
yS Subjective -^ Getting Info
<^""^
Vv \ \^ on Getting Get lnfo m Behavior
I_I
^
Download Delay ' \. V. /

\
^^ .i
.-. Controllability / j ^^
^^~V^^^ p4
Time Resources (GettingInfo) ^. /
j_-A?2fc Jt^
I-1 ^^ PBC M' j

Website Navigability ^*^ ^^^V \ _ .
) (Getting Info)
'' j
!^! Self-Efficacy j
i Information
Getting Skills I-U lnfo)^^
(Getting
EXTERNAL BELIEFS (GETTING INFO) PERCEIVED BEHAVIORAL CONTROL

(GETTING INFORMATION)
Trust- Product Purchasing

L______________^^
"
?,, I r _ '. Attitudetoward
PUofPurchasing |\ _fc.
j ^y-??^ Purchasing k
PEOU of Purchasing
' %T\ ^^^^ ^\
\\ -V^^ i-;-1 ~I_ " _ ~ ~~
I Z\Jk><\
77T, \ Subjective NormTl . . ^^ I I- I
ProductValue -* lntentlont0 Purchasing
!\ \ \ on Purchasinq
y
'-1 \ \ \ Purchase * Behavior
^
MonetaryResources /
ivi-1
' '
^---^\ j/
/ y^
Controllability
^^\^-JjJ
Product 9' ^ ^s./
Diagnosticity-V?^ y/
I I^^-^^"^ ; i PRP- ' )f
k^ V \
Information
Protection Past
S]
j ._. Experience
\ I_I :\ j y'\/*\ (Purchasing)
Habit ; j J?
; ,
^ - Self-Efficacy 1/ | Web Vendor Reputation
Purchasing Skills -\>l (Purchasing) \ ProductPrice
I-?-1 j I_I Consumer Demographics
EXTERNAL BELIEFS (PURCHASING) PERCEIVED BEHAVIORAL CONTROL '-'
CONTROL VARIABLES
(PURCHASING)
Figure 3. The Proposed Research Model

effectsbetween beliefs and perceptions (Taylor and Todd 1995a). Web vendor eases fears that purposely false information may expose
both a consumer to adverse outcomes In sum, trust for
For example, trusting beliefs may simultaneously impact (Gefen 2002).
describes a consumer's belief that the Web
attitude and PBC for each behavior. getting information
vendor will provide valid, accurate, and timely information.
Trusting Beliefs
Trusting Belief: Product Purchasing
Trust haslong been a central defining feature of economic and
of authority, and Trust is important for product purchasing since online consumers are
social interactions where uncertainty, delegation
fears of opportunism are present Trust is the vulnerable in several ways (e.g., not receiving the right product,
(Luhmann 1979).
belief that the trustee will act cooperatively to fulfill the trustor's becoming victims of fraud). A trusted Web vendor must have
without its vulnerabilities. A detailed competence, integrity, and benevolence. Competence refers to "the
expectations exploiting
discussion on the nature and role of trust in e-commerce can be expectation of technically competent role performance" (Barber
found inGefen et al. (2003),McKnight and Chervany (2002), and 1983, p. 14). Integrity provides assurance that the vendor will keep
Pavlou (2003). Benevolence ensures that the vendor will act fairly and
promises.
stand behind its product, even if new conditions arise. In sum, for
In general, trust is viewed as a three-dimensional product purchasing, trust describes the belief that the vendor will
construct,
et al. deliver, fulfill, and stand behind its product.
composed of competence, integrity, and benevolence (Gefen properly
2003). Competence is thebelief in the trustee'sability toperform

as expected by the trustor. Integrity is the belief that the trustee will
be honest and keep its promises. Benevolence is the belief that the
Trusting Beliefs and Attitude
trustee will not act opportunistically, even given the chance. In sum,
trust gives the trustor the confidence that the trustee will behave Trust is proposed as an attitudinal belief for both getting information
capably (ability),ethically (integrity),and fairly(benevolence).6 and purchasing. The relationship between trust and attitude draws
from the notion of perceived consequences (Triandis 1979). Trust
To be placed in a TPB-based model, trust must be defined with enables favorable that no harmful outcomes will occur
expectations
respect to a behavior through a well-specified target, action, context, if a trustor undertakes a behavior (Barber 1983). Trust also refers to
and timeframe(Ajzen 2002a). The targetof trustis theWeb ven optimistic expectations that the trustee will protect the trustor's
dor, the action is getting informationor purchasing, and the context interests (Hosmer In sum, trust creates favorable
1995). perceptions
is the online environment. In terms of time frame, the impact of about the outcomes of the vendor's actions, thus creating
positive
trust is observed for a specific window during which the consumers attitudes. In terms of getting information, trust creates
positive
are making their decisions. This view is consistent with the trust that the vendor will post credible information. For
expectations
literature where trust is considered with respect to a specific trustor trust engenders confident expectations that the
product purchasing,
(Mayer et al. 1995), context (Lewicki and Bunker 1995), and time Web vendor will fulfill its promises. a similar
Using logic,
window (Tan and Thoen 2001). Jarvenpaaetal. (2000),McKnight andChervany (2002), and Pavlou
(2003) show that trust has an impact on intentions by creating
The practical utilityof placing trustin theproposed TPB model positive attitudes. Therefore,
stems from the fact thatWeb vendors have a considerable influence
on trust through their reputation and size (Jarvenpaa et al. 2000), and H4a: beliefs in a Web vendor
Trusting regarding getting
institutional factors (Pavlou and Gefen 2004), among others. information influence attitude toward
positively
getting product information from that Web vendor.
Trusting Belief: Getting Information H4b: Trusting beliefs in a Web vendor regarding product
purchasing positively influence attitude toward
Trust is important for getting information since consumers assess product purchasing from theWeb vendor.
whether the information on a website is valid, credible, and accurate
(Choudhury et al. 2001). Therefore, competence and integrity are

themost relevant dimensions for getting information as they reflect Trusting Beliefs and Perceived Behavioral Control
theWeb vendor's ability to provide credible information. A trusted
Trust is also placed in the nomological structure of the TPB as a
control belief. The trust literature assumes that the trustor lacks
Trust has also been viewed as a four-dimensional construct, comprising of control over the trustee's behavior, but trust builds the trustor's
ability, integrity,benevolence, and predictability (McKnight and Chervany confidence to depend on the trustee (Fukuyama 1995). The
2002). However, the literature on buyer-seller relationships has focused on
relationship between trust and PBC draws from Luhmann's (1979)
credibility (competence and integrity) and benevolence (Ba and Pavlou 2002;
and Cannon 1997). Therefore, predictability or consistency is omitted. notion that trust reduces social uncertainty, which refers to all
Doney
MIS Quarterly Vol. 30 No. 1/March 2006 123

unforeseen contingencies. In doing so, trust decreases efforts to Perceived Ease of Use
copiously account for all potential contingencies (Gefen 2002).
Following this logic,Zand (1972) concludes thatby reducingsocial PEOU is theextenttowhich a personbelieves thatusing thesystem
uncertainty, trust results in a greater controllability over the will be effortless (Davis 1989). Applied to online consumer
behavior. Therefore, trust facilitates not by behavior, perceived ease of getting information is defined as the
trusting behaviors,
theWeb vendor's actions (such as in agency theory), but extent towhich a consumer believes that getting product information
controlling
barriers to engaging in a behavior. from a website would be free of effort. Similarly, perceived ease of
by overcoming psychological
Trust thus acts as an uncertainty resource that enables the purchasing is defined as the extent towhich a consumer believes that
absorption
trustor to better cope with social uncertainty. In terms of getting purchasing products from a Web vendor would be free of effort.
information, trust rules out due to the Similar to PU, the role of PEOU on intentionsismediated by
negative contingencies
information that the vendor on its website. In terms of attitude(Davis 1989; Taylor and Todd 1995b). Hence, we propose
provides
trust reduces the uncertainty of product
thefollowinghypotheses:
product purchasing, delivery
and fulfillment.
We therefore
propose thefollowinghypotheses:
H7a: Perceived ease of getting information positively
influencesattitude toward gettingproduct information
H5a: Trusting beliefs in a Web vendor regarding getting
from a Web vendor.
information positively influence controllability over
getting product information from that Web vendor.

H7b: Perceived ease of product purchasing positively
influences attitude toward from a
Trusting beliefs in aWeb vendor regarding product
H5b: product purchasing
Web vendor.
purchasing positively influence controllability over
product purchasing from that Web vendor.

In addition to theattitudinalrole of PEOU, the instrumental
aspect
of PEOU (Davis 1989) isviewed as a controlbelief thatfacilitates
a behavior with lower personal effort (Lepper For example,
1985).
Davis argued thatSE isone of themeans bywhich PEOU influences
TAM Beliefs behavior. to e-commerce, a website from which it is
-
Applied
perceived as being easy to get information and make a purchase is
Following the TRA, TAM asserts that the intention to use a system
likely to increase the consumer's ability and confidence in getting
isdeterminedby twogeneralizedbeliefs: perceivedusefulness (PU) information and purchasing, respectively.
and perceived ease of use (PEOU) (Davis 1989). The twoTAM
variables have been used to predict Internet purchasing behavior
Similarly, an easy to use website removes the cognitive impediments
(e.g.,Gefen et al. 2003; Koufaris 2002; Pavlou 2003). of using the website, making getting information and purchasing
more accessible to the consumer. It causes the perception of these
online behaviors as being under the consumer's full control, thus
Perceived Usefulness and purchasingcompletelyup

making gettingproduct information
to consumer. Thus, the following hypotheses are offered:
PU is the extentto which one believes that using a system will

enhance her performance H8a: Perceived ease of getting information positively
(Davis 1989). PU of getting information
is defined as the extent towhich a consumer believes that a website influences (1) self-efficacyand (2) controllabilityover
would enhance her effectiveness getting product information from that Web vendor.
in getting product information.
Perceived purchasing usefulness is defined as the extent towhich a
H8b: Perceived ease of product
consumer believes that a vendor would enhance her purchasing positively
specific
in purchasing
influence (1) self-efficacyand (2) controllability over
effectiveness products. PU has been shown to
product from that Web vendor.
influencebehavioral intentionthroughattitude(Davis 1989;Taylor purchasing
and Todd 1995b). Therefore, the following hypotheses are

proposed:
Consumer Resources
H6a: Perceived usefulness of getting informationpositively
influencesattitude toward gettingproduct information Time Resources
from a Web vendor.
Leisure time has been considered a critical resource for getting

H6b: Perceived usefulness of product purchasing positively information (Bellman et al. 1999). Having the time needed to
influences attitude toward product purchasing from a
browse for product information is a prerequisite for getting
Web vendor. since time is a key resource
information for time-consuming tasks.
124 MIS QuarterlyVol. 30 No. 1/March

2006

theTPB
We thus hypothesize that time resource is a facilitating condition easily accessible to consumers, navigability makes getting informa
that increases the controllability over a behavior. tion be completely under the consumer's control. We thus propose
H9a: Time resources positively influencecontrollabilityover Hll: Website navigability positively influences controlla
getting product information from a Web vendor. bility over getting product informationfrom a Web
vendor.
Monetary Resources
Information Protection
Purchasing a product necessitates an outlay of monetary resources.
Having the required monetary resources is a prerequisite for Concerns about information security and privacy have made
a product. the financial to consumers
purchasing By overcoming impediments skeptical about online transactions (George 2002), and
purchasing, consumers increase their controllability over purchasing. they have been termed as key e-commerce obstacles (Hoffman et al.
1999; Rose et al. 1999). Information security refers to the

H9b: Monetary resources positively influencecontrollability consumers' belief about theWeb vendor's ability to fulfill security
over product from a Web vendor. and non-repudiation)
purchasing requirements (e.g., authentication, encryption,
(Cheung and Lee 2001). Information privacy refers to the
consumers' belief about theWeb vendor's ability to protect their
personal information from unauthorized use or disclosure (Cassell
Technological Characteristics and Bickmore 2000). Informationprotection is defined as the
consumer's belief about
theWeb vendor's ability to safeguard her
Download Delay information
from and
personal security privacy breaches.7 When
consumers feel comfortable with theway aWeb vendor will protect
Download delay is defined as the amount of time it takes for a their personal information, they overcome any psychological
website to display a requested page from a Web server (Rose et al.
barriers to purchasing from that vendor. Thus,
1999). Download delay relates to a website's response time, a factor
associated with lower intentions to use a system (Ives et al. 1983). H12: Information protection positively influences
Download delay is also negatively related to the time needed to over product
controllability purchasing from a Web
perform a task, which has been shown to negatively impact inten vendor.
tions to use a system (Mawhinney and Lederer 1990). Download
delay is thusexpected tonegatively impactattitudetowardgetting

information since having to wait too long for information creates
negative expectations about the behavior. Product Characteristics
Rose et al. (1999) identifieddownload delay as a key e-commerce Product Diagnosticity

barrier. Since download delay acts as an impediment to receiving
information quickly, it reduces the availability of time resources for Product is the extent towhich a consumer believes that
diagnosticity
consumers, thus making itmore difficult for them to get product a website is helpful in termsof fullyevaluating a product (Kempf
information. The and Smith Product is driven and
preceding arguments suggest 1998). diagnosticity by virtual
functional control (Jiang and Benbasat 2004). Virtual control refers
HlOa: Download influences attitude toward to allowing a consumer tomanipulate a product
delay negatively image to see it from
information from a Web vendor. control allows a consumer
getting product multiple angles and distances. Functional
to try different product functions. Since online consumers must
rely
HI Ob: Download delay negatively influences controllability on limited product representations (as opposed to traditional
over from a Web vendor. a real feel for the
product purchasing commerce), by providing product and enabling
adequate product evaluation,
product diagnosticity overcomes the
barrier created by the lack of physical inspection of products and
Website Navigability
Navigability as the natural sequencing

is defined of web pages, well
organized layout, and consistency of navigation protocols (Palmer

2002). A useful navigational structure facilitates traffic and sales on
aWeb siteby increasinginformation
availability (Lohse and Spiller
While information security and privacy can be viewed as distinct constructs,
1998). Navigability enables consumers to find the right set of
we propose a unitary view of information protection. The
unidimensionality
products and compare among alternatives. By making information of information protection was validated during the pilot studies.

2006 125

causes product purchasing to be under the consumer's full control.8 H15b: Purchasing skillspositively influenceself-efficacyover
Accordingly, we propose product from a Web vendor.
purchasing
H13: Perceived diagnosticity positively influences con

trollability over product purchasing from a Web
vendor.
Control Variables
The following variables are controlled for in this study:
Product Value
Past Experience: Studies have shown that past behavior
influences future behavior (Conner and Armitage 1988), and
Product value refers to a product that offers an attractive com
online experience is a key factor in online behavior (Hoffman
bination of quality and price. Price discounts are examples where
et al. 1999). Hence, this study controls for the role of past
the consumer can save money by getting a product at a lower price,
experience on both intentions and behaviors.
and they have been shown to influence purchase intentions (Alford
and Biswas 2002). Product value favorably predisposes consumers
Habit: Habit represents a variable thatmeasures the frequency
by allowing them to expect a high quality product at a low cost.
of repeated performance of behavior, and ithas been shown to
This suggests influencebehavioral intentions(Limayem andHirt 2003). In
e-commerce, Liang and Huang (1998) found that consumers'
H14: Product value positively influences attitude toward had a
prior experience moderating effect in predicting their
from a Web vendor.
product purchasing acceptance of Internetshopping (including the twobehaviors
we consider). Therefore, the role of habit (both for getting
information and purchasing) is controlled for its impact on
information and purchasing,
Consumer Skills getting respectively.
Web Vendor Reputation: The reputationof aWeb vendor

prerequisiteof engaging in a behavior is tohave the
An important
to be an antecedent
has been shown of transaction behavior
necessary personal skills and knowledge to undertake the behavior
(Jarvenpaaet al. 2000), and it is thus controlled for in this
(Koufaris2002). Following Bandura (1986), SE isnot equivalent to
as to whether study.
personal skills; SE deals with subjective judgments
one has the personal skills needed to accomplish a behavior (p.
Product Price: Since both focal behaviors are based on a
391). In contrast, "consumer skills" specifically describes the
specific product, product selection may differ across users and
knowledge and expertise a consumer has to undertake a behavior,
lead to different degrees of uncertainty due to price (Ba and
and it is thus a potential predictor of whether a certain behavior can
Pavlou 2002). To account for product characteristics, we
be accomplished.
control for product monetary price.
Applied to e-commerce, getting information skills captures a

Demographics: Finally, we also control for age, gender,
consumer's knowledgeability in getting product information from a
income, education, and Internet experience.
vendor's website and making product evaluations. Having such
skills is likely to increase consumers' judgments of how well they
can get information from a vendor's website, thus increasing their
SE for getting information. skills refer to the
Similarly, purchasing
Research _ ___________
consumer's about purchasing online and Methodology
knowledgeability products
sound purchasing which are likely to increase
making decisions,
Measurement Development
consumers' judgments of their efficacy to purchase products online,
leading tohigherSE. We thuspropose thefollowing: the TPB framework, each behavior must be defined
Following
within a well-specified target, action, context, and time frame
HI5a: information skills influence self
Getting positively
(TACT) (Ajzen 2002a). Throughoutthestudy,thetargetis the
Web
efficacy over getting information from aWeb vendor.
vendor, the action is either getting information or purchasing a
specific product, the context is the online environment, and the time
frame is a specific window of time, set at 30 days after the
behavioral intentions were assessed.
Product diagnosticity is not hypothesized to influence getting information

All measurement items (Appendix A) were drawn from the litera
since consumers can still get information about a product, but theymay not
ture, and they were then adapted using standard psychometric scale
purchase itonline until they have fully evaluated the product in a traditional
development (Boudreau et al. 2001) and a refinement
setting. procedures

2006

theTPB
procedurebased on thepilot studies. All scales followedAjzen's a $250 draw and a report that summarized the study results. The
(2002a) recommendationsfordesigning a TPB survey. invitees were assured that the results would be reported in aggregate
to assure their anonymity.
A indicator (criterion variable) was used to assess PBC
single
(Taylorand Todd 1995b). The SE measures are based on Compeau Similar to the pilot studies, the respondents were asked to choose a
and Higgins (1995). The controllabilitymeasures are based on specific product about which they were seriously considering getting
Taylor and Todd (1995b). Attitude and SN were adapted from information and purchasing online within the next 30 days. Having
Karahanna etal. (1999). selected a product, they were then asked to select and report a
specificWeb vendor thattheyhad recentlyvisited thatoffers this

In accordance with Ajzen and Fishbein's (1980) expectancy- value product. They were then asked to respond to the survey questions
formulation, belief-based measures are obtained by multiplying based on their selection. Thirty days after completing the first
belief strengthand power (equations 1 through3). Attitudinal survey, the respondents were contacted again. Following Blair and
beliefs aremeasured as theproductofbehavioral belief strength(b) Burton (1987), theywere asked to indicate if theyhad acted on
and outcome evaluation (e). Control beliefs are measured as the "getting information" and "purchasing" their selected product from
productof controlbelief strength(c) and controlbelief power (p). theWeb vendor of their choice.
Trustwas based onMcKnight andChervany (2002). Trust (getting

information) captures the vendor's honesty and competence in terms
of posting credible information.Trust (purchasing) captures the Results ^^^^^ i
Web vendor's integrity, and benevolence
competence, in fulfilling
product orders. PU and PEOU were adapted from Gefen et al. We used least square to analyze our data. PLS
partial (PLS)
(2003). Time and monetary resources were based on Bellman et al. a for estimation
employs component-based approach purposes (e.g.,
(1999), download delay on Rose et al. (1999), and website Lohmoller 1989) and can handle formative factors, unlike LISREL.
navigability on Palmer (2002). Information protection was based on PLS on measurement
places minimal restrictions scales, sample size,
the scales of perceived privacy and securitydeveloped by Cheung and residualdistributions(Chin et al. 2003). PLS was thuschosen
andLee (2001) and Salisburyetal. (2001). Productvaluewas based to accommodate the presence of formative factors and the large
on Chen andDubinsky (2003), productdiagnosticityon Jiangand number of constructs.
Benbasat (2004), and consumer skills on Koufaris (2002). Habit
was from Limayem and Hirt (2003), and Web vendor Based on Chow's (1960) test statistic9andWilk's
adapted lambda,10the
reputation from Jarvenpaa et al. (2000). Past behavior used standard results from the student and consumer not signi were
samples
items for past activities. Product price was ex post captured as a
ficantly different. To double check, we performed a separate data
binary (high/low)variable. analysis on each sample and got virtually identical results.
Therefore, the results reported here are based on the statistical
analysis of the combined data from both samples. Demographic

information is shown in Table 4.
Survey Administration
the development of the constructs and their opera
The total number of completed responses was 312. Out of the 1,000
Following
consumers we contacted, 84 e-mails were undeliverable, and 134
tionalization, several small-scale pretests (including personal
responses were obtained (15 percent response rate). The response
interviews) were conducted with a total of 75 respondents to
rate is comparable to recent online consumer surveys (e.g., Koufaris
enhance the psychometric properties of the measurement scales.
2002; Pavlou 2003). Out of the 290 students, 179 responses were
Given the large number of constructs in the proposed model, the
was to have a small number of items per construct while
obtained (62 percent response rate). The follow-up study was
goal
completed by 267 (86 percent) of the original respondents (77
retaining adequate measurement properties. Finally, a larger-scale
percent of consumers and 91 percent of students).
pretest with 214 students was also contacted to confirm the
measurement properties of the final items and provide preliminary
Nonresponse bias was assessed by verifying that (1) respondents'
evidence for the proposed model. All pilot tests were conducted
demographics were similar to those of other Internet consumers
following the same procedure as the subsequent actual data
collection (Churchill 1979). infoplease.com/ipa/A0901651.html),and (2) earlyand
(http://www.
This study's main sample comprised 312 Internet consumers drawn

from two populations. The first sample was selected from students, The Chow test compares the sum of squared errors from three
and the second regressions?one for each sample period and one for the pooled data. The F
sample consisted of Internet consumers. All
value is .27 (p > .99).
respondentswere asked to click on theWeb URL linkprovided in
an invitation e-mail message, which linked to an online survey The Wilk's lambda criterion measures the difference between groups, and
instrument. The respondents were offered incentives in the form of
itwas .99, implying virtually no difference.

2006 127

Table 4. Demographic Characteristics

Internet
Gender Age Education Income Experience
Variables ' (%Male) (Years) (Years) ($1,000s) (Years)
"Ivlean/Median
(STD) 50/50(50) 31.6/30(15) 2^9/21(4.2) 31.6/29(6Z5) 4.4/4.7(2.1)
Table 5. Descriptive Statistics for Principal TPB Perceptions
Getting Information Purchasing

Coefficient of
Mean (STD) Coefficientof Internal Mean (STD) Variation Internal
Principal Construct [Scale 1-7] Variation (%) Consistency [Scale 1-7] (%) Consistency
Actual Behavior .65 (.47) 72 1.0 .27 (.45) 166 1.0
Behavioral Intent 5.3(1.7) 32 .84 4.2(1.9) 45.97
Attitude toward ,_n /A AS Â Â
D . . v ' v ' 5.8(1.4) 24 .94 4.8(1.6) 33 .92
Behavior
Subjective Norm 4.6(1.6) 34 .77 4.6(1.6) 34.83

PBC (Indicator) 5.9(1.3) 22 1.0 5.3(1.6) 30 1.0
Self-Efficacy 5.8(1.3) 21 .92 5.7(1.5) 26 .93

.86
Controllability_5.7(1.4)_24_.74_5.6(1.5)_26
Table 6. Descriptive Statistics for External Beliefs
External Beliefs Mean (STD)

(belief strength x [Scale: 1-49 Coefficient of Internal
Behavior beliefpower) (7 * 7)] Variation(%) Consistency
Trust - Getting Information 34.4 (10.4) 29 .88
PU - Getting Information 35.7 (10.8) 30 .89
PEOU - Getting Information 35.6 (10.6) 29 .83
,Gettmg TimeResources 33.7(12.7) 37 !<H)
Information-._:_
Download Delay 35.6(11.1) 31 .89
Website Navigability 35.9 (11.6) 32 .82
Getting InformationSkills 33.7 (12.6) 37 .84

Trust - Purchasing 35.9(11.4) 31 .87
PU - Purchasing 35.2(11.8) 33 .88
PEOU - Purchasing 36.1(11.9) 32 .89
Product Value 30.9(14.1) 45 .85
Purchasing
Monetary Resources 31.7(14.2) 44 .88
Product Diagnosticity 32.8 (11.8) 35 .83
InformationProtection 36.9(12.3) 33 .90
_PurchasingSkills_34.9(11.2)_32_?7_

theTPB
late respondentswere not significantlydifferent(Armstrongand orderPBC factorfullymediates the impactof SE and controllability
Overton 1977). The first set of tests compared gender, age, educa on intentionand behavior, and (2) SE and controllabilityfully
tion, income, and Internet experience. The second set of tests mediate the impactof all externalbeliefs on PBC.
compared these characteristics, plus all principal constructs for the
two groups. All possible t-test comparisons between the means of
thetwogroups inboth setsof testsshowed insignificant
differences
Measurement Validation
(p<0.1 level).
Measure reliability was assessed using internal consistency scores,
calculated by the composite reliability scores (Werts et al. 1974).11
Descriptive Statistics Internal consistencies of all variables are considered acceptable
since they exceed .70, signifying tolerable reliability. Convergent
statistics for the principal constructs are shown inTables and discriminantvalidity is inferredwhen the PLS indicators
Descriptive
(1) loadmuch higher on theirhypothesized factor thanon other

5 and 6. Since the respondents self-selected the focal product and
theWeb vendor, social desirability bias could be present. However, factors are higher than cross-loadings), and (2) when
(own-loadings
the coefficients of variation (STD/Mean ratio) attest to substantial the square root of each construct's average variance extracted (AVE)
variability. is larger than its correlations with other constructs (Chin 1998a).
The first test was

performed using the CFA procedure in PLS.12 As
Test forHigher-Order Factors shown in Appendix B, all items loaded well on their respective
factors, which are much higher than all cross loadings. Second, as
In PLS, factors can be two shown in Tables 7 and 8, the square root of all AVEs are above .80,
higher-order approximated using
common et al. 2003). The first uses repeated which are much larger than all the cross-correlations These tests
procedures (Chin
indicatorsfollowingLohmoller's (1989) hierarchical component suggest that all measures have adequate convergent and discriminant
model the higher-order constructs using all validity. Common method bias was assessed using Harman's one
by directly measuring
items of its lower-order constructs 130-133). The second factortest (Podsakoffand Organ 1986). Each principal construct
(pp.
models the paths from the lower order to the higher order construct explains roughlyequal variance (omittedforbrevity),indicatingthat
(Edwards 2001). The latterapproachwas chosen for this study
our data do not suffer from high common method variance. Finally,
because itspecifies therelativeweight of SE and controllabilityon multicollinearity among the external beliefs was not a serious
PBC. These were derived a principal concern since none of the checks (eigen analysis, tolerance values,
weights using components
factoranalysis (Diamantopoulos andWinklhofer 2001, p. 270): VIF) indicatedany problem.
PBC = Yj x SE + y2x Controllability (4)
Where: yl and y2 are the parameters of the impact of SE and The Structural Model
controllabilityon the latentvariable PBC.
The PLS path coefficients are shown in Figure 4. For clearer
exposition, the item loadings of each construct are omitted since

The existence of a higher-order model was tested with a set of tests
they are all above .80. All control variables were initially included
following Chin (1998a) and Diamantopoulos and Winklhofer
in the model, but since none were they were dropped.
(2001). First, we examined the correlations between the lower- and significant,
With respect to the control variables of past experience and habit,
higher-order factors.
For getting information (Table 7), the
correlations between
SE and controllability and the aggregate PBC
this findingis consistentwith Ajzen (1991), who argues thatthe
main constructs of TPB should account for both because
factor are .72 and .63 (p < .01), respectively. For purchasing (Table past
< experiences are captured via PBC.
8), the correlations are .74 and .66 (p .01) for SE and
controllability, respectively. Second, to insure content validity,

or criterion Getting information has a significant impact on purchasing. There
indicator items were used to assess whether the
is also a significant impact of purchase intention on intention to get
aggregate latent factor is highly correlatedwith a direct PBC
information. Together with attitude, PBC is a significant predictor
indicator. The correlation between the aggregate four-item PBC
factorand the singlePBC indicatoritemis .74 (p < .01) forgetting of intention to get information (R2= .55). Also, intention and PBC
information, and .76 (p < .01) for purchasing. This suggests that the
aggregate factor captures the content of PBC for each behavior.
Finally,we testedwhether theaggregatePBC factorfully

mediates 11
+
The composite reliability score is (lXi)2/[(lXif ZiVar(er)], where Xx is
the impact of the underlying formative factors on intentions and
the indicator loading, and Varfo^l-Xi2.
behavior, and externalbeliefs influencePBC only throughSE and
controllability.All mediation tests (Baron and Kenny 1986) for 1
Confirmatory factor analysis in PLS was the
performed following
both behaviors (omittedforbrevity)confirmedthat(1) thehigher
procedure of Agarwal and Karahanna (2000).

2006 129

Table 7. Correlation Matrix and Average Variance Extracted for Principal Constructs (Getting
Information)
- - oh ?8 * _ d
?<C/)CL(flOl-Q-Q-t-QZ(0
? ?
_O
Getting Info 1.0
__
Intention .35* .93
_
"Attitude .31* .66* .97
Subjective Norm .07 .24* .36*

.90_
PBC (Aggregate) .27* .25* "".34* .20*' .86
.27* .42* .20* .72*
Self-Efficacy_.25* .96_
.24* .26* .19* .63* .69*
Controllability_.24* .89_
Trust -Getting .22* .35* .44* .24* .41* !39* .45* .91
Info_
PU -Getting Info .13* .33* .42* .25* .38* .42* .39* .61*
.91_
PEOU-Getting .19* .36* .44* .26* .43* .47* .40* .62* .71* .90
Info_
Time Resources .15 .29* ~.38* .24* .42* ~~47*~~ .40* .40* .47* ".46*
.94_
Download Delay .23* .35* .37* .22* .55* .58* .52* .52* .52* .58* .50*
.93_
Navigability_.17*
.37* .42* .17* .47* .48* .44* .57* .53* .61* .50* .71*
.90_ '
Consumer
Skills | .12 | .23*| .32*| .19*| .35*I .45*| .45*| .49*I .51*I .52*| .64*| .48*I .52*| .89
*
Note: denotes significant correlations at the p < .01 level. The diagonal elements (in bold) represent the square root of the AVE.
Table 8. Correlation Matrix and Average Variance Extracted for Principal Constructs (Purchasing)
3t;i-zcoiuOtt:=>LiJ<o<(**:
_Q-g<(OQ.(OOHQ-Q->SQQ-(0
Purchasing 1.0
Intention .28* ' .98

~~
"Attitude .23* .74* .94
Norm .03 .38* .45* .89
Subjective ' " ~~
PBC (Aggregate) .22* .36*~41*~
" .39* .87 ~ "" "
Self-Efficacy .21*" .34* .37* .36*~ .74* .96 ~
.32* ' .66*
Controllability .19* .29* ~30*~ .73* .91
~
Trust - ~2 .23* .33* .36*' .43* .40* 42* .85
Purchasing ~
PU - Purchasing """iT" .34* .37* .29* .45* .39* 42* .67* .91
PEOU- !l6* ^33* 42* ^36* ^2* 45* 47* JV W* ^94
Purchasing ' "
"Product Value .08 .34* ~36*~ .38* .32* .33*~ .32* .39* .35* .43* ~85~
"Monetary Res. .18*' .52* ~44*~ .41* ".42* .47* .43* """.42* .45* .43* .37* .93
"Diagnosticity .21*' .28* .35* .13 .35* .41* .40* "~~50* .55* .44* .29* .40* .90
'
Info Protection ~18*" .17* ".28* .25*~ .31* .35* .40* .54* .40* ~50*"~ .24* .30* .41*" .93
Consumer Skills | .10 | .22* I .34*" .25* | .38* | .43* | .41* .58* I .61* | .58*~34*"~ .50* "".64* I .47*" .92
*
Note: denotes significant correlations at the p < .01 level. The diagonal elements (in bold) represent the square root of the AVE.
130 MIS QuarterlyVol. 30 No. 1/March2006

'
Trust-Getting Information ~1 Attitude toward
- ^______?
?
.19* Getting Info
\[__ ^^^W
PU of Getting Info l\ ?""Ô^-' !.23*
_| .36 .57**
\|_ ^^
!^V^ ^ I-1
"16* -11* - - ^-1 Intention to ? I-1
PEOU of Getting Info Ci Subjective Norm .03 30.* ^ Getting Info
onGettingInfo | Get Info
I_l\k J< j ^j | ^p| Behavior
iL *55
I-\JfK.02 - \ / -22*
.22
Download Delay Nr.-. /".
^^""
'
ZlP^^ ?!** i^4 Controllability "
Time Resources _l_ .17* (Getting Info) ^ ^,*^^
*""Hfc ^^>*^
08+ _ H^l-' X_'-^^""^^"~? PBC
I-1 L-~ oi* -37 ^ ;
Website Navigability ^<^ _ ^j (Getting Info)
-2] ;
;
24*
^* Self-Efficacy -60** '-'
(Getting Info)
I Getting InformationSkills M? .29** -V
I_i ' 33
. .
EXTERNALBELIEFS(GETTING INFO)
_ PERCEIVED BEHAVIORAL CONTROL
i-1 (GETTING INFORMATION)
Trust - Product Purchasing __^_
I_L'- .24*
.19*
" .11*-_^_
Z~~.
I TZ |\ ! .Qi, _> Attitude toward
PU of
| ??"? Purchasing \
Purchasing_|J^--.18*
-61"
PEOU of Purchasing
i-1__\---2r -41^<JB-1
c-jA^ 2Q.^^^ X^
' "
.21*/^ Subjective Norm nA
"" _T . . . I nA? I Z . . I
\\^^\ ***
I-1J\ oo. ?4- Intention to _ .21* ?
22 ^V 14 nn
on D..r^haei
Purchasing Purchasing
Monetary Resources <J i-1 Purchase^r Behavior
24** /\ n \ y\___i
nÎ . "
! Product Value -^ 19
17. 17.
vNc}. ./ Controllability ./
-. \ . sS \ ~-~--^_v 23?
-
I Product Diagnosticity .15* , (Purchasin9) .50** S^
K- ! /'
-17*c1^ ^*T1* pbc
InformationProtection i-\J>
-""'^ - __p?^
30** :^^
I_I .20* ; _ ^ j (Purchasing) \ "Significant at p < .01
'-' *
S\ Self-Efficacy 58**
" Significant at p < .05
I -14* + Significant at p < .10
Purchasina Skills r-+- -^' (Purchasing)
Variance explained inbold
I_Purchasing Skills_|? H |
E X T E R~NA L BELIEFS(PURCHASING) .:.?.?....,
PERCEIVED BEHAVIORAL CONTROL
(PURCHASING)
Figure 4. PLS Results for Getting Information and Purchasing
influence getting information (R2= .22). The antecedents of attitude diagnosticity, and purchasing skills significantly influence
= are trust, PU, PEOU,
(R2 .36) toward getting information and purchasing SE(R2=.33).
download delay. With the exception of PEOU and website
= <
navigability (b .08, p .10), all hypothesized control beliefs are To examine the predictive power of the proposed model, we com
significant, explaining R2= .37 of controllability over getting infor pare ittofourmodels in termsofR2 adjusted'(1) a TRA model (PBC
mation. Finally, PEOU, skills, and website navigability are omitted), (2) a TAM model (onlyPU and PEOU), (3) a TAM-trust
significant predictors of SE (R2= .33). integrated model, and (4) a direct model (attitude, SN, and PBC
omitted as mediators), using Cohen's (1988) formula for calculating
Product purchasing (R2= .19) is predicted by purchase intentions effect size (f2) (the degree towhich the phenomenon is present in the
and PBC. Attitude and PBC explained R2= .59 of the variance of population) (Chin 1998b):
=
purchasing intentions. Attitude toward purchasing (R2 41) is
= " "
predicted by trust, PU, PEOU, monetary resources, product value, f~ (R includedR excluded)/ (1 R (5)
included)
and product diagnosticity. Controllability over purchasing (R2= .34)
is impacted by trust, PEOU, monetary resources, product value and 1. Dropping PBC significantly reduces the variance explained in
= . 12 = .
diagnosticity, and information protection. Finally, PEOU, product getting information to R2 (f2 13) and in purchasing to

2006 131

theTPB
R2 =. 11 (f2=. 10), and also substantially decreases the variance recommend selecting five to nine beliefs thataremost likely to
=
explained in intentions to get information to R2 = .42 (f2 .29) influence each behavior. Given that the literature has offered
and purchase to R2 = .46 (f2 = .32). numerous variables to predict e-commerce adoption, this study
identifiedthemost accessible ones by freelyeliciting(notarbitrarily
2. A TAM model predictsR2 = .23of thevariance in intentionsto selecting) consumer
responses through open-ended questions. The
= = =
get information (f2 .71), and R2 .29 (f2 .73) in intentions resulting are
beliefs presumably foremost in consumers' minds.
to purchase. Most of these beliefs recur in the IT adoption and use literature (e.g.,
PU, PEOU, navigability), or within the domain of IS (e.g., trust,

3. TAM-trustmodel (Gefenet al. 2003) explainsR2
An integrated information protection, product diagnosticity, user skills),
= =
.28 (f2 .69) of the variance in intentions to get information, highlightingthefundamentalroleof IT inonline consumerbehavior,
= =
and R2 .35 (f2 .59) in purchasing intentions. beyond existing variables (e.g., reputation, product value). It is
shown that these beliefs can adequately predict the two proposed e
= = commerce
4. A direct model explains R2 .41 (f2 .40) of the variance in behaviors and, as a consequence and to a large extent, e
= =
intentions to get information, and R2 .46 (f2 .32) in commerce adoption.
purchase intentions. Finally, a test of mediation (Baron and
Kenny 1986) validates thatattitudeand PBC fully

mediate the
impact of external beliefs on intentions for both behaviors
(omittedforbrevity).
Implications for Theory and Research
The results have implications for the e-commerce, IS, TPB, and trust
In sum, the four competing models have
significantly lower
literatures.
predictive validity compared to the original model, as shown by the
theoriginal
substantialeffectsizes (Cohen 1988).13Most important,
model explicates most accessible factors that underlie online
consumer over simpler models.
Implications for Electronic Commerce Research
behavior, establishing its superiority
Most e-commerce studies follow the TRA or TAM, implicitly

assuming that behavior is volitional. However, online consumers
face several new constraints, such as the impersonal nature of the
Discussion MHHi online environment, the extensive use of IT, and the uncertainty of
the open Internet infrastructure. These issues call for the inclusion
This paper aims light on the phenomenon
to shed of consumer of PBC in e-commerce adoption models (and the use of TPB rather
of B2C e-commerce using an extended version of thanTRA orTAM). Indeed,neglectingPBC and relyingon simpler
adoption
TPB. The study draws upon theories from information systems, models lead to e-commerce models that are
may adoption
social psychology, marketing, and economics to propose, and potentially
incomplete misleading.
and empirically examine a comprehensive, yet
operationalize,
parsimonious model that explains and predicts two key online B2C e- commerce as a two-stage process not only
Viewing adoption
consumer behaviors: getting information and purchasing products. a more of online consumer behavior,
yields complete understanding
but it also enhances the predictive power of the e-commerce
adoption model, since the two behaviors are interrelated. In doing
so, this study provides a comprehensive, model to

Key Findings and Insights
yet parsimonious
describe and predict online consumer behavior and e-commerce
as a monolithic adoption.
This study does not view e-commerce adoption
behavior (product purchasing), but rather as consisting of at least
one contingent behavior These behaviors are
(getting information).
related: information influences while intention Implications for Information Systems
getting purchasing,
to purchase triggers intention to get information.
The setof accessible beliefs identifiedinour studywas empirically
shown to draw from the IT adoption and use literature (e.g., PU,
The well-established TPB was extended to predict these two
behaviors and to derive the set of their respective accessible beliefs. PEOU, download delay, navigability) or come fromwithin the
domain IS
of trust, information
The derivation is consistentwith Ajzen and Fishbein (1980), who (e.g., protection, product
diagnosticity, user skills), confirming the increasingly important role

of IT in online consumer behavior. It is important to reiterate that
these beliefs were not

arbitrarily chosen, but they were freely
For comparing correlation-based coefficients (such as the R2 adjusted), elicited. These have for the emerging role of
findings implications
effect sizes of 0.1 are considered small, 0.3 are considered medium, and 0.5
IS as a reference discipline for online consumer behavior. While
are considered large. However, no specific significance test has been
traditional consumer behavior is well described by marketing and
proposed.

2006

theTPB
economic theories, overwhelming evidence suggests that IT-related Finally, this study provides empirical evidence on whether past
variables have become at least as important as traditional factors in behaviors should be included in the TPB model as control variables.
predicting consumer behavior on the Internet (e.g., Gefen et al. By accountingforbothdeliberate (priorexperience) and automated
2003; Jarvenpaaet al. 2000;McKnight andChervany 2002; Pavlou (habit) past behaviors as control variables, we find empirical
2003). Existing literature on consumer behavior has largely ignored evidence for the adequacy of TPB perceptions to reflect past
IT issues, and justifiably so, since consumers did not face any IT activities,
validating Ajzen's (1991) theoretical assertion, at least for
issues in physical markets. However, online consumers are intrin the two e-commerce behaviors.
sically active users of IT, and IT considerations take center stage

(Stewart and Pavlou 2002). Rather than viewing e-commerce as a
marketing issue influenced by IT use, it is perhaps more accurate to Implications for the Trust Literature
view e-commerce as an IS phenomenon where an IT user interacts
with a complex IT system(Koufaris 2002). This system includes Perhaps themost theoretically interesting and empirically influential
not only a website, but also the supporting services and the people belief associatedwith gettinginformation
and purchasingproducts
and proceduresbehind thoseservices (TaylorandTodd 1995b). An is trust. By integrating trust as an external belief in the TPB model,
IS view would not only help better understand B2C e-commerce, but we make two key contributions to the trust literature.
itmay also shed light on how marketing, economic, and other
factors integrate with IS concepts to better explain other complex IT An contribution is the placement of trust in the
important
phenomena. nomological network of TPB. Although trusthas already been
hypothesized and shown to influence online transaction behavior,
previous views were incomplete. They either considered trust as
Implications for the Theory of Planned Behavior directly affecting intentions (e.g., McKnight and Chervany 2002),
or as influencing intentions through attitude (e.g., Jarvenpaa et al.
In striving to fully understand and simultaneously predict two 2000). Our view delineates the process by which trust influences
distinct, contingent, and non-volitional behaviors, this paper behavior by acting as both an attitudinal and control belief, and thus
contributes to the social psychology literature by extending TPB in places trust as an antecedent of both attitude (due to confident
three key ways. expectations) and controllability (due to uncertainty reduction).
First,thisstudysheds lighton thenatureand role of PBC, which is A second important contribution lies in the proposed con
still not well understood. We theorize and empirically show that ceptualization of trusting beliefs that is consistent with the behavior
PBC acts as a second-order formative structure, formed by two specific nature of TPB. Compared with most trust studies, trust is
distinct dimensions: SE and controllability. This structure conceptualized at a more granular level, namely, as distinct beliefs
maintains the parsimonious unitary view of PBC, while allowing the about getting information and purchasing. Given the increasing
role of its two underlying dimensions to vary depending on the importance of trust in e-commerce, such a thorough to the
approach
relativeimportanceof SE and controllabilityfordifferent
behaviors. nature and role of trust becomes necessary for predicting specific
The proposed second-order formative structure of PBC should be consumer behaviors.
applicable to virtually any behavior, even if the impact of SE or

controllability could vary across behaviors. Finally, a formative
structure permits a more
Implications forPractice
detailed prediction of external control
beliefs a distinct prediction of SE and controllability,
by allowing and Public Policy
thus leading to better prediction of PBC, intention, and behavior.
The proposed e-commerce adoption model describes a concrete set

Second, while TPB is commonly used to model behaviors of factors that managers
might manipulate to facilitate consumer
independently, this study extends TPB to allow
modeling the
browsing and purchasing. It also suggests that managers should
association between two related behaviors. The behaviors are linked
examine interventions to improve consumer attitudes and enhance
at the intention and behavior stages, while perceptions and beliefs PBC over online activities. The external beliefs (and their relative
remain strictly behavior-specific (consistent with TPB). One on attitude
impact and PBC) represent specific factors on which
behavior can thus influence another without violating TPB. In
managers should focus their attention, efforts, and investments,to
addition, by employing Gollwitzer's (1999) notion of imple online consumer behavior and increase transaction volume.
shape
mentation intentions, a certain
goal-directed intention can trigger
another intention if this serves as a means to accomplishing the goal.
The proliferation of B2C e-commerce has been a priority formany
Allowing two related behaviors to be simultaneously modeled opens
governments. Public policy officials could support e-commerce by
new avenues for future research. It also paves the way for a more
instituting mechanisms for influencing the proposed consumer
complete explanation and prediction of behaviors beyond TPB's
beliefs. For example, trust can be engendered through institution
original constructs.
based mechanisms and prosecution of online fraud (Pavlou and

2006 133

theTPB
Gefen 2004). Laws unsolicited e-mail and In our survey design, the respondents were asked to self-select a
against (spam)
establishment of security guidelines can increase perceived product they were seriously considering purchasing. This might
information protection. Education and training could influence have downgraded thepotential impactof PBC by encouraging the
consumer online skills. participants to self-select familiar products that they consider easily
accessible. While this downward bias further stresses the prominent
role of PBC in online consumer behavior, future research should
experimentwith differentsurvey designs that could completely

Limitations and Suggestions self-selection bias.
prevent
forFuture Research
In contrast to TAM, whose two beliefs (PU and PEOU) aim to
Despite the comprehensiveness of the proposed model and the
predict "system use," the external beliefs in our extended TPB
empirical support for it, we acknowledge some theoretical and
model are specific to each e-commerce behavior. Therefore, our PU
empirical limitations, which call for additional research.
and purchasingare not identical
and PEOU forgettinginformation
to TAM's variables for system use. For example, productivity does
First, e-commerce adoption cannot be fully assessed with only two not make much sense in e-commerce, and itwas therefore dropped
behaviors. Future research could examine additional consumer
as a component of PU. Purchasing PU has higher emphasis on
behaviors, such as giving information and post-purchase interaction.
enhancing the consumer's effectiveness in purchasing products,
Our TPB-based framework a blueprint for identifying the
provides
while getting informationPU focused on obtaining useful and
key accessible beliefs for such behaviors.
valuable information. Also, even ifprevious research on TAM (e.g.,
A major obstacle in our study was the large number of survey items Taylor and Todd 1995b) has shown that attitudedoes not fully
mediate the impact of PU and PEOU on intentions to use IT, these
(over 100) we had to pose to our respondents in order to simul
on the "IT usage" behavior should not necessarily
assess two behaviors in a fashion using the empirical findings
taneously comprehensive
to the focal e-commerce behaviors.
and belief power foreach externalbelief.
productof belief strength
generalize
We werethus limited to the use of mostly two-item scales. To over

come this limitation, we used multiple pretests to judiciously reduce Consistentwith Ajzen's (2002a) recommendationsand following
the number of items per scale without weakening the underlying suggestions from the pilot tests, our survey items were grouped
measurement we took care to under meaningful categories, and the "belief strength" and "belief
construct's properties. Furthermore,
base our measures on well-validated scales with excellent psycho power" items were presented in pairs. While such survey pre
The of 312 participants was sentation may create higher construct reliabilities, our internal
metric properties. sample size large
the largest number of structural paths directed at results are similar to previous TPB studies, and are not
enough to capture consistency
et al. 2003).
research might succeed in et al. 2004). Nevertheless, to entirely
any construct (Chin Future exceedingly high (Straub
when considering account for such bias, future research could randomize all items.
reducing the number of survey items (especially
two or more related behaviors) by questioning the validity or
of TPB's formula where both belief Following TPB, all constructs in the proposed model reflect
necessity expectancy-value
strength and belief power must be measured. assessments for a specific product and Web vendor. Extrapolating
beyond these parameters requires additional research since TPB

Another is the possible presence of social desirability bias
limitation cautions the generalization from one behavior to others. Finally,
or "ceiling effects" due to the self-selection of products and Web the longitudinal assessment of actual behavior, the remaining
despite
vendors. While this may result in relatively high means for the factors were measured at a single point in time, which prevents
variables, there was enough (relative) variability in our measures causal interpretation.
consumers
(Tables 4 and 5) tomake testing possible. Theoretically,
tend to get information and purchase products from Web vendors
toward whom they have positive convictions. Therefore, the
of this to most e-commerce
findings study may generalize
Conclusion H__H_H_____HHHM___H
transactions. Nevertheless, further research is required to test this
assertion.
This study represents a systematic approach to understanding and
online consumer behavior using an extended version of

Self-selection of products and vendors also introduces complications predicting
TPB. The development of a large set of accessible e-commerce
because the focal behaviors may vary in their level of inherent
factors is in response to several IS researchers (e.g., Ives et al. 1983)
uncertainty. Even if product price and Web vendor reputation are
some relationships,
who encouragedbuilding a cumulativeknowledge in emerging IS
controlled for, risk variability can moderate such
areas, such as online consumer behavior. In doing so, we aim to
as the impact of trust on PBC. Such moderated relations require a
encourage IS researchers to view online consumer behavior as a new
distinct conceptualization on their own right, which were beyond the
research area where IS can serve as a reference discipline.
scope of the study.
134 2006

theTPB
Acknowledgments Cassell, J., and Bickmore, T. "External Manifestations of

Trustworthiness of the Interface," Communications of the ACM
The authors would like to thank the senior editor, Ritu Agarwal, the (43:12), 2000, pp. 50-59.
associate editor, Elena Karahanna, and the three reviewers, David Chen, Z., and Dubinsky, A. J. "A Conceptual Model of Perceived
Gefen, Harrison and Jonathan Palmer, for the excellent Customer Value in E-commerce: A Preliminary Investigation,"
McKnight,
reviews that have significantly improved the manuscript's quality. Psychology andMarketing (20:4), 2003, pp. 323-347.
The authors also thank Professor leek Aizen for valuable feedback Cheng, S. F., and Chan, K. S. "The Role of Perceived Behavioral
on criticalpointspertainingto thedevelopmentof theproposedTPB Control in Predicting Human Behavior: A Meta-Analytic

extension model. Both authors contributed equally to this manu Review of Studies on theTheory of PlannedBehavior,Working
names has been randomly selected. Chinese of Hong Kong, 2000.
script and the order of authors' Paper, University
Cheung, and Lee, M.
C, "Trust in Internet Shopping: Instrument
Development and Validation through Classical and Modern Ap

proaches," Journal of Global Information Management (9:3),
References 2001, pp. 23-35.
Chin, W. W. "Issues and Opinion on Structural Equation
and Karahanna, E. "Time Flies When You're
Agarwal, R., Having Modeling," MIS Quarterly (22:1), 1998a, pp. 7-16.
Fun: Absorption and Beliefs about Information to Structural
Cognitive Chin, W. W. "The Partial Least Squares Approach
TechnologyUsage," MIS Quarterly (24:4), 2000, pp. 665-694. Equation Modeling," inModern Methods for Business Research,
I. "Constructing a TpB Questionnaire: Conceptual and G. A. Marcoulides Lawrence Erlbaum
Ajzen, (ed.), Associates,
Considerations," of
Methodological Working Paper, University Mahwah, NJ, 1998b, pp. 295-336.
Massachusetts, Amherst, 2002a online at
September (available Chin, W. W., Marcolin, B. L., and Newsted, P. R. "A Partial Least
http://www-unix.oit.umass.edu/~aizen/pdf7tpb.measurement.pdf). Squares Latent Variable Modeling Approach forMeasuring
Ajzen, I. "Perceived Behavioral Control, Self-Efficacy, Locus of Interaction Effects: Results from a Monte Carlo Simulation
Control, and the Theory of Planned Behavior," Journal of Study and an Electronic Mail Emotion/Adoption Study,"
Applied Social Psychology, (32), 2002b, pp. 665-683. Information Systems Research (14:2), 2003, pp. 189-217.
Ajzen, I. "The Theory of Planned Behavior," Organizational Chow, G. "The StandardF Test for theEquality of Two Sets of
Behavior & Human Decision Processes (50), 1991, pp. 179-211. Coefficients in Linear Models," Econometrica
Regression (28),
Ajzen, I., and Fishbein, M. Understanding Attitudes and Predicting 1960, pp. 591-605.
Social Behavior, Prentice-Hall, Englewood Cliffs, NJ, 1980. Churchill, G. "A Paradigm for Developing Better Measures of
Alford, B., and Biswas, A. "The Effects of Discount Level, Price Constructs," Journal Research
Marketing ofMarketing (16:1),
Consciousness and Sale Proneness on Consumers' Price 1979, pp. 64-73.
Perception and Behavioral Intention," Journal of Business V., Karahanna, E., and Dumm, R. "The Relative
Choudhury,
Research (55:9), 2002, pp. 775-783. Advantage of Electronic Channels: A Conceptual and
Armstrong, J. S., and Overton, T. "Estimating Nonresponse Bias in Operational Definition," Working Paper, University of
Mail Surveys,"Journal ofMarketing Research (19), 1977, pp. Cincinnati, 2001.
396-402. Cohen, J. Statistical Power Analysis for the Behavioral Sciences,
Ba, S., and Pavlou, P. A. "Evidence of the Effect of Trust in Lawrence Erlbaum Associates, Hillsdale, NJ, 1988.
Electronic Markets: Price Premiums and Buyer Behavior," MIS Compeau, D., and Higgins, C. "Application of Social Cognitive
Quarterly (23:4), 2002, pp. 243-268. Theory to Training for Computer Skills," Information Systems
Bandura, A. Social Foundations of Thought and Action: A Social Research (6:2), 1995, pp. 118-143.
Cognitive Theory, Prentice-Hall, Englewood Cliffs, NJ, 1986. Conner, M., and Armitage, C. J. "Extending the Theory of Planned
Barber, B. The Logic and Limits of Trust, Rutgers University Press, Behavior: A Review and Avenues for Future Research," Journal
New Brunswick, NJ, 1983. ofApplied Social Psychology (28) 1998,pp. 1429-1464.
Baron, R.,and Kenny, D. "The Moderator-Mediator Variable Davis, F. D. "Perceived Usefulness, Perceived Ease of Use and
Distinction in Social Research: User Acceptance of Information Technology," MIS Quarterly
Psychological Conceptual,
Strategic, and Statistical Considerations," Journal of Personality (13:3), 1989,pp. 319-340.
and Social Psychology (51:6), 1986,pp. 1173-1182. Diamantopoulos, A., and Winklhofer, H. M. "Index Construction
Bellman, S., Lohse, G. L., and Johnson, E. J. "Predictors of Online with Formative Indicators: An Alternative to Scale Devel
Buying Behavior," Communicationsof theACM (42: \2), 1999, opment," Journal of Marketing Research (38:2), 2001, pp.
269-277.
pp. 32-38.
and Burton, S. Processes Used Doney, P. M., and Cannon, J. P. "An Examination of the Nature of
Blair, E., "Cognitive by Survey
to Answer Behavioral Trust inBuyer-Seller Relationships," Journal ofMarketing (61),
Respondents Frequency Questions,"
Journal of Consumer Research (14), 1987, 141-154. April 1997, pp. 35-51.
and D. "Validation in IS Edwards, J. R. "Multidimensional Constructs in Organizational
Boudreau, M.-C, Gefen, D., Straub,
Research: A State-of-the-Art MIS Behavior Research: An Integrative Analytical Framework,"
Assessment," Quarterly
(25:1), 2001, pp. 1-16. Organizational Research Methods (4:2), 2001, pp. 144-192.

2006 135

Engel, J.,Kollat, D., and Blackwell, R. Consumer Behavior, Holt, Lewicki, R. J., and Bunker, B. B. "Trust inRelationships: A Model
Rinehart, and Winston, New York, 1973. of Trust Development and Decline," in Conflict, Cooperation,
Fukuyama, F. Trust: The Social Virtues and the Creation of and Justice, B. B. Bunker and J. Z. Rubin (eds.), Jossey-Bass,
Prosperity, The Free Press, New York, 1995. San Francisco, 1995, pp. 133-173.
Gefen, D. "Reflections on the Dimensions of Trust and Trust Liang, T. P., and Huang, J. S. "An Empirical Study on Consumer
worthiness among Online Consumers," DataBase (33:3), 2002, of Products in Electronic Markets: A Transaction
Acceptance
pp. 38-53. Cost Model," Decision Support Systems (24:1), 1998, pp. 29-43.
Gefen, D., Karahanna, E., and Straub, D. W. "Trust and TAM in M., and Hirt, S. G. "Force of Habit and Information
Limayem,
Online Shopping: An Integrated
Model,"MS Quarterly (27:1), SystemsUsage: Theory and InitialValidation," Journal of the
2003, pp. 51-90. Association of Information Systems (4:3), 2003, pp. 65-97.
Gefen, D., and Straub, D. "The Relative Importance of Perceived J.B. "The PLS Program System: Latent Variables Path
Lohmoller,
Ease-of-Use in IS Adoption: A Study of e-Commerce Analysis with Partial Least Squares Estimation," Multivariate
Adoption," Journal ofAIS (1:8), 2000, pp. 1-30. Behavioral Research (23), 1989, pp. 125-127.
George, J. "Influences on the Intent toMake Internet Purchases," G. and P. "Electronic Com
Lohse, L., Spiller, Shopping,"
Internet Research (12:2), 2002, pp. 165-180. munications of theACM (41:7), 1998,pp. 81-87.
Gollwitzer, P. M. "Implementation Intentions: Strong Effects of N. Trust and Power, John Wiley and
Luhmann, Sons, London,
Simple Plans," American Psychologist (54) 1999, pp. 493-503. 1979.
Hoffman, D., Novak, T., and Peralta, M. "Building Consumer Trust
Mathieson, K. User Intentions: the Tech
"Predicting Comparing
Communications of the ACM'(42:4), 1999, pp. 80-85.
Online,"
nologyAcceptanceModel with theTheory ofPlannedBehavior,"
Hosmer, T. "Trust: The Connecting Link between Organizational Research 173-191.
Information Systems (2:3), 1991, pp.
Theory and Philosophical Ethics," Academy of Management and Lederer, C. "A Study of Personal
Mawhinney, C, Computer
Review (20:2), 1995,pp.379-403. Utilization and Management
by Managers," Information (18:5),
Ives, B., and Learmonth, G. P. "The Information System as a
1990, pp. 243-253.
Competitive Weapon," Communications of the ACM (21), 1984, R. C, F. D., and Davis, J.H. "An
Mayer, Schoorman, Integrative
pp. 1193-1201. Model of Organizational Trust," Academy of Management
M. and Baroudi, J. J. "The Measurement of
Ives, B., Olson, H., Review (20:3), 1995,pp. 709-734.
User Information Satisfaction," Communications of the ACM
McKnight, D. H., and Chervany, N. L. "What Trust Means in E
(26:10), 1983,pp. 785-793. Commerce Customer Relationships: An Interdisciplinary Con
Jarvenpaa, S. L., Tractinsky, N., and Vitale, M. "Consumer Trust
ceptual Typology," International Journal of Electronic Com
in an Internet Store," Information and Management
Technology merce (6:2), 2002, pp. 35-60.
(1:12), 2000, pp. 45-71. Olson, J. S., and Olson, G. M. "i2i Trust in e-commerce,"
Jiang, Z., and Benbasat, I. "Virtual Product Effects of
Experience: Communicationsof theACM(43:12), 2000, pp. 41-44.
Visual and Functional Control of Products on Perceived and Performance
Palmer, J. "Web Site Usability, Design, Metrics,"
Diagnosticity in Electronic Shopping," Journal ofManagement
Information Systems Research (13), 2002, pp. 151-167.
Information Systems (21:3), 2004, pp. 111-147.
Pavlou, P. A. "Consumer Acceptance of Electronic Commerce:
Kalakota, R., and Whinston, A. Electronic Commerce: A
Integrating Trust and Risk with the Technology Acceptance
Manager's Guide, Addison-Wesley, Reading, MA, 1997.
Model," International Journal of Electronic Commerce (7:3),
Karahanna, E., Straub, D. W., and Chervany, N. L. "Infoimation
2003, pp. 69-103.
Technology Adoption across Time: A Cross-Sectional
Pavlou, P. A., and Gefen, D. "Building Effective Online
Comparison of Pre-Adoption and Post-Adoption Beliefs," MIS
Marketplaces with Institution-Based Trust," Information Systems
Quarterly (23:2) 1999,pp. 183-213. Research (15:1), 2004, pp. 35-62.
Kempf, D. S., and Smith, R. E. "Consumer Processing of Product
Podsakoff, P. M., and Organ, D. W. "Self-Reports inOrganizational
Trial and the Influence of Prior Advertising: A Structural
Research: Problems and Prospects," Journal of Management
Modeling Approach," Journal ofMarketing Research (35:3),
(12:4), 1986,pp. 531-544.
1998, pp. 325-338.
Povey, R., Conner, M., Sparks, J., James, R., and Shepherd, R.
Kim, D., and Benbasat, I. in Internet
"Trust-Related
Arguments
Journal
"Applicationof theTheory of Planned Behavior toTwo Dietary
Stores: A Framework for Evaluation," of Electronic
Behaviours: Roles of Perceived Control and Self-Efficacy,"
Commerce Research (4:2), 2003, pp. 49-64.
BritishJournal ofHealth Psychology (5:2), 2000, pp. 121-139.
Koufaris,M. "Applying theTechnology Acceptance Model and and D. W. "Current
to Online Rose, G., Khoo, H., Straub, Technological
Flow Theory Consumer Behavior," Information
Impediments to Business-to-Consumer Electronic Commerce,"
Systems Research (13:2), 2002, pp. 205-223.
W. a Taxonomy
Communications of theAIS (1:16), 1999,pp. 1-73.
Law, K. S., Wong, C. S., and Mobley, H. "Toward
Salisbury, W., Pearson, R., Pearson, A., and Miller, D. "Identifying
of Multidimensional Constructs," Academy of Management
Review (23:4), 1998, pp. 741-755.
Barriers That Keep Shoppers Off theWorld Wide Web:
Developing a Scale of Perceived Web Security," Industrial
Lepper, M. R. "Microcomputers in Education: Motivational and
American 1-18. Management & Data Systems (101:4), 2001, pp. 165-176.
Social Issues," Psychologist (40:1), 1985, pp.

2006

theTPB
Sheeran, P., and Orbell, S. "Implementation Intentions and Zajonc, R. B. "The Attitudinal Effects ofMere Exposure," Journal
Repeated Behavior: Augmenting thePredictiveValidity of the ofPersonality and Social Psychology (9:1), 1968,pp. 1-27.
Theory of Planned Behavior," European Journal of Social Zand, D. "Trust and Managerial Problem Administrative
Solving,"
Psychology (29:2), 1999, pp. 349-369. Science Quarterly (17:2), 1972,pp. 229-239.
Sheppard, B. H., Hartwick, J., and Warshaw, P. "The Theory of
Reasoned Action: A Meta-Analysis of Past Research with
Recommendations for Modifications and Future Research,"
Journal of Consumer Research 1988, pp. 325-343.
(15:3),
About theAuthors
Stewart, D. W., and Pavlou, P. A. "Substitution and Comple
mentarity: Measuring the Effectiveness of Interactive Marketing

Paul A. Pavlou is an assistant professor of Information Systems at
Communications," Journal of theAcademy ofMarketing Science
theUniversity of California atRiverside. He received his Ph.D.
(30:4), 2002, pp. 376-396.
D. W. in MIS
from theUniversity of SouthernCalifornia inMay of 2004. His
Straub, "Validating Instruments Research," MIS
research focuses on electronic commerce, institutional trust building,
Quarterly (13:2), 1989,pp. 147-169.
and information systems strategy in turbulent environments. His
Straub, D. W., Boudreau, M. C, and Gefen, D. "Validation
research has appeared in MIS Quarterly, Information Systems
Guidelines for IS Positivist Research," Communications of the
AIS (13), 2004, pp. 380-427. Research, Journal of the Academy of Marketing Science,
a Generic International Journal of Electronic Commerce, Journal
Tan, Y. H., and Thoen, W. "Toward Model of Trust for of Strategic
Electronic International Journal the proceedings of the International Con
Commerce," of Electronic Information Systems,
Commerce (5:2), 2001, pp. 61-74. ference on Information Systems, and the Best Paper Proceedings of
Taylor, S., and Todd, P. A. "Decomposition and Crossover Effects the Academy of Management conference. Paul received theMIS
in the Theory of Planned Behavior: A Study of Consumer Quarterly Reviewer of the Year award for 2003. He also received
Adoption Intentions," International Journal of Research in the Best Doctoral Dissertation Award of the 2004 International
Marketing (12), 1995a, pp. 137-156. Conference on Information Systems.

Taylor, S., and Todd P. A. "Understanding Information Technology
Usage: A Test of CompetingModels," InformationSystems Mendel Fygenson is an associate professor of Statistics in the
Research (6:3), 1995b,pp. 144-176. Information & Operation at theMarshall
Management department
Trafimow, D., Sheeran, P., Conner, M., and Finlay, K. "Evidence School of Business, the University of Southern California. He
that Perceived Behavioral Control is a Multi-Dimensional received his Ph.D. from theUniversity ofMichigan in 1989. His
Construct: Perceived Control and PerceivedDifficulty"British research focuses on statistical methodologies and their applications.
Journal ofSocial Psychology (41:1), 2002, pp. 101-121. His major are
methodological contributions in testing hypothesis,
Triandis, H. C. "Values, Attitudes, and Interpersonal Behavior," in
semi-parametric regression models and in categorical data analysis.
Nebraska Symposium on Motivation, 1979, H. E. How, Jr. (ed.),
He has multiple publications in Journal of the American Statistical
University of Nebraska Press, Lincoln, NE, 1979, pp. 195-259.
Association and in Annals of Statistics.
Werts, C. E., Linn, R. L., and Joreskog, K. G. "Interclass Reliability
Estimates: Testing Structural Assumptions," Education and Psy
chological Measurement (34:1), 1974, pp. 25-33.

2006 137

theTPB
Appendix A
Measurement Items forPrincipal Constructs
GETTING INFORMATION
During the last30 days, Igot Informationabout thisproduct from thiswebsite (Yes/No)
Intentions to Get Information
I intendto get informationabout thisproduct fromthiswebsite within the next 30 days: (Extremely unlikely/likely)
Iplan to get informationabout thisproduct fromthiswebsite within the next 30 days: (Stronglydisagree/agree
Attitude toward Getting Information
For me, getting informationabout thisproduct fromthiswebsite within the next 30 days would be: (1: a very bad/good idea,
2: very foolish/wise)
Subjective norm on Getting Information

Most people who are importanttome think it isa good idea to get informationabout this product fromthiswebsite: (Not at
all/Completely true)
Most people who are importanttome would get information about this product from thiswebsite: (Not at all/Completely true)
Perceived Behavioral Control over Getting Information (Indicator)
of you getting informationabout this product fromthiswebsite within the next 30 days:
Please rate the difficulty (Extremely
difficulteasy)
Self-Efficacy over Getting Information

If Iwanted to, Iwould be able to get informationabout thisproduct from thiswebsite within the next 30 days: (Strongly
disagree/agree)
If Iwanted to, Iam confident Icould get informationabout this product fromthiswebsite within the next 30 days: (Strongly
disagree/agree)
Controllability over Getting Information
All necessary resources forgetting informationabout thisproduct fromthiswebsite will be accessible tome within the next
30 days: (Stronglydisagree/agree)
Getting informationabout this product fromthiswebsite within the next 30 days is completely under my control: (Strongly
disagree/agree)
Trust - Getting Information
b. This Web vendor would be competent inproviding objective informationabout thisproduct: (Extremely unlikely/likely)
e. For me, getting objective product informationfroma website is: (Not at all/Extremely important)
b. This Web vendor would be honest inproviding accurate informationabout thisproduct: (Extremely unlikely/likely)
e. For me, getting accurate product informationfroma website is: (Not at all/Extremely important)
Perceived Usefulness of Getting Information
b. This website would be useful forgetting valuable informationabout this product: (Extremely unlikely/likely)
e. For me, getting valuable informationabout a product froma website is: (Not at all /Extremely important)
b. This website would enhance my effectiveness ingetting useful informationabout this product: (Extremely unlikely/likely)
e. For me, getting useful informationabout a product froma website is: (Not at all /Extremely important)
Perceived Ease of Getting Information
b. Getting informationabout thisproduct from thiswebsite would be easy: (Extremely unlikely/likely)
e. For me, getting product informationeas//y froma website is: (Not at all /Extremely important)
b. Learning how to get informationabout thisproduct fromthiswebsite would be easy: (Extremely unlikely/likely)
e. For me, learninghow to get product information easily froma website is: (Not at all /Extremely important)_
138 2006

Time Resources
[
c. Iexpect to have the time needed to get informationfromthiswebsite within the next 30 days: (Stronglydisagree/agree)
p. Having the time needed would make it(much more difficult/easier)forme to get informationabout this product,
c. There would always be time forme to get informationfromthiswebsite within the next 30 days: (Stronglydisagree/agree)
p. Finding timewould make it(much more difficult/easier)forme to get information about this product.
I Download Delay
c. Iexpect the speed bywhich thiswebsite would provide informationto be fast enough: (Stronglydisagree/agree)
p. The speed bywhich a website provides information would make it(much more difficult/easier)forme to get information
about this product,
c. Iexpect the rate at which the informationwould be displayed on thiswebsite to be fast enough: (Stronglydisagree/agree)
would make it (much more difficult/easier)forme to get information
p. A fast rate at which websites display information
about this product.
Website Navigability
c. Iexpect the sequencing of hyperlinks in thiswebsite to be clear. (Stronglydisagree/agree)
p. Having a clear sequence of hyperlinkswould make it(much more difficult/ easier) forme to get information
about this
product,
c. Iexpect the layout of thiswebsite to be intuitive: (Stronglydisagree/agree)
p. A website with an intuitivelayoutwould make it(much more difficult/ easier) forme to get information
about thisproduct.
Getting Information Skills

c. If Iwanted to, Icould become skillfulat comparing and evaluating products on thiswebsite: (Stronglydisagree/agree)
J
p. Becoming skillfulwould make it (much more difficult/easier)forme to get informationabout thisproduct.
I c. If Iwanted to, Icould easily become knowledgeable about getting all relevant informationabout products fromthis
website: (Stronglydisagree/agree),
p. Becoming knowledgeable about getting information would make it(much more difficult/easier)forme to get all relevant
informationabout this product fromthisWeb vendor.
I Getting Information Habit
Getting product informationfromthis vendor's website has become a habit forme: (Stronglydisagree/agree)
Getting product informationfromthis
website has become natural forme: (Stronglydisagree/agree)
Past Experience - Getting Information
How long have you been using the Internetforgetting informationabout products?
_years.
During the last30 days, how much timedid you spend on the Internetgetting product informationingeneral?
_hours.
During the lastyear, how many times have you made product purchases fromthe selected Web vendor? _times.
PURCHASING
During the last30 days, Ipurchased this product fromthisWeb vendor (Yes/No).
Purchasing Intentions
I intendto purchase this product fromthiswebsite within the next 30 days. (Extremely unlikely/likely)
Iplan to purchase this product fromthiswebsite within the next 30 days. (Stronglydisagree/agree)
I Purchasing Attitude
For me, purchasing thisproduct fromthisWeb vendor within the next 30 days would be: (1: a very bad/good idea, 2: very
I undesirable/desirable)_

2006 139

Purchasing Subjective Norm

Most people who are importanttome thinkthat it is fine to purchase a product fromthisWeb vendor within the next 30 days:
(Not at all/Completely true)
Most people who are importanttome would purchase this product fromthisWeb vendor: (Not at all/Completely true)
Perceived Behavioral Control over Purchasing
Please rate the difficulty
of you purchasing thisproduct fromthisWeb vendor within the next 30 days: (Extremely
difficult/easy)I
Purchasing Self-Efficacy
If Iwanted to, Iwould be able to purchase this product from thisWeb vendor within the next 30 days: (Strongly I
disagree/agree)
If Iwanted to, Iam confident Icould purchase this product from thisWeb vendor within the next 30 days: (Strongly
disagree/agree)
j Purchasing Controllability
All necessary resources forpurchasing thisproduct from thisWeb vendor will be accessible tome within the next 30 days: I
(Stronglydisagree/agree) j
Purchasing this product fromthisWeb vendor will be completely under my controlwithin the next 30 days: (Strongly
disagree/agree) I
Trust - Purchasing
b. This Web vendor would be competent indelivering this product ina timelyfashion: (Extremely unlikely/likely)
e. For me, product delivery ina timelyfashion is: (Not at all/Extremely important)
b. This Web vendor would be honest in itsdealings when Ipurchase this product from it: (Extremely unlikely/likely)
e. For me, a Web vendor that ishonest in itsdealings with itscustomers is: (Not at all/Extremely important)
b. This Web vendor would not seek to take advantage ofme ifIpurchase this product from it: (Extremely unlikely/likely)
e. For me, a Web vendor thatdoes not seek to take advantage of itscustomers is: (Not at all/Extremely important)
Perceived Purchasing Usefulness
b. This website would be useful inpurchasing this product: (Extremely unlikely/likely)
e. For me, a website that is useful inpurchasing products is: (Not at all /Extremely important)
b. This website would enhance my effectiveness inpurchasing this product: (Extremely unlikely/likely)
e. For me, a website thatenhances my effectiveness inpurchasing products is: (Not at all /Extremely important)
Perceived Ease of Purchasing
b. Purchasing this product from this
website would be easy. (Extremely unlikely/likely)
e. For me, purchasing products eas//yfroma Web vendor is: (Not at all /Extremely important)
b. Learning how to purchase thisproduct fromthisWeb vendor would be easy. (Stronglydisagree/agree)
e. For me, learninghow to purchase products easily froma Web vendor is: (Not at all /Extremely important)
Product Value
b. Purchasing this product from this
Web vendor would save me money within the next 30 days: (Extremely unlikely/likely)
e. For me, saving money within the next 30 days is: (Not at all /Extremely important)
b. Iwould purchase this product fromthisWeb vendor at a bargain price within the next 30 days: (Extremely unlikely/likely)
e. For me, getting products at bargain prices within the next 30 days is: (Not at all /Extremely important)
Monetary Resources
c. Iexpect to have themoney needed to purchase this product from thisWeb vendor within the next 30 days: (Strongly
disagree/agree)
p. Having themoney needed to purchase products would make it(much more difficult/easier)forme to purchase this
product fromthisWeb vendor.
Ic. Itwould be withinmy budget to purchase this product from thisWeb vendor within the next 30 days: (Strongly
disagree/agree)
p. Being withinmy budget would make it(much more difficult/easier) forme to purchase this product fromthisWeb vendor.
140 2006

theTPB
Perceived Diagnosticity
c. Iexpect thiswebsite to help me get a real feel forthisproduct: (Stronglydisagree/agree)
p. Being able to get a real feel fora productwould make it (much more difficult/easier) forme to purchase this product from
thisWeb vendor,
c. Iexpect thiswebsite to help me carefullyevaluate this product: (Stronglydisagree/agree)
p. Being able to carefullyevaluate a productwould make it(much more difficult/easier)forme to purchase this product from
thisWeb vendor.
Perceived Information Protection
c. Iexpect my personal informationto be adequately protected when Ipurchase this product fromthisWeb vendor:
(Stronglydisagree/agree)
p. An adequate protection ofmy personal information would make it(much more difficult/easier)forme to purchase this
product fromthis vendor,
c. Ifeel secure thatmy personal informationis kept privatewhen Ipurchase thisproduct fromthisWeb vendor: (Strongly
disagree/agree),
p. Feeling secure thatpersonal informationis kept privatewould make it(much more difficult/easier)forme to purchase this
Purchasing Skills
c. If Iwanted to, Icould become skillfulat making good product purchasing decisions on theWeb: (Strongly
disagree/agree),
p. Becoming skillfulat making good purchasing decisions on theWeb would make it(muchmore difficult/easier)forme to
purchase this product fromthisWeb vendor,
c. If Iwanted to, Icould easily become knowledgeable about purchasing products on theWeb: (Stronglydisagree/agree),
p. Becoming knowledgeable about Web purchasing would make it(much more difficult/easier)forme to purchase this
Purchasing Habit
Getting product informationfromthisvendor's website has become a habit forme: (Stronglydisagree/agree)
Getting product informationfromthiswebsite has become natural forme: (Stronglydisagree/agree)
Past Experience - Purchasing
During the lastyear, how many times have you made product purchases fromthe Internetingeneral?_times.
During the lastyear, how much have you approximately spent on Internetpurchases? $_.
During the lastyear, how many times have you made product purchases fromthe selected Web vendor? _times.
Web Vendor Reputation
This Web vendor has a good reputation in themarketplace: (Stronglydisagree/agree)
LEGEND: b: attitudinalbelief strength

e: belief power (outcome evaluation)
c: control belief strength
p: belief power (perceived facilitation)

2006 141

o c: 9^ OCD
1PLS
Confirmatory
Factor
for
Analysis
Discriminant
and
Convergent
Validity
BM_______________________________I
s
^|GET1
^9"~1Q
H"~08"~ToT~T3
-13"
"^
~**~
~^
?19"~04'
~^"
~^~
-10
~~^~
16
05
11
03
01
34
24
12
O^<C0C0OHfl-C_HOZC0____<C0C/)Ol-____:>______tt>
ITEMS
o
.38
.15
.46
.30
.25
.33
.32
.35
INT1
.24
.27
.31
.32
.19
.16
-.03
.29
.14
.21
-.03
.92
.06
.09
.23
.25
.10
.05
.25
.20
.15
.12
.16
SN1
.22
.14
.17
.10
.02
-.04
.19
.34
.12
.08
.10
.13
.11
.81
.14
.04
.03
.12 .12
.04
.15
.07
.25
.19
.28
.35
SK1
.33
.49
.39
.36
-.03
.05
.04
.09
.08
.13
.17
.45
.14
.05
.20
.10
.08
.92
|I|I[SK2
|I|-.02
I.02
|-06
|.30
.28
I.25
.33
|.12
09
.34
.39
.27
|.40
.02
|.12
.11
.10
.13
.03
.14
.02
.40
.94
INT2
.36
.88
.39
.11
.29
.32
.34
.30
.30ATT1
.35
.19.35
.26
.25.94
.18.28
.22.40
.40.30
.33.44
-.04.37
.11
.19.44
.32
.20.30
.41
.17.25
.05.13
.03
.10.26
ATT2
.33
.20.44
.38
.20.14
.93
.01.31
.20
.23
.31.21
.32.30
.41.26
.39.15
.39.17
.33.25
.29
.37.21
.36.14
.17
.29
.46
.10
.16
.18
.20
.22SN2
.01
.28.06
.21.32
.25.76
.20.10
.17.14
.18.18
.19
.15
.10
.11
.05
.15
.04
.06
zSE1
.14.10
.33.40
.33.07
.16.04
.90.05
.47.10
.30.05
.33.06
.40.09
.42.05
.45
.37
O .08
SE2
.30
.10.05
.27
.20
.38
.12
.09
.05
.94
.49
.39
.47
.26
.35
.23
.38
.22
.40
.30
.38
.41
.12
.19
.35
.20
.08
.12
.15
.21
.17
.07
.30
.44
.28
.26
.25
.17
.12
.22
.08
.18 2TR2
.11
.21
.30
.24
.33
.25
.89
.48
.37
.33
.41
.35
.37
.06
.10
ZPU1
.17.20
.30.06
.39.16
.10
.11
.30.45
.20.32
.42.28
.87.16
.48.17
.36.14
.40
.39.12
.38
PU2
.13.09
.10
.16
.18
.27.05
.45.07
.12
.20
.32
.29.40
.45.38
.91.13
.44Z
.18
PE1
.13
.39.21
.28
.43
.37.09
.28
.40.17
.20
.05.39
.13.39
.12.36
.04.45
.10
.15.90
.30
.39
.44
.30.35
.33.07
.20
H
.15
PE2.16
.16
.15.16
.11
.33.05
.13
.40.25
.30
.25
.30.23
.30.25
.33.40
.40.20
.89
.32.15
.33
.40.11
.05
-.04
.36
.05
.20
.14
.08
.19
.20
.28
.21
.42
.12
OTIM2
.16
.09
.15
.12
.29
.02
.14
.30
.33
.36
.41
.29
.95
.39
.35
DEL1
.42
.18
-.01
.20
.13
.33
.12
.16
.04
.40
.16
.40
.12
.13
.38
.36
.37
.19
.36
.11
.95
.49
.07
.15
.40
.08
.10
.11
.12
.07
.21
DEL2
.16
.22.11
.36.21
.25
.09
.38.12
.31.18
.39.15
.38.06
.31NAV1
.04
.09
.40.13
.19
.94
.47.06
.42
.36.15
.06.35
.08.18
.23.41
.09.37
.19.40
.20.42
.14.45
.90
.25NAV2
.39
.06
.20.04
.24
.09.37
.10
.10.18
.22
.08.40
.10
-.04
.19.32
.15
.12
.36
.41
.20
.38
.21
.45
.13
.43
.88
.10
.06
.37
-.01
.10
.06
.13
.24
.20
.11
.21
.15
.17
.14
.18
.12
.09
.05
.10
.24
I-
CON1
.19
.31
.29
.12
.40
.84
.39
.26
.34
.29
.44
.30
.38
.10
_5.24
.17
CON2
.16.08
.27
.36.41
.30
.18
.45.17
.80.20
.35.17
.32.12
.30.22
.33.14
.33
.28
0_ .10
TR1
.29
.09.10
.15
.25
.20
.41
.26
.20
.12
.28
.30
.29
.37
.88
.42
.40
.15
.21
.29
.19
.39
.40
.13
.20
.29
.05
.04
.05
.15
.24
.10
.20
.16
.40
.30
.32
.20
.15
.10
.14
.11 ?TIM1
.08
.23
.25
.14
.29
.25
.30
.35
.34
.93
.38
.40
.47
.03
.15
.15
.07
.15
.10
.13
.16
.14
.10
.19
.21
.15
.14
?I?I?i?I?i?I?I?I?i?i?I?I-i?I?I?I?I?I?I?I?I?m?i?i?
i
^tl=ZmOrr3ml-J<^3^HZ_jOc_3LiJ<0<Q___
itcmc
o
?Getting
Information
Purchasing
&>
IiBAppendix
^,-,-,-,
2fe Ii

^-
ctT s CD 00
W
-13 -13 -10 ~QQ~ ?6 ?5

"01~--05
-04 .40"
.47 .12 .30 .36 .09 "~10~ .18 "~04~ .15 IT .14 ToT"
^ TIM:
Ease
Use;
Time
Download
DEL:
SK:
Perceived
Navigability;
of
Resources;
NAV:
Delay;
Skills;
PUR:
^Consumer
VAL:
Purchasing;
MON:
Value;
Product
Monetary
To~
INT1
.25
.30
.23
.09
.18
.20
.11
.15
.18
.10
.13
.03
.48
.98
.49
.32
.30
.35
.24
.30
.33
.36
.40
.23
.16
.26
ATT1
.24
.31
.38
.10
.11
.16
.21
.17ATT2
.15.30
.16
.24
.20.37
.08
.39.32
.47
.91.02
.36
.35.15
.37.20
.30
.33.20
.35.20
.17
.35
.41.13
.19
.29.26
.20
.27.11
.44
.48
.93
.39
.30
.38
.31
.29
.32
.26
.33
.32
.28
.17 SE1 .16
.11
.25
.14
.36
.32
.15
.16
.26
.11
.15
.20
.10
.28
.27
.32
.28
.98
SE2
.18.47
.36
.41
.13.40
.16
.10.30
.28
.35.25
.33.22
.29
.12
.12
.23
.15
.12
.23
.16
.29
.36
.31
.92
.44
.35
CON1
.12
.20.34
.36
.25
.06.25
.33
.30.26
.30.35
.18
.10
.28CON2
.14
.20.16
.19.18
.14
.28.22
.30
.35.09
.30
.51.29
.93.34
.30.17
.40
.35.16
.22
.30
.35.12
.18
.36.17
.30
.35.17
.31
.39
.34
.48
.92
.32
.42
0TR1
.29
.11
.15
.19
.28
.03
.31
.20
.24
.38
.42
.32
.12
.39
.19
.10
.15
.24
.15
.13
.15
.29
.33
.37
.33
.91
.46
.40
.30
.51
.36
.45
.43 1PU1
.18
.16
.26
.10
.20
.17
.33
.35
.26
.17
.20
.18
.09
.30
.36
.27
.39
.33
.40
.40
?.90
PU2
.30
.46
.25
.19
.10
.21
.25
.11
.44
.28
.23
.37
.30
.22
.20
.17
.18
.12
.15
.29
.333
PE1
.10
.22
.35.05
.29
.36
.41.08
.25
.35
.88
.31
.50
.29.15
.30
.28
.35
.14
.41.21
.10
.23
.14
.27
.25
.30
.40
.41
.45
.47
.39
.33
.95
.30
.44
.46 VAL1
.12
.08
.12
.12
.15
.10
.22
.15
.19
.15
.09
.14
VAL2
.07
.11
.13
.35
.04
.34
.30
.23
.26
.33
.09
.18
.32
.20
.17
.19
.89
.25
.10
.23
.20
.18
.09
.08
.11
.03
MON1
.10
.18
.34
.32
.06
.35
.30
.12
.35
.00
.26
.12
.30
.20
.12
.87
.23
.17
.20
.14
.16
.24
.11
.09
.05
.10
.04
.38
.39
.36
.26
.30
.30
.36
.35
.30
.96
.30
.20
.40 | PR1
.19
.17
.14
.02
.10
.09
.18
.10
^.06
.14
.07
.05
.09
PR2
.15
.22
.23
.30
.19
.20
.24
.19
.27
.03
.38
.40
.07
.28
.12
.13
.22
.19
.14
.30
?.09
.95
.30
.17
J
.11
.02
.15
SK1
.11
.05
.15
.19
-.06
.15
.27
.09
.31
.06
.43
.42
.25
.31
.10
.09
.17
.15
.19
.14
.03
.90
.33
.10
?.23
g.43
.00
.28
_.25
.20
.33
.40
SK2
.08.40
.19
.03.47
.45
.08
.05.35
.16.40
.03.90
.14?.23
.16
.01
.08
.20
.46
.05
.23
.21
.22
.31
.37
.44
.45
.40
.25
.31
.46
.29
.84
Oi
itcmc
wti=zwOr?3wlili<*D^tzwOor:=>iJLJ<Q<or*:
ITEMS INT2
.17
.35
.18
.05
.16
.25
.06
.11
.21
.16
.11
-.01
.42
.96
.46
.29
.25
.29
.24
.19
.26
.32
.36
.26
.20
.21 SN1
O^<a)C0OHQ-Q-l-QZa)Q-^<C0a)Ol-CLCL>2QQ.C0 -.02
.00
.12
.37
.04
.10
.08
-.03
.04SN2
.08
.09.04
.06.02
.02
.10
.30.07
.35.32
-.02
.84
.30.15
.28.07
.35
.37.00
.26.01
.33
.39.05
.14
.12
.10
.21-.03
.06
.35
.36
.80
.27
.32
.26
.22
.31
.31
.35
.20
.18
.24 5TR2
.09
.13
.20
-.04
.29
.20
.47
.11
.21
.15
.13
.15
JJ
.12
TR3
.15
.19
.14
.26
.13
.31
.33
.16
.26
.89
.06
.35
.26
.29
.47
.39
.21
.39
.40
.36
.14
.15
.15
.23
.13
.16
.10
.20
.30
.24
.32
.29
.46
.79
.42
.30
.26
.41
.35
.44 ?"
PE2
.14
.03
.22
.06
.24
.25
.28
.34
.32
.17
.16
.18
.13
.07
.32
.31
.27
.42
.38
.43
.45
.91
.35
.29
.35
.29
.43 5MON2
.13
.09
.15
.01
.17
.18O
.18
.19DIA1
.18.26
.19
.10
.06.10
.06
.10.21
.40
.37.10
.25
.33.16
.36.10
.40.09
.32.20
.20.10
.92.16
.13
.10
?.25
.17.12
.07
.25
.28.33
.14
.26
.40
.39
DIA2
.20
.12.37
.39
.25
.07
.20.29
.15.86
.23
.18.44
.15
.12
.20
.08
.14
.09
.10
.20
.35
.18
.23
.33
.35
.38
.37
.25
.27
.90
.17
.45
^ Legend:
GET:
Getting
Information;
INT:
Intention;
ATT:
Attitude;
SN:
Subjective
Self-Efficacy;
SE:
Norm; CON:
Controllability;
TR:Trust;
PU:
Perceive
Use
Getting Purchasing
Information
^ Resources;
DIA:
ProductPR: Information
Protection
S1
Diagnosticity;
CQ
0>
1 ^ I8 3O


Extending the Theory of Planned Behavior to Predict Electronic Commerce Adoption

Uploaded by

Document Information

Original Title

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

Extending the Theory of Planned Behavior to Predict Electronic Commerce Adoption

Uploaded by

Copyright:

Available Formats

Understanding and Predicting Electronic Commerce Adoption: An Extension of the Theory of

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

Understanding and Predicting Electronic

Anderson Graduate School of behavior.

University of California, Riverside

Abstract of planned behavioral

materialize, consumers must first adopt online activities, such as

MIS Quarterly Vol. 30 No. 1, pp. 115-143/March 2006 115

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

1. of B2C e-commerce is not viewed as a monolithic

underlying dimensions {self-efficacyand controllability),

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

purchase, and post-purchase interactions. Prepurchase interaction are to obstacles,

product purchasing?that largely determine e-commerce adoption.3

Behavioral Intentions and

MIS QuarterlyVol. 30 No. 1/March

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

_ Subjective Intention Behavior

Figure 1. The Theory of Planned Behavior (Adapted from Ajzen 1991)

Connecting Getting Information

product purchasing is contingent upon getting information. This

getting information facilitates purchases. For example, Kim and

To link behavioral intentions between

118 MIS Quarterly Vol. 30 No. 1/March 2006

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

Underlying Dimensions of PBC

Self-Efficacy: Following Bandura (1986), we define SE as

Perceived Behavioral Control behavior. Applied to e-commerce, SE describes consumers'

judgments of their own capabilities to get product information

PBC is defined as a person's perception of how easy or difficult it

behavior toward getting product information from

MIS QuarterlyVol. 30 No. 1/March

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

Normative -^ Subjective -^^^fc Intentions Behavior

Figure 2. The Proposed Extension of the Theory of Planned Behavior

TPB can aggregate beliefs to create measures of attitude, SN, and

We conducted a belief elicitation study using an

120 MIS QuarterlyVol. 30 No. 1/March2006

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

Table 1. Questionnaire for Eliciting External Salient Beliefs

Table 2. Frequency of Elicited Beliefs (Getting Information)

MIS QuarterlyVol. 30 No. 1/March

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

Table 3. Frequency of Elicited Beliefs (Purchasing)

Download Delay ' \. V. /

I-1 ^^ PBC M' j

EXTERNAL BELIEFS (GETTING INFO) PERCEIVED BEHAVIORAL CONTROL

Trust- Product Purchasing

Figure 3. The Proposed Research Model

122 MIS Quarterly Vol. 30 No. 1/March 2006

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

2003). Competence is thebelief in the trustee'sability toperform

(Choudhury et al. 2001). Therefore, competence and integrity are

Trust is also placed in the nomological structure of the TPB as a

MIS Quarterly Vol. 30 No. 1/March 2006 123

This content downloaded on Mon, 14 Jan 2013 17:03:39 PM

getting product information from that Web vendor.

product purchasing from that Web vendor.

Perceived Usefulness and purchasingcompletelyup

PU is the extentto which one believes that using a system will

and Todd 1995b). Therefore, the following hypotheses are

Leisure time has been considered a critical resource for getting

124 MIS QuarterlyVol. 30 No. 1/March