302 Int. J. Electronic Business, Vol. 8, Nos. 4/5, 2010
Copyright © 2010 Inderscience Enterprises Ltd.
The effects of website design on purchase intention
in online shopping: the mediating role of trust
and the moderating role of culture
Chandragupt Institute of Management,
Hindi Bhawan, Chajjubagh, Patna 800001, India
Satya Bhusan Dash
Indian Institute of Management,
off Sitapur Road, Prabandh Nagar,
Lucknow 226013, India
Simon Fraser University,
Central City 250 – 13450 102nd Avenue Surrey,
BC Canada V3T 0A3
DeGroote School of Business McMaster University Hamilton,
Ontario, Canada L8S 4M4
Abstract: Lack of trust in online transactions has been cited as the main reason
for the abhorrence of online shopping. We have tested the mediating role
of trust in online transactions to provide empirical evidence that trust in the
online store represents the generic mechanism through which the focal
independent variables of website design are able to positively influence
purchase intention and reduce the perceived risk. We have further demonstrated
the moderating effect of the individual’s culture in e-commerce and thereby
offered insights into the relative importance of website design factors
contributing to trust for customers of different cultural values.
Keywords: trust; online purchase intention; perceived risk; online shopping;
moderators; mediators; website design; culture; e-commerce; B2C e-commerce.
The effects of website design on purchase intention in online shopping 303
Reference to this paper should be made as follows: Ganguly, B., Dash, S.B.,
Cyr, D. and Head, M. (2010) ‘The effects of website design on purchase
intention in online shopping: the mediating role of trust and the moderating role
of culture’, Int. J. Electronic Business, Vol. 8, Nos. 4/5, pp.302–330.
Biographical notes: Boudhayan Ganguly is an Assistant Professor of
Management Information Systems at Chandragupt Institute of Management
Patna. He has been conferred “Fellow of the Indian Institute of Management
Lucknow” in the year 2010. His area of research is focused on trust building
in e-commerce, cross cultural and gender issues in the context of online
consumer behavior and e-governance. He has also published a paper in
Journal of Information Science and Technology.
Satya Bhusan Dash is currently working as an Assistant Professor of Marketing
in Indian Institute of Management Lucknow. He has published in several major
national and international journals including Academy of Marketing Science
Review, Journal of International Consumer Marketing, Marketing Intelligence
and Planning, International Journal of Bank Marketing, Online Information
Review and Journal of Information Science and Technology. He has also
coauthored a book titled “Marketing Research: An applied orientation”.
His current areas of teaching and research interest have been in relationship
marketing, buyer behaviour in insurance and pharmaceutical industries, rural
marketing, lifestyle influences in food retailing and e-business.
Dianne Cyr is a Professor in the Faculty of Business at Simon Fraser University
in Vancouver. Her research is focused on how trust, satisfaction, and loyalty
are built in online business environments through website design. More specific
consideration of these topics is related to cross-cultural and gender issues.
She is the author of five books and over 85 research articles. Select publications
appear in MIS Quarterly (Best Paper 2009), Journal of Management
Information Systems, Information and Management, and Journal of the
American Society for Information Science and Technology. Website:
Milena Head is a Professor of Information Systems at the DeGroote School of
Business, McMaster University, Canada. Specialising in eBusiness and Human
Computer Interaction, she has published over 70 papers in academic journals,
books, and conferences. Her research interests include trust and adoption
in electronic commerce, interface design, mobile commerce, identity theft,
cross-cultural issues in electronic commerce and human computer interaction,
e-retailing, and web navigation. Further details on her work can be found at
Over the years the evolution of the internet as a marketing medium has become a global
phenomenon, leading to a rapid escalation of e-commerce in the past decade. The rise in
the number of households possessing computers and the ease of internet access has
led to this widespread acceptance of B2C e-commerce. According to Jupiter Corporation,
e-commerce in the USA will reach $144 billion by 2010. The penetration of e-commerce
is also quite high in the developing economies of Asia. However e-commerce as a
percentage of internet penetration continues to be very small. An I-Cube Report (2008)
304 B. Ganguly et al.
shows that in 2008 the percentage of internet users who use the internet for
e-commerce is only 7%. According to Tan and Guo (2005) the internet is viewed
by customers as a world of chaos. Purchase is made only if the benefits are more than
the risks. According to Grabner-Krauter and Kaluscha (2003) lack of trust is cited
as the main reason for not shopping online. Hence, establishing trust in online shopping
is one of the most crucial factors for success in online business environments. Academic
researchers are eager to discover important website design factors that develop trust
in online shopping. Past studies such as Cyr (2008) or Yoon (2002) provided empirical
evidence on how various design factors build trust in the context of B2C ecommerce.
However, there has been no consensus among researchers on the constituents for website
design factors. In this study we have carried an extensive review of the literature
on website design in order to determine a comprehensive model on website design
and its influence on trust.
Culture has been defined in different ways by various scholars. It can be loosely
defined as ‘group psychographics’. From past studies (Marcus and Gould, 2000;
Cyr, 2008), it has been noted that online companies are increasingly operating in multiple
countries, each having a different cultural orientation. Past studies by Singh et al. (2005a,
2006) revealed that customers from different cultural backgrounds place different
emphasis on various aspects of the website. We further note that none of the scholarly
works have comprehensively studied the influence of culture on the website design, trust
and the consequences of trust in online shopping. Although, this issue was partially
addressed by Cyr (2008), even in her study country was used as a surrogate for culture.
In the current study we discuss the limitations of using country as a surrogate for culture
and advocate for measurement of the cultural values of the individual participants.
We further, conceptualise that trust is the generic mechanism through which the
website design factors affect purchase intention and perceived risk. In this paper we
propose that the website design factors affect trust and trust, in turn, increase purchase
intention and reduce perceived risk. We note that none of the scholarly works have tested
the mediating role of trust in relationship between design factors and purchase intention
in online shopping, although Sultan et al. (2005) tested the mediating role of trust with
some limited antecedent factors such as user and website characteristics.
The objectives of this study are as follows. Firstly, to review past literature and
identify and empirically validate the website design factors that affect trust in the online
store; secondly, to test the role of cultural variables in the relationship between website
design factors and trust, and also between trust and its consequences; and thirdly, to test
the role of trust in the relationship between the website design factors and purchase
Our paper is divided into five parts. After the introductory Section 1, we review the
relevant literature related to trust, website design, and culture in Section 2 and propose a
model depicting the major website design factors as antecedents of online trust and also
the consequences of online trust. Following the discussion, we propose a number of
research hypotheses that hypothesise how the cultural values would affect the
relationship between trust and website design factors and also between trust and its
consequences. In Section 3, we discuss the methodology that has been used in the study.
In Section 4, we present the results and findings from our study. In the final section we
discuss the theoretical contributions, managerial implications of the empirical model and
provide suggestions for future research.
The effects of website design on purchase intention in online shopping 305
2 Theoretical background and research model
2.1 Website design
The quality of website design is very important for any online store to attract customers.
Cho and Park (2001) have found in their study that customer satisfaction in e-commerce
is related to the quality of website design. According to Ranganathan and Grandon
(2002), website design represents the way in which the content is arranged in the website.
Wolfinbarger and Gilly (2003) argued that when customers interact with an online store
they prefer to do so via a technical interface and not through any employee. Therefore the
design of the website, which acts as the interface, would play an important role in
influencing customer satisfaction. Lee and Lin (2005) had empirically found that website
design positively influences overall customer satisfaction and perceived service quality.
Besides, Ranganathan and Ganapathy (2002) have empirically established that website
design positively affects purchase intention.
However, there has been little consensus among the scholars on the factors that
constitute website design. In continuation of the above works, in this study, we
investigate the nature of the website design factors. We believe that website design
factors are too many and the development of the taxonomy of such factors would help us
to better understand how it affects trust in B2C ecommerce. Kim and Lee (2002)
suggested two perspectives for analysing design of a website: process and architecture.
In the process perspective the market transactions are supposed to consist of a number
of processes. The system is regarded as a sequence of processes. The architectural school
of thought, on the other hand, considers the system as a collection of webpage
In the architecture perspective the four components of design are content, structure,
interaction and presentation. Content represents the information that is put up on the
webpage. Structure represents that the way in which the information is arranged; for
example, hierarchical, network etc. According to Park and Kim (2000) interaction
represents the way the user can surf the web pages with maximum ease. The presentation
aspect of the design represents the emotional appeal of the website, like the presence of
visual aids etc.
In the present study we shall follow the architecture perspective as it deals with the
system implementation details. In the online shopping context, Cyr (2008) had classified
design factors as information design, navigation design and visual design. This is
analogous to the architecture perspective of website design. Information design consists
of content and structure of information. Navigation design is the interaction component
and visual design is the presentation component of the website design.
2.2 Trust on the online store
Trust has been defined in different ways by various authors. In our study we limit
ourselves to the domain of B2C online shopping. Online Trust is a multi-dimensional
construct. Authors such as Lee and Turban (2001) and Yoon (2002) had noted that there
is no agreement among the authors on the definition of online trust. A detailed study on
online trust, carried out by Tan and Sutherland (2004), had conceptualised online trust to
be a multi-dimensional construct consisting of institutional trust, dispositional trust and
interpersonal trust. Institutional trust comes from the internet and the concerns related to
306 B. Ganguly et al.
the medium of online shopping. Dispositional trust deals with the individuals’ openness,
agreeableness, neuroticism, conscientiousness, and extroversion. Interpersonal trust deals
with the trust between the two parties doing business. In the context of B2C online
shopping the two participating parties in the business are the customer and the online
vendor and would fall under the category of interpersonal trust. Interpersonal trust
consists of predictability, integrity, credibility, and benevolence. Predictability is
concerned with the vendors reputation of providing a good service, Integrity is the belief
that the online vendor shall be honest and follow standards. We further believe that as
predictability and integrity deals with honesty and consistency and they give rise to
credibility of the online vendor. So, we conceptualise trust on the online store as the
perceived credibility and benevolence of the online store in the eyes of consumer.
Further, this definition has been used in the domain of online shopping by some authors
like Stephens (2004) and Dash and Saji (2007). Credibility refers to the buyer’s belief in
the seller’s expertise to do the job effectively, while benevolence is based on the buyer’s
belief in the positive intention of the seller (Ganesan, 1994).
In the next sub-section we discuss the role of culture in the relation between website
design and trust.
Culture is a multi-dimensional construct. Various authors have defined culture in various
ways. According to Kroeber and Kluckhohn (1952) “Culture consists of patterns, explicit
and implicit, of and for behaviour acquired and transmitted by symbols, constituting the
distinctive achievements of human groups, including their embodiments in artifacts”.
Kluckhohn and Kelly (1945) have said that “culture is historically created designs for
living”. To study the influence of cultural value, we turn to the seminal work of Hofstede
(1980, 1991, 2001). Hofstede defines culture as the “collective programming of the mind
which distinguishes the members of one group or category of people from those of
another” (p.5, 1991). In the most exhaustive cross-cultural study to date, Hofstede (1980,
1991) established five dimensions of national culture. This development of a cultural
dimensions typology is one of the major frameworks for understanding the influence
culture on consumer behaviour in several marketing studies. Masculinity is the degree to
which achievement, competition, assertiveness and performance are emphasised in a
society. Uncertainty avoidance is the extent to which the members of a culture feel
threatened by uncertain or ambiguous situations. Long term orientation refers to the time
orientation of a culture; i.e., whether that culture tends to operate in a long-term or
short-term context. Individualism refers to the ties between individuals. These ties tend to
be relatively loose in individual cultures, and individual interests are emphasised
over group interests. Collectivism, on the other hand, emphasises group welfare.
Power distance refers to the extent to which the less powerful members of the society
expect and accept that power is distributed unequally.
For more than 25 years researchers have relied on the work by Hofstede (1980)
to make meaningful comparisons between national groups. Although there have been
questions regarding the validity of using Hofstede’s findings, the results of his work have
been supported quantitatively and qualitatively by numerous studies in various disciplines
(Sondergaard, 1990; Straub, 1994). Ford et al. conclude:
The effects of website design on purchase intention in online shopping 307
“In spite of criticisms from some quarters about the validity and
generalisability of Hofstede’s results, papers published in leading journals have
established its usefulness in theory development and testing and have found
support for its contributions.” (Ford et al., 2003, p.10)
In keeping with the preceding endorsement, in this research Hofstede’s framework has
been used for this study. Besides, Hofstede’s framework has been applied to investigate
the influence of cultural value on consumer behaviour in the online context in several
past studies in online settings such as Information, Visual, Navigation Design
(Cyr, 2008), Privacy (Singh, 2002; Singh et al., 2005b; Singh and Matsuo, 2004),
Security (Marcus and Gould, 2000).
In cross cultural studies, country has been used as a surrogate for culture by several
past scholars such as Lee et al. (2007), Tan et al. (2007) and Cyr (2008). However, there
is no consensus among authors on the issue as there are arguments for measuring cultural
values at the individual levels. The notion of national culture construct is based on the
postulate that there are larger cultural differences between countries than within
countries. However, scholars such as Ng et al. (2007) argued that culture does not
necessarily correspond to national boundaries, but often follows linguistic, ethnic, or
religious divides. Yoo et al. (2001) had argued that members of a society need not have
the same cultural values. Some individuals may develop ‘unique’ values that are different
from the other members of the society. They argue that the learning opportunities of the
individuals are ‘patchy’ and ‘constrained’ by the social structure. Besides, considering a
nation-state as a surrogate for a culture raised problems because within-country
heterogeneity may be greater than between-culture heterogeneity (Hofstede 1980;
Samiee and Jeong, 1994). Srite and Karahanna (2006) argued that national culture is a
macro-level phenomenon whereas acceptance of a technology such as online shopping is
an individual level decision and, therefore, advocated the measuring of individuals’
cultural values by personality tests. Besides, McCoy et al. (2005) argued that it is
inappropriate to use country scores on cultural dimensions because different individuals
identify with the national culture in varying degrees. Besides, scholars like Srite and
Karahanna (2006), Dash and Saji (2006), Dash et al. (2009) have done moderator
analysis with the individuals’ cultural values and empirically established moderator
effects of culture. Hofstede (2001) cautioned that his cultural metric is not valid for
computing cultural scores for individuals. Realising these issues we did not focus on
individual nation as a surrogate for culture. The cultural value identity of an individual
person in this study were conceptualised at the individual level and not according to his
or her national association.
In the next sub-sections we discuss the influence of these cultural values on the
relationship between website design and trust and also between trust and its
2.4 Hypotheses development
2.4.1 Trust in online store, its consequences and the influence of culture
From the literature the two most commonly identified consequences of trust were
perceived risk and purchase intention. Purchase intention is concerned with the likelihood
to purchase products online. In order to increase the acceptance of e-commerce it is
308 B. Ganguly et al.
indispensable for the consumer to intend to use a retailer’s website to obtain and provide
information in order to complete a transaction by purchasing a product or service.
Purchase intention is the final consequence of a number of cues for the e-commerce
customer. Jarvenpaa and Tractinsky (1999) have argued that a customer’s willingness to
buy from the online store shall increase if the seller is able to evoke the customer’s trust.
Several studies (Gefen et al., 2003; Kim and Kim, 2005; Salam et al., 2005; Suh and Hun,
2003; Sultan et al., 2005) have empirically shown that increase in customer trust on the
online vendor increases purchase intention.
Customers from high individualistic societies have a loosely knit social
framework. According to Singh et al. (2003) personal freedom is valued more in
individualistic cultures and individual decision making is emphasised. In collectivist
societies the individual members form cohesive groups in which each individual
expects the others in the group to look after his interest in exchange of unquestionable
loyalty (Gong et al., 2007). Gong et al. (2007) further argued that in societies that are
high on individualism, the individuals have a penchant to develop and try out new things.
This is because individualistic societies favour more uniqueness and differentiation.
Further, Yaveroglu and Donthu (2002) found that the individuals who are high on
individualism are also high on innovativeness. So, customers high on individualism
would require more trust from the online vendor in order to remain loyal and keep
purchasing from him.
H1a: Higher perception of customer trust in the online store results in higher
H1b: Collectivism negatively moderates the relationship between trust and purchase
Perceived risk has been defined by Chellappa (2005) as the uncertainty that the customers
face when they cannot foresee the consequences of their purchase decisions. As the
internet is a virtual and global channel for buying and selling goods the seller cannot be
physically felt it creates a perception of uncertainty in online transactions and,
therefore, perceived risk in online shopping is high. Jarvenpaa and Tractinsky (1999)
argued that there is no assurance that the customer will get what he sees on the internet.
If there are technical problems during transactions, then the seller is not bound to bear the
expenses. According to Yoon (2002), online trust is different from offline trust in three
ways. First of all there is a huge distance between the buyer and seller; secondly, the
absence of a sales person and thirdly there is no physical contact between the buyer and
the product. So trust with the online vendor is indispensable for reducing perceived risk.
Besides, it has been empirically shown by several studies (Jarvenpaa and Tractinsky,
1999; Pavlou, 2003; Harridge-March, 2006) that trust with the online vendor reduces
Customers from high uncertainty avoidance societies have less tolerance for
uncertain and ambiguous situations. Tan and Guo (2005) noted that the internet
is viewed as a world of chaos, so we believe that customers who are high on uncertainty
avoidance have higher perceived risk from using the internet. Nath and Murthy (2004)
found empirically that customers who are high on uncertainty avoidance are risk averse
and are resistant to the use of the internet. Similarly, Lim et al. (2004) prescribed
that those countries that are high on uncertainty avoidance need to stress on trust
The effects of website design on purchase intention in online shopping 309
in online stores to reduce the perceived risk associated with online transactions.
So the online stores need to create higher trustworthiness in the minds of high
uncertainty-avoidance-customers in order to overcome perceived risk. Thus, we propose
H2a: Higher perception of customer trust in the online store results in lower
perception of perceived risk in online shopping.
H2b: Uncertainty avoidance positively moderates the relationship between trust and
Perceived risk and purchase intention
For online customers increased level of perceived risk is likely to reduce purchase
intention. Consumers are reluctant to provide information on the internet because they
fear their private information may be misused by some unauthorised person. Jarvenpaa
and Tractinsky (1999) argued that a customer may be willing to buy from an online store
if it is perceived to be of low risk even if he does not have a highly positive attitude
towards the store. Choi and Lee (2003) and Jarvenpaa and Tractinsky (1999) have
empirically shown that increased level of perceived risk reduces purchase intention.
Thus, we propose the following.
H3: Higher perception of perceived risk in online shopping results in lower purchase
2.4.2 Trust in online store, website design factors as its antecedents
and the influence of culture
Website design is concerned with how information is put up on the website.
It is primarily concerned with the information that is put up on the site, the aesthetic
beauty of the website, the ease of navigation through the site and the time taken for
navigation for improving usability. The relationship of the website design factors:
information design, visual design, and navigation design with trust are discussed
Culture, information design and trust
Information design deals with the information that is placed on the site and how the
information is organised. Park and Stoel (2005) had argued that more information on the
website leads to higher purchase intention. Kim and Eom (2002) had found that
information about the firm, its products and services, promotions etc. positively affect
customer satisfaction. Mithas et al. (2006) have found empirically that information
content generates loyalty with online customers if the information is accurate, relevant
and current. Further, Corritore et al. (2005) have argued that if the customer gets the
relevant information in the website then the trustworthiness of the site increases.
Furthermore, Cyr (2008) has provided empirical evidence that information design had
strong positive effect on trust.
There have been a number of studies that have looked into the cultural values of
customers and their impact on website acceptance. Tai and Chan (2001), in the domain
of advertising, found empirically that customers high on masculinity give more
preference to information cues for evaluating the quality and performance of products.
310 B. Ganguly et al.
According to Singh et al. (2004), Japanese sites show masculine themes like online
games and realism. Singh (2002) argued that masculine customers would emphasise
sportiness and ambition. Further, there have to be decision making aids available on the
site like explicit comparisons of products. It was pointed out by Newman and Nollen
(1996) that masculine cultures have been viewed as “doing and acquiring rather than
thinking and observing”. According to Ranganathan and Ganapathy (2002) information is
more concerned with comparing alternatives, having more information about the firm,
and having decision making aids on the site. These features are more concerned with
‘doing’ and ‘acquiring’ rather than ‘observing’. This is a masculine characteristic as per
our discussion, so a customer from a masculine culture will pay more importance to
information design. Further, since there is very little difference in Hofstede’s scores
between Canada (52), India (56), and the USA (61) we believe that there will not be any
significant difference in perception of information design between the Canadians, Indians
and Americans. Thus, we propose the following
H4a: Higher perception of information design in the website results in higher
customer trust in the online store.
H4b: Masculinity positively moderates the relationship between information design
Culture, navigation design and trust
Navigation Design of a website is concerned with the navigation of the website.
Navigational schemes help the users to browse the site with ease. Cyr (2008) argued that
even if detailed information is put on the site the customer may leave the site if he finds it
difficult to search for the information he wants. Harridge-March (2006) had argued that
proper navigation helps the customer to save time and overcome financial and
performance risks and, therefore, leads to trust. Besides, Lim and Dubinsky (2004) had
shown empirically that navigation properties of the website generate a favourable attitude
towards the online store. Yoon (2002) has empirically shown that navigation design
results in trust.
Singh (2002) had argued that guided navigation will reduce uncertainty for customers
who are high on uncertainty avoidance. Marcus and Gould (2000), and Singh et al.
(2005a) had argued that customers from high uncertainty avoidance cultures need to be
given better navigational schemes so that they do not get lost in the website. Besides, Cyr
(2008) found empirical evidence that customers who are high on uncertainty avoidance
give more preference to navigational design for generating trust. Thus, we propose the
H5a: Higher perception of navigation design in the website results in higher
customer trust in the online store.
H5b: Uncertainty avoidance positively moderates the relationship between
navigational design and trust.
Culture, visual design and trust
Visual design of the website deals with the aesthetic beauty and the emotional
appeal of the website. It is concerned with the graphical aspects of the website.
This includes the use of graphics, colours, photographs, and various font types to improve
the look and feel of the site. Karvonen (2000) had shown that ‘aesthetic beauty’ of a
The effects of website design on purchase intention in online shopping 311
website positively affects trust. Cyr (2008) argued that visual design gives
‘overall enjoyment’ to the user because it beautifies the look and feel of the website.
So, we believe that visual design is a key element that represents website usability.
Thus, improving the visual design of a site should result in better usability of the website
and this, in turn, would reduce ambiguities and increase the trustworthiness of the
online vendor. Besides, Cyr (2008) also empirically established that the visual design of
the website positively affects trust.
Sun (2001) argued that users from collectivist cultures such as China have a strong
preference for visuals, whereas users who are more individualistic prefer a logical and
structured page layout. Customers who are high on individualism avoid getting into
groups and a more logical flow of information would allow them to perform the
transaction on their own, without asking for any help from anybody. Besides, Cyr (2008)
found empirically that visual design resulted in trust for users from China, which is high
on collectivism, but not for users from Germany and Canada who are high on
individualism. So, visual design as an antecedent to trust should be preferred more by
collectivist customers. Hence, we hypothesise
H6a: Higher perception of visual design in the website results in higher customer
trust in the online store.
H6b: Collectivism positively moderates the relationship between Visual design and
We present the conceptual models derived from the above hypotheses in Figure 1.
Figure 1 Customers’ cultural values and its moderating effects on the relation between the
website design factors and trust, between trust and perceived risk and trust and purchase
intention in online stores
312 B. Ganguly et al.
2.4.3 Trust as a mediator
Some past studies provided empirical evidence that website design factors (Ranganathan
and Ganapathy, 2002) develop purchase intention directly. However, there have been a
plethora of papers (Yoon, 2002; Dash and Saji, 2006; Chen and Barns, 2007) that
conceptualised and empirically verified that the antecedent factors generate trust, and
trust, in turn, generates purchase intention. Baron and Kenny (1986) in their seminal
paper defined that a mediator variable is a variable that represents the generic mechanism
through which the focal independent variables are able to positively influence the
outcome variable. Further, in the context of relationship marketing, Morgan and Hunt
(1994) had suggested and empirically established that trust would mediate the
relationship between commitment and its antecedents such as communication and
opportunistic behaviour. Auh (2005), in the context of service marketing, had divided
the attributes that generate loyalty into soft and hard attributes. Drawing inspiration
from social exchange theory he argued that soft attributes involve more human
interactions like social and relational attributes, whereas hard attributes are related to the
core of the service such as competence, functionality, and reliability. Auh (2005) further
established that trust is a mediator in the relationship between the soft attributes and
We also note that, in the context of online shopping, attributes like information
design, visual design, navigation design are ‘soft’ attributes as they deal with social
and relational attributes such as human contact, warmth, attentiveness, care etc. So we
propose that these soft attributes would influence their outcome variable ‘purchase
intention’ through the key mediating variable, trust. Therefore, in this study
we hypothesise that trust represents the generic mechanism through which these focal
independent variables are able to positively influence purchase intention. Furthermore,
Sultan et al. (2005) had empirically established the mediating role of trust in
online context. In their study trust was established as a mediator between the focal
independent variables (e.g., website characteristics and consumer characteristics) and
purchase intention. In a similar vein, we conceptualised the mediating role of trust in our
H7: Trust in the online store mediates
(a) the positive effect of perceived information design on purchase intention.
(b) the positive effect of perceived visual design on purchase intention.
(c) the positive effect of perceived navigation design on purchase intention.
Similarly we hypothesise that the website design factors are able to reduce perceived risk
by developing trust toward online store. Therefore, we propose the following
H8: Trust in the online store mediates:
(a) the negative effect of perceived information design on perceived risk.
(b) the negative effect of perceived visual design on perceived risk.
(c) the negative effect of perceived navigation design on perceived risk.
The effects of website design on purchase intention in online shopping 313
3.1 Context of study and sampling
A questionnaire was designed to measure trust and website design. The sample consisted
of students chosen randomly from various premier B-schools in India, USA and
Canada. The countries were chosen because it has been established by Hofstede (1980)
that India and western countries such as Canada and the USA differ significantly along
a number of cultural dimensions. In order to comprehend the influence of culture
on the website design factors that generate trust we focus on three of the dimensions
used by Hofstede – collectivism, uncertainty avoidance and masculinity. From Hofstede’s
research it is clear that Indians are low on uncertainty avoidance, with an index score of
40 compared to the USA (46) and Canada (48). Similarly, the Americans and Canadians
are high on individualism, with scores of 91 and 80, respectively, as compared to
India, which has an index of 48. However, in masculinity there is not much difference
between the countries as the Indians feature moderately high on masculinity with an
index score of 56 and so do Canada (52) and the USA (62). We believe that by using the
data from these countries it would be possible for us to determine whether country could
be used as a surrogate for culture or whether cultural values operate at individual levels.
Choi and Lee (2003), in their study on online shopping behaviour among Americans and
Koreans, argued that a student sample is a very effective and easy way of achieving
sample equivalence. Besides, researchers from Pew Internet and American Life Project
(2008) have revealed that the online shoppers in the USA are primarily from younger age
groups – 26% of online shoppers are in the age group 18–29 and a further 46% are in the
age group 30–49. Similarly, in India most of the internet users and shoppers are from
younger age groups – only 6% of the internet users are men in the age group 36–58
(I-Cube Report, 2008). Further, students are online shoppers from younger age groups
and are heavy users of the internet. Besides, the students from the premier B-schools
of India, USA and Canada have continuous internet access from their institutes and,
hence, served our purpose.
The scales for the constructs were taken from existing literature in the domain of
e-commerce. The scale for measuring Information design, Navigation design and Visual
design were adapted from Cyr (2008). Trust was measured by the scales used by
Chellappa (2005) and Suh and Han (2003). The scales for purchase intention was chosen
from Suh and Han (2003) and that of perceived risk from Chan and Lu (2004). The
cultural constructs masculinity-femininity, collectivism-individualism and uncertainty
avoidance were measured by the scale developed by Yoo et al. (2001). This scale was
empirically tested by several past studies related to culture such as Yoo and Donthu
(2002), Donthu and Yoo (1998), Yoo and Donthu (2001) and Dash et al. (2009). The
scales have been presented in tabular form in Appendix 1. All variables were measured
on a 5-point Likert scale from “strongly disagree (1) to strongly agree (5)”. Data
collection was carried out in two phases. In the first phase the participants were asked
whether they do online shopping and whether they wanted to participate in the survey. In
the next phase, the responses from the affirmative students were considered for the
survey. For collecting data from the Indian, Canadian and American respondents, an
online questionnaire was created and links to the survey were sent randomly to the
students. The questionnaire was distributed to 1200 Indian students randomly chosen
from the affirmative list of students, out of which 305 responses were received. After
314 B. Ganguly et al.
eliminating unfilled and partially filled responses the final sample size was 290.
Similarly, the questionnaire was distributed randomly among the students from B-schools
in Canada and the USA who had consented to participate in the survey. The survey was
distributed to 900 B-school students in Canada and 300 B-school students from the USA.
337 responses were received. After eliminating unfilled and partially filled responses the
final sample size was 292. The sample characteristics are shown in Table 1.
Table 1 The sample characteristics
Sample Indian Western Pooled
Size 290 292 582
Male 240 136 376
Female 50 156 206
Years of internet experience 8.38 9.81 9
Average age 27 21 23.95
Average number of transactions with the portal in the last year 11 5.89 8.37
There was not much variation in the underlying demographics across the two samples
except for gender, where the western sample had a much higher proportion of women
than the Indian one. However, we believe that the psychographic orientations of the
respondents are more important than simply the gender differences. Since the cultural
values of all the respondents were measured at the individual level the ‘gender effect’
was taken care of.
4 Results and data analysis
4.1 The measurement model
Before using the inferential statistical analysis we assessed the validity and reliability
of the constructs. Confirmatory Factor Analysis (CFA) was used to test discriminant
and convergent validity and reliability of the questionnaire items and the constructs.
The results of CFA are presented in Tables 2 and 3.
The fit of the six factor measurement model consisting of online shopping constructs
on a correlation matrix of 19 measures was acceptable:
2 (137) = 422.099 (p < 0.001);
CFI = 0.94; IFI = 0.94; RMSEA = 0.06. Although the
2 statistics is significant
(p < 0.001), other goodness-of-fit indices indicated a good fit. The CFI and IFI of 0.93
satisfied the recommended cut-off criterion. The RMSEA for the model is below
the cut-off criterion of 0.08. Convergent validity is achieved if the loading of each of the
individual items on a construct is greater than 0.5. With the exception of 1 item each from
trust and purchase intention, all other items displayed high convergent validity with
factor loading greater than 0.5. Hence, convergent validity was achieved. In a similar
vein, for the four cultural variables (constructs) the goodness of fit measures such as
2(87) = 182.748 (p < 0.001); CFI = 0.93; IFI = 0.93; RMSEA = 0.06 were acceptable.
Even here, the
2 statistics was significant (p < 0.001), but other goodness-of-fit indices
indicated a good fit.
The effects of website design on purchase intention in online shopping 315
Table 2 The factor loadings and reliability for the predictor and criteria variables (Indian and
Indian sample Western sample
Variable Items Loadings Mean (SD) CR Loadings Mean (SD) CR
ID1 0.698 3.97 (0.806) 0.726 3.94 (0.859) Information design
ID2 0.852 3.88 (0.831)
0.789 4.00 (0.823)
VD1 0.810 3.96 (0.847) 0.817 4.16 (0.814) Visual design
VD2 0.633 3.82 (0.819)
0.770 3.98 (0.859)
ND1 0.824 3.97 (0.907) 0.828 4.10 (0.905)
ND2 0.769 3.97 (0.753) 0.723 4.12 (0.866)
ND3 0.716 3.83 (0.846)
0.541 3.84 (0.815)
T1 0.647 3.78 (0.844) 0.707 3.76 (0.816)
T2 0.722 3.81 (0.823) 0.733 3.79 (0.879)
T3 0.633 3.58 (0.924) 0.620 3.59 (0.855)
T4 0.523 3.69 (0.772) 0.610 3.77 (0.756)
T5* 0.692 3.79 (0.741) 0.484 3.45 (0.713)
T6 0.696 3.60 (0.810) 0.675 3.59 (0.792)
T7 0.795 3.89 (0.770)
0.778 3.84 (0.782)
PR1 0.596 2.60 (1.02) 0.671 2.68 (0.990)
PR2 0.619 2.41 (0.970) 0.781 2.55 (0.916)
PR3 0.652 2.09 (0.885) 0.686 2.27 (0.811)
PR4 0.733 2.20 (0.923)
0.733 2.34 (0.886)
PI1 0.731 4.13 (0.840) 0.800 3.91 (0.843)
PI2 0.795 4.08 (799) 0.790 3.85 (0.860)
PI3* 0.482 3.97 (1.11)
0.608 3.66 (1.06)
*Items with factor loadings less than 0.5 were dropped not considered for computation
Table 3 The factor loadings and reliability for the cultural variables (Indian and Western
Indian sample Western sample
Variable Items Loadings Mean (SD) CR Loadings Mean (SD) CR
MF1 0.541 2.41 (1.2) 0.652 2.22 (1.27)
MF2 0.637 2.83 (1.1) 0.727 2.79 (1.15)
MF3 0.822 2.59 (1.107) 0.757 2.61 (1.09)
MF4 0.514 3.17 (1.12)
0.539 3.05 (1.24)
CI1 0.579 3.37 (0.922) 0.653 3.08 (0.958)
CI2 0.619 3.74 (0.785) 0.581 3.51 (0.843)
CI3 0.854 3.52 (0.893) 0.646 3.31 (0.920)
CI4 0.775 3.58 (0.866) 0.729 3.24 (0.936)
CI5 0.630 3.41 (0.861) 0.611 3.09 (0.978)
CI6 0.500 3.14 (0.880)
0.577 2.98 (0.930)
316 B. Ganguly et al.
Table 3 The factor loadings and reliability for the cultural variables (Indian and Western
Indian sample Western sample
Variable Items Loadings Mean (SD) CR Loadings Mean (SD) CR
UA1 0.500 3.40 (0.940) 0.531 3.52 (0.891)
UA2 0.584 3.64 (0.824) 0.780 3.78 (0.791)
UA3 0.745 3.74 (0.830) 0.695 3.74 (0.733)
UA4 0.719 3.76 (0.747) 0.565 3.72 (0.724)
UA5 0.678 3.78 (0.736)
0.671 3.87 (0.722)
The assessment of discriminant validity was conducted for all the correlated constructs.
The correlation matrices for the latent variables presented in Table 4 show that the
correlation coefficient between any two constructs was significantly below unity, which
supports the discriminant validity of the model. However, a stringent criterion for testing
discriminant validity, suggested by Bagozzi and Phillips (1982) is to fix the correlation
between two constructs as 1.0 and then employ a
2 difference test for the constrained
and unconstrained models. A significantly lower
2 value for the model in which
construct correlations are not constrained to unity would indicate that the constructs are
not perfectly correlated and discriminant validity is achieved. Our results indicated that
with an additional degree of freedom there was an increase in
2 value ranging from
92 (with navigation design and visual design constrained) to 818 (with trust and
perceived risk constrained). So our model demonstrated improved model fits when the
constructs were separated and, hence, discriminant validity was achieved.
Table 4 The correlation between the various constructs in the model
Information design 1.00
Visual design 0.567 1.00
Navigation design 0.583 0.690 1.00
Trust 0.436 0.366 0.365 1.00
Perceived risk –0.313 –0.236 –0.228 –0.746 1.00
Purchase intention 0.464 0.359 0.338 0.684 –0.515 1.00
In assessing measurement reliability, Fornell and Larcker (1981) stressed the importance
of reliability of each measure (individual item), and the internal consistency of the
composite reliability of each construct. Composite reliability is calculated as the squared
sum of the individual item loadings divided by the squared sum of loadings plus the sum
of error variances for the measures. The composite reliability of each construct should be
more than 0.6 for measurement reliability. The results in Tables 2 and 3 indicate that
reliability of the measurement scales for the criteria and predictor variables and cultural
variables respectively was achieved.
The effects of website design on purchase intention in online shopping 317
4.2 Measurement invariance
In any cross national study instruments can be compared when cross national
measurement equivalence is achieved. In this study we restricted ourselves to metric
invariance i.e., invariance of factor loadings, which indicates that the respondents from
different countries interpret and respond to the measures in an identical manner (Bagozzi
and Yi, 1988; Steenkamp and Baumgartner, 1998). Measurement invariance was tested
by using a hierarchical ordering of two nested models. The first model tested whether the
pattern of salient and non-salient factor loadings was equal across countries. This is
called configural invariance. Our results indicate that our data fit well with the a-priori
2(274) = 616.224 (p < 0.001), CFI = 0.932, RMSEA = 0.046.
In the second model we tested whether the factor loadings were equal across the countries
or not. This is known as metric invariance. Our results indicate that our data fit well
with the a-priori hypothesised model:
2(287) = 640.343 (p < 0.001), CFI = 0.93,
RMSEA = 0.046. The increase in
2(13) = 24.119 is insignificant at p=.001 level.
Further, there were no changes in the IFI, CFI and RMSEA values, indicating that metric
invariance had been achieved. Similarly, measurement invariance was tested for the
cultural variables. Our results indicate that our data fit well with the a-priori hypothesised
2(174) = 419.220 (p < 0.001), CFI = 0.90, RMSEA = 0.049. In the second model
we tested metric invariance, i.e., whether the factor loadings were equal across the
countries or not. Our results indicate that our data fit well with the a-priori hypothesised
2(186) = 449.938 (p < 0.001), CFI = 0.89, RMSEA = 0.049. The increase
2(12) = 30.718 is significant at p = 0.001 level. However, there were no changes in
the CFI and RMSEA values, indicating that metric invariance had been achieved.
4.3 Results of research hypotheses
After achieving a satisfactory fit in the measurement model, the structural model based
on a path analysis was then estimated. Path analysis using AMOS 4.0 was performed
with trust, perceived risk and purchase intention as the dependent variables and
information design, visual design and navigation design as the independent variables.
The goodness of fit indices were then evaluated to determine if the model could be
considered reliable in testing the hypotheses. The path model (
2(6) = 43.58 (p < 0.001);
IFI = 0.98, CFI = 0.98, RMSEA = 0.05), yielded a reasonable fit to the data. Although the
2 statistics are significant (p < 0.001), the other goodness-of-fit indices also indicated a
good fit. The Comparative Fit Index (CFI) and Incremental Fit Index (IFI) were above
the guideline of 0.90. The RMSEA was also below 0.08. Therefore, the model was
considered fit enough to proceed with further analysis.
The results from the path analysis shown in Table 5 and Figure 2 indicate that
information design (
= 0.308, p < 0.01), visual design (
= 0.121, p < 0.05) and
navigation design (
= 0.102, p < 0.05) are significant predictors of trust in online stores.
Information design of the website was considered to the most important factor for
generating trust, followed by visual design and navigation design. However, the
relationship between visual design, navigation design and trust were found to be
insignificant for the Indian sample. As indicated in Table 5, trust was also found to be a
significant predictor of purchase intention (
= 0.676, p < 0.001) and perceived risk
= –0.746, p < 0.001). Thus, the hypotheses H1a, H2a, H4a, H5a and H6a were
supported, whereas H3a was rejected.
318 B. Ganguly et al.
Figure 2 The structural path model
*P < 0.05; **P < 0.01; ***P < 0.001.
4.4 Test for moderator effects of culture
Next, the significant predictor variables of trust (and also of perceived risk and purchase
intention) were taken into analysis and the moderator effects of culture were verified.
To test the moderator effect of the cultural variables, moderator regression analysis
was performed, as prescribed by Sharma et al. (1981). Stepwise linear regression analysis
was carried out. First, the independent variables were entered. Second, moderators were
entered and the change in R2 was noted. In each of the subsequent steps, the interaction of
the independent variables and moderators were entered and the change in R2 value noted.
The results of the moderator regression analysis with cultural variables as moderators are
presented in Tables 5–7. The proposed moderators masculinity (
= –0.106, p < 0.01)
= 0.069, p < 0.1) and uncertainty avoidance (
= 0.100, p < 0.05) were
found to be significant predictors of trust. The moderator regression analysis results show
that the interaction terms of masculinity and information design (
= 0.463, p < 0.05),
and navigation design and uncertainty avoidance (
= 0.656, p < 0.05), were found to be
significant predictors of trust, indicating that masculinity is a quasi moderator in the
relationship between information design and trust, whereas uncertainty avoidance is a
quasi moderator between navigation design and trust. Similarly, uncertainty avoidance
was a significant predictor of purchase intention (
= 0.088, p < 0.01). Our results
indicate that the interaction of trust and collectivism is significant with a negative effect
= –0.588, p < 0.01) on purchase intention. Collectivism negatively moderates the
relationship between trust and purchase intention. Thus, the hypotheses H1b, H4b, and
H5b were supported and H2b, H6b were rejected.
Table 5 Moderator regression analysis between website design and trust with culture
Independent variable R2 R
2 change Standardised coefficient
0.215 – 0.308***,
0.240 0.025 –0.106***
Information design × Masculinity 0.245 0.005 0.463**
Visual design × Collectivism 0.245 0.000 –0.048
Navigation design × Uncertainty avoidance 0.250 0.005 0.656**
*P < 0.05; **P < 0.01; ***P < 0.001.
The effects of website design on purchase intention in online shopping 319
Table 6 Moderator regression analysis between trust and perceived risk with culture
Independent variable R2 R
2 change Standardised coefficient
Trust 0.557 – –0.746***
Uncertainty avoidance 0.558 0.001 0.014
Trust × Uncertainty avoidance 0.558 0.000 0.294
*P < 0.05; **P < 0.01; ***P < 0.001.
Table 7 Moderator regression analysis between trust and purchase intention with culture
Independent variable R2 R
2 change Standardised coefficient
Trust 0.467 – 0.684***
Collectivism 0.468 0.001 0.033
Trust × Collectivism 0.476 0.008 –0.580**
*P < 0.05; **P < 0.01; ***P < 0.001.
We checked for country level variations of cultural values and summarised the results in
Table 8. Table 8 shows that, except for collectivism, there is no significant difference
in the cultural values between the Indian and the Western samples. The mean score
for collectivism among the Indians was 3.46 and that for the western sample was 3.2.
This suggests that cultural variations exist at the individual level although at the country
level the cultural variations may be less. This result is in agreement with the findings of
McSweeney (2002) who stated that
“If a national culture were common to all national individuals then there would
not have been significant intra-country differences in individuals’ responses.
But the IBM survey responses within each country were characterised by
This result shows that country should not be used as a surrogate for culture.
Table 8 The cultural scores of Indian and Western samples
Cultural measure India Canada and USA T value
Uncertainty avoidance 3.66 3.73 –1.27
Collectivism 3.46 3.20 4.88***
Masculinity 2.75 2.67 1.14
***Indicates significance at p < 0.001 level.
To determine whether country can be used as a surrogate for culture we carried out path
analysis for each sample. As the western sample had a greater proportion of women
than the Indian sample we wanted to ensure that the results of the country level analysis
is not skewed by gender effects. So, we investigated the path coefficients for each gender.
Our results showed that there is no difference in path coefficients between the two
genders. Our results in Table 9 show that the westerners gave more importance to trust to
develop purchase intention (
= 0.793, p < 0.01) compared to the Indians (
320 B. Ganguly et al.
p < 0.01). However, we found that the westerners gave less importance to trust to reduce
perceived risk (
= –0.742, p < 0.01) compared to the Indians (
= –0.758, p < 0.01).
Our results indicate that the Indian customers give more emphasis to information design
to generate trust (
= 0.401, p < 0.01) compared to the westerners (
= 0.180, p < 0.1).
Similarly, westerners gave more importance to navigation design as an antecedent of trust
= 0.146, p < 0.1) compared to the Indians (
= 0.069, ns). Further, we found the
westerners gave more importance to visual design as an antecedent of trust (
p < 0.05) compared to the Indians (
= 0.045, ns).
Table 9 Effect of the website design factors on trust
Information design → Trust 0.401*** 0.180* 0.308***
Visual design → Trust 0.045 0.224** 0.121**
Navigation design → Trust 0.069 0.146* 0.102**
Trust → Perceived risk –0.758*** –0.742*** –0.746***
Trust → Purchase intention 0.518*** 0.793*** 0.676***
Perceived Risk → Purchase intention –0.027 0.151** –0.010
*P < 0.05; **P < 0.01; ***P < 0.001.
4.5 Test for mediator effect of trust
In the next step we tested the mediator effect of trust on the relationship between the
website factors and purchase intention/perceived risk. We compared the Direct Effect
Model (D-E-M) when trust and purchase intention /Perceived risk are constrained
(i.e., trust not linked to purchase intention to transact/Perceived risk) with the free model
when the mediating path from trust to intention/Perceived risk was not constrained.
The Direct Effects model (Shown in Figure 4) included the additional direct path from
website factors to intention to transact/ Perceived risk, in addition to the mediating paths
from trust to purchase intention/perceived risk, as shown in Figure 3. The comparison of
the proposed constrained D-E-M model with the free D-E-M model allowed us to test
whether trust fully or partially mediates the effect of the website factors on purchase
intention/perceived risk. To fulfil the condition of full/partial mediation, the effect of the
antecedent variables on the dependent variables (purchase intention and perceived risk)
should be significant in the constrained model and the previous significant effect should
not be significantly/insignificantly reduced in the free model. This procedure of testing
the mediating effect is consistent with the one suggested by Baron and Kenny (1986).
Our review of the conditions for mediation (Baron and Kenny, 1986) suggested that the
mediating effects of trust were indeed present between the website factors (information
design, visual design and navigation design) on purchase intention. First the F-M-M
showed that some of the antecedent variables information design (
= 0.308, p < 0.001),
visual design (
= 0.121, p < 0.05), and navigation design (
= 0.102, p < 0.05) had a
significant direct effect on trust. When the mediating path, through trust, to purchase
intention was constrained (i.e., trust was not linked to purchase intention) the direct
effects of information design (
= 0.31, p < 0.001), visual design (
= 0.27, p < 0.001)
and navigation design (
= 0.23, p < 0.001) on purchase intention was significant. Fourth,
The effects of website design on purchase intention in online shopping 321
the previously direct effects of information design (
= 0.20, p < 0.001), visual design
= 0.16, p < 0.05) and navigation design (
= 0.10, p < 0.05) were significantly reduced
when the mediating path from trust to purchase transact was opened (see Table 10). From
this result we can conclude that trust partially mediates the relationship between
information design, visual design, navigation design and purchase transaction. Therefore
H7a, H7b and H7c are partially supported.
Figure 3 The mediator effect model with website design factors as antecedents and perceived
risk and purchase intention as consequences of trust in online stores
Figure 4 The direct effect model with website design factors as antecedents and purchase
intention (and perceived risk) as consequences of trust in online stores
In order to test H8 which predicted that trust mediates the negative effect of the website
design factors on perceived risk, we followed the same procedure as outlined above.
First, the F-M-M showed that the antecedent variables information design, visual design
and navigation design had a significant effect on trust. Second, the F-M-M also noted that
trust had a significant negative effect on perceived risk. Third, when the mediating path
through trust to perceived risk is constrained (i.e., not linked to perceived risk) the direct
effects of information design (
= –0.31, p < 0.001), visual design (
= –0.24, p < 0.001)
and navigation design (
= –0.23, p < 0.001) were significant. Fourth, the previously
direct effect of information design (
= 0.02), visual design (
= 0.4) and navigation
= –0.05) were not significant when the mediating path from trust to perceived
risk was opened (See Table 11). From this result we can conclude that trust fully
322 B. Ganguly et al.
mediates the relationship between information design, visual design, navigation design
and perceived risk.
To test the mediator role of trust we used the mediator analysis prescribed by Baron
and Kenny (1986). Our results in Tables 10 and 11 show that trust fully mediates the
relationship between the all website design factors and perceived risk and partially
mediates the relationship between the all website design factors and purchase intention.
Therefore hypothesis H7 is partially supported and H8 is fully supported.
We discuss the possible implications of the support and counter support for the
hypotheses in the next section.
Table 10 The mediating effect of trust between website design factors and purchase intention
Path Path via trust Path coefficient
2 (DF) GFI IFI CFI RMSEA
Not constrained 0.16* 19.42 (5) 0.99 0.99 0.99 0.04
Visual design →
Purchase intention Constrained 0.27** 119.37(6) 0.94 0.93 0.93 0.15
Not constrained 0.10* 30.72 (5) 0.98 0.98 0.98 0.06
Navigation design →
Purchase intention Constrained 0.23** 135.63 (6) 0.92 0.92 0.92 0.17
Not constrained 0.20** 8.54 (5) 0.99 0.99 0.99 0.00
Information design →
Purchase intention Constrained 0.31** 100.24 (6) 0.95 0.94 0.94 0.14
*P < 0.05; **P < 0.001.
Table 11 The mediating effect of trust between website design factors and perceived risk
Path Path via trust Path coefficient
2 (DF) GFI IFI CFI RMSEA
Not constrained 0.4 37.06 (5) 0.98 0.98 0.98 0.08
Visual design →
Perceived risk Constrained –0.24* 518.51 (6) 0.83 0.68 0.68 0.36
Not constrained 0.05 36.26 (5) 0.98 0.98 0.98 0.07 Navigation design
→ Perceived risk Constrained –0.23* 520.58 (6) 0.83 0.68 0.68 0.36
Not constrained 0.02 38.75 (5) 0.98 0.98 0.98 0.08 Information design
→ Perceived Risk Constrained –0.31* 491.86 (6) 0.83 0.69 0.69 0.35
*P < 0.001.
5 Theoretical contributions, managerial implications and future directions
We identified that trust in online stores is one of the key obstacles of online transactions.
In order to operate a successful e-business, an online company requires a deep
understanding of how trust is developed and how it affects purchase intention in online
stores. In this study we empirically demonstrated that website design factors, which
constitute the drivers of trust, eventually contribute to online purchase decision. Online
stores should use effective implementation of website design factors such as information
design, visual design and navigation design as marketing tools by which trust in the
website can be created and subsequently, purchase intention can be enhanced. This is in
accord with the work of Sultan et al. (2005) and Dash and Saji (2006) who pointed out
that trust mediates the relation between website design elements and purchase intention.
In summary, we have empirically found that specific website design factors generate trust
The effects of website design on purchase intention in online shopping 323
in online shopping. We comprehensively tested the mediator role of trust in online
shopping and our results corroborate the results of Sultan et al. (2005) who tested the
mediator role of trust in the relation between website design, customer characteristics,
and purchase intention.
Besides, we believe that the emphasis attached to each website design factor would
vary depending on the cultural values of the customer. According to Yoo et al. (2001),
culture has been measured mainly using two approaches: culture-centred and
personality-centred. Culture-centred approaches are qualitative methods of culture
assessment, whereas the personality-centred approaches are based on quantitative
measures. These approaches have been widely used in business studies using proxies and
values inference (Yoo et al., 2001). There are two types of value inferences: direct values
inference which uses primary data and indirect values inference which uses secondary
data (Lenartowicz and Roth, 1999). One major drawback of the use of proxies is that they
can only provide nominal measures. The indirect values inference, on the other hand,
may result in measurement errors. So, the method of direct values inference is the most
important for personality-cantered ways to measure culture. Therefore, in our study we
used the scale developed by Yoo et al. (2001) to measure culture at the individual level.
To the best of our knowledge no past study has used moderator regression analysis,
with culture as a moderator in the relation between website design and trust in the context
of B2C online shopping. Our results illuminate that culture acts as moderator in the
relationship between website design factors and trust, and also between trust and
purchase intention. As such, our work provides a framework for online customer
segmentation based on cultural values of customers. Using our model, consumer
segmentation can be made at the individual level i.e., culture-centric segmentation across
cultures and countries. This is a meaningful framework for global product management
strategies because there may be equivalent market segments across countries based on
consumer groupings instead of the country-level segmentation. In this case, a similar
marketing programme may be applied to equivalent market segments in different
countries under the assumption that the segments’ responses would be similar across
countries. For example, an online company may develop an online product/service that
may be targeted at very highly masculine, individualistic, and low uncertainty avoiding
consumers. Instead of identifying a few countries that meet this profile the online store
can identify customer segments within each country.
Our results have shown that culture at the individual level acts as a moderator in the
relationship between website design factors and trust and also between trust and purchase
intention. Alternately, there were cases when culture did not moderate the relation
between two variables at the individual level, but at the country level there were
differences. For example, it was believed that collectivism would result in difference
across the two samples with respect to visual design as a predictor of trust. But we found
no moderator effect of collectivism, although visual design was a significant predictor of
trust in USA and Canada, but not in India. This suggests that culture operates at the
individual level rather than at the country level. Western customers were found to place
more stress on visual design than Indian customers. This could be because of the fact that
out of the Rs. 9210 crore e-commerce market in India, online travel related purchases
account for Rs. 7000 crore. We believe that for purchasing tickets, or any intangible
goods, one does not require better visual design. However, better visual design implies
use of applets and more add-ons that make the webpage download even slower and make
324 B. Ganguly et al.
the online shopping process more time consuming. Therefore, the Indians give less
importance to visual design.
From our study we found that customers who are high on masculinity lay more
emphasis on information design. This means that online stores selling products to
masculine customers should present information logically so as to help them in assertive
and quick decision making. This is in agreement with the conceptual works by Singh
(2002) where he stated that masculine customers would like to have features that would
help in assertiveness. Although, we did not find any significant difference in masculinity
scores between India, Canada and the USA we found that Indian customers gave more
importance to Information Design. This has an important implication, as it corroborates
further that there could be cultural variations at the individual level and country should
not be used as a surrogate for culture. The Indians use online shopping primarily for
convenience and this could be the reason for the excessive importance given to
Similarly, customers who are high on uncertainty avoidance were found to pay more
importance to navigation design to generate trust. So, online stores selling products to
such customers need to place greater emphasis on navigation features of the website. Our
results give empirical support to the arguments of Marcus and Gould (2000) who had
pointed out that the customers who are high on uncertainty avoidance would need more
guided navigation. Further, we found that western customers gave more importance to
navigation design than Indians, although there was no significant difference between
Indians and the westerners for uncertainty avoidance scores. This further suggests that
there could be other issues that influence the importance attached to navigation design
From our hypotheses we found that collectivism negatively moderates the relation
between trust and purchase intention and that western customers place more emphasis on
trust to generate purchase intention. However, we found no moderator effect of
uncertainty avoidance in the relationship between trust and perceived risk. Further, trust
was given more importance by Indians (compared to the Canadians and Americans) to
reduce perceived risk. This could be because Indian e-commerce is in a nascent stage
compared to the USA and Canada. According to internet world stats report, the
penetration of the internet in India was less than 1% prior to 2002. This, in turn,
means that Indian customers have not been using the internet for long and the
institutional trust is low. So, Indians emphasise trust more with the online vendor as
means to reduce perceived risk.
One drawback of this study is that we have only used website design factors as
antecedents of trust. Jarvenpaa and Tractinsky (1999) had conceptualised that there
are other factors such as vendor repute and vendor size that affect purchase intention.
Tan et al. (2007) had shown empirically that external and internal norms affect purchase
intention. We believe that future studies should include the effect of other antecedents
of trust such as vendor repute, subjective norms etc on trust in online stores. This would
be a possible extension for future research. Further, only the moderator effect of culture
has been tested in the study. There could be moderator effects of product type, customers’
personal values like demographics and psychographics which we have not tested in this
Finally, this study constitutes an early effort to empirically test the effect of website
design factors on online trust and intention in online context. Consequently, it raises more
questions than it answers. As e-commerce become more and more global in its strategic
The effects of website design on purchase intention in online shopping 325
thinking, it behoves managers and researchers to grapple with the complex interplay
between website design and cultural values. To ignore the fact that culture influences
website design is synonymous with overlooking the reality that advertising influences
consumer choices or lower prices typically lead to higher sales, or higher trust typically
leads to more purchase intention. The challenge is apparent-culture is a factor that both
researchers and managers must become cognisant of in trying to develop website design
over the long term.
Auh, S. (2005) ‘The effects of soft and hard service attributes on loyalty: the mediating role of
trust’, Journal of Services Marketing, Vol. 19, No. 2, pp.81–92.
Bagozzi, R.P. and Yi, T. (1988) ‘On the evaluation of structural equation models’, Journal of the
Academy of Marketing Science, Vol. 16, No. 1, pp.74–94.
Bagozzi, R.P. and Phillips, L.W. (1982) ‘Representing and testing organizational theories: a holistic
construal’, Administrative Science Quarterly, Vol. 27, pp.459–489.
Baron, R.M. and Kenny, D.A. (1986) ‘The moderator-mediator variable distinction in social
psychological research: conceptual, strategic and statistical considerations’, Journal of
Personality and Social Psychology, Vol. 51, pp.1173–1182.
Chan, S. and Lu, M. (2004) ‘Understanding internet banking adoption and use behavior’, Journal
of Global Information Management, Vol. 12, No. 3, pp.21–43.
Chellappa, R. K. (2005) ‘Consumers’ trust in electronic commerce transactions: the role of
perceived privacy and perceived security’; pp.1–46 in Fulk et al. (1987).
Chen, Y.H. and Barns, S. (2007) ‘Initial trust and online behaviour’, Industrial Management and
Data Systems, Vol. 107, No. 1, pp.21–36.
Cho, N. and Park, S. (2001) ‘Development of electronic commerce user – consumer satisfaction
index (ECUSI) for internet shopping’, Industrial Management and Data Systems, Vol. 101,
No. 8, pp.400–405.
Choi, J. and Lee, K.H. (2003) ‘Risk perception and e-shopping: a cross cultural study’, Journal of
Fashion Marketing, Vol. 7, No. 1, pp.49–64.
Corritore, C.L., Wiedenbeck, S, Marble, R.P., Kracher, B. and Chandran, A. (2005) ‘Measuring
online trust of websites: credibility, perceived ease of use, and risk’, Proceedings of the
Eleventh Americas Conference on Information Systems, Omaha, NE, USA, 11–14th August.
Cyr, D. (2008) ‘Modeling website design across cultures: relationships to trust, satisfaction, and
e-loyalty’, Journal of Management Information Systems, Vol. 24, No. 4, pp.47–72.
Dash, S.B., and Saji, K.B.(2006) ‘Role of effective website-design in online shopping: a large scale
empirical study in the Indian context’, in Panda, T.K. and Donthu, N. (Eds.): Marketing in the
New Global Order, Challenges and Opportunities, Excel Books, pp.435–451.
Dash, S.B. and Saji, K.B. (2007) ‘Role of self-efficacy and website social-presence in customers’
adoption of b2c online shopping: an empirical study in the Indian context’, Journal of
International Consumer Marketing, Vol. 20, No. 2, pp.33–48.
Dash, S.B., Bruning, E. and Guin, K.K. (2009) ‘A cross-cultural comparison of individualism’s
moderating effect on bonding and commitment in banking relationships’, Marketing
Intelligence and Planning, Vol. 27, No. 1, pp.146–169.
Donthu, N. and Yoo, B. (1998) ‘Cultural influences on service quality expectations’, Journal of
Service Research, Vol. 1, No. 2, pp.178–186.
326 B. Ganguly et al.
Ford, D.P., Connelly, C.E. and Meister, D.B. (2003) “Information systems research and Hofstede’s
culture’s consequences: an uneasy and incomplete partnership’, IEEE Transactions on
Engineering Management, Vol. 50, No. 1, pp.8–25.
Fornell, C. and Larcker, D. (1981) ‘Structural equation models with unobservable variables and
measurement error’, Journal of Marketing Research, Vol. 18, No. 1, pp.39–50.
Fulk, J., Schmitz, J. and Power, G.J. (1987) ‘A social information processing model of media use
in organizations’, Communication Research, Vol. 14, No.5, pp.520–552.
Ganesan, S. (1994) ‘Determinants of long-term orientation in buyer-seller relationship’, Journal of
Marketing, Vol. 58, No. 2, pp.1–19.
Gefen, D., Karahanna, E. and Straub, D.W. (2003) ‘Trust and TAM in online shopping:
an integrated model’, MIS Quarterly, Vol. 27, No. 1, pp.51–90.
Gong, W., Li Z.G. and Stump R.L. (2007) ‘Global internet use and access: cultural considerations’,
Asia Pacific Journal of Marketing and Logistics, Vol. 19, No. 1, pp.57–74.
Grabner-Krauter, S. and Kaluscha, E.A. (2003) ‘Empirical research in online trust: a review and
critical assessment’, International Journal of Human Studies, Vol. 58, pp.783–812.
Harridge-March, S. (2006) ‘Can the building of trust overcome consumer perceived risk online?’,
Marketing Intelligence & Planning, Vol. 24, No. 7, pp.746–761.
Hofstede, G. (1980) Culture’s Consequences: International Differences in Work-related Values,
Sage Publications, Beverly Hills, CA.
Hofstede G. (1991) Culture and Organizations: Software of the Mind, McGraw Hill, NewYork,
Hofstede G. (2001) Culture’s Consequences: Comparing Values, Behaviors, Institutions, and
Organizations Across Nations, 2nd ed., Sage Publications, Thousand Oaks, CA.
Jarvenpaa, S.L. and Tractinsky, N. (1999) ‘Consumer trust in an internet store a cross-cultural
validation’, Journal of Computer-Mediated Communication, Vol. 5, No. 2, pp.1–35.
Karvonen, K. (2000) ‘The beauty of simplicity’, ACM Proceedings on the Conference on Universal
Usability, University of Helsinki, Finland, pp.85–90.
Kim, E.B. and Eom, S.B. (2002) ‘Designing Effective Cyber Store User Interface’, Industrial
Management and Data Systems, Vol. 102, No. 5, pp.241–251.
Kim, Y.H. and Kim, D.J. (2005) ‘A study of online transaction self efficacy, consumer trust and
uncertainty reduction in electronic commerce transactions’, Proceedings from 38th Hawaii
International Conference on System Sciences, Hawaii.
Kim, J. and Lee, J. (2002) ‘Critical design factors for successful e-commerce systems’, Behaviour
and Information Technology, Vol. 21 No. 3, pp.185–199.
Kluckhohn, C. and Kelly, W.H. (1945) ‘The concept of culture’, In the Science of Man in the
World Crisis, Linton, R. ed. Columbia University Press, NY, pp.78–106.
Kroeber, A. and Kluckhohn, C. (1952). Culture, Meridian Books, New York.
Lee, G.G. and Lin, H.F. (2005) ‘Customer perceptions of e-service quality in online shopping’,
Journal of Retail and Distribution Management, Vol. 33, No. 2, pp.161–176.
Lee, J.A., Garbarino, E. and Lerman, D. (2007) ‘How cultural differences in uncertainty avoidance
affect product perceptions’, International Marketing Review, Vol. 24, No. 3, pp.330–349.
Lee, M.K.O. and Turban, E. (2001) ‘A trust model for consumer internet shopping’, International
Journal of Electronic Commerce, Vol. 6, No. 1, pp.75–91.
Lenartowicz, T. and Roth, K. (1999) ‘A framework for culture assessment’, Journal of
International Business Studies, Vol. 30, No. 4, pp.781–798.
Lim, K.H., Leung, K., Sia, C.H. and Lee, M.O. (2004) ‘Is e-commerce boundary less? Effects of
individualism-collectivism and uncertainty avoidance on internet Shopping’, Journal of
International Business Studies, Vol. 35 pp.545–559.
The effects of website design on purchase intention in online shopping 327
Lim, K.H. and Dubinsky, A.J. (2004) ‘Consumers’ perceptions of e-shopping characteristics:
an expectancy-value approach’, Journal of Services Marketing, Vol. 18, No. 7, pp.500–513.
Marcus, A. and Gould, E.W. (2000) ‘Cultural dimensions and global web user-interface design’,
Interactions, Vol. 7, No. 4, pp.32–46.
McCoy, S., Galletta, D.F. and King, W.R. (2005) ‘Integrating national culture into IS research: the
need for current individual level measures’, Communications of the Association for
Information Systems, Vol. 15, pp.211–224.
McSweeney, B. (2002) ‘Hofstede’s model of national cultural differences and their consequences:
a triumph of faith – a failure of analysis’, Human Relations, Vol. 55, No. 1, pp.89–118.
Mithas, S., Ramasubbu, N., Krishnan, M.S. and Fornell, C. (2006) ‘Designing websites for
customer loyalty across business domains: A multilevel analysis’, Journal of Management
Information Systems, Vol. 23, No. 3, pp.97–127.
Morgan, R.M. and Hunt, S.D. (1994) ‘The commitment trust theory of relationship marketing’,
Journal of Marketing, Vol. 58, pp.20–38.
Nath, R. and Murthy, N.R.V. (2004) ‘A study of the relationship between Internet diffusion and
culture’, Journal of Information Technology and Information Management, Vol. 13, No. 2,
Newman, K.L. and Nollen, S.D. (1996) ‘Culture and congruence: the fit between management
practices and national culture’, Journal of International Business Studies, (4th Qtr.),
Ng, S.W., Lee J.A. and Soutar G.N. (2007) ‘Tourists’ intention to visit a country: the impact of
cultural distance’, Tourism Management, Vol. 28, pp 1497–1506.
Park, J.H. and Stoel, L. (2005) ‘Effect of brand familiarity, experience and information on online
apparel purchase’, International Journal of Retail and Distribution Management, Vol. 33,
No. 2, pp.148–160.
Park, J. and Kim, J. (2000) ‘Contextual navigation aids for two world wide web systems’,
International Journal of Human Computer Interaction, Vol. 12, pp.193–217.
Pavlou, P.A. (2003) ‘Consumer acceptance of electronic commerce: integrating trust and risk with
the technology acceptance model’, International Journal of Electronic Commerce, Vol. 7,
No. 3, pp.101–134.
Pew Internet Research Report (2008) Available at http://www.pewinternet.org/topics/Digital-
Ranganathan, C. and Ganapathy, S. (2002) ‘Key dimensions of business-to-consumer websites’,
Information and Management, Vol. 39, pp.457–465.
Ranganathan, C. and Grandon, E. (2002) ‘An exploratory examination of factors affecting online
sales’, Journal of Computer Information Systems, Vol. 42, No. 3, pp.87–93.
Salam, A.F., Iyer, L., Palvia, P. and Singh, R. (2005) ‘Trust in e-commerce’, Communications of
the ACM, Vol. 48, No. 2, pp.73–77.
Samiee, S. and Jeong, J. (1994) ‘Cross-cultural research in advertising: an assessment of
methodologies’, Journal of the Academy of Marketing Science, Vol. 22, No. 3, pp.205–217.
Sharma, S., Durand, R.M. and Gur-Arie, O. (1981) ‘Identification and analysis of moderator
variables’, Journal of Marketing Research, Vol. 18, No. 3, pp.291–300.
Singh, N. and Matsuo, H. (2004) ‘Measuring cultural adaptation on the web: an exploratory study
of US and Japanese web sites’, Journal of Business Research, Vol. 57, No. 8, pp.464–472.
Singh, N., Zhao, H. and Hu, X. (2005b) ‘Analyzing the cultural content of websites: a cross-
national comparison of China, India, Japan, and US’, International Marketing Review,
Vol. 22, No. 2, pp.129–146.
328 B. Ganguly et al.
Singh, N., Zhao, J. and Hu, X. (2003) ‘Cultural adaptation on the web: a study of American
companies’ domestic and Chinese websites’, Journal of Global Information Management,
Vol. 11 No. 3, pp.63–80.
Singh, N., Furrerr, O. and Ostinelli, M. (2004) ‘To localize or to standardize on the web: empirical
evidence from Italy, India, Netherlands, Spain, and Switzerland’, Multinational Business
Review, Vol. 12, No. 1, pp.69–87.
Singh, N., Kumar, V. and Baack, B (2005a) ‘Adaptation of cultural content: evidence from B2C
e-commerce firms’, European Journal of Marketing, Vol. 39 Nos. 1–2, pp.71–86.
Singh, N., Fassot, G., Mike C.H.C. and Hoffmann J.A. (2006) ‘Understanding international website
usage’, International Marketing Review, Vol. 23, No. 1, pp.83–97.
Singh, N. (2002) ‘Analyzing cultural sensitivity of websites’, Journal of Practical Global Business,
Sondergaard, M. (1990) ‘Hofstede’s consequences: a study of reviews, citations and replications’,
Organization Studies, Vol. 15, No. 3, pp.447–456.
Srite, M. and Karahanna, E. (2006) ‘The role of espoused national cultural values in technology
acceptance’, Management Information System Quarterly, Vol. 30, No. 3, pp.679–704.
Steenkamp, E.M. and Baumgartner, T. (1998) ‘Assessing measurement invariance in cross-national
research’, Journal of Marketing Research, Vol. 25 No. 1, pp.78–90.
Stephens, R.T. (2004) ‘A framework for the identification of electronic commerce design elements
that enable trust within the small hotel industry’, ACMSE ’04, 2–3 April, Huntsville, Alabama,
Straub, D.W. (1994) ‘The effect of culture on IT diffusion: e-mail and fax in Japan and the US’,
Information Systems Research, Vol. 5, No. 1, pp.23–47.
Suh, B. and Han, I. (2003) ‘The impact of customer trust and perception of security control on
acceptance of electronic commerce’, International Journal of Electronic Commerce, Vol. 7,
No. 3 pp.135–164.
Sultan, F., Urban, G.L., Shankar, V. and Bart, I.Y. (2005) ‘Determinants and role of trust in
e-commerce: a large scale empirical study’, Journal of Marketing, Vol. 69, pp.133–152.
Sun, H. (2001) ‘Building a culturally-competent corporate web site: an explanatory study of
cultural markers in multilingual web design’, Proceedings of the 19th Annual International
Conference on Computer Documentation, Santa Fe, NM, 21–24 October, pp.95–102.
Tai, S.H.C. and Chan, R.Y.K. (2001) ‘Cross cultural studies on the information content of service
advertising’, Journal of Services Marketing. Vol. 15, No. 7, pp.547–564.
Tan, F.B. and Sutherland, P. (2004) ‘Online consumer trust: a multi-dimensional model’, Journal
of Electronic Commerce in Organizations, July–September, Vol. 2, No. 3, pp.40–58.
Tan, F.B., Yan, L. and Urquhart, C. (2007) ‘The effect of cultural differences on Attitude, Peer
influence, external influence and Self-efficacy in actual online shopping behavior’, Journal of
Information Science and Technology, Vol. 4, No. 1, pp.3–26.
Tan, H. and Guo, J. (2005) ‘Some methods to depress the risks of the online transactions’,
ICEC’05, 15–17 August, Xi’an, China.
Wolfinbarger, M. and Gilly, M.C. (2003) ‘eTailQ: dimensionalizing, measuring and predicting etail
quality’, Journal of Retailing, Vol. 79, pp.183–198.
Yaveroglu, I. and Donthu N. (2002) ‘Cultural influences on the diffusion of new products’,
Journal of International Consumer Marketing, Vol. 14, No. 4, pp.49–64.
Yoo, B. and Donthu N. (2002) ‘The effects of marketing education and individual cultural values
on marketing ethics of students’, Journal of Marketing Education, Vol. 24, pp.92–103.
Yoo, B. and Donthu, N. (2001) ‘Cultural orientation and consumer ethnocentrism’, in Rao, C.P.
(Ed.): Marketing and Multicultural Diversity, Greenwood Publishing, Westport, CT.
The effects of website design on purchase intention in online shopping 329
Yoo, B., Donthu, N. and Lenartowicz, T. (2001) Measuring Cultural Values: Development and
Validation of CVSCALE, Working paper, Georgia State University, Atlanta, GA, Available at:
Yoon, S-J. (2002) ‘The antecedents and consequences of trust in online purchase decisions’,
Journal of Interactive Marketing, Vol. 16, No. 2, pp.47–63.
I-Cube Report (2008) Available at www.iamai.in/Upload/Research/I-Cube_2008_Summary
Jupiter Corp Report, Available at http://www.e-commerce-guide.com/news/trends/article.php/
Table 1 The Questionnaire along with the measurement scales for the variables used
in the study
Construct Items Questions Source
ID1 In my online store, I find the information to be logically
ID2 In my online store, I find the information on this site to be well
VD1 The website of my online store looks professionally designed
and well presented
VD2 The screen design on the website (i.e., colours, boxes, menus,
navigation tools etc.) of my online store is harmonious and
ND1 I can easily navigate the website of my online store
ND2 I find the website of my online store easy to use
ND3 The site of my online store provides good navigation facilities
to search the information content
T1 Based on my past experience I do believe that the transaction
through my online store is always safe
T2 Based on my past experience I do believe that the transaction
through my online store is always reliable
T3 Based on my past experience I do not think that things may go
wrong with my transaction through my online store
T4 Based on my past experience I am confident that my online
store will promptly inform me if at all any problem occur with
any of my transactions
T5 Based on my past experience I am confident that my
transaction through my online store will always be transparent
T6 Based on my past experience I do believe that my online store
always protects my best interest
T7 Based on my past experience, I can say that my online store is
Suh and Han
330 B. Ganguly et al.
Table 1 The Questionnaire along with the measurement scales for the variables used
in the study (continued)
Construct Items Questions Source
PR1* I do not perceive any risk by sharing my personal information
concerning my transaction with the online store
PR2* I am confident that others can not tamper with information
concerning my transaction with the online store
PR3* I believe that advanced technology can provide the desired
security for my transaction with the online store
PR4* I do not think that my money will get stolen whenever
I transact through my online store
Chan and Lu
PI1 I intend to continue using my online store for purchasing
a product or service in future
PI2 I would strongly recommend others to use my online store Suh and Han
PI3* I shall not transact with my online store in the near future Chen and
*Indicates that the items are coded in reverse.