Technology in Society 62 (2020) 101280

Contents lists available at ScienceDirect

Technology in Society
journal homepage: http://www.elsevier.com/locate/techsoc

Understanding the attitude and intention to use smartphone chatbots

for shopping
Dharun Lingam Kasilingam
MICA - The School of Ideas, Ahmedabad, India

A R T I C L E I N F O A B S T R A C T

Keywords: Using the technology acceptance model and diffusion of innovations theory, this study evaluated the intention of
Mobile applications consumers to use chatbots on smartphones for shopping. Chatbot is a relatively new technology and is expected
Chatbots to dominate mobile commerce and shopping applications in future. Hence, this study aimed to determine the
Technology acceptance model (TAM)
association of perceived usefulness, perceived ease of use, perceived enjoyment, price consciousness, perceived
Diffusion of innovations (DOI)
Attitude
risk, trust, and personal innovativeness with attitude and intention to use chatbots for shopping. Respondents
Intention to use were asked to fill a questionnaire after using a Facebook e-commerce chatbot that was specifically created for this
study. In total, 350 responses were analyzed using partial least squares structural equation modeling. Results
indicated that attitude toward chatbots was considerably influenced by the variables perceived usefulness,
perceived ease of use, perceived enjoyment, price consciousness, perceived risk, and personal innovativeness.
However, intention to use was directly influenced only by trust, personal innovativeness, and attitude. Mediation
analysis indicated that full mediation occurs through the attitude variable for most direct relationships.
Moderation analysis by using age, gender, and prior experience with mobile shopping applications indicated
considerable differences between the groups in terms of the strength of certain relationships and the mean re­
sponses between the variables.

1. Introduction Chatbots have been around for many years but are making a come­
back due to the rise of artificial intelligence and internet of things.
Mobile commerce was considered the technology that revolutionized ELIZA, a computer program created by Joseph Weizenbaum in 1966, is
e-commerce. However, with smartphone evolution, the traditional web- generally recognized as the first chatbot. Earlier, chatbots were
based mobile commerce was replaced by mobile shopping applications. considered an intriguing but a mostly useless technology. Today, chat­
Furthermore, with the advancement of mobile shopping applications, bots are poised to revolutionize our communication method and are
the big question is “Which technology will replace mobile shopping expanding into all possible fields requiring human interaction, from
applications?” Mobile chatbot is one of the technologies that could simple weather chatbots for daily forecast (the most notable example
revolutionize mobile commerce once again. American Marketing Asso­ being Facebook’s Poncho chatbot) to news chatbots for updates on
ciation and various other sources claim that chatbots are the future of current events, or for finding a ride (such as the Uber chatbot, which was
marketing, and as they can be integrated with messaging applications created as an alternative to the app). Furthermore, chatbots are used in
such as Facebook, WhatsApp, and Skype, they can replace mobile ap­ most of the standard e-commerce fields (e.g., delivering flowers,
plications for shopping [1]. Consumers can interact with a chatbot ordering pizzas, or booking flights) as well as in information procure­
through a chat interface by using written or verbal statements. Chatbot ment on various topics (mortgage, health care, food recipes, and so on)
is considered the hottest development trend of 2018; since the last [4].
quarter of 2015, the use of messaging apps has surpassed that of social The literature on mobile chatbot is scarce in the fields of marketing
networking sites in terms of monthly active users [2]. While using a and behavioral sciences as it is a relatively new technology. Eeuwen [5]
mobile, an estimated 90% of the time is spent on email and messaging was the first to evaluate user attitude toward chatbot technology from a
platforms [3]. Millennials are especially keen on using chatbots, as this mobile shopping perspective. He categorized chatbots as “conversa­
technology suits their habits and routines [4]. tional commerce” (c-Commerce), which was coined by Messina [6], and

E-mail addresses: [email protected], [email protected].

https://doi.org/10.1016/j.techsoc.2020.101280
Received 25 June 2019; Received in revised form 5 February 2020; Accepted 22 May 2020
Available online 30 May 2020
0160-791X/© 2020 Elsevier Ltd. All rights reserved.
D.L. Kasilingam Technology in Society 62 (2020) 101280

defined it as “utilizing chat, messaging, or other natural language in­ must be taken into account to make these rational decisions, which are
terfaces (i.e., voice) to interact with people, brands, or services.” Of the considered determinants to user attitude and behavior [10].
five variables considered by Eeuwen [5], perceived usefulness, Venkatesh and Davis [12] were two of the most influential scholars
compatibility, and internet privacy concern were considerably associ­ to expand the model, making a new standardized model of technology
ated with and perceived ease of use and attitude toward mobile adver­ acceptance, namely TAM2, through introduction of new factors and
tising were insignificantly associated with consumer attitudes toward exclusion of attitude. The unified theory of acceptance and use of
chatbot use for mobile shopping. Eeuwen’s study directed future re­ technology (UTAUT) is another standardized model developed through
searchers to focus on other variables such as enjoyment, risk, and trust. comparison and unification of existing models of information systems
Chatbots have various advantages such as availability, [13]. Furthermore, this model excludes attitude and introduces four key
cost-effectiveness, customer interaction, automation, and personal factors for determining usage intention and behavior. These factors are
assistance. However, it has the disadvantage of reduced flexibility in effort expectancy, performance expectancy, social influence, and facil­
comparison with products across different e-commerce vendors. Mobile itating conditions, which are connected to the moderating constructs of
shopping applications have reduced flexibility, and chatbots decrease age, gender, use voluntariness, and experience [13].
the flexibility further. This factor may play a major role in influencing Although both these models continue to be extended and improved
attitudes and usage intentions of users in developing countries, which with new research, which further legitimate these models, TAM is more
are highly priced sensitive markets. Hence, price consciousness of con­ influential and popular than UTAUT [11]. Moreover, in contrast to
sumers is an exciting study area. UTAUT, TAM includes attitude as a major determinant of user accep­
This study aimed to identify factors that influence attitude and tance, which has proven to have a significant effect on mobile applica­
intention to use chatbots for mobile shopping. The chatbot considered in tion acceptance earlier [14–16]. Hence, it was used as the core model in
this study is a conversational smartphone interface (through Facebook this research. TAM has several benefits when determining factors for
messenger) that can interact with users and assist in reading reviews, technology acceptance. First, it possesses consistent tools of measure­
browsing for and researching of products, comparing products, access­ ment, empirical soundness, and conciseness [14,17]. Second, it explains
ing saved coupons, buying products, keeping a track of orders, and a major part of variance in usage intentions [14,18,19]. Third, as it has
receiving rewards and loyalty points. A literature review of research been applied in many studies, it offers a wide range of questions related
articles that used grand theories such as diffusion of innovations (DOI), to each factor, adding reliability to the relevance of the questions asked
theory of planned behavior (TPB), and technology acceptance model in the questionnaire. Although TAM is useful in terms of explaining the
(TAM) was performed to identify factors that affect intention to use behavioral intention of using a technology, extended variables related to
various technologies such as mobile commerce, mobile applications, the specific technology must be addressed to clearly understand the its
mobile payments, and gaming. Thereafter, a generic model was created. acceptance [10].
The significance of user demographics is prevalently studied for tech­
nology adoption, and this has been widely researched in the fields of e- 2.2. Diffusion of innovations
commerce, mobile commerce, and mobile applications. The prognostic
power of these models must be well understood, and with readily An innovation can be defined as something that is perceived as new
available demographic variables, demographic-based models can be by an individual or a social system. Conversely, diffusion is explained as
quickly formulated and used to provide solutions for emerging problems “the process by which an innovation is communicated through certain
in businesses [7–9]. Chatbots for mobile shopping is a relatively new channels over time among the members of a social system” [20]. This
technology promising revolution in marketing [1]. Hence, demographic theory shows the innovation–decision stages from invention to the wide
variables must be studied at the initial stage of the study to efficiently use of a new technology, as well as differences among the categories of
design and tailor services for consumers with increased personalization. adopters [21]. According to Rogers [20], people go through five stages
Furthermore, this study aims at identifying differences among genders, before they accept an innovation: knowledge, persuasion, decision,
age groups, and users with different levels of experience with mobile implementation, and confirmation. Additionally, five characteristics of
applications in terms of chatbot adoption. In line with the objectives of technology influence acceptance: relative advantage, compatibility,
the study and gaps identified in the literature, the following research complexity, trialability, and observability [20]. Finally, five adopter
questions were proposed: categories are available: innovators, early adopters, early majority, late
Research Question 1: What factors significantly influence attitude majority, and laggards [20]. In summary, the DOI model describes how
and intention of chatbot users to use mobile messenger e-commerce? users create beliefs regarding innovation characteristics based on which
Research Question 2: Do gender, age group, and prior experience in an innovation is adopted or rejected by the user [22–24], and this pro­
using mobile shopping applications significantly moderate adoption of vides an excellent platform for researchers to study the adoption of new
chatbots? technologies such as chatbots.
The remainder of the paper is organized as follows. Section 2 deals
with the hypotheses and model development in line with the research 2.3. Perceived usefulness
questions. The methodology, data collection, and analysis to validate the
research model are presented in Section 3. Results are presented in Venkatesh and Davis believe that “perceived ease-of-use play an
Section 4 followed by discussion and conclusion in Sections 5 and 6, important role and get more attention, while the perceived usefulness is
respectively. believed as equally important as ease-of-use and lean toward service-
dependent” [25]. Perceived usefulness of an information system is
2. Theoretical model and hypothesis development defined as “the extent to which individuals believe that using the new
technology will enhance their task performance” [26].
2.1. Technology acceptance model As initially argued by Davis and subsequently supported by research,
usefulness is central to the acceptance of information technology. This
TAM, which was first proposed by Davis [10], is the most influential theory was extrapolated and applied to mobile shopping. Research
of research models explaining information technology adoption and is confirmed that usefulness has a substantially positive influence on
considered useful for studying acceptance in various contexts related to behavioral intention to mobile shopping [27]. If consumers believe that
information technology [11]. The central message of this model is that they can gain from this shopping channel, they are likely to adopt it.
technology users make rational decisions regarding using a technology Gains in mobile shopping might include saving time, comparing pricing,
[11]. Perceived usefulness and perceived ease of use are the factors that obtaining promotional information, or receiving customized offers [28].

2
D.L. Kasilingam Technology in Society 62 (2020) 101280

Preliminary studies have empirically proven that in the technological influenced by attitude toward mobile commerce, subjective norm, and
field, perceived usefulness is an essential factor that stimulates the use of perceived enjoyment” [46]. The more an individual enjoys the shopping
a specific form of technology [29–38]. Thus, studying the effect of experience through chatbots, the more likely he/she will continue using
perceived usefulness on attitude and the intention to use chatbots for this channel. In this study, the following hypotheses were proposed:
mobile shopping is appropriate. Simply put, individuals who find chat­ H3a: Perceived enjoyment has a significant positive influence on
bots useful will adopt the technology. As the chatbot technology can attitude toward chatbots.
save time and offer flexibility and convenience, its usefulness must be H3b: Perceived enjoyment has a significant positive influence on the
understood. The following hypotheses were therefore proposed: intention to use chatbots.
H1a: Perceived usefulness has a significant positive influence on
attitude toward chatbots. 2.6. Price consciousness
H1b: Perceived usefulness has a positive influence on the intention to
use chatbots. Prices of products and services have always been a significant
influencer of consumer attitudes and intentions [47–50]. According to
2.4. Perceived ease of use Goldsmith and Newell [51], price sensitivity indicates the reactions of
consumers to varying price levels. Price consciousness can influence a
The perceived ease of use of an information system is defined as “the customer’s decision when choosing or buying a product or service if that
degree to which an individual believes that using a particular technol­ customer has high levels of price consciousness [52]. Price is a critical
ogy will be free of mental effort” [10]. Moreover, Davis suggests that factor when purchasing products online [53]. The notion of price is
ease of use is an indicator for technology acceptance. Consumers are essential in the purchase of products using a new technology. Unlike
more likely to adopt mobile shopping using chatbots if the technical traditional e-commerce or even mobile commerce, chatbots for shop­
infrastructure for it already exists. Infrastructure includes ping do provide enough flexibility to compare or verify prices between
internet-enabled phone plans, user-friendly interfaces, and messenger various platforms for the same product. According to many studies on
apps that are compatible with various phones. Making it easy for con­ the importance of price, price is a deciding factor for using an online
sumers to shop on their mobile devices can prevent any barrier-to-entry service by a customer. It is a bothersome issue [54] and relates to the
that may be caused by technological confusion [28]. In other words, intention of behavior [55]. Hence, studying this construct in our model
consumers will willing accept a technology they can easily understand is crucial and in line with the objectives of the study. The following
and use. Shoppers are likely to use the technology if they can merely add hypotheses were proposed:
a contact to their traditional messaging apps, such as WhatsApp, Face­ H4a: Price consciousness has a significant negative influence on
book, or Skype, easily and navigate interfaces without hassle. attitude toward chatbots.
Many researchers [10,39,40] have indicated that users of technolo­ H4b: Price consciousness has a significant negative influence on the
gies have predefined assumptions regarding how easy or difficult it will intention to use chatbots.
be to use a technology. Thus, researchers must study the perceived ease
of use from users to understand their expectations. Studies have shown 2.7. Perceived risk
that the ease of use largely affects users’ perceptions toward techno­
logical devices and must be seriously considered [29,30,32–38]. The concept of perceived risk and its different constructs affect
Natarajan, Balasubramanian and Kasilingam [18] observed that acceptance and adoption of new technologies [29,34,35,37,56–60]. The
perceived ease of use affects the intention to use mobile shopping ap­ concept of perceived risk is a group of several risk components, and one
plications. As chatbots are closely associated with the technology in of the most common ways to break down the concept of perceived risk
terms of functionality, perceived ease of use is a core variable in our was introduced by Jacoby and Kaplan [61], where they listed five
study. different components: financial, performance, psychological, social, and
H2a: Perceived ease of use has a significant positive influence on physical risk. As physical risk is not applicable to the context of infor­
attitude toward chatbots. mation technology adoption, it is usually excluded from perceived risk,
H2b: Perceived ease of use has a significant positive influence on the whereas privacy risk is introduced as it affects customers online [62,63].
intention to use chatbots. Psychological and social risks have been grouped together as social risk
as they address similar areas of risk. Adding privacy risk is a natural step
2.5. Perceived enjoyment while analyzing risks online as technological evolution has introduced
security concerns regarding identity thefts online and misuse of financial
For many years, perceived enjoyment was believed to influence the information [62]. Moreover, privacy risk is a barrier in the adoption of
behavior of individuals [41]. Kim [42] discovered that smartphone users mobile banking as consumers are not in full control of their information,
are more likely to accept a technology and use it comprehensively than for example, credit card numbers [64]. Recently, concerns have arisen
others if they experience pleasure or delight in using the technology. The regarding privacy risks in terms of phishing, where criminals manage to
TAM was extended across various technologies as it possesses one core obtain user information and carry out financial transactions [63].
variable, namely perceived enjoyment. According to Ha, Yoon and Choi Financial risk can be a consequence of privacy risk when mobile bank
[43], perceived enjoyment is one of the most essential variables to users are subjected to monetary losses due to phishing or hacking [63].
determine the adoption of mobile games by users. This necessitates the Moreover, financial risk can occur because of fraudulent behavior by the
urge to determine whether this perceived enjoyment is also a factor that recipient as the transaction occurs online, where the consequence is
influences chatbot users because it affects internet usage and web-based monetary loss in this case as well [62]. Performance risk can be
commerce. described as the risk due to failure of a product or service, which leads to
Teo [44] observed that in terms of using the internet for activities a loss in performance, and predicts overall perceived risk best [65].
such as browsing, messaging, and downloading, the idea of “perceived Social risk refers to the perception of others while a consumer adopts
enjoyment” plays an important part in the intention to use the internet and uses products or services. Depending on how that usage is perceived
among consumers in Singapore. Other studies in the field of technology by others, one’s self-esteem could be affected positively or negatively
adoption have revealed the positive influence of perceived enjoyment on [62].
user attitudes and behavior [32,34,38,45]. Enjoyment strongly affects As chatbot users could be exposed to all of the aforementioned risk
one’s engagement in mobile shopping [27]. One study found that dimensions, and as the variable of perceived risk has been widely used as
“consumers’ shopping intentions via mobile technology are positively one of the extensions of the TAM, this research includes it as one of the

3
D.L. Kasilingam Technology in Society 62 (2020) 101280

variables in our model. If customer expectations are not met with the intention to use chatbots.
chatbots from a risk perspective, their relationship with that retailer will
be affected. Hence, the following hypotheses were formulated in the 2.10. Attitude and intention to use
area of perceived risk:
H5a: Perceived risk has a significant negative influence on attitude Human beings have an evaluation scheme for the consequences of
toward chatbots. performing a particular behavior, which is called attitude [88]. Briefly,
H5b: Perceived risk has a significant negative influence on the on the basis of TPB [89,90], which is the extended version of the theory
intention to use chatbots. of reasoned action (TRA) [91], behavior is a direct function of elements
like behavioral intention, which is formed by the person’s attitude.
2.8. Trust Attitude is the users’ negative or positive feeling regarding chatbots use
on a mobile phone for shopping. From TRA and TAM, individuals’ belief
From a web retailer’s perspective, trust can be defined as “the sub­ regarding the consequences of their behavior significantly affects their
jective probability by which the consumers expect that a Web retailer attitude toward behaving in that manner. Both models posit that attitude
will perform a given transaction by their confident expectation” [57]. significantly influence individual intentions toward behavior. This
Because of the nature of internet shopping, consumers always face some relationship has been studied and found to be significant in various
level of risk in the shopping process. In fact, actions by other parties can technologies such as e-banking [92], smart homes [93], virtual worlds
be favorable (as in the case of a well-known web vendor) or harmful (as [94], academic social networking sites [95], and mobile payment ser­
in the case of hackers). Once consumers face uncertain situations, trust vices [96]. Thus, the relationship between attitude and intention to use
can solve the risk conundrum [66]. When risk is absent, trust is not was included in the research model [97].
needed. Risk brings uncertainty into the buying process, and trust can be H8: Attitude toward chatbots has a significant positive influence on
a successful strategy in dealing with uncertain future [57]. the intention to use.
The importance of trust and its crucial role in online transactions and Fig. 1 shows the research model for the study.
buying behavior in creating expected positive results has been studied
by many researchers [67–71]. Trust may be a fundamental element that
shapes a customer’s belief system, ensuring that online vendors would 2.11. Gender
not behave opportunistically [72].
Trust in incorporated into the TAM in several ways. The results of Gender is defined as the “biological differences between sexes,” and
earlier studies place trust as a determinant of perceived ease of use [67], this definition is widely used in today’s society [98]. Researchers in all
perceived usefulness [67,73], attitude [74,75], and behavioral intention disciplines of management always report significant differences between
[67,74,76]. According to the findings of Dahlberg, Segesten, Nystro €m, genders in terms of attitudes and intentions [99]. Gender-based differ­
Suserud and Fagerberg [73], the model of trust-enhanced technology ences have been observed in the different facets of consumer behavior
acceptance explains technology adoption of customers [73]. Trust af­ among which are attitudes toward shopping behavior and purchasing
fects intention both directly and indirectly. Its indirect effect is through behavior [100,101]. It is generally considered the one most common
the formation of a positive attitude toward shopping online [67]. and a primary variable used by marketers for segmentation. Venkatesh
Chatbots are new technologies in the market, and trust is a crucial factor and Morris [102] proved that women’s perception of usefulness of a
in explaining users’ attitudes and intentions to use. technology influences adoption weakly compared with men. Moreover,
H6a: Trust has a significant positive influence on attitude toward various recent studies have revealed considerable differences between
chatbots. the genders in numerous forms of technology, such as mobile commerce,
H6b: Trust has a significant positive influence on the intention to use mobile payment, and mobile shopping applications [19,37,59,83,103,
chatbots. 104]. When studying risk- and trust-related topics in online shopping,
gender is used as a determinant [57]. Chatbot is a new technology in the
2.9. Personal innovativeness market, and hence, the adoption of the technology itself may vary across
gender. To develop strategies catering to each gender, which have high
Rogers and Shoemaker [77] define innovativeness as “the degree to success rates, it is necessary to understand differences between women
which an individual is relatively earlier in adopting new ideas than other and men regarding adoption of the mobile messenger chatbot. Hence,
members of his social system.” Personal innovativeness regarding a new the following hypotheses were proposed in the context of chatbots for
technology has been studied in different technology contexts and
countries. AbuShanab, Pearson and Setterstrom [78] studied internet
banking in Jordan, while Xu and Gupta [79] studied the acceptance of
location-based services in Singapore, both including the aspect of per­
sonal innovativeness in their research.
Prior research in the field of consumer behavior has always given
increased attention to the innovativeness of individuals [23,80,81]. It is
considered a key variable in the online environment [82]; in the context
of omnichannel retail, it significantly affects the purchase intention of
consumers [81]. Moreover, it is one of the most significant variables in
driving technology acceptance and adoption [31,32,34,56,58,83–86].
Herrero Crespo and Rodríguez del Bosque [87] and Rogers [20] argue
that highly innovative individuals take a high level of risk while
adopting new technologies. All in all, personal innovativeness is an
interesting and a crucial variable to explore and may have an effect on
attitude and intention to use new technologies like chatbots; hence, the
following hypotheses were proposed:
H7a: Personal innovativeness has a significant positive influence on
attitude toward chatbots.
H7b: Personal innovativeness has a significant positive influence on Fig. 1. Research model.

4
D.L. Kasilingam Technology in Society 62 (2020) 101280

mobile shopping: TAM model and found moderating effects [12]. In addition to modera­
H9: There is a significant difference between genders in “a: perceived tion, this variable directly influences attitudes and intentions of the
usefulness, b: perceived ease of use, c: perceived enjoyment, d: price users [42,118]. Some studies have revealed that an individual’s ac­
consciousness, e: perceived risk, f: trust, g: personal innovativeness, h: quaintance and perceived benefit of use increase with increase in
attitude, and i: intention to use” of chatbots for mobile shopping. experience with the internet [119–121], ultimately leading to the
H10: Gender moderates the relationships in H1 to H8. intention of use [122]. Moreover, the reviewed literature showed that
regarding the moderating influence of experience, new users who
2.12. Age possess less levels of experience would need a higher ease of use than
users with more experience. Thus, users with less experience are likely to
Younger users, who are de facto more familiar with newly introduced be driven both intrinsically and extrinsically. Studies by Gefen, Kar­
technologies and more eager to accept innovative tools, would have ahanna and Straub [123] and Moon and Kim [124] conclude that users
fewer issues with acclimatization than older users and would consider with limited experience are likely to be highly attentive to variables
chatbots for mobile shopping as useful. This notion is empirically backed associated with the user interface itself and hence pay less attention to
by a study regarding mobile and internet usage within this age group the purpose of their visit.
and shows that an affinity exists for adopting an online, continually Moreover, Miyazaki and Fernandez [125] support the notion that
evolving platform, such as chatbots for mobile shopping with little to no experience with internet use can diminish perceived risk in the context
amount of perceived effort from their part [105]. Conversely, older adult of online shopping, as well as security and privacy concerns, and this
users would perceive the need for more efforts to familiarize themselves concept can also be extended to chatbots and mobile shopping appli­
with chatbot technologies for mobile shopping in general and are more cations. Chatbots for mobile shopping is a new technology that most of
likely to find its use as not being commensurate to the effort than the the digital population has not used yet or is even unaware of. However,
younger adult group. Moreover, compared with youngsters, the older this technology is expected to storm the mobile applications market [1].
group is likely to be affected by external influences more explicitly [106] As both these technologies serve the same purpose and users of mobile
and would be less willing to accept change without considerable in­ applications would mainly adopt the chatbot technology, it is logical to
vestment on behalf of the developers and representatives of the chatbot determine whether the previous experience of users in using mobile
[107]. shopping applications moderates the relationships.
However, older adults tend to trust the endorsement made by their H13: A significant difference exists between experience in using
online contacts and other early adopters within the context of their mobile shopping applications in “a: perceived usefulness, b: perceived
personal and online social circle, remaining mostly unaffected by third ease of use, c: perceived enjoyment, d: price consciousness, e: perceived
parties for a new technology or product. By contrast, younger adults, risk, f: trust, g: personal innovativeness, h: attitude, and i: intention to
who are more familiar and therefore more comfortable in their online use” of chatbots for mobile shopping.
surroundings than the older ones, would be more likely to believe the H14: Experience in using mobile shopping applications moderates
promises of a marketing campaign or third-party developer or marketer, the relationships in H1 to H8.
owing to their subjective perception of the effectiveness of mobile
payment in general [105]. 2.14. Methodology, data collection, and analysis
A study by Plaza, MartíN, Martin and Medrano [108] on the effects of
mobile phone use on older adults in the United States, Japan, and For this study, a custom Facebook e-commerce chatbot was devel­
Europe revealed that the most efficient way to allow this age group to oped for performing most e-commerce functions such as searching for
adopt new technologies is through the implementation of design choices information, recommend products, process orders and transactions,
and approaches that are tailored according to their needs and are cus­ track shipping locations, and track orders. With an expected value of 150
tomizable to some degree to suit their lifestyle [108]. Further research billion USD by 2022 and 93% of internet traffic coming from mobile
carried out in the Netherlands cemented the notion that age played a devices [126], Indian e-commerce market is mobile driven and is perfect
significant role when determining online trends among users. Specif­ for testing the research model. Verified e-commerce users, whose de­
ically, older mobile and online users deviated considerably from the mographic details matched with the digital population of India, were
norm established by younger adults by focusing mostly on shopping and selected for the study [127] to ensure that the selected sample represents
other online ventures and information research in general [109]. the population. The participants were asked to use the chatbot to search
Numerous researchers have, in fact, proven that a clearly defined “dig­ for a product of their choice and complete an order placement with the
ital divide” exists among users based almost entirely on age, with entire cash on delivery option. The study participants were asked to use the
platforms or technologies explicitly adapted for use by older adults to chatbot for at least 5 min to ensure that they understood the technology
facilitate their adoption. Information, enterprise, and commerce follow properly. After using the chatbot, the participants were asked to fill a
a similar pattern, gravitating toward the demographic that possesses a questionnaire with 46 questions followed by their demographic details.
steadier, more disposable income [110–115]. Van den Broeck, Zarouali and Poels [128] used a similar method that
H11: A significant difference exists between age groups in terms of involved deploying a real-time chatbot followed by data collection from
“a: perceived usefulness, b: perceived ease of use, c: perceived enjoy­ the users to understand chatbot advertising effectiveness.
ment, d: price consciousness, e: perceived risk, f: trust, g: personal The items of the questionnire were adapted from the literature. The
innovativeness, h: attitude, and i: intention to use” of chatbots for mo­ items of the perceived risk construct were adopted from Featherman and
bile shopping. Pavlou [62], perceived usefulness and perceived ease of use from Davis
H12: Age moderates the relationships in H1 to H8. [10], perceived enjoyment from Davis, Bagozzi and Warshaw [129],
personal innovativeness from Goldsmith and Hofacker [130], trust from
2.13. Experience in using mobile shopping applications Chong, Chan and Ooi [131] and Tsu Wei, Marthandan, Yee Loong
Chong, Ooi and Arumugam [33], price consciousness from Dickinger
Fishbein and Ajzen [91] posit that a consumer’s current behavior and Kleijnen [132], Lichtenstein, Ridgway and Netemeyer [133], and
toward an item is largely influenced by that individual’s positive deal­ Swaminathan and Bawa [134], attitude form Taylor and Todd [135],
ings in the past with the same item. Prior experience in using a form/tool and intention to use from Kim, Mirusmonov and Lee [11]. A seven-point
of technology moderates variables and relationships in various tech­ Likert scale was used to measure the constructs. In total, 399 responses
nology fields including mobile and internet messaging [116,117]. TAM2 were obtained of which 34 had missing data and were hence rejected.
had, in fact, included the moderating effect of experience in the original Furthermore, 15 responses were eliminated due to non-engagement

5
D.L. Kasilingam Technology in Society 62 (2020) 101280

while filling the survey. Finally, 350 responses were analyzed. The youth convergent validity were met by the measurement model. Discriminant
of India aged �35 years were categorized as the younger age group, as validity of the constructs was assessed using cross-loadings, Fornell and
reported by the Ministry of Statistics and Programme Implementation, Larcker [144] criterion, and heterotrait monotrait ratios (HTMT). Ta­
Government of India [136]. Hence, on the basis of the report, partici­ bles 3 and 4 display indication for the model’s discriminant validity. As
pants were categorized as young (�35 years) and old (>35 years). For shown by the correlation matrix, all pairs have a correlation <0.70
experience in using mobile shopping applications, people with experi­ [145]. Similarly, all the inter-item correlations reported below the di­
ence of <3 years were categorized as “Low Experience” group and those agonal are not more than the square roots of the average variance
pffiffiffiffiffiffiffiffiffi
with experience >3 years were categorized as “High Experience” group. extracted ( AVE) for the corresponding factors [144]. Furthermore,
Demographic details of the respondents are given in Table 1. Table 4 shows that all the HTMT ratios were <0.85, indicating excellent
The partial least squares structural equation modeling (PLS-SEM) discriminant validity [146]. SRMR was found to be 0.057, which was
tool was used for testing the hypotheses. PLS-SEM was mainly used for lesser than the threshold of 0.08, indicating an excellent model fit [147].
the analysis because it is flexible regarding data requirements, model Additionally, tests for skewness, kurtosis, common method bias, multi­
complexity, and relationship specifications [137]. Another major collinearity, and invariances between groups (prior to moderation
advantage of using this technique is the ability to compare two or more analysis) were conducted, and the model passed all tests with excep­
groups by specifying a permutation-based analysis of variance approach, tional statistical significance and standard thresholds [140,148].
which helps in maintaining the familywise error rates without relying on
distributional assumptions and at the same time exhibiting an adequate
3.2. Partial least squares structural equation modeling
level of statistical power [138]. SPSS 23.0 and Smart-PLS 3.0 were used
as statistical software tools for different analysis. An independent sam­
The statistical significance of the relationships in the model was
ple t-test using SPSS was conducted to compare group responses for the
tested by using bootstrapping procedure with 5000 subsamples and no
variables. Smart-PLS 3.0 was used to analyze the structural model and to
sign changes with a confidence interval method of bias-corrected and
conduct the multi-group moderation analysis to compare the strengths
accelerated bootstrap and two-tailed test with a significance of 0.05 in
of the relationships between the groups. The percentile bootstrap pro­
Smart PLS 3.0. On analyzing the structural model (Table 5), Hypothesis
cedure using the PLS algorithm was performed on 5000 samples. For
H3a relating to perceived enjoyment to attitude was found to have a
analyzing the structural model, a minimum confidence level of 95% for
significance of 0.001 and was the strongest relationship in the model.
both independent sample t-test and moderation analysis and a minimum
Hypotheses H6b, H7b, and H8 relating to trust, personal innovativeness,
significance of 0.1 or a confidence level of 90% were considered.
and attitude, respectively, with intention to use chatbots for mobile
shopping also had a significance of 0.001. Hypotheses H1a, H2a, H4a,
3. Results
H5a, and H7a relating to perceived usefulness, perceived ease of use,
price consciousness, perceived risk, and personal innovativeness with
3.1. Reliability and validity
attitude toward chatbots had a significance of 0.05. All the significant
relationships in the model were positive except for the effect of
The measurement model was assessed using two validity tests,
perceived risk on attitude. All remaining hypotheses, namely H1b, H2b,
namely discriminant validity and convergent validity. Convergent val­
H3b, H4b, H5b, and H6a, were rejected although the effect of trust on
idity generally refers to the condition that verifies whether the in­
attitude was close to the significance of 0.05. The main reason for
dicators of a multi-item construct are indeed measuring the same
rejecting these hypotheses could be because attitude may mediate them.
construct [60,139]. To accept measurement items, the factor loadings of
Hence, the total effects and indirect effects in the model must be studied
the individual items belonging to a construct must be greater than the
to conclude.
threshold value of 0.50 and reported as significant [140], and the
The total effects in Table 6 for the rejected hypotheses involving
average factor loading for the particular construct must be > 0.70 [141].
indirect relations indicate that the strengths of the regression co­
The items PI4, PS3, and PEU1 were removed from the analysis as they
efficients change. Hypothesis H3b relating to the effect of perceived
showed loadings <0.50 or cross loadings.
enjoyment on the intention to use was significant at a confidence level of
Table 2 shows that for all the variables, the Cronbach’s alpha value
99%. Hypotheses H4b and H5b, although insignificant, were still close
is > 0.70, which proves that the measurement is reliable [142]. More­
to a significance of 0.05.
over, composite reliability was greater than the desirable value of 0.60
Analysis of the indirect effects presented in Table 7 indicate media­
in all the cases [143]. The average variance extracted, reported in
tion through attitude. Considering a significance of 10%, attitude me­
Table 3, was >0.50 for all the factors, thereby fulfilling the conditions of
diates all constructs that relate to attitude. The strongest mediation is in
Fornell and Larcker [144]. Therefore, the conditions for reliability and
the indirect relationship between perceived enjoyment and intention to
use at a level of 99% confidence. This is followed by the indirect re­
Table 1 lationships between perceived ease of use and perceived usefulness with
Respondent demographics. the intention to use at 0.05 significance. Finally, all the remaining re­
No. Percentage lationships leading to intention to use from price consciousness, per­
sonal innovativeness, perceived risk, and trust were also mediated by
Gender
Male 242 69.14
attitude weakly at a significance of 0.10. Overall, the model had an
Female 108 30.86 explained variance (R2) of 62.1% for the attitude variable and 70.8% for
Experience of Mobile Shopping the intention to use variable.
High (>36 months) 169 48.29
Low (�36 months) 181 51.71
Awareness of Chatbot Technology 3.3. Independent sample t-test
Aware 194 55.43
Unaware 156 44.57
3.3.1. Gender
Age
Younger (�35 years) 176 50.29 The results of the independent sample t-test to compare the mean
Older (>35 years) 174 49.71 responses of the variables between respondents based on gender are
Frequency of Mobile Shopping given in Table 8. Hypotheses H9e and H9g were supported by a high
High (Daily to Weekly) 186 53.14 significance of 5%. A significant difference in the mean responses was
Low (Monthly to Occasionally) 164 46.86
observed between genders for the variables perceived risk and personal

6
D.L. Kasilingam Technology in Society 62 (2020) 101280

Table 2
Reliability and convergent validity.
Factors Individual Factor Average Loading Cronbach’s Alpha Composite Reliability Average Variance Extracted
Loading (AL) (CA) (CR) (AVE)

Attitude (ATT) AT1 0.931 0.943 0.938 0.938 0.960

AT2 0.947
AT3 0.952
Intention to Use (IU) IU1 0.906 0.864 0.888 0.923 0.753
IU2 0.914
IU3 0.913
IU4 0.722
Price Consciousness (PC) PC1 0.747 0.704 0.809 0.856 0.501
PC2 0.780
PC3 0.717
PC4 0.566
PC5 0.649
PC6 0.763
Perceived Enjoyment (PE) PE1 0.926 0.924 0.915 0.946 0.854
PE2 0.942
PE3 0.905
Perceived Ease of Use PEU1 0.706 0.793 0.853 0.895 0.633
(PEU) PEU2 0.722
PEU3 0.830
PEU4 0.878
PEU5 0.827
Personal Innovativeness PI1 0.566 0.759 0.818 0.875 0.588
(PI) PI2 0.831
PI3 0.840
PI4 0.829
PI5 0.731
Perceived Risk (PR) PR1 0.723 0.809 0.868 0.905 0.657
PR2 0.840
PR3 0.849
PR4 0.777
PR5 0.856
Perceived Usefulness (PU) PU1 0.824 0.868 0.918 0.939 0.754
PU2 0.859
PU3 0.903
PU4 0.889
PU5 0.864
Trust (TR) TR1 0.884 0.872 0.947 0.957 0.762
TR2 0.899
TR3 0.877
TR4 0.911
TR5 0.905
TR6 0.778
TR7 0.849

Table 3
Discriminant validity.
ATT IU PC PE PEU PI PR PU TR

ATT 0.943
IU 0.761 0.868
PC 0.237 0.246 0.708
PE 0.706 0.640 0.169 0.924
PEU 0.517 0.462 0.130 0.543 0.796
PI 0.531 0.604 0.249 0.480 0.539 0.767
PR 0.478 0.504 0.054 0.434 0.310 0.286 0.811
PU 0.653 0.609 0.137 0.725 0.487 0.439 0.454 0.868
TR 0.614 0.705 0.192 0.562 0.369 0.461 0.608 0.590 0.873
pffiffiffiffiffiffiffiffiffi
Elements on the diagonal in bold indicate ( AVE).

innovativeness. Chatbots for shopping were considered less risky by significance of 10%. The variables perceived usefulness and perceived
male respondents than by female respondents. Moreover, men were enjoyment were significantly different between age groups in the mean
more innovative than women. Innovators tend to take risks, and this responses. The young respondents believe that chatbots are useful and
may also be the case in the context of chatbots. All remaining hypotheses increase the level of enjoyment in shopping using a mobile phone. Hy­
were rejected. potheses H11g, H11h, and H11i were strongly supported by a signifi­
cance of 0.05. The personal innovativeness, attitude, and the intention
3.3.2. Age to use chatbots for mobile shopping were significantly high for the
The results of the independent sample t-test comparing the mean younger population. Hence, the results prove that age will always be a
responses of the variables between the age groups of the respondents are crucial variable for segment users. Hypotheses H11d and H11f were
given in Table 9. Hypotheses H11a and H11c were supported by a low close to being accepted by a significance of 10%. All remaining

7
D.L. Kasilingam Technology in Society 62 (2020) 101280

Table 4
HTMT ratios.
ATT IU PC PE PEU PI PR PU TR

ATT
IU 0.835
PC 0.256 0.269
PE 0.761 0.703 0.187
PEU 0.576 0.525 0.156 0.607
PI 0.608 0.703 0.297 0.559 0.643
PR 0.527 0.562 0.081 0.481 0.349 0.334
PU 0.702 0.669 0.149 0.790 0.549 0.505 0.503
TR 0.649 0.763 0.220 0.603 0.404 0.526 0.673 0.630

Table 5 Table 7
PLS-SEM path coefficients. PLS-SEM indirect effects path coefficients.
Hypothesis Regression T p Result Indirect Path Regression T Statistics p Values Result
Coefficient Statistics Values Coefficient
H1a PU → 0.167* 2.546 0.011 Supported
PC → ATT → IU 0.03* 1.811 0.070 Supported
ATT
PE → ATT → IU 0.122*** 2.762 0.006 Supported
H1b PU → IU 0.032 0.529 0.597 Not
PEU → ATT → IU 0.04** 1.957 0.050 Supported
Supported
PI → ATT → IU 0.053* 1.833 0.067 Supported
H2a PEU → 0.106* 2.005 0.045 Supported
PR → ATT → IU 0.039* 1.699 0.089 Supported
ATT
PU → ATT → IU 0.063** 1.960 0.049 Supported
H2b PEU → IU 0.026 0.703 0.482 Not
TR → ATT → IU 0.056* 1.884 0.060 Supported
Supported
H3a PE → 0.305*** 4.347 0.000 Supported ***p < 0.01, **p < 0.05, *p < 0.1.
ATT
H3b PE → IU 0.063 1.004 0.315 Not
Supported with a high experience of >3 years in using mobile shopping applica­
H4a PC → 0.082* 2.132 0.033 Supported tions found chatbots for mobile shopping easy to use and showed a
ATT significantly high intention to use the technology than did respondents
H4b PC → IU 0.032 0.931 0.352 Not
with less experience. Hypothesis H13e was strongly supported by a
Supported
H5a PR → ¡0.101* 2.207 0.027 Supported significance of 0.05. The perceived risk in using chatbots for mobile
ATT shopping was significantly high for the respondents with low experience
H5b PR → IU 0.044 1.11 0.267 Not in using mobile shopping applications. Hence, the results prove that
Supported experience level in using mobile shopping applications are useful for
H6a TR → 0.141 1.835 0.067 Not
ATT Supported
segmenting users, and strategies must be implemented accordingly.
H6b TR → IU 0.291*** 4.018 0.000 Supported Hypothesis H13g was accepted by a significance of 10%. All remaining
H7a PI → ATT 0.138* 2.311 0.021 Supported hypotheses were rejected.
H7b PI → IU 0.211*** 4.274 0.000 Supported
H8 ATT → IU 0.388*** 3.984 0.000 Supported
3.4. Multigroup moderation analysis
***p < 0.001, **p < 0.01, *p < 0.05.
3.4.1. Gender
Table 11 shows the results of moderation analysis using gender as the
Table 6
grouping variable. The effect of price consciousness on attitude (H10-
PLS-SEM total effects path coefficients.
4a) was moderated by using gender at a significance of 5% and was
Hypothesis Regression T p Result
significant only for female respondents. The effects of personal inno­
Coefficient Statistics Values
vativeness (H10-7a) and perceived risk (H10-5a) on attitude toward
H1b PU → IU 0.098 1.542 0.123 Not chatbots were moderated by using gender at a significance of 10%. Both
Supported
H2b PEU → 0.014 0.196 0.845 Not
these relationships hold true only for the male respondents and are
IU Supported statistically insignificant for the female population. Moreover, the ef­
H3b PE → IU 0.183** 2.933 0.003 Supported fects of price consciousness (H10-4b) and personal innovativeness (H10-
H4b PC → IU 0.064 1.723 0.085 Not 7b) on the intention to use chatbots for shopping applications were
Supported
moderated by using gender at a significance of 10%. The relationship
H5b PR → IU 0.084 1.878 0.060 Not
Supported between trust and intention to use is stronger among men than among
women. However, although gender moderates the relationship between
***p < 0.001, **p < 0.01, *p < 0.05.
price consciousness and intention to use, it was insignificant for both
cases. All remaining hypotheses were rejected.
hypotheses were rejected.
3.4.2. Age
3.3.3. Experience in using mobile shopping applications Table 12 shows the results of moderation analysis using age as the
Independent sample t-test are given in Table 10 to compare the mean grouping variable. From the results, moderation effect is evident. The
responses of the variables between the respondents with different effects of perceived risk (H12-5b) and trust (H12-6b) on the intention to
experience levels in using mobile shopping applications. Hypotheses use were moderated by using age at a significance of 5%. The rela­
H13b and H13i were supported by a low significance of 10%. The var­ tionship between perceived risk and intention to use was only significant
iables perceived ease of use and intention to use were significantly for the younger age group, whereas the relationship between trust and
different in the mean responses between the respondents with different intention to use was significant for the older age group. The effect of
experience levels in using mobile shopping applications. Respondents perceived usefulness on attitude toward chatbots (H12-1a) was

8
D.L. Kasilingam Technology in Society 62 (2020) 101280

Table 8
Comparison of Mean Responses of Constructs between Gender (Independent Sample t-test Results).
H. Construct Gender (Mean Response) Result
No.
Male Female Mean t-value p Value
Difference

H9a Perceived Usefulness 4.219 4.190 0.029 0.263 0.793 Not Supported
H9b Perceived Ease-of-use 4.703 4.649 0.053 0.585 0.559 Not Supported
H9c Perceived Enjoyment 5.065 5.037 0.028 0.205 0.837 Not Supported
H9d Price Consciousness 3.223 3.114 0.110 1.430 0.154 Not Supported
H9e Perceived Risk 1.425 1.666 0.242** 2.508 0.013 Supported
H9f Trust 3.995 3.912 0.082 0.629 0.530 Not Supported
H9g Personal Innovativeness 3.332 3.179 0.153** 2.277 0.023 Supported
H9h Attitude 4.991 4.855 0.137 1.077 0.282 Not Supported
H9i Intention to Use 5.580 5.407 0.173 1.216 0.225 Not Supported

Note: *p < 0.10; **p < 0.05; ***p < 0.01.

Table 9
Comparison of Mean Responses of Constructs between Age Groups (Independent Sample t-test Results).
H. Construct Age (Mean Response) Result
No.
Younger Elder Mean t-value p Value
Difference

H11a Perceived Usefulness 4.298 4.121 0.177* 1.718 0.087 Supported

H11b Perceived Ease-of-use 4.709 4.662 0.047 0.557 0.578 Not Supported
H11c Perceived Enjoyment 5.174 4.938 0.236* 1.890 0.060 Supported
H11d Price Consciousness 3.244 3.134 0.110 1.554 0.121 Not Supported
H11e Perceived Risk 1.497 1.502 0.005 0.055 0.960 Not Supported
H11f Trust 4.064 3.873 0.191 1.585 0.114 Not Supported
H11g Personal Innovativeness 3.352 3.216 0.136** 2.187 0.029 Supported
H11h Attitude 5.067 4.830 0.237** 2.031 0.043 Supported
H11i Intention to Use 5.668 5.383 0.285** 2.180 0.030 Supported

Note: *p < 0.10; **p < 0.05; ***p < 0.01.

Table 10
Comparison of Mean Responses of Constructs between Experience in Using Mobile Shopping Applications (Independent Sample t-test Results).
H. Construct Experience (Mean Response) Result
No.
Low High Mean t-value p Value
Difference

H13a Perceived Usefulness 4.158 4.266 0.108 1.049 0.295 Not Supported
H13b Perceived Ease-of-use 4.615 4.762 0.147* 1.750 0.081 Supported
H13c Perceived Enjoyment 5.055 5.058 0.004 0.032 0.978 Not Supported
H13d Price Consciousness 3.171 3.209 0.038 0.538 0.591 Not Supported
H13e Perceived Risk 1.586 1.406 0.180** 2.018 0.044 Supported
H13f Trust 3.883 4.061 0.178 1.469 0.143 Not Supported
H13g Personal Innovativeness 3.237 3.336 0.099 1.588 0.113 Not Supported
H13h Attitude 4.870 5.034 0.164 1.402 0.162 Not Supported
H13i Intention to Use 5.422 5.639 0.217* 1.656 0.099 Supported

Note: *p < 0.10; **p < 0.05; ***p < 0.01.

moderated by age group at a significance of 10%. This relationship was insignificant for the other group.
significant only for the older age group. All remaining hypotheses were
rejected. 4. Discussion

3.4.3. Experience in using mobile shopping applications As introduced in our theoretical framework, the TAM model [129]
Table 13 shows the results of moderation analysis based on experi­ concludes that consumers are likely to accept technology based on fac­
ence in using mobile shopping applications. The effects of perceived ease tors of enjoyment, usefulness, and ease of use. Consumers are willing to
of use (H14-2a), perceived enjoyment (H14-3a), perceived risk (H14- adopt a technology that is new and exciting, deemed as an enjoyable
5a), and trust (H14-6a) on attitude were significantly moderated by experience. Retailers adopt a different approach while adopting a
users’ experiences in using mobile shopping applications. The effect of technology. Online retailers invest in technology to gain a competitive
perceived enjoyment on attitude toward chatbots was significantly advantage, appeal to their consumer base, and ultimately thrive
strong for respondents with increased experience in using mobile economically. Whether the consumer will like or benefit from the new
shopping applications. The relationship between perceived risk and technology is not the primary aim. Instead, retailers must maintain
attitude was significant only for users with considerable experience. The healthy margins and increase topline numbers. This influences tech­
effect of perceived ease of use, trust, and price sensitivity on attitude nology adoption in two ways. First, it creates hesitation in taking risks
toward chatbots was significantly strong for respondents with low and, second, it gives large, well established companies a high control.
experience in using mobile shopping applications. Moreover, all three The second fact is evident with companies such as Apple, Google, and
relationships were true only for users with low experience and was Amazon, who are known for their large R&D budgets and their

9
D.L. Kasilingam Technology in Society 62 (2020) 101280

Table 11 Table 12
Multigroup analysis with gender as a moderator. Multigroup analysis with age as a moderator.
H. Hypothesis Gender Result H. Hypothesis Age Result
No. No.
Path Coefficient Path Path Coefficient Path
Coefficient Coefficient
Male Female Younger Elder
Difference Difference

H10- PU → ATT 0.170** 0.108 0.062 Not H12- PU → ATT 0.080 0.271*** 0.192* Supported
1a Supported 1a
H10- PU → IU 0.010 0.046 0.036 Not H12- PU → IU 0.052 0.038 0.013 Not
1b Supported 1b Supported
H10- PEU → 0.065 0.203* 0.138 Not H12- PEU → 0.101 0.089 0.012 Not
2a ATT Supported 2a ATT Supported
H10- PEU → IU 0.026 0.041 0.015 Not H12- PEU → IU 0.019 0.030 0.011 Not
2b Supported 2b Supported
H10- PE → ATT 0.335*** 0.315** 0.02 Not H12- PE → ATT 0.265** 0.325*** 0.060 Not
3a Supported 3a Supported
H10- PE → IU 0.015 0.183 0.169 Not H12- PE → IU 0.039 0.097 0.059 Not
3b Supported 3b Supported
H10- PC → ATT 0.012 0.182** 0.169** Supported H12- PC → ATT 0.040 0.112** 0.072 Not
4a 4a Supported
H10- PC → IU 0.004 0.119 0.115* Supported H12- PC → IU 0.026 0.053 0.027 Not
4b 4b Supported
H10- PR → ATT ¡0.144** ¡0.007 0.137* Supported H12- PR → ATT 0.106 0.088 0.017 Not
5a 5a Supported
H10- PR → IU 0.021 0.044 0.023 Not H12- PR → IU ¡0.163*** 0.037 0.201** Supported
5b Supported 5b
H10- TR → ATT 0.098 0.301* 0.203 Not H12- TR → ATT 0.223 0.084 0.139 Not
6a Supported 6a Supported
H10- TR → IU 0.356*** 0.227* 0.128 Not H12- TR → IU 0.112 0.405*** 0.293** Supported
6b Supported 6b
H10- PI → ATT 0.189*** ¡0.002 0.191* Supported H12- PI → ATT 0.156* 0.115 0.041 Not
7a 7a Supported
H10- PI → IU 0.150*** 0.325*** 0.175* Supported H12- PI → IU 0.244*** 0.183*** 0.061 Not
7b 7b Supported
H10- ATT → IU 0.475*** 0.199 0.276 Not H12- ATT → IU 0.449*** 0.310*** 0.139 Not
8a Supported 8a Supported

Note: *p < 0.10; **p < 0.05; ***p < 0.01. Note: *p < 0.10; **p < 0.05; ***p < 0.01.

willingness and ability to invest in innovative technologies. rates. This can be achieved through keeping the number of taps to a
From the results of the analysis, it is evident that the factors price minimum, making the information easy to read, and minimizing input
consciousness, perceived enjoyment, perceived ease of use, personal from the user. A strenuous and slow process can often be an obstacle for
innovativeness, perceived risk, and perceived usefulness influence the the customer. Therefore, consumers must find the transaction faster and
intention to use chatbots indirectly through forming attitudes. Trust more convenient than the mobile shopping application. A significant
does not influence their attitudes but is one of the core variables influ­ number of consumers would, for example, use chatbots to shop on their
encing the intention to use chatbots for mobile shopping. mobiles if they did not have to input their information such as credit
The majority of the respondents in the survey thought that all card numbers and shipping location with every purchase. Therefore,
questions relative to usefulness were essential, although it did not storing relevant information for the consumer is recommended while
strongly influence attitude and intention to use chatbots for shopping considering the importance of privacy and safety of personal
when compared with the enjoyment variable. Kim, Chan and Gupta information.
[149] explained that the consumer’s knowledgeable valuation of the Attitude toward using chatbots for shopping was the most signifi­
quality or supremacy of a technology defines its usefulness. Thus, if a cantly influenced by perceived enjoyment. Perceived enjoyment is an
consumer does not consider chatbots for mobile shopping better or su­ intrinsic motivation. If individuals enjoy shopping on their mobile de­
perior to traditional means of mobile shopping, he/she may not consider vice using chatbots and ultimately reap the rewards from it, they will
it useful [149]. Chatbots for mobile shopping must have more perceived continue using that technology. If the interface is attractive, customers
usefulness than their substitutes to create an intention to adopt. would be happy and would be interested in using the technology in
Improving the attitude of consumers toward perceived usefulness can future. Different aspects of technology interface design include its
enhance adoption and approval rating of chatbots by consumers [149]. layout, color, background pictures, and text size. Moreover, chatbot
Furthermore, respondents want a technology to be useful according to developers can provide customers options to customize the interface to
the effectiveness usability criteria by Seffah and Metzker [150], con­ match their requirements or personality. This includes the capability of
cerning the extent to which it helps the respondents to achieve and changing the background color, background picture, and text size. This
complete their goals or task with the mobile system [150]. enhances users’ experience, and they would continue using chatbots for
Ease of use is an essential factor that affects the adoption intent of mobile shopping in future.
consumers in the TAM model. In the survey, it was crucial for re­ Analysis of essential dimensions of the level of innovativeness of
spondents to know that the chatbot technology is easy to use. Under­ customers shows that self-perception of the customers is an essential
standing the prejudice of the consumers is vital because technicality can aspect as well. If customers consider themselves unconventional, they
negative affect adoption intentions. A unified and cohesive experience will tend to use new technologies that are introduced in the market­
must be created for consumers, and they must be enabled to access all place. This self-perception of customers may encourage them to leave
requisite aspects such as offers, discounts, and coupons directly within old or routine technologies like web-based mobile commerce or mobile
the chatbot. Streamlining the checkout process and making it easy for shopping applications. Such people generally tend to be unique, and so,
the consumers to use the technology would certainly influence adoption they do not hesitate to let go of popular technologies to explore new

10
D.L. Kasilingam Technology in Society 62 (2020) 101280

Table 13 Moreover, developers can focus on improving information privacy as it

Multigroup analysis with experience in using mobile shopping applications as a can also help in reducing risks and ultimately lead to adoption [57].
moderator. Trust is one of the most crucial variables, which directly influences
H. Hypothesis Experience Result the intention to use chatbots for mobile shopping. Trust and risk
No.
Path Coefficient Path
generally go hand in hand. When trust toward a technology increases,
Coefficient the perceived risks associated with it significantly reduces. However, the
Low High
Difference importance of perceived ease of use and perceived usefulness is high­
H14- PU → ATT 0.137* 0.181* 0.043 Not lighted in this work; people rely on trust and perceived risk when using
1a Supported chatbots for online shopping. Moreover, trust plays a major role in the
H14- PU → IU 0.018 0.089 0.107 Not process of decision-making for customers who want to purchase some
1b Supported items online [149]. Furthermore, many studies on online consumer
H14- PEU → 0.172*** 0.010 0.163* Supported
2a ATT
behavior have indicated that perceived risk negatively influences con­
H14- PEU → IU 0.007 0.063 0.069 Not cepts such as attitude toward internet shopping [97,154,155]. This
2b Supported research proves that the relationship is also true in the case of chatbots
H14- PE → ATT 0.199** 0.454*** 0.255* Supported for online shopping.
3a
Previously developed theories posit that an attitude is an educated
H14- PE → IU 0.057 0.065 0.008 Not
3b Supported inclination to evaluate an object to be favorable or unfavorable. A
H14- PC → ATT 0.108** 0.062 0.046 Not positive attitude is recognized by a favorable evaluation of an object
4a Supported [156]. The overall means of the items within the factors display an
H14- PC → IU 0.003 0.066 0.07 Not overall positive attitude toward the chatbot technology for mobile
4b Supported
H14- PR → ATT ¡0.012 ¡0.191*** 0.178** Supported
shopping. It can be then presumed that the respondents have positive
5a views about the technology based on their positive attitudes [91].
H14- PR → IU ¡0.059 ¡0.009 0.05** Supported This study aimed to identify significant differences in the strengths of
5b the relationships between gender, age groups, and prior experience in
H14- TR → ATT 0.240** 0.011 0.230* Supported
using mobile shopping applications. Moreover, it explored differences in
6a
H14- TR → IU 0.274*** 0.301*** 0.027 Not the mean responses for the variables between gender, age groups, and
6b Supported experience in using mobile applications for shopping. The mean com­
H14- PI → ATT 0.139* 0.129* 0.009 Not parison using independent sample t-test revealed that the younger
7a Supported population is more innovative and perceives chatbot technology for
H14- PI → IU 0.237*** 0.209*** 0.028 Not
7b Supported
mobile shopping to be more useful and enjoyable than the older one.
H14- ATT → IU 0.395*** 0.383*** 0.012 Supported Furthermore, they have a significantly favorable attitude and the
8a intention to use the technology when compared with the elderly popu­
Note: *p < 0.10; **p < 0.05; ***p < 0.01.
lation. This again supports findings of the previous literature on the age
effects [157–161].
Regarding gender, men were significantly more innovative and
technologies such as chatbots, which are not popular among the mass.
consider new chatbot technology significantly less risky than women
Furthermore, findings related to the level of innovativeness of customers
did, consistent with earlier studies on technology adoption [57]. As
show that if the level of innovation among customers is high, then they
chatbot technology is going to replace mobile applications for shopping,
tend to experiment new things. As chatbots for mobile shopping is a new
the influence of prior experience in using mobile application must be
technology in the market, a significant majority are yet to try it for the
understood. Results showed that people with more than 3 years of
first time. Generally, when retailers promote their technologies, they
experience in using mobile applications perceived chatbots to be easy to
highlight benefits associated with the user. It would create a situation
use and less risky. Moreover, they seemed to have a significantly high
where customers would abandon traditional methods of shopping or
intention to use chatbots for mobile shopping. This proves that mobile
start using the new chatbot technology for mobile shopping in addition
applications have been in the market long enough, and people are
to traditional mobile commerce technologies. Such people continue
looking for newer technologies for their shopping activities. Younger
browsing information about new products, services, and technologies. If
people with significant experience in using mobile shopping applica­
any technology matches their interests and requirements, they will tend
tions can be used as a sample to pilot study the chatbot technology
to use it. Findings of this research are consistent with that of the study by
before launching it in the market by companies as this population pri­
Siu and Chang [151]. They highlighted that if a person is curious, then
marily has a favorable attitude and intention to use the technology.
he/she would attempt new things in life. Likewise, if a person intends to
Moreover, they will act as innovators and positively influence the pop­
experiment new things, he/she is innovative.
ulation to use chatbots through word of mouth.
Moreover, findings of this research corroborate the work of Lin,
Moderation analysis revealed few exciting findings. The variables
Marshall and Dawson [152] who suggested that customers purchase
perceived usefulness, perceived risk, and personal innovativeness in­
products and use technologies to promote their self-image. This study
fluence attitude toward chatbot technologies for mobile shopping only
found that customers want to use chatbots for mobile shopping as they
among men. Similarly, perceived ease of use, price consciousness, and
perceive themselves to be risk takers and innovative. Likewise, findings
trust on attitude toward chatbots significantly influenced only men.
of this research were aligned to the work of Siu and Chang [151] who
Attitude significantly influenced the intention to use chatbots for
suggested that if a person tends to try new technologies, then his level of
shopping only among male consumers. Hence, marketers could tailor
innovativeness is high.
variables to cater the chatbots to the targeted gender. When using
Furthermore, this study shows that perceived risk is negatively
chatbot technologies for mobile shopping, attitude and intention to use
associated with attitude toward chatbots for shopping. Although this
of the older population can be influenced by concentrating on the var­
relationship is not significant, the risk is still a crucial variable. Hence,
iable perceived usefulness and trust, whereas these relationships remain
marketers must reduce the risk involved in using chatbots to the greatest
insignificant for the younger population. The influence of the perceived
extent. A perception of a high level of security in transactions would lead
risk on the intention to use is significant only among younger users, and
to a high intention to purchase from that e-commerce business [153],
hence, implementing risk-reducing actions is necessary to attract these
and this can be expected in the case of chatbots for mobile shopping also.
target consumers. The price consciousness of older users significantly

11
D.L. Kasilingam Technology in Society 62 (2020) 101280

influences attitude toward chatbots, whereas the relationship between variations. Although the study explains >72% of the variance associated
attitude and personal innovativeness remains significant only for the with adoption, variables can differ over time as people begin to adopt
younger age group. Furthermore, prior experience in using mobile and use the technology, and hence, a longitudinal research would
shopping applications moderates the relationships. The effect of certainly provide interesting results. Although the sample was presumed
perceived risk on attitude was significant only for people with signifi­ to represent the population, a large sample size can reduce the proba­
cant experience in using mobile shopping applications. The effect of bility of error in the estimation, and hence, future researchers can
perceived ease of use and trust on attitude was significant only for attempt to validate the model with larger datasets. Chatbot application
people with little experience in using mobile shopping applications. in business is truly immense, and this study analyzed chatbot application
Comparing the mean responses of common variables in the studies in mobile-based commerce. Identifying applications of mobile chatbots
by Natarajan, Balasubramanian and Kasilingam [19] and Natarajan, in a B2B perspective and understanding adoption would indeed be
Balasubramanian and Kasilingam [18] with the those in this study interesting. Variables such as internet speed and mobile device used for
revealed that the perceived ease of use, usefulness, and enjoyment are chatbot functions may significantly moderate the relationships in the
low with chatbot technologies compared with mobile shopping appli­ model, and hence, future researchers can explore these dimensions.
cations. However, the intention to use chatbots for shopping is higher Overall, this research is one of the first to propose a generic model that
than to use mobile shopping applications. Hence, consumers are ready to captures a vast variance in technology adoption specific to a futuristic
adopt new technologies for mobile shopping. Mobile applications are technology. Organizations can hence use this model to develop and
undoubtedly the best mode of purchasing products and services on pivot their strategies accordingly.
mobile in the current market. However, the high difference in the in­
tentions toward the technologies suggests a market shift soon, and Appendix A. Supplementary data
chatbots may replace mobile applications for shopping.
The implications of this research to the management is that managers Supplementary data to this article can be found online at https://doi.
and marketers must pay special attention to perceived usefulness, ease org/10.1016/j.techsoc.2020.101280.
of use, enjoyment, risk, and trust variables and should ensure that these
factors are met with respect to the demographic profile of the users to Appendix
increase the acceptance and usage of chatbots for mobiles shopping. As
perceived ease of use is one of the most significant factors affecting the Perceived Risk—Featherman and Pavlou [62].
intention to use chatbots for mobile shopping, marketers and retailers
must, when designing their chatbots, focus on features that would pro­ Chatbots for shopping may not perform well and process payments
vide users with reliable and quality information regarding products and incorrectly.
trends and exploit to the fullest unique features, such as the ability to Using chatbots would lead to payment uncertainty.
shop and browse anytime and anywhere, use the bar scanner, use GPS to The security systems built into the chatbots may not be strong
set shipping address, provide auto pay options, help locate the nearest enough to protect my account.
stores, and provide personalized chat notifications regarding the latest Internet hackers (criminals) might take control of my account if I
products. This can attract women users and consumers with low expe­ used chatbots for shopping.
rience as these groups are tricky to handle and low in numbers. More­ My decision to use chatbots for shopping involves a high risk.
over, doing so will naturally increase the intentions and attitudes of the
other groups as they are more innovative than the aforementioned Perceived Usefulness—Davis [10].
groups. The better the chatbot support in shopping activities, the more
attractive they would be. Chatbots for shopping will be useful to me.
Managers should ensure that chatbots add value to their users by Using chatbots for shopping will enable me to accomplish trans­
guaranteeing that they optimize their output by saving shopping time, actions quickly.
obtaining promotion information, and obtaining customized products Using chatbots for shopping will increase my productivity.
information. Furthermore, previous experience in using mobile shop­ Using chatbots for shopping will enhance my effectiveness.
ping is an essential factor. Therefore, managers could refer to the design Using chatbots for shopping would enable me to accomplish shop­
of a prevalent computer software and popular shopping websites as the ping tasks fast.
more familiar the people are with using a particular technology, the
quicker it becomes a habit. Chatbot being a relatively new technology, Perceived Ease of Use—Davis [10].
marketers should provide online and offline support for using the apps,
tutorials, and FAQs as these facilitate chatbot use. Likewise, for Shopping using chatbots does not require great mental effort.
perceived enjoyment construct, managers could enforce appropriate Ithink I will be able to shop using chatbots without the help of an
functions on chatbots to create an enjoyable and entertaining shopping expert.
experience. For instance, enjoyment can be enhanced through animated Learning to operate chatbots for shopping will be easy for me.
features, fun content, and non-conventional interaction with the chat­ Overall, I believe that using chatbots for shopping is easy.
bots. As chatbots use advanced data mining and big data analytics as an Working with chatbots is not complicated; it is easy to understand
integral part, respondents can be profiles to create a personalized and what is going on.
self-catered experience for them, thereby retaining user adoption levels.
Perceived Enjoyment—Davis, Bagozzi and Warshaw [129].
5. Conclusion
Ifind using chatbots for shopping enjoyable.
In the context of chatbots for mobile shopping, a relatively new The actual process of using chatbots for shopping is pleasant.
technology that is yet to be released in the market, TAM and DOI Iwill have fun while using chatbots for shopping.
continue to be robust theories for explaining attitude and usage in­
tentions. The study results can be generalized across the county, mainly Personal Innovativeness—Goldsmith and Hofacker [130].
because of the diversity of the sample. Moreover, it can be generalized
for developing markets. However, empirical validation may be neces­ Ithink I know more about chatbots than my circle of friends
sary for developed markets, mainly due to cultural and economic

12
D.L. Kasilingam Technology in Society 62 (2020) 101280

If I heard about a new application/technology like chatbots, and I [8] G.L. Lilien, A. Rangaswamy, Modeled to bits: decision models for the digital,
networked economy, Int. J. Res. Market. 17 (2) (2000) 227–235.
would look for ways to experiment with it.
[9] V.W. Mitchell, Questioning the role of descriptor variables in market
segmentation, J. Target Meas. Anal. Market. 5 (1996) 95–103.
Among my peers, I am usually the first to try out new applications/ [10] F. Davis, Perceived usefulness, perceived ease of use, and user acceptance of
technologies such as chatbots. information technology, MIS Q. 13 (3) (1989) 319–340.
[11] C. Kim, M. Mirusmonov, I. Lee, An empirical examination of factors influencing
the intention to use mobile payment, Comput. Hum. Behav. 26 (3) (2010)
In general, I am hesitant to try out new applications/technologies 310–322.
such as chatbots. [12] V. Venkatesh, F.D. Davis, A theoretical extension of the technology acceptance
model: four longitudinal field studies, Manag. Sci. 46 (2) (2000) 186–204.
Ilike to experiment with new applications/technologies such as [13] V. Venkatesh, M.G. Morris, G.B. Davis, F.D. Davis, User acceptance of information
chatbots. technology: toward a unified view, MIS Q. (2003) 425–478.
[14] P.G. Schierz, O. Schilke, B.W. Wirtz, Understanding consumer acceptance of
mobile payment services: an empirical analysis, Electron. Commer. Res. Appl. 9
Trust—Chong, Chan and Ooi [131] and Tsu Wei, Marthandan, Yee (3) (2010) 209–216.
Loong Chong, Ooi and Arumugam [33]. [15] S.T. Chuang, Y.R. Lin, J.H. Huang, Elucidating user behavior of mobile learning: a
perspective of the extended technology acceptance model, Electron. Libr. 25 (5)
(2007) 585–598.
Ibelieve payments made through chatbots will be processed securely. [16] P. Eunil, J.K. Ki, User acceptance of long-term evolution (LTE) services: an
Ibelieve transactions conducted through chatbots will be secure. application of extended technology acceptance model, Program 47 (2) (2013)
Ibelieve my personal information will be kept confidential while 188–205.
[17] S. Kim, G. Garrison, Investigating mobile wireless technology adoption: an
using chatbots.
extension of the technology acceptance model, Inf. Syst. Front 11 (3) (2009)
Iam confident regarding the security measurements offered by 323–333.
chatbots. [18] T. Natarajan, S.A. Balasubramanian, D.L. Kasilingam, The moderating role of
Privacy chatbots are well protected. device type and age of users on the intention to use mobile shopping applications,
Technol. Soc. 53 (2018) 79–90.
Ido not worry about providing credit card information to chatbots. [19] T. Natarajan, S.A. Balasubramanian, D.L. Kasilingam, Understanding the
Chatbots is as secure as any e-commerce or m-commerce website. intention to use mobile shopping applications and its influence on price
sensitivity, J. Retailing Consum. Serv. 37 (2017) 8–22.
[20] E.M. Rogers, Diffusion of Innovations, Simon and Schuster2010.
Price Consciousness—Dickinger and Kleijnen [132], Lichtenstein, [21] A. Dillon, M.G. Morris, User Acceptance of New Information Technology:
Ridgway and Netemeyer [133], and Swaminathan and Bawa [134]. Theories and Models, Annual Review of Information Science and Technology,
Information, Medford, NJ, 1996. Today.
[22] R. Agarwal, E. Karahanna, Time flies when you’re having fun: cognitive
Ifind myself checking prices in stores even for small items. absorption and beliefs about information technology usage, MIS Q. 24 (4) (2000)
A person can save ample money by shopping around for bargains. 665–694.
Iam not willing to put extra efforts to find items with lower prices. [23] R. Agarwal, J. Prasad, A conceptual and operational definition of personal
innovativeness in the domain of information technology, Inf. Syst. Res. 9 (2)
Itry to buy inexpensive products. (1998) 204–215.
A low price is the most important factor influencing my purchasing [24] Y.-H. Lee, Y.-C. Hsieh, C.-N.J.J.o.E.T. Hsu, Society, Adding innovation diffusion
decisions. theory to the technology acceptance model: supporting employees’ intentions to
use, e-learning systems 14 (4) (2011).
Ivalue low prices for various products.
[25] K. Phan, T. Daim, Exploring technology acceptance for mobile services, J. Ind.
Eng. Manag. 4 (No 2) (2011), https://doi.org/10.3926/jiem..v4n2.p339-360,
Attitude—Taylor and Todd [135]. 2011.
[26] L. Ya Ching, An empirical investigation into factors influencing the adoption of an
e-learning system, Online Inf. Rev. 30 (5) (2006) 517–541.
Using chatbots for shopping is a good idea. [27] H.P. Lu, P. Yu Jen Su, Factors affecting purchase intention on mobile shopping
Ilike using chatbots for shopping. web sites, Internet Res. 19 (4) (2009) 442–458.
Using chatbots for shopping would be pleasant. [28] K. Yang, Determinants of US consumer mobile shopping services adoption:
implications for designing mobile shopping services, J. Consum. Market. 27 (3)
(2010) 262–270.
Intention to Use—Kim, Mirusmonov and Lee [11]. [29] J.H. Wu, S.C. Wang, What drives mobile commerce?: an empirical evaluation of
the revised technology acceptance model, Inf. Manag. 42 (5) (2005) 719–729.
[30] M. Khalifa, K. Ning Shen, Explaining the adoption of transactional B2C mobile
Now I intend to use chatbots to shop and procure products. commerce, J. Enterprise Inf. Manag. 21 (2) (2008) 110–124.
Assuming that I have access to chatbots, I intend to use it. [31] J. Ald�
as Manzano, C. Lassala Navarr� e, C. Ruiz Maf�e, S. Sanz Blas, The role of
During the next 6 months, I intend to use chatbots for shopping. consumer innovativeness and perceived risk in online banking usage, Int. J. Bank
Market. 27 (1) (2009) 53–75.
Iintend to use chatbots for shopping after 5 years. [32] H. Dai, P.C. Palvi, Mobile commerce adoption in China and the United States: a
cross-cultural study, SIGMIS Database 40 (4) (2009) 43–61.
[33] T. Tsu Wei, G. Marthandan, A. Yee Loong Chong, K.B. Ooi, S. Arumugam, What
drives Malaysian m-commerce adoption? An empirical analysis, Ind. Manag. Data
References
Syst. 109 (3) (2009) 370–388.
[34] L. Zhang, J. Zhu, Q. Liu, A meta-analysis of mobile commerce adoption and the
[1] K. DuHadway, Why Chatbots Are the Future of Market Research, 2017. https:// moderating effect of culture, Comput. Hum. Behav. 28 (5) (2012) 1902–1911.
www.ama.org/publications/eNewsletters/Marketing-News-Weekly/Pages/wh [35] A. Goyal, R. Thakur, M. Srivastava, Customer usage intention of mobile
y-chatbots-are-the-future-of-market-research.aspx. commerce in India: an empirical study, Journal of Indian Business Research 5 (1)
[2] DMU, Why Will Super Messaging App Overpass Classic Mobile Application in the (2013) 52–72.
Next Few Years?, 2018. http://digital-me-up.com/2018/02/09/will-super-mess [36] P. Zahir Irani, S. Mohamed Fadel Bukhari, A. Ghoneim, C. Dennis, B. Jamjoom,
aging-app-overpass-classic-mobile-application-next-years/. The antecedents of travellers’ e-satisfaction and intention to buy airline tickets
[3] Medium, the Next Bot: the Future of Chatbot, AI and Messaging, 2017. https://me online, J. Enterprise Inf. Manag. 26 (6) (2013) 624–641.
dium.com/iquii/the-next-bot-the-future-of-chatbot-ai-and-messaging-f42b41b3e [37] S. Yuan, Y. Liu, R. Yao, J. Liu, An investigation of users continuance intention
00c. towards mobile banking in China, Inf. Dev. 32 (1) (2014) 20–34.
[4] D. Holt, 5 Reasons Millennials Love Chatbots, 2016. https://venturebeat.com/ [38] S. Agrebi, J. Jallais, Explain the intention to use smartphones for mobile
2016/08/18/5-reasons-millennials-love-chatbots/. shopping, J. Retailing Consum. Serv. 22 (2015) 16–23.
[5] M.v. Eeuwen, Mobile Conversational Commerce: Messenger Chatbots as the Next [39] A.U. Jan, V. Contreras, Technology acceptance model for the use of information
Interface between Businesses and Consumers, University of Twente, 2017. technology in universities, Comput. Hum. Behav. 27 (2) (2011) 845–851.
[6] C. Messina, Will Be the Year of Conversational Commerce, 2016, 2016, https [40] W.R. King, J. He, A meta-analysis of the technology acceptance model, Inf.
://medium.com/chris-messina/2016-will-be-the-year-of-conversational-comme Manag. 43 (6) (2006) 740–755.
rce-1586e85e3991. [41] T.S. Teo, V.K. Lim, R.Y. Lai, Intrinsic and extrinsic motivation in Internet usage,
[7] P.S.H. Leeflang, D.R. Wittink, Building models for marketing decisions:: past, Omega 27 (1) (1999) 25–37.
present and future, Int. J. Res. Market. 17 (2) (2000) 105–126.

13
D.L. Kasilingam Technology in Society 62 (2020) 101280

[42] S.H. Kim, Moderating effects of Job Relevance and Experience on mobile wireless [77] E.M. Rogers, F.F. Shoemaker, Communication of Innovations, A Cross-Cultural
technology acceptance: adoption of a smartphone by individuals, Inf. Manag. 45 Approach, 1971.
(6) (2008) 387–393. [78] E. AbuShanab, J.M. Pearson, A. Setterstrom, Internet banking and customers’
[43] I. Ha, Y. Yoon, M. Choi, Determinants of adoption of mobile games under mobile acceptance in Jordan: the unified model’s perspective, Commun. Assoc. Inf. Syst.
broadband wireless access environment, Inf. Manag. 44 (3) (2007) 276–286. 26 (1) (2010) 23.
[44] T. Teo, Is there an attitude problem? Reconsidering the role of attitude in the [79] H. Xu, S. Gupta, The effects of privacy concerns and personal innovativeness on
TAM, Br. J. Educ. Technol. 40 (6) (2009) 1139–1141. potential and experienced customers’ adoption of location-based services,
[45] A. Dickinger, M. Arami, D. Meyer, The role of perceived enjoyment and social Electron. Mark. 19 (2) (2009) 137–149.
norm in the adoption of technology with network externalities, Eur. J. Inf. Syst. [80] A. Varma Citrin, D.E. Sprott, S.N. Silverman, D.E. Stem, Adoption of Internet
17 (1) (2008) 4–11. shopping: the role of consumer innovativeness, Ind. Manag. Data Syst. 100 (7)
[46] C. Kim, W. Li, D.J. Kim, An empirical analysis of factors influencing M-shopping (2000) 294–300.
use, Int. J. Hum. Comput. Interact. 31 (12) (2015) 974–994. [81] E. Juaneda-Ayensa, A. Mosquera, Y. Sierra Murillo, Omnichannel customer
[47] E.C. Chang, Y.F. Tseng, Research note: E-store image, perceived value and behavior: key drivers of technology acceptance and use and their effects on
perceived risk, J. Bus. Res. 66 (7) (2013) 864–870. purchase intention, Front. Psychol. 7 (1117) (2016).
[48] E. Young Kim, Y.K. Kim, Predicting online purchase intentions for clothing [82] H. San Martín, A.� Herrero, Influence of the user’s psychological factors on the
products, Eur. J. Market. 38 (7) (2004) 883–897. online purchase intention in rural tourism: integrating innovativeness to the
[49] J. Moon, D. Chadee, S. Tikoo, Culture, product type, and price influences on UTAUT framework, Tourism Manag. 33 (2) (2012) 341–350.
consumer purchase intention to buy personalized products online, J. Bus. Res. 61 [83] Y.B. Chiu, C.P. Lin, L.L. Tang, Gender differs: assessing a model of online purchase
(1) (2008) 31–39. intentions in e-tail service, Int. J. Serv. Ind. Manag. 16 (5) (2005) 416–435.
[50] S.-J. Yoon, The antecedents and consequences of trust in online-purchase [84] M.C. Hung, H.G. Hwang, T.C. Hsieh, An exploratory study on the continuance of
decisions, J. Interact. Market. 16 (2) (2002) 47–63. mobile commerce: an extended expectation-confirmation model of information
[51] R.E. Goldsmith, S.J. Newell, Innovativeness and price sensitivity: managerial, system use, Int. J. Mobile Commun. 5 (4) (2007) 409–422.
theoretical and methodological issues, J. Prod. Brand Manag. 6 (3) (1997) [85] J. Ald�as Manzano, C. Ruiz Maf�e, S. Sanz Blas, Exploring individual personality
163–174. factors as drivers of M-shopping acceptance, Ind. Manag. Data Syst. 109 (6)
[52] C. Ofir, Reexamining latitude of price acceptability and price thresholds: (2009) 739–757.
predicting basic consumer reaction to price, J. Consum. Res. 30 (4) (2004) [86] J. Lu, Are personal innovativeness and social influence critical to continue with
612–621. mobile commerce? Internet Res. 24 (2) (2014) 134–159.
[53] L. Lu, D. Bock, M. Joseph, Green marketing: what the Millennials buy, J. Bus. [87] A.
� Herrero Crespo, I. Rodríguez del Bosque, The effect of innovativeness on the
Strat. 34 (6) (2013) 3–10. adoption of B2C e-commerce: a model based on the Theory of Planned Behaviour,
[54] D. Papies, M. Clement, Adoption of new movie distribution services on the Comput. Hum. Behav. 24 (6) (2008) 2830–2847.
Internet, J. Media Econ. 21 (3) (2008) 131–157. [88] A. Athiyaman, Internet users’ intention to purchase air travel online: an empirical
[55] S. Plowman, S. Goode, Factors affecting the intention to download music: quality investigation, Market. Intell. Plann. 20 (4) (2002) 234–242.
perceptions and downloading intensity, J. Comput. Inf. Syst. 49 (4) (2009) 84–97. [89] I. Ajzen, The theory of planned behavior, Organ. Behav. Hum. Decis. Process. 50
[56] C. Park, J.K. Jun, A cross-cultural comparison of Internet buying behavior: effects (2) (1991) 179–211.
of Internet usage, perceived risks, and innovativeness, Int. Market. Rev. 20 (5) [90] I. Ajzen, From Intentions to Actions: A Theory of Planned Behavior, Action
(2003) 534–553. control, Springer1985, pp. 11-39.
[57] D.J. Kim, D.L. Ferrin, H.R. Rao, A trust-based consumer decision-making model in [91] M. Fishbein, I. Ajzen, Belief, Attitude, Intention and Behavior: an Introduction to
electronic commerce: the role of trust, perceived risk, and their antecedents, Theory and Research, 1975.
Decis. Support Syst. 44 (2) (2008) 544–564. [92] S. Ahmad, S.H. Bhatti, Y. Hwang, E- Service Quality and Actual Use of E-Banking:
[58] R. Thakur, M. Srivastava, Adoption readiness, personal innovativeness, perceived Explanation through the Technology Acceptance Model, Information
risk and usage intention across customer groups for mobile payment services in Development vol. 17.
India, Internet Res. 24 (3) (2014) 369–392. [93] A. Shuhaiber, I. Mashal, Understanding users’ acceptance of smart homes,
[59] Y. Chen, X. Yan, W. Fan, M. Gordon, The joint moderating role of trust propensity Technol. Soc. 58 (2019) 9.
and gender on consumers’ online shopping behavior, Comput. Hum. Behav. 43 [94] N. Ahmad, H. Abdulkarim, The impact of flow experience and personality type on
(2015) 272–283. the intention to use virtual world, Int. J. Hum. Comput. Interact. 35 (12) (2019)
[60] A.C. Teo, G.W.H. Tan, K.B. Ooi, B. Lin, Why consumers adopt mobile payment? A 1074–1085.
partial least squares structural equation modelling (PLS-SEM) approach, Int. J. [95] M.S. Rad, M. Nilashi, H.M. Dahlan, O. Ibrahim, Academic researchers’
Mobile Commun. 13 (5) (2015) 478–497. behavioural intention to use academic social networking sites: a case of
[61] J. Jacoby, L.B. Kaplan, The Components of Perceived Risk, ACR Special Volumes, Malaysian research universities, Inf. Dev. 35 (2) (2019) 245–261.
1972. [96] J. Park, J. Ahn, T. Thavisay, T. Ren, Examining the role of anxiety and social
[62] M.S. Featherman, P.A. Pavlou, Predicting e-services adoption: a perceived risk influence in multi-benefits of mobile payment service, J. Retailing Consum. Serv.
facets perspective, Int. J. Hum. Comput. Stud. 59 (4) (2003) 451–474. 47 (2019) 140–149.
[63] M.C. Lee, Factors influencing the adoption of internet banking: an integration of [97] H.P. Shih, An empirical study on predicting user acceptance of e-shopping on the
TAM and TPB with perceived risk and perceived benefit, Electron. Commer. Res. Web, Inf. Manag. 41 (3) (2004) 351–368.
Appl. 8 (3) (2009) 130–141. [98] J. Shannon, Gender differences or gendered differences: understanding the power
[64] T. Pikkarainen, S. Pahnila, H. Karjaluoto, K. Pikkarainen, Consumer acceptance of of language in training and research in supervision, Int. J. Adv. Counsell. 41
online banking: an extension of the technology acceptance model, Internet Res. (2019) 598–608, https://doi.org/10.1007/s10447-019-09380-y.
14 (3) (2004) 224–235. [99] S.M. Edwards, C. La Ferle, Does gender impact the perception of negative
[65] L.B. Kaplan, G.J. Szybillo, J. Jacoby, Components of perceived risk in product information related to celebrity endorsers? J. Promot. Manag. 15 (1–2) (2009)
purchase: a cross-validation, J. Appl. Psychol. 59 (3) (1974) 287–291. 22–35.
[66] N. Luhmann, Law as a social system, Nw, UL Rev 83 (1988) 136. [100] J. Pahl, Household spending, personal spending and the control of money in
[67] P.A. Pavlou, Consumer acceptance of electronic commerce: integrating trust and marriage, Sociology 24 (1) (1990) 119–138.
risk with the technology acceptance model, Int. J. Electron. Commer. 7 (3) (2003) [101] M. Jansen Verbeke, Women, shopping and leisure, Leisure Stud. 6 (1) (1987)
101–134. 71–86.
[68] S.Y. Yousafzai, J.G. Pallister, G.R. Foxall, A proposed model of e-trust for [102] V. Venkatesh, M.G. Morris, Why don’t men ever stop to ask for directions?
electronic banking, Technovation 23 (11) (2003) 847–860. Gender, social influence, and their role in technology acceptance and usage
[69] D. Gefen, D.W. Straub, Consumer trust in B2C e-Commerce and the importance of behavior, MIS Q. 24 (1) (2000) 115–139.
social presence: experiments in e-Products and e-Services, Omega 32 (6) (2004) [103] K.M.S. Faqih, M.-I.R.M. Jaradat, Assessing the moderating effect of gender
407–424. differences and individualism-collectivism at individual-level on the adoption of
[70] J.J. Wu, Y.S. Chang, Towards understanding members’ interactivity, trust, and mobile commerce technology: TAM3 perspective, J. Retailing Consum. Serv. 22
flow in online travel community, Ind. Manag. Data Syst. 105 (7) (2005) 937–954. (2015) 37–52.
[71] C. Flavi�an, M. Guinalíu, Consumer trust, perceived security and privacy policy: [104] R.A. Spreng, S.B. MacKenzie, R.W. Olshavsky, A reexamination of the
three basic elements of loyalty to a web site, Ind. Manag. Data Syst. 106 (5) determinants of consumer satisfaction, J. Market. 60 (3) (1996) 15–32.
(2006) 601–620. [105] F.L. Cabanillas, El papel de los medios de pago en los nuevos entornos
[72] D. Gefen, V.S. Rao, N. Tractinsky, The conceptualization of trust, risk and their electronicos, Universidad de Granada, 2012.
electronic commerce: the need for clarifications, System Sciences. Proceedings of [106] M. Van De Watering, The impact of computer technology on the elderly, 2005,
the 36th Annual Hawaii International Conference on, IEEE, 2003, p. 10, 2003. p. 12. Retrieved June 29(2008).
[73] K. Dahlberg, K. Segesten, M. Nystr€ om, B.-O. Suserud, I. Fagerberg, Att F€
orstå [107] J.E. Chung, N. Park, H. Wang, J. Fulk, M. McLaughlin, Age differences in
Vårdvetenskap, Studentlitteratur2003. perceptions of online community participation among non-users: an extension of
[74] B. Suh, I. Han, Effect of trust on customer acceptance of Internet banking, the Technology Acceptance Model, Comput. Hum. Behav. 26 (6) (2010)
Electron. Commer. Res. Appl. 1 (3–4) (2002) 247–263. 1674–1684.
[75] L.D. Chen, J. Tan, Technology adaptation in E-commerce:: key determinants of [108] I. Plaza, L. MartíN, S. Martin, C. Medrano, Mobile applications in an aging society:
virtual stores acceptance, Eur. Manag. J. 22 (1) (2004) 74–86. status and trends, J. Syst. Software 84 (11) (2011) 1977–1988.
[76] D. Gefen, D.W. Straub, Managing user trust in B2C e-services, e Serv. J. 2 (2) [109] A.J. Van Deursen, J.A. van Dijk, M. Peter, Increasing inequalities in what we do
(2003) 7–24. online: a longitudinal cross sectional analysis of Internet activities among the

14
D.L. Kasilingam Technology in Society 62 (2020) 101280

Dutch population (2010 to 2013) over gender, age, education, and income, [138] S. Marko, H. J€ org, M.R. Christian, Multigroup Analysis in Partial Least Squares
Telematics Inf. 32 (2) (2015) 259–272. (PLS) Path Modeling: Alternative Methods and Empirical Results, Measurement
[110] F. Cruz Jesus, T. Oliveira, F. Bacao, Digital divide across the European union, Inf. and Research Methods in International Marketing, Emerald Group Publishing
Manag. 49 (6) (2012) 278–291. Limited2011, pp. 195-218.
[111] O. Kyriakidou, Gender, management and leadership, equality, diversity and [139] C.S. Shirish, S.H.T. Thompson, E-Government Payoffs, Evidence from cross-
inclusion, Int. J. 31 (1) (2011) 4–9. country data, J. Global Inf. Manag. 15 (4) (2007) 20–40.
[112] M.R. Vicente, A.J. L� opez, Assessing the regional digital divide across the [140] R.B. Kline, Principles and Practice of Structural Equation Modeling, Guilford
European Union-27, Telecommun. Pol. 35 (3) (2011) 220–237. publications2015.
[113] P.B. Brandtzæg, J. Heim, A. Karahasanovi�c, Understanding the new digital [141] J. Joseph, F. Hair, R.E. Anderson, R.L. Tatham, W.C. Black, Multivariate Data
divide—a typology of Internet users in Europe, Int. J. Hum. Comput. Stud. 69 (3) Analysis, in: With Readings, fourth ed., Prentice-Hall, Inc., 1995.
(2011) 123–138. [142] J.C. Nunnally, I.H. Bernstein, J.M.t. Berge, Psychometric Theory, JSTOR1967.
[114] B. Kim, J. Oh, The difference of determinants of acceptance and continuance of [143] R.P. Bagozzi, Y. Yi, On the evaluation of structural equation models, J. Acad.
mobile data services: a value perspective, Expert Syst. Appl. 38 (3) (2011) Market. Sci. 16 (1) (1988) 74–94.
1798–1804. [144] C. Fornell, D.F. Larcker, Evaluating structural equation models with unobservable
[115] G. Lee, J. Lee, S. Kwon, Use of social-networking sites and subjective well-being: a variables and measurement error, J. Mark. Res. 18 (1) (1981) 39–50.
study in South Korea, Cyberpsychol., Behav. Soc. Netw. 14 (3) (2011) 151–155. [145] J.C. Anderson, D.W. Gerbing, Structural equation modeling in practice: a review
[116] H.H. Chang, S.W. Chen, The impact of customer interface quality, satisfaction and and recommended two-step approach, Psychol. Bull. 103 (3) (1988) 411.
switching costs on e-loyalty: internet experience as a moderator, Comput. Hum. [146] J. Henseler, T.K. Dijkstra, M. Sarstedt, C.M. Ringle, A. Diamantopoulos, D.
Behav. 24 (6) (2008) 2927–2944. W. Straub, J. David J. Ketchen, J.F. Hair, G.T.M. Hult, R.J. Calantone, Common
[117] Z. Deng, Y. Lu, K.K. Wei, J. Zhang, Understanding customer satisfaction and beliefs and reality about PLS:comments on r€ onkk€o and evermann (2013), Organ.
loyalty: an empirical study of mobile instant messages in China, Int. J. Inf. Manag. Res. Methods 17 (2) (2014) 182–209.
30 (4) (2010) 289–300. [147] J. Henseler, T.K. Dijkstra, M. Sarstedt, C.M. Ringle, A. Diamantopoulos, D.
[118] P.A. Dabholkar, Consumer evaluations of new technology-based self-service W. Straub, D.J. Ketchen Jr., J.F. Hair, G.T.M. Hult, R.J. Calantone, Common
options: an investigation of alternative models of service quality, Int. J. Res. beliefs and reality about PLS: comments on r€ onkk€o and evermann (2013), Organ.
Market. 13 (1) (1996) 29–51. Res. Methods 17 (2) (2014) 182–209.
[119] P. Gerrard, J. Barton Cunningham, The diffusion of internet banking among [148] H.H. Harman, Modern Factor Analysis, University of Chicago, 1976. Press.
Singapore consumers, Int. J. Bank Market. 21 (1) (2003) 16–28. [149] H.W. Kim, H.C. Chan, S. Gupta, Value-based adoption of mobile internet: an
[120] E.J. Lee, K.N. Kwon, D.W. Schumann, Segmenting the non-adopter category in the empirical investigation, Decis. Support Syst. 43 (1) (2007) 111–126.
diffusion of internet banking, Int. J. Bank Market. 23 (5) (2005) 414–437. [150] A. Seffah, E. Metzker, Usability Engineering Methods Plethora, Adoption-centric
[121] V. Nui Polatoglu, S. Ekin, An empirical investigation of the Turkish consumers’ Usability Engineering, Springer2009, pp. 15-33.
acceptance of Internet banking services, Int. J. Bank Market. 19 (4) (2001) [151] N.Y. Siu, L.M. Chang, A Study of Service Quality, Perceived Risk and Personal
156–165. Innovativeness in Internet Banking, Revolution in Marketing: Market Driving
[122] J. Ristola, M. Sc dissertation, Information Technology Service Management for Changes, Springer2015, pp. 78-83.
Cloud Computing, Aalto University School of Science and Technology, 2010. [152] C.Y. Lin, D. Marshall, J. Dawson, How does perceived convenience retailer
[123] D. Gefen, E. Karahanna, D.W. Straub, Inexperience and experience with online innovativeness create value for the customer? Int. J. Bus. Econ. 12 (2) (2013) 171.
stores: the importance of TAM and trust, IEEE Trans. Eng. Manag. 50 (3) (2003) [153] W.D. Salisbury, R.A. Pearson, A.W. Pearson, D.W. Miller, Perceived security and
307–321. world wide web purchase intention, Ind. Manag. Data Syst. 101 (4) (2001)
[124] J.-W. Moon, Y.-G. Kim, Extending the TAM for a world-wide-web context, Inf. 165–177.
Manag. 38 (4) (2001) 217–230. [154] S.L. Jarvenpaa, P.A. Todd, Is there a future for retailing on the Internet, Electronic
[125] A.D. Miyazaki, A. Fernandez, Consumer perceptions of privacy and security risks marketing and the consumer 1 (12) (1997) 139–154.
for online shopping, J. Consum. Aff. 35 (1) (2001) 27–44. [155] A. O’Cass, T. Fenech, Web retailing adoption: exploring the nature of internet
[126] N. Mathur, India’s Internet Base Crosses 500 Million Mark, Driven by Rural India, users Web retailing behaviour, J. Retailing Consum. Serv. 10 (2) (2003) 81–94.
in: Livemint, 2019. [156] I. Ajzen, Nature and operation of attitudes, Annu. Rev. Psychol. 52 (1) (2001)
[127] Beaconstac, The State of Indian Mobile Commerce, 2016. 27–58.
[128] E. Van den Broeck, B. Zarouali, K. Poels, Chatbot advertising effectiveness: when [157] J.W. Lian, D.C. Yen, Online shopping drivers and barriers for older adults: age and
does the message get through? Comput. Hum. Behav. 98 (2019) 150–157. gender differences, Comput. Hum. Behav. 37 (2014) 133–143.
[129] F.D. Davis, R.P. Bagozzi, P.R. Warshaw, Extrinsic and intrinsic motivation to use [158] S.I. Wu, The relationship between consumer characteristics and attitude toward
computers in the workplace, J. Appl. Soc. Psychol. 22 (14) (1992) 1111–1132. online shopping, Market. Intell. Plann. 21 (1) (2003) 37–44.
[130] R.E. Goldsmith, C.F. Hofacker, Measuring consumer innovativeness, J. Acad. [159] T.F. Stafford, A. Turan, M.S. Raisinghani, International and cross-cultural
Market. Sci. 19 (3) (1991) 209–221. influences on online shopping behavior, J. Global Inf. Technol. Manag. 7 (2)
[131] A.Y.L. Chong, F.T.S. Chan, K.B. Ooi, Predicting consumer decisions to adopt (2004) 70–87.
mobile commerce: cross country empirical examination between China and [160] K.A. Passyn, M. Diriker, R.B. Settle, Images of online versus store shopping: have
Malaysia, Decis. Support Syst. 53 (1) (2012) 34–43. the attitudes of men and women, young and old really changed? J. Bus. Econ. Res.
[132] A. Dickinger, M. Kleijnen, Coupons going wireless: determinants of consumer 9 (No 1) (2011), 2011.
intentions to redeem mobile coupons, J. Interact. Market. 22 (3) (2008) 23–39. [161] P.E. Pedersen, Adoption of mobile internet services: an exploratory study of
[133] D.R. Lichtenstein, N.M. Ridgway, R.G. Netemeyer, Price perceptions and mobile commerce early adopters, J. Organ. Comput. Electron. Commer. 15 (3)
consumer shopping behavior: a field study, J. Mark. Res. 30 (2) (1993) 234–245. (2005) 203–222.
[134] S. Swaminathan, K. Bawa, Category-specific coupon proneness: the impact of
individual characteristics and category-specific variables, J. Retailing 81 (3)
Dr. Dharun is an Assistant Professor of Digital Platform and Strategies at Mudra Institute of
(2005) 205–214.
Communications Ahmedabad (MICA), Ahmedabad, India. He holds Masters of Engineer­
[135] S. Taylor, P.A. Todd, Understanding information technology usage: a test of
ing in Industrial Engineering and a Doctorate in Marketing and Analytics. His research
competing models, Inf. Syst. Res. 6 (2) (1995) 144–176.
extensively revolves around advanced technologies and behavior, and unstructured data
[136] D. Verma, P. Dash, M. Sain, S.P. Kumar, C. Prakash, R. Kumar, Kesan, Youth in
analytics. He has published in journals like Journal of Retailing and Consumer Services,
India, Central Statistics Office, Ministry of Statistics and Programme
Technology in Society, etc. His research has appeared in the proceedings of several
Implementation, Government of India, 2017.
renowned international conferences in India, Australia, Dubai, and Malaysia. He is an
[137] M. Sarstedt, C.M. Ringle, D. Smith, R. Reams, J.F. Hair Jr., Partial least squares
expert in software packages like R, Python, Minitab, SPSS, Stata, AMOS, Smart-PLS, etc.
structural equation modeling (PLS-SEM): a useful tool for family business
researchers, Journal of Family Business Strategy 5 (1) (2014) 105–115.

15