Corresponding Author: Meilisa Rumetor
E-mail:
[email protected]
INTRODUCTION
In conjunction with the development of technology,
several service industries have experienced significant innovation and
digitalization renewal. Thus, value creation in the digital era has become a
collaboration between customers and enterprises (Hosseini et al., 2022). Internet and technological advancement have altered
the delivery and utilization of financial services (Hwang & Kim, 2021). The digital revolution has also changed the
landscapes of numerous industries (Wang et al., 2021). Many financial institutions provide innovative
alternative electronic channels to maintain a competitive advantage and meet
consumer expectations (S.-N. Zhang et al.,
2019). Like many other industries, the banking sector is
evolving due to changing consumer behavior, rising expectations, the adoption
of new technology, and the general digitization of business and society (Rahmayati, 2021). Significantly, the onset of the COVID-19 pandemic
has induced a transformative change in how communication and work are
conducted, emphasizing the importance of automated chat functions, specifically
Chatbots, in the operations of diverse companies. (Nguyen et al., 2021) brahi. Banking must become digitally adapted to
survive in this new technological era. Failure to respond and adapt to the new
environment will inevitably lead to catastrophic losses and failure.
Chatbots appeal to bank customers because they can
communicate quickly and anytime. Customers utilize these messaging services to
obtain information (e.g., product specifications) or technical support (e.g.,
problem resolution). This service provides customers with real-time assistance
and services (Adam et al., 2021). The immediate responsiveness of messaging services
has altered customer service into an interactive dialogue, exerting notable
impacts on consumer trust, satisfaction, repurchase intent, and loyalty. (Mero, 2018) As with most industries affected by digitization,
technological advances have infiltrated the Banking Industry (Rahmayati, 2021). Consumers who have experienced convenience and
comfort when obtaining services from one bank frequently desire the same
services from other banks. The perception of safety and comfort generates
tension, such as contagious social disease. Social Contagion is a person's
intentions and behaviors that are influenced and modeled by their observations
and social learning. (Chaouali & El
Hedhli, 2018) This means that Banking Businesses that have yet to
adopt Chatbot services must be able to adapt and develop by employing
technology to modify their business model to continue to exist and compete with
other Banking Businesses. The functions of the most popular Chatbot services
are interaction, amusement, problem-solving, style, and customization (Misischia
et al., 2022). (Richad et
al., 2019) have already highlighted the importance
of studying Chatbot acceptance in the context of the banking industry by
researching the technology acceptance model and customer experience for
consumers banking with Chatbot technology, respectively. Existing literature
has investigated the significance of Chatbot acceptance and user experience in
the banking industry.
Numerous
prior studies have voiced concerns regarding implementing Chatbots due to their
widespread use. The term "Chatbot" combines "Chat" and
"Robot" (Hwang & Kim, 2021). According to (Lui & Lamb, 2018), chatbots are computer programs propelled by artificial
intelligence that engage in conversations or interactions with real people via
messaging apps and websites. Text-based exchanges and verbal interactions
between humans and Chatbots are possible regardless of time or location (Karri & Kumar, 2020). Both forms of machine-based engagement are deftly camouflaged as
human agent assistance, allowing consumers to initiate conversations more easily
(Prasetya et al., 2018). The technology acceptance model (TAM) is essential to the study
of consumer behavior concerning the adoption of technologies. (Carranza et al.,
2021). According to the Technology Acceptance Model (TAM), if usability and
ease of use influence an individual's attitude towards technology, it can alter
technology usage intentions. When Chatbots provide real-time banking services,
many clients enjoy their advantages. Chatbot is part of the banking industry's
initiative to innovate and improve the quality of banking services (Carranza et al., 2021). The factors influencing Chatbot adoption have been studied using
technology acceptance models, and the consumer experience has been considered
in the context of banking with Chatbot technology. Nevertheless, more research
is needed to investigate the influence of social contagion in the acceptance of
Chatbot services. This study aims to thoroughly comprehend the impacts of
attitudes and social contagion on the adoption of Chatbots. The focus and
objectives of this research involve assessing the impact of Chatbot adoption on
clients' engagement in private banks through the application of the Technology
Acceptance Model (TAM) and Social Contagion.
Technology Acceptance Model
(TAM)
The Technology Acceptance Model (TAM) is a paradigm for examining and
comprehending the elements that influence the adoption of technology use. TAM
illustrates the causal relationship between ideas about the benefits of a
technology or information system, perceived ease of use, perceived usefulness,
attitude towards using the technology, and intention to keep using it (Carranza
et al., 2021). Researchers have widely utilized TAM models to assess behavioral
intentions and user satisfaction (Duarte et
al., 2018). The TAM theory was developed to analyze
user acceptance behavior concerning information service systems, grounded in
the perspective of social psychology (T. Zhang
et al., 2018). According to �(Ilhamalimy
& Ali, 2021), TAM is the best predictive model for
analyzing consumer behavior regarding its use of information technology. TAM
utilizes the explanatory variables of "Perceived Usefulness" (PU) and
"Perceived Ease of Use" (PEOU) to explain a user's attitude toward
the adoption of a specific technology and their intention to persist in its
use. (Ashfaq et
al., 2020)
Perceived Ease of Use
(PEOU) and Perceived Usefulness (PU)
Perceived Ease of Use (PEOU) pertains to "the extent to which an
individual believes that utilizing a specific system would entail minimal
effort." (Davis et al., 1989). Previous research has consistently
demonstrated a substantial correlation between PEOU and PU. (Ashfaq et
al., 2020). PEOU represents an individual's
perception of the ease of technology use (Carranza
et al., 2021). In the context of Chatbot services,
ensuring ease of learning and usage is crucial to minimize user intimidation. (Jain et
al., 2018) This suggests that a positive impact on
users' PU in their engagement with Chatbot services will likely result from the
construction of PEOU. Perceived usefulness (PU) is defined as the extent to
which an individual believes using a specific system would improve their job
performance." (Davis et al., 1989). The preceding explanation can be
expressed as the following hypothesis:
H1: PEOU
has a positive effect on PU on Chatbot.
Perceived Usefulness (PU) and Attitude Towards Using
(ATU)
Perceived Usefulness (PU) can be
characterized as individuals' perception of the enhancement in performance
achieved by using a particular technology. (Foroughi
et al., 2019) Limited research has indicated that PU
stands out as the most robust cognitive factor influencing the acceptance of
technology, as consumers emphasize the potential benefits of an innovation (Zarouali
et al., 2018). Furthermore, it has been established
that PU plays a crucial role in shaping positive attitudes among consumers
toward novel digital technologies. Consequently, the perceived usefulness of a
Chatbot will exhibit a positive correlation with consumers' attitudes toward
utilizing the Chatbot.
H2: PU has
a positive effect on the ATU Chatbot.
Perceived Ease of Use (PEOU) and Attitude Towards
Using (ATU)
In the realm of technology, PEOU within TAM has been
recognized as a pivotal element in bolstering Continuous Intention. (Ashfaq et al., 2020) PEOU is delineated as the degree to which consumers
perceive the utilization of the Chatbot as straightforward and uncomplicated (Zarouali et al., 2018). Given its focus on the efficacy of action, PEOU has
been acknowledged as a significant inherent motivator for consumers, forming a
direct link with their attitude. Attitude toward using (ATU) signifies an
individual's inclination or desirability to employ the system. (Emaran et al.,
2020). Following this reasoning, we expect that:
H3: PEOU has a positive effect on
the ATU Chatbot.
Perceived Usefulness (PU) and Continuous Intention (CI)
In the first version of TAM, Davis et al. (1989)
identified perceived usefulness (PU) as a pivotal determinant influencing users
to adopt and sustain their usage of novel technologies. The significance of PU
extends across various research contexts, particularly in e-commerce studies,
where it consistently emerges as a noteworthy factor impacting shoppers'
behavioral decisions (He et al., 2018). In the e-commerce domain, PU has demonstrated a
consistent and significant association with continuance intention (CI),
denoting users' intent to persist in using the information system (Gupta et al., 2021). Continuous intention (CI) is defined as users'
willingness to maintain their engagement with the information system over time (Ahmad et al., 2020). The essence of PU lies in its ability to instill a
belief among users that utilizing the system will enhance their overall
performance. A heightened perception of usefulness leads consumers to believe
that the technology will improve their performance, positively influencing
their intention to continue using it. (Gupta et al., 2021) Ahmad et al. (2020) defined continuous intention (CI) as an individual's
inclination to persist in utilizing a specific technology beyond the initial
adoption phase; the same article emphasizes that the perceived usefulness (PU)
of the technology significantly influences an individual's intent to continue
using it. A higher PU of Chatbot services corresponds to a more positive
inclination for their sustained usage, thereby increasing the likelihood of
continued intention.
H4: PU has a positive effect on
the CI of Chatbot.
Attitude Toward Using (ATU) and Continuous Intention (CI)
Attitude toward using (ATU) represents a user's
disposition towards incorporating a system into their daily life, reflecting
their acceptance or rejection of technology usage. According to TAM, attitudes
shape users' behavioral tendencies concerning technology usage. Numerous
studies have affirmed a positive correlation between attitude and the intention
to persist in using technology (Foroughi et al., 2019). A different research defines ATU as an individual's
inclination or desire to engage with the system. (Al-Emran et al., 2020). Earlier research Prior studies showcased a
significant association between ATU and CI (Al-Emran et al., 2020). Moreover, attitude is delineated as the degree to
which an individual holds positive or negative sentiments regarding executing a
specific behavior (Davis et al., 1989). Additional research outlines crucial
factors that impact attitudes regarding technology use (Kemp et al., 2019). Moreover, according to (Alsharo et al., 2020), ATU positively influences the sustained utilization
of technology. Consequently, attitude is anticipated to be a notable
determinant in predicting users' intentions regarding using Chatbot services.
In line with this, the following hypothesis is suggested:
H5: ATU has a positive effect on
the CI of Chatbot
Social Contagion Theory
Social contagion is how consumers impact each other's
decisions to adopt or utilize a product. It embodies the concept that
individuals choose products when they observe others who have already adopted
them, akin to the spread of innovations resembling epidemics (Kiogothe, 2018). Social contagion pertains to an individual's
intentions and behaviors, molded and influenced by observations and social
learning (Chaouali & Hedhli, 2019).
Social contagion emanates from the societal framework,
leading individuals in analogous social positions to assess the advantages and
risks of adoption comparably (Kiogothe, 2018). When individuals interact with various social
factors, they are prone to internalize implicit choice norms, which become the
fundamental basis for their future decisions. (e.g., individuals, institutions,
and organizations) (Chaouali & Hedhli, 2019). Individuals embrace novel
entities such as products, services, and technology either involuntarily to
meet the expectations of perceived authoritative entities like service
providers, suppliers, and public authorities or willingly through observational
learning and positive reinforcement (Chaouali & Hedhli, 2019). Past
research has indicated that continuous intention (CI) is notably influenced by
the impact of social contagion (Al-et al., 2020). The key expressions of this
form of social contagion include normative pressure, mimetic pressure, and
coercive pressure.
Coercive Pressure
Social contagion is characterized by the extent to
which an individual believes that influential or similar others endorse the
idea that they should utilize the new system (Chaouali & El
Hedhli, 2018). Coercive pressures encompass formal and informal
influences exerted on social actors, whether individuals or organizations,
compelling them to adopt similar attitudes, behaviors, and practices under the
influence of more influential actors (Fauzi, 2019). Informal pressures may stem from the cultural
dynamics and expectations within a community or an organization's environment (Masocha & Fatoki,
2018). Coercive pressures, exemplified by government and
regulatory bodies, play a crucial role in influencing behaviors for
sustainability (Masocha & Fatoki,
2018). Coercive pressure denotes alterations in individual
behavior prompted by directives from more influential social actors. This
leaves communities or individuals with no recourse but to conform to the
desired conduct of entities possessing greater power (Hiqmah, 2020). For instance, entities like banks or government
bodies may enforce specific channels for operations, such as electricity
billing, leading individuals to adopt technology to consistently comply with
more potent authorities. In this context, the formulated hypothesis is as
follows:
H6: Coercive pressure has a
positive effect on Chatbot's CI.
Normative Pressure
Normative pressure entails shifts in an individual's
behavior driven by an unconscious inclination to conform to social norms,
especially when a particular behavior becomes widely accepted or popular among
relatives or society (Hiqmah, 2020). Such pressures may lead social actors who have yet
to embrace the innovation to experience dissonance and consequent discomfort
when witnessing peers whose approval they value adopt the innovation (Fauzi, 2019). Furthermore, a similar study also asserted that
social actors are more inclined to imitate a specific action if numerous others
have already taken that same action. The adoption of certain behaviors by
social actors is compelled by their anticipation of legitimacy rather than a
strict consideration of appropriateness. Normative pressure within a society
significantly shapes individuals' behavior and choices. However, the motivation
to adhere to these social norms can range from profound intrinsic involvement
to a superficial identification with a social group and mere extrinsic
compliance (Nguyen et al., 2021). Individuals not conforming to these behaviors may
encounter frustration and discomfort as they lag behind those who have embraced
the new norms (Hiqmah, 2020). Consequently, individuals generally continuously use
technology to adapt to their environment. Building on this rationale, our
expectation is that:
H7: Normative Pressure positively
affects the CI's decision to use a Chatbot.
Mimetic Pressure
Mimetic pressure refers to intentional and voluntary
efforts to imitate the conduct of individuals perceived to be more successful
and possessing a higher social standing (Hiqmah, 2020). Social actors are prompted by mimetic pressures to
imitate the established behaviors and practices of other influential and
effective actors deliberately and willingly. Essentially, if an influential
figure accomplishes a task effectively, it motivates others to replicate it (Fauzi, 2019). In innovation adoption, the mimetic influence acts
as a catalyst or connection that motivates the non-adopter group to feel
compelled to undertake the action, as it has already been executed by someone
else (Hiqmah, 2020). This will lead to an uncertain reaction to a
specific regulation. The uncertainty arises from various factors within the
organization, including shifts in governmental politics and culture and
frequent changes in legislation, among others. Inadequate organizational
readiness for a rule will result in a lack of organizational comprehension
regarding implementing new regulations (Ridha & Basuki, 2019). Pressure
where users feel superior to obtain banking services and observe the
convenience of other customers who use Chatbots, resulting in a solid intent to
continue using the Chatbot application. Based on these findings, we expect
that:
H8: Mimetic pressure positively
affects the CI's ability to use a Chatbot.
Figure 1.
Conceptual Framework
METHOD
This research will employ a quantitative approach, necessitating
the collection of data. Furthermore, the sampling method uses nonprobability
sampling of the purposive type. The study utilized a questionnaire survey
methodology to obtain pertinent data, using a questionnaire as the instrument
to test our research model. We conducted a literature review in the initial
phase to identify relevant reference variables. The measurements can be found
in the Appendix. The questionnaire responses were evaluated on a five-point
Likert scale, ranging from 1 to 5, where 1 corresponds to "strongly
disagree," 3 indicates "neutral," and 5 represents
"strongly agree." The research model comprises seven structural
variables, with their measures drawn from existing literature or tailored to
suit the Chatbot implementation context for content validity assurance. The
measurement of PU and PEOU is based on the study of �(Foroughi et al., 2019). ATU and CI are measured based on the analysis of (Foroughi et al., 2019) and (Chaouali & El Hedhli, 2018). The measurement of the three pressures of social contagion,
which are CP, NP, and MP, is based on the study of (Chaouali & El Hedhli, 2018), Lin, Luo & Luo (2020). Each construct within the model is
operationalized in the form of reflective constructs.
To collect
relevant data, the study targeted users of chatbots in the Internet banking
applications of private institutions in Manado who have interacted with a
chatbot at least once and used Chatbot services within the previous six months.
The survey questionnaire was distributed online through Google Form Survey
using WhatsApp platforms, and 300 responses were included for analysis.
Subsequently, SmartPLS was employed for statistical testing and validation of
the proposed model. Several motivational factors influenced the selection of
SmartPLS for this study. Firstly, SmartPLS is known for its user-friendly
interface. Secondly, it has emerged as a prominent structural equation modeling
(SEM) analysis method. Lastly, SmartPLS is widely recognized and accepted in
academic circles, particularly within diverse research domains related to
Management Information Systems (MIS).
Table 1. Data Analysis
And Result
|
Characteristics
|
Options
|
Frequency
|
Percentage
|
|
Gender
|
Male
|
146
|
49
|
|
Female
|
154
|
51
|
|
Age
|
<17
years
|
5
|
2
|
|
17
- 25 years
|
15
|
5
|
|
26
- 35 years
|
181
|
60
|
|
36
- 45 years
|
84
|
28
|
|
>
45 years
|
15
|
5
|
|
Education
|
High
School and Equivalent
|
12
|
4
|
|
College
|
23
|
8
|
|
Undergraduates
|
226
|
75
|
|
Graduate
|
39
|
13
|
|
Postgraduate
|
0
|
0
|
|
Job Status
|
Employed
|
278
|
93
|
|
Unemployed
|
22
|
7
|
|
Occupation
|
Office
Employee
|
230
|
77
|
|
Government
Employee
|
15
|
5
|
|
Students
|
5
|
2
|
|
Entrepreneur
|
11
|
4
|
|
Pensioner
|
3
|
1
|
|
Housewives
|
3
|
1
|
|
Others
|
30
|
10
|
|
Name of the Bank where I use Chatbot Services
|
Bank
BNI
|
78
|
26
|
|
Bank
Mandiri
|
59
|
20
|
|
Bank
BCA
|
81
|
27
|
|
Bank
Lain
|
92
|
31
|
|
Purpose of using Chatbot Services
|
Products
and services provided by the Bank
|
108
|
36
|
|
Open
a Bank Account
|
52
|
17
|
|
Lodging
a complaint
|
92
|
31
|
|
Others
|
48
|
16
|
|
How long I�ve been using the Chatbot services
|
<1
Year
|
114
|
38
|
|
<2
Years
|
93
|
31
|
|
<3
Years
|
55
|
18
|
|
>3
Years
|
38
|
13
|
|
I am actively using Chatbot services to fulfill
the objective of
|
Products
and services provided by the Bank
|
108
|
36
|
|
Open
a Bank Account
|
52
|
17
|
|
Lodging
a complaint
|
92
|
31
|
|
Others
|
48
|
16
|
|
How often I have used the Chatbot services over
the last six months
|
1
time
|
95
|
32
|
|
2
times
|
86
|
29
|
|
3
times
|
42
|
14
|
|
>3
times
|
77
|
26
|
As seen in
Table 1, among all the participants, 51% of them are female, while 49% are
male. Furthermore, the age categories of respondents were as follows: 60% were
noted to be between 26 to 35 years, followed by 28% were 36 to 45 years, the
rest, 5% were above 45 years, and 5% were 17 to 25 years. The sample is
well-educated since 75% of the respondents were undergraduates, 13% graduated,
and 12% were high school graduates. Based on employment status, around 93% of
the respondents are employees, with the most common occupation being bank
employees. Moving away from demographic factors and more into practical
analysis, we quantify the frequency of chatbot service usage. In total, 87% of
them use chatbot services less than 3 times, and only 13 % are the ones who
often use the services. Regarding respondent usage, the result shows that
chatbot services are used to obtain information about bank products.
Table 2. Result for reliability and convergent validity
|
Construct
|
Indicator
|
Factor Loading
|
Cronbach's
|
CR
|
AVE
|
|
Perceived usefulness
|
PU1
|
0.873
|
0.809
|
0.875
|
0.638
|
|
|
PU2
|
0.787
|
|
|
|
|
|
PU3
|
0.765
|
|
|
|
|
|
PU4
|
0.765
|
|
|
|
|
Perceived Ease of Use
|
PEOU1
|
0.727
|
0.762
|
0.846
|
0.580
|
|
|
PEOU2
|
0.802
|
|
|
|
|
|
PEOU3
|
0.723
|
|
|
|
|
|
PEOU4
|
0.791
|
|
|
|
|
Attitude Toward Using
|
ATU1
|
0.827
|
0.850
|
0.899
|
0.689
|
|
|
ATU2
|
0.831
|
|
|
|
|
|
ATU3
|
0.820
|
|
|
|
|
|
ATU4
|
0.843
|
|
|
|
|
Continuous Intention
|
CI1
|
0.825
|
0.849
|
0.898
|
0.689
|
|
|
CI2
|
0.750
|
|
|
|
|
|
CI3
|
0.879
|
|
|
|
|
|
CI4
|
0.861
|
|
|
|
|
Coercive Pressure
|
CP1
|
0.810
|
0.872
|
0.913
|
0.723
|
|
|
CP2
|
0.838
|
|
|
|
|
|
CP3
|
0.881
|
|
|
|
|
|
CP4
|
0.871
|
|
|
|
|
Normative Pressure
|
NP1
|
0.794
|
0.875
|
0.913
|
0.726
|
|
|
NP2
|
0.862
|
|
|
|
|
|
NP3
|
0.892
|
|
|
|
|
|
NP4
|
0.857
|
|
|
|
|
Mimetic Pressure
|
MP1
|
0.852
|
0.852
|
0.902
|
0.700
|
|
|
MP2
|
0.915
|
|
|
|
|
|
MP3
|
0.885
|
|
|
|
|
|
MP4
|
0.674
|
|
|
|
This study initially assessed the measurement
model to validate the scale's reliability and validity before examining the
structural model. Internal consistency was evaluated using Cronbach's α
and Composite Reliability (CR) values, detailed in Table 2. The NP construct exhibited the highest CR and AVE values (CR =
0.913, AVE = 0.726), while the PEOU construct displayed the lowest values (CR =
0.846, AVE = 0.580). Consequently, all constructs in this research surpassed
the recommended thresholds. Indicator loading factors or correlations with respective latent variables
in Table 2 indicate individual reliability. Latent variable reliability
employed Cronbach's α coefficient, with values above 0.7 considered
acceptable.
Additionally, composite reliability was
calculated to assess unidimensionality. Convergent validity was examined
through Average Variance Extracted (AVE), where values above 0.5 were deemed
acceptable. Table 2 presents Cronbach's α coefficient, composite
reliability, and AVE. Discriminant validity
was confirmed as the square root of AVE for each latent variable exceeded
correlations with other latent variables. The recommended reliability threshold
of 0.7 was met.
Validity assesses data accuracy and is
determined by calculating the Average Variance Extracted (AVE). In the validity
test results, the AVE values for each variable surpassed 0.5. Furthermore, all
factor loadings exceeded 0.7, signifying robust convergent validity of the
model. Convergent validity results are presented in
Table 2. The square root of the AVE for each construct exceeded the
inter-correlations among all constructs, confirming the model's discriminant
validity. Discerning validity outcomes are detailed in
Table 3.
Table 3. Discriminant Validity
Discriminant Validity - Fornell-Lacker
criterion
|
|
ATU
|
|
CI
|
CP
|
MP
|
NP
|
PEOU
|
PU
|
|
ATU
|
0.830
|
|
|
|
|
|
|
|
|
CI
|
0.760
|
|
0.830
|
|
|
|
|
|
|
CP
|
0.577
|
|
0.703
|
0.850
|
|
|
|
|
|
MP
|
0.404
|
|
0.509
|
0.685
|
0.836
|
|
|
|
|
NP
|
0.529
|
|
0.635
|
0.780
|
0.650
|
0.852
|
|
|
|
PEOU
|
0.647
|
|
0.592
|
0.474
|
0.399
|
0.444
|
0.762
|
|
|
PU
|
0.732
|
|
0.697
|
0.398
|
0.256
|
0.403
|
0.676
|
0.799
|
Notes: Bold values on the diagonal are the square roots of AVE,
and the off-diagonal values are correlations. ATU: Attitude Towards Using; CI:
Continuous Intention; CP: Coercive Pressure; MP: Mimetic Pressure; NP:
Normative Pressure; PEOU: Perceived Ease of Use; PU: Perceived Usefulness.
Table 4.
Discriminant Validity � Heterotrait�monotrait ratio (HTMT)
|
|
ATU
|
CI
|
CP
|
MP
|
NP
|
PEOU
|
PU
|
|
ATU
|
|
|
|
|
|
|
|
|
CI
|
0.888
|
|
|
|
|
|
|
|
CP
|
0.662
|
0.811
|
|
|
|
|
|
|
MP
|
0.464
|
0.583
|
0.789
|
|
|
|
|
|
NP
|
0.591
|
0.711
|
0.878
|
0.741
|
|
|
|
|
PEOU
|
0.775
|
0.715
|
0.547
|
0.464
|
0.508
|
|
|
|
PU
|
0.881
|
0.844
|
0.474
|
0.304
|
0.461
|
0.846
|
|
Notes:
ATU: Attitude Towards Using; CI: Continuous Intention; CP: Coercive Pressure;
MP: Mimetic Pressure; NP: Normative Pressure; PEOU: Perceived Ease of Use; PU:
Perceived Usefulness.
Table 3 affirms the discriminant
validity of this study's utilized variables through the Fornell-Larcker
criterion and HTMT ratio (Lee et al., 2021). The Fornell-Larcker results demonstrate that the square root of
the AVE exceeded the corresponding columns and rows. According to (Rohmatulloh & Nugraha, 2022), discriminant validity is considered valid if the HTMT value is
below 0.9. All HTMT ratio results meet this criterion, as determined by the
bootstrapping algorithm. Moreover, reliability values ranging from 0.60 to 0.70
are deemed "acceptable in exploratory research," while values between
0.70 and 0.90 are considered "satisfactory to good" (Purwanto, 2021).
Figure 2. Path Modelling
The
significance of path modeling (Figure 2) was measured using the bootstrapping
technique (Hair Jr et al., 2021) with a p-value of 0.05 and the
variance of the dependent variable.
Table 5. Results for
the hypothesis test.
|
Hypothesis
|
Path
|
Original Sample
|
T Statistic
|
P Values
|
Comment
|
Result
|
|
H1
|
PEOU -> PU
|
0.676
|
13.081
|
0.000
|
Significant
|
Supported
|
|
H2
|
PU -> ATU
|
0.543
|
8.426
|
0.000
|
Significant
|
Supported
|
|
H3
|
PEOU -> ATU
|
0.280
|
4.358
|
0.000
|
Significant
|
Supported
|
|
H4
|
PU -> CI
|
0.321
|
5.789
|
0.000
|
Significant
|
Supported
|
|
H5
|
ATU -> CI
|
0.281
|
3.826
|
0.000
|
Significant
|
Supported
|
|
H6
|
CP -> CI
|
0.325
|
6.122
|
0.000
|
Significant
|
Supported
|
|
H7
|
NP -> CI
|
0.078
|
1.313
|
0.190
|
Not Significant
|
Not supported
|
|
H8
|
MP -> CI
|
0.040
|
0.926
|
0.355
|
Not Significant
|
Not supported
|
The proof for the final hypothesis can be determined by examining the t-statistic value and p-value.
The hypothesis exhibits a noteworthy positive impact when the
t-statistic value equals or surpasses 1.96, and the p-value is below 0.05.
Conversely, if the t-statistic value falls below 1.96 and the p-value exceeds
0.05, it can be inferred that the hypothesis lacks an effect. We can scrutinize
Table 4 for relevant findings in this study.
Table 5
presents the outcomes of the hypothesis testing. The findings from the PLS-SEM
analysis reveal a significant positive impact of PEOU on PU (t-statistic = 13.081, p-value
= .00), thereby supporting H1. Additionally,
PU has a notable positive effect on ATU (t-statistic = 8.426 p-value = .00) and PEOU on ATU (t-statistic = 4.358, p-value
= .00), confirming hypotheses H2 and H3. The
significant positive effect of PU on CI is evident, with a t-statistic of 5.789
and a p-value of .00, supporting H4. Furthermore, ATU substantially positively
impacts CI (t-statistic = 3.826, p-value =
.00), helping H5. Moreover, CP exhibits a significant positive impact on CI,
with a t-statistic of 6.122 and a p-value of 0.000, backing H6. Conversely,
hypotheses H7 and H8 lack support as their t-statistics and p-values do not
meet the criteria for hypothesis confirmation. The t-statistics are below 1.96
in both cases, and the p-values exceed 0.05. Therefore, these hypotheses cannot
be accepted, indicating no positive and meaningful relationship between NP and
CI and MP and CI.
RESULTS AND DISCUSSION
Research introduces a conceptual model for examining users'
intention to use chatbots, incorporating the Technology Acceptance Model (TAM)
and the influential effects of social pressure. The author notes a need for
more empirical studies investigating the continuous intention to use chatbots
based on TAM and social pressure, limiting direct comparisons with earlier
research. However, meaningful comparisons can be drawn with relevant
literature. The study, utilizing multivariate statistical analysis techniques on
a dataset of 300 responses collected through an online questionnaire survey,
tested eight hypotheses outlined in the proposed model, employing SmartPLS
software. The results supported six hypotheses (H1-H6), while the remaining two
hypotheses (H7-H8) lacked support from the data (see Table 4). The findings
contribute significantly to both theoretical understanding and practical
insights into the continuous intention of using chatbots.
The path modeling analysis revealed several significant
findings that shed light on the factors influencing users' continuous intention
to use chatbot services. Notably, Perceived Usefulness (PU) significantly
affected users' attitudes towards chatbot usage. This supports the idea that
users are likelier to continue using chatbots if they perceive them as valuable
tools that enhance their performance and productivity. Moreover, the study
found that Perceived Ease of Use (PEOU) plays a vital role in shaping users'
perceptions and attitudes. A positive impact on Perceived Usefulness and
Attitude Toward Using (ATU) was observed, indicating that chatbots that provide
immediate responses can lead to greater satisfaction and a higher intention to
continue using the service. Coercive Pressure (CP), characterized by external
influences pushing users to adopt Chatbot services, was also identified as a
significant factor impacting users' continuous intention. The results suggest
that certain external factors can influence users' decisions to continue using
Chatbots, highlighting the importance of considering external drivers in
adopting Chatbot technology. Furthermore, the influence exerted by the external
environment and the opinions of relatives and friends is not sign needs to be
enough to have a continuous intention to use Chatbot. Hence, the research
hypotheses about normative pressure and mimetic pressure on the constant
intention of using Chatbot have yet to prove their validity based on the data
collected from the respondents.
The pressure to conform to social norms and expectations,
specifically regarding adopting Chatbot services, does not significantly
influence bank clients' continuous intention to use Chatbots. The
non-significant effects of NP and MP on CI highlight the need for further
investigation into the specific circumstances and contexts under which these
factors may or may not influence customer adoption behavior. Normative
pressure, which involves individuals conforming to social norms or behaviors
considered normal or popular within their society or social circles, is a
well-documented psychological phenomenon. Nevertheless, within the context of
continuous intention to use Chatbots, empirical investigations have suggested
that the impact of normative pressure does not play a significant role in
prompting bank clients to continuously use Chatbot services. When the user
perceives that the Chatbot does not effectively fulfill their needs or deliver
the desired outcomes, their reliance on the technology is primarily guided by
this assessment rather than conforming to social norms. In the context of
mimetic pressure, the tendency to imitate or follow the actions of others in
adopting Chatbot services does not significantly impact bank clients'
continuous intention to use Chatbots, particularly when the users perceive the
Chatbot as lacking in utility. In such circumstances, the user's evaluation.
This finding suggests that clients' adoption decisions are not heavily
influenced by observing others' behavior or actions related to Chatbot usage.
The lack of significant influence of mimetic pressure highlights that clients'
perceptions of the value and effectiveness of Chatbots play a more critical
role in shaping their continuous intention to use these services. This result
underscores the importance of considering other factors in understanding
customer acceptance and adoption of banking technologies. In contrast, the
usage of Chatbot is significant and approved due to the beneficial influence of
the results of utilizing chatbots, both in terms of perceived usefulness and
perceived ease of use.
Theoretical Implications
This is one of the early studies that we are aware of
that apply TAM and social contagion theory to the context of Chatbot Service's
continuous usage intention to determine the coexistence of the aspects in these
theories that define Chatbot Service's ongoing usage intention. Utilizing the
TAM and Social Contagion theories, this study established hypotheses to
investigate their direct influence on Chatbot Continuous Intention. However,
it's worthwhile to note that the antecedents of adoption from TAM and Social
Contagion function better when they are separately used in the context of
adoption and continuous intention, respectively. Additionally, the examination
asserts that Perceived Usefulness is positively related to Attitude Toward
Using, as supported by (Zarouali et al., 2018), while Attitude Toward Using further predicts Chatbot
Continuous Usage Intention (Ashfaq et al., 2020). A significant number of studies in the field of Chatbot
Services have supported this assertion. Revealing these two predictors is still
salient in research on determining Chatbot Continuous Intention.
This present investigation expands on existing research
that explores the impact of social contagion on the ongoing intention to use
Chatbots. Previous studies have predominantly concentrated on the influence of
social acceptance, specifically others' approval, in utilizing Chatbots.
Managerial Implication
The results have practical consequences for providers of
chatbot services. In practice, it was discovered that Perceived Usefulness positively
influences Chatbot Service Continuous Use Intention. This calls for chatbot
service providers, particularly banks, to provide chatbot services to let users
use chatbot services while performing various tasks. Similarly, the Perceived
ease of use of chatbot service significantly impacts the Chatbot service�s
Continuous Usage Intention of users. This also calls for banks to ensure
chatbots provide services that give users a good experience while using
chatbots. In that, customers are more likely to continue using Chatbot service.
Furthermore, as customers enjoy the continuous usage of chatbot services, it
motivates bankers to continue with innovative services and provide constant
updates to enhance and promote better services.
Prior research has been less active in finding social
contagion to improve our understanding of the predictors of intent to use a
chatbot service. In essence, how might the impact of social contagion be
significant for necessary technologies like Chatbots? Addressing this question
emphasizes the importance of considering attitude and social contagion as
pivotal factors in the inference-drawing process. As the authors know, the
intertwined functions of attitude transferability across a bank's distribution
channels and the contagion of social influences in technology usage have yet to
be specifically explored. By utilizing their established self-service
technology and examining social contagion, the present study introduces a
comprehensive research model to assist banks in effectively encouraging their
clients' ongoing utilization of chatbot services.
The study's findings indicate that banks can enhance the
continuous usage of Chatbot services through advertising campaigns, adhering to
practical and valuable criteria. As clients encounter inquiries about banking
services for transactions, there is an opportunity to improve services by
establishing more effective methods, contingent upon banks' profound
understanding of the combined contagious effects that arise in such situations.
Moreover, financial institutions can leverage diverse
manifestations of social contagion to impact the rates of continuous intention
toward chatbot services. As previously mentioned, coercive pressure exerts a
more substantial and comparatively more significant influence than normative
and mimetic pressures in shaping customers' inclination to utilize chatbot
services persistently. Banks are advised to tap into social networks
facilitated by mobile phone users, fostering word-of-mouth dissemination (e.g.,
peer recommendations via emails) to enhance adoption rates among non-adopters.
This underscores banks' reliance on authentic mobile banking customers to
sustain the operational effectiveness of chatbots. For instance, this reliance
could involve enlisting well-known figures (e.g., celebrities, prominent
business figures, politicians, athletes, etc.) as advocates for chatbot usage.
In the context of coercive pressure, banks may explore partnerships and
alliances with other private or public entities, such as service providers or
public authorities, to compel their clients to persist in using chatbot
services. Customers are presented with the option to continually utilize
chatbot services for financial transactions entirely accessible through mobile
phones.
CONCLUSION
The conclusion of this research is to
introduce a conceptual model for studying users' intention to use chatbots by
combining the Technology Acceptance Model (TAM) and social influence. This
study addresses the gap in empirical research on the continuous intention to
use chatbots by testing the proposed model on data from 300 respondents through
an online survey. The analysis results indicate that factors such as Perceived
Usefulness, Perceived Ease of Use, and Coercive Pressure significantly
influence users' intention to continue using chatbots. However, normative and
mimetic pressures were not found to have significant effects. These findings
significantly contribute to both theoretical understanding and practical
applications in continuous chatbot usage.
The implications of this research
suggest that understanding the drivers behind users' continuous intention to
use chatbots is crucial for the successful implementation and utilization of
chatbot technology in the banking industry. Banks can leverage the insights
from this study to tailor their chatbot services and marketing strategies more
effectively, focusing on enhancing perceived usefulness and ease of use while
considering the influence of coercive pressures. Additionally, the findings
highlight the importance of addressing social influence factors in the design
and promotion of chatbot services to encourage sustained user engagement.
REFERENCES
Adam, M., Wessel, M., & Benlian, A. (2021). AI-based
chatbots in customer service and their effects on user compliance. Electronic
Markets, 31(2), 427�445.
Ahmad, A., Rasul, T., Yousaf, A., & Zaman, U. (2020).
Understanding factors influencing elderly diabetic patients' intention to use
digital health wearables: extending the technology acceptance model (TAM). Journal
of Open Innovation: Technology, Market, and Complexity, 6(3), 81.
Al-Emran, M., Arpaci, I., & Salloum, S. A. (2020). An
empirical examination of continuous intention to use m-learning: An integrated
model. Education and Information Technologies, 25, 2899�2918.
Alsharo, M., Alnsour, Y., & Alabdallah, M. (2020). How
habit affects continuous use: evidence from Jordan�s national health
information system. Informatics for Health and Social Care, 45(1),
43�56.
Ashfaq, M., Yun, J., Yu, S., & Loureiro, S. M. C. (2020).
I, Chatbot: Modeling the determinants of users� satisfaction and continuance
intention of AI-powered service agents. Telematics and Informatics, 54,
101473.
Carranza, R., D�az, E., S�nchez-Camacho, C., &
Mart�n-Consuegra, D. (2021). e-Banking adoption: an opportunity for customer
value co-creation. Frontiers in Psychology, 11, 621248.
Chaouali, W., & El Hedhli, K. (2018). Toward a
contagion-based model of mobile banking adoption. International Journal of
Bank Marketing, 37(1), 69�96.
Duarte, P., e Silva, S. C., & Ferreira, M. B. (2018). How
convenient is it? Delivering online shopping convenience to enhance customer
satisfaction and encourage e-WOM. Journal of Retailing and Consumer Services,
44, 161�169.
Fauzi, A. A. (2019). CRITICAL FACTORS ON SME MANAGEMENT OF
ONLINE DELIVERY SERVICE APPLICATION. International Journal of Business and
Society, 20(3), 1130�1148.
Foroughi, B., Iranmanesh, M., & Hyun, S. S. (2019).
Understanding the determinants of mobile banking continuance usage intention. Journal
of Enterprise Information Management, 32(6), 1015�1033.
Gupta, P., Prashar, S., Vijay, T. S., & Parsad, C.
(2021). Examining the influence of antecedents of continuous intention to use
an informational app: the role of perceived usefulness and perceived ease of
use. International Journal of Business Information Systems, 36(2),
270�287.
Hair Jr, J. F., Ringle, C. M., Danks, N. p, Hult, G. T.,
Sarstedt, M., Bolyai.Babes, & Ray, S. (2021). Review of Partial Least
Squares Structural Equation Modeling (PLS-SEM) Using R: A Workbook. In Structural
Equation Modeling: A Multidisciplinary Journal.
https://doi.org/10.1080/10705511.2022.2108813
Hiqmah, F. (2020). Determinants of the Intention of Social
Aid Beneficiaries to Use Banking Self-Service Technology (SST). The
International Technology Management Review, 10(1), 12�18.
Hosseini, M., Shajari, S., & Akbarabadi, M. (2022).
Identifying multi-channel value co-creator groups in the banking industry. Journal
of Retailing and Consumer Services, 65, 102312.
Hwang, S., & Kim, J. (2021). Toward a chatbot for
financial sustainability. Sustainability, 13(6), 3173.
Ilhamalimy, R. R., & Ali, H. (2021). Model perceived risk
and trust: e-WOM and purchase intention (the role OF trust mediating IN online
shopping IN shop Indonesia). Dinasti International Journal of Digital
Business Management, 2(2), 204�221.
Jain, M., Kumar, P., Kota, R., & Patel, S. N. (2018).
Evaluating and informing the design of chatbots. Proceedings of the 2018
Designing Interactive Systems Conference, 895�906.
Karri, S. P. R., & Kumar, B. S. (2020). Deep learning
techniques for implementation of chatbots. 2020 International Conference on
Computer Communication and Informatics (ICCCI), 1�5.
Kemp, A., Palmer, E., & Strelan, P. (2019). A taxonomy of
factors affecting attitudes towards educational technologies for use with
technology acceptance models. British Journal of Educational Technology,
50(5), 2394�2413.
Kiogothe, M. W. (2018). Factors influencing adoption of
mobile banking in Kenya: a case of commercial banks� customers in Nairobi
County. Strathmore University.
Lee, C. M. J., Che-Ha, N., & Alwi, S. F. S. (2021).
Service customer orientation and social sustainability: The case of small,
medium enterprises. Journal of Business Research, 122, 751�760.
Lui, A., & Lamb, G. W. (2018). Artificial intelligence
and augmented intelligence collaboration: regaining trust and confidence in the
financial sector. Information & Communications Technology Law, 27(3),
267�283.
Masocha, R., & Fatoki, O. (2018). The impact of coercive
pressures on sustainability practices of small businesses in South Africa. Sustainability,
10(9), 3032.
Mero, J. (2018). The effects of two-way communication and
chat service usage on consumer attitudes in the e-commerce retailing sector. Electronic
Markets, 28, 205�217.
Misischia, C. V., Poecze, F., & Strauss, C. (2022).
Chatbots in customer service: Their relevance and impact on service quality. Procedia
Computer Science, 201, 421�428.
Nguyen, D. M., Chiu, Y.-T. H., & Le, H. D. (2021).
Determinants of continuance intention towards banks� chatbot services in
Vietnam: A necessity for sustainable development. Sustainability, 13(14),
7625.
Prasetya, S. A., Erwin, A., & Galinium, M. (2018).
Implementing Indonesian language chatbot for e-commerce sites using artificial
intelligence markup language (AIML). Prosiding Seminar Nasional Pakar,
313�322.
Purwanto, A. (2021). Partial least squares structural
equation modeling (PLS-SEM) analysis for social and management research: a
literature review. Journal of Industrial Engineering & Management
Research.
Rahmayati, R. (2021). Competition Strategy In The Islamic
Banking Industry: An Empirical Review. International Journal Of Business,
Economics, And Social Development, 2(2), 65�71.
Richad, R., Vivensius, V., Sfenrianto, S., & Kaburuan, E.
R. (2019). Analysis of factors influencing millennials' technology acceptance
of chatbots in the banking industry in Indonesia. International Journal of
Civil Engineering and Technology, 10(4), 1270�1281.
Rohmatulloh, I. H., & Nugraha, J. (2022). Penggunaan
Learning Management System di Pendidikan Tinggi Pada Masa Pandemi Covid-19:
Model UTAUT. Jurnal Pendidikan Administrasi Perkantoran (JPAP), 10(1),
48�66.
Wang, Y. M., Ahmad, W., Arshad, M., Yin, H. L., Ahmed, B.,
& Ali, Z. (2021). Impact of coordination, psychological safety, and job
security on employees� performance: The moderating role of coercive pressure. Sustainability,
13(6), 3175.
Zarouali, B., Van den Broeck, E., Walrave, M., & Poels,
K. (2018). Predicting consumer responses to a chatbot on Facebook. Cyberpsychology,
Behavior, and Social Networking, 21(8), 491�497.
Zhang, S.-N., Li, Y.-Q., Liu, C.-H., & Ruan, W.-Q.
(2019). Critical factors in identifying word-of-mouth enhanced with travel
apps: the moderating roles of Confucian culture and the switching cost view. Asia
Pacific Journal of Tourism Research, 24(5), 422�442.
Zhang, T., Lu, C., & Kizildag, M. (2018). Banking
�on-the-go�: examining consumers� adoption of mobile banking services. International
Journal of Quality and Service Sciences, 10(3), 279�295.
