ArticlePDF Available

Abstract

Technological developments are very rapidly making changes in consumer behaviour where there is a transition from offline transactions to online in Yogyakarta. The purpose of this study is to evaluate and validate the influence of online trust and the factors that influence it is e-commerce knowledge, perceived reputation, perceived risk, perceived technology, prior purchase experience, to the intention of purchasing online. This study also examines the effect of perceived technology and prior purchase experience on online purchase intentions. Respondents in this study are about 260 respondents. Respondents must be domiciled in Yogyakarta and have conducted online transactions in the last 3 months. Analysis of data on research using Smart PLS 3.0. The test of measurement model that is convergent validity, discriminant validity, and internal consistency reliability is done to ensure the validity and reliability of the questionnaire and then tested the structural model to test the hypothesis, besides the fit model, predictive relevance and effect size of each latent variable. This study found that perceived risk is the most influencing factor of consumer confidence followed by prior purchase experience, perceived technology, and perceived reputation. The study also found that perceived technology and trusts influence online purchase intentions. While e-commerce knowledge has no effect on consumer trust and prior purchase experience has no effect on the intention of purchasing online.
International Journal of Industrial Engineering & Production Research September 2022 Vol. 33, No. 3: 1-9
DOI: 10.22068/ijiepr.33.3.8
Determinants of Online Trust and Their Impact on Online Purchase
Intention In Yogyakarta
Budi Suprapto1*, Paulus Dian Wicaksana2 & Mohd Fazli Mohd Sam3
Received 1 9 April 2022; Revised 21 May 2022; Accepted 31 May 2022;
© Iran University of Science and Technology 2022
ABSTRACT
Technological developments are very rapidly making changes in consumer behaviour where there is a
transition from offline transactions to online in Yogyakarta. The purpose of this study is to evaluate
and validate the influence of online trust and the factors that influence it is e-commerce knowledge,
perceived reputation, perceived risk, perceived technology, prior purchase experience, to the intention
of purchasing online. This study also examines the effect of perceived technology and prior purchase
experience on online purchase intentions. Respondents in this study are about 260 respondents.
Respondents must be domiciled in Yogyakarta and have conducted online transactions in the last 3
months. Analysis of data on research using Smart PLS 3.0. The test of measurement model that is
convergent validity, discriminant validity, and internal consistency reliability is done to ensure the
validity and reliability of the questionnaire and then tested the structural model to test the hypothesis,
besides the fit model, predictive relevance and effect size of each latent variable. This study found that
perceived risk is the most influencing factor of consumer confidence followed by prior purchase
experience, perceived technology, and perceived reputation. The study also found that perceived
technology and trusts influence online purchase intentions. While e-commerce knowledge has no effect
on consumer trust and prior purchase experience has no effect on the intention of purchasing online.
KEYWORDS: E-Commerce knowledge; Perceived risk; Perceived technology; Prior purchase
experience; Online trust; Online purchase intention.
1. Introduction1
According to data from the Association of
Internet Service Providers Indonesia there are
98.6% or 130.8 million people know that the
internet as a place of buying and selling goods
and services, as many as 63.5% or 84.2 million
people have done transactions online,
onlineshop or buying and selling sites into
commercial content most frequentl y visited by
Internet users with a percentag e of 62% or 82.2
million inhabitants.
From these data proves that the level of
Corresponding author: Budi Suprapt o
*
budi@staff.uajy.ac.id
1. Faculty of Economics, Universitas Atma Jaya Yogyakarta, Jl .
Babarsari no.43, Yogyakarta, 55281, Indonesia.
2. Master of Management Program, Universitas Atma Jaya
Yogyakarta, Jl. Babarsari no.43, Yogyakarta, 55281,
Indonesia..
3. Faculty of Techno logy Management and Technopreneurshi p,
Universiti Teknikal Malaysia M elaka, 7 5450 Melaka,
Malaysia . & Cent re for Robotics and Industrial Automation
(CeRIA), Fakulti Kejuruteraan Elekt rik (FKE), Universit i
Teknikal Malaysi a Mel aka, 75450 Me laka, Malaysia.
enthusiasm of the community on online trading
sites in Indonesia is quite high. Many previous
studi es have examined the factors affecting
online purchasing intentions (Lim et al, 2016;
Wijoseno and Ariyanti, 2015; Kim et al., 2011;
Ling et al., 2011). Lack of trust proved to be an
important factor affecting the intentions of
online purchases (Wijoseno and Ariyanti, 2015;
Kim et al., 2011; Ling et al., 2011; Trust
becomes an important aspect of online
shopping because consum ers will not shop if
they do not believe in seller sites that
consumers do not trust (Ponte, 2015).
According to the past studies, most of the effort
has been invested in examining the trust issues
in online shopping and indicated t hat
determinants of online trust has a i mpact on
online purchase intention(Yusi et al., 2016;
Wijoseno and Ariyanti, 2015; Assegaff, 2015;
Paramita et al.,2014; Sahir et al.,2014; Ling et
al.,2011). Ther efore, theories and studies have
been put forth to determine that what causes a
cust omerto trust and its impa ct to online
RESEARCH PAPER
[ Downloaded from ijiepr.iust.ac.ir on 2022-06-27 ]
1 / 9
2 Determinants of Online Trust and Their Impact on Online Purchase Intention In Yogyakarta
International Journal of I ndustrial Engineering & Production Research, September 2022, Vol. 33, No. 3
purchase intention.According to study
conducted by Yusi et al. (2016), found t hat
previous online purchasing exp erience has
positive effect on online trust, and then online
trust has a positive effect on purchasing
decision, as well as previous online purchasing
experience have a positive effect on purchasing
decision.Indeed, Ling et al. (2010) state t hat
prior online purchas e experience and online
trust do have a positive effect on online
purchase intention. Wijos eno and Ariyanti
(2015) identified factors like e-commerce
knowledge, perceived risk , perceived
reputation, perceived technology affect
cust omer trust on purchase intention.This study
found that e-commerce knowl edge, perceived
risk, perceived technolog y influence consu mer
trust but perceived reputation does not affect
consumer trust. Moreover, the study also
shows that online trust have a significant effect
to the intention of consumer purchase online.
Sahir et al. (2014) highlighted the effects of risk
perception, perceptions of convenience and the
perception of benefits to online purchasing
decisions. According to his study risk
perception, perception of ease and perception of
benefits affect the purchase decisions online.
Similarly, Ling et al. (2011) highlighted on the
influence of perceived risk (perceived risk), the
ease of usin g technology (perceived
technolog y) against online trust. In addition,
this research also exa mines the effect of
perceived technol og y on online purchasing
intention medi ated by online trust and examines
how online trusts relate to online pur chase
intentions. This study found that perceived risk
and perceived technology have positi ve effect
on online trust and online trust hav e positive
effect on online purchase intention and
perceived technology have positive effect on
online purchase intention mediated by online
trust.
After man y studies in the past have evaluated
the behaviour of online consumers. Researchers
have not yet developed a comprehensive
understanding of the determinants of consumer
trust in the online shopping and its relationship
withconsumers online purchase
intentionespeciall y for consumers bu ying and
selling sites in Yogya karta. Another gap that
could be pointed outis that most of the previous
researches have used regression analysis to
check the impact of thedeterminants. Therefore,
this study aims to research about determinants
of online trust and their impact on online
purchase intention using PLS -SEM algorithm
to help formulat e better strategies to incr ease
the level of consumer confidence in the
intention of bu ying online and help companies
online trading sites to cr eate online buying and
selling sites better by understanding the
determinants of trust can help adding business
value to company through understanding
consumers perception towards online
purchasing better.
2. Shop Online Factor
According to Monsuwe et al. (2004) i n
Wijoseno an d Ariyanti (2015) the limitations o f
time, distance, the need for scarce goods, and
the attractiveness of alternatives in shopping is
a situation that encoura ges one to make online
purchases. Consumers of online trading sites
tend to pay attention to perception information
about the product, delivery, payment, privacy,
security, visual appeal, entertainment,
convenience, and convenience. The perception
of these variables in previous exp erience will
affect the intent of the next purchase.
The ease of operating the bu ying and selling
site will add to the level of consumer
confidence in the trading site (Ling et al.,
2011). Farag et al. (2006) in Yuliati and
Simanjuntak (2011) show that online search
and benefit perception have a p ositive effect on
expenditure frequency, and this has a positive
effect on online purchases.According to
Jarvenpaa et al. (1999) in Wijoseno and
Ariyanti (2015) reputation is generall y
suggest ed to be a fa ctor that contributes to
consumer confidence t o the seller within an
organization.P avalo et al., (2013) found that the
perceived risk consumers ha ve a negative effect
on the intentions of consum er transactions so
that consumers tend to be motivated to avoi d
transactions. According Suhir et al., (2014)
perceived risk can be interpreted as a subjective
judg ment by a person against the possibility of
an accident event and how concerned the
individual with the consequences or the i mpact
of the event. When consumers feel that there is
a risk on the activity of buying and selling
online it will affect consumer confidence.
According to Ling et al. (2010) online quality-
buying sites will rely heavily on the quality of
consumer experience that can be gained
through previous exp erience. If the quality of
the previous p urchase experience is good then it
will affect the level of consumer confidence.
Past purchase experience will affect future
purchase intentions (Weisberg et al., 2010).
Trust becomes an important aspect of online
shopping because consumers will not shop if
they do not believe in seller sites that
[ Downloaded from ijiepr.iust.ac.ir on 2022-06-27 ]
2 / 9
3 Determinants of Online Trust and Their Impact on Online Purchase Intention In Yogyakarta
International Journal of I ndustrial Engineering & Production Research, September 2022, Vol. 33, No. 3
consumers do not trust (Pont e, 2015).
According to previous study, prior discussion
and several literature review has led to formed
eight hypotheses in this study.
H1: Perceived Technolog y (the ease and
benefits of online buying and selling sites) will
have a positive effect on consumer confidence
in online trading sites.
H2: Perceived Technology will positivel y
influence consumer purchase intention.
H3: Reputation of online trading sites will
affect consumer confidence in online buying
sites.
H4: The risk of buying and selling transactions
on the site negatively affects consumer
confidence in online trading sit es.
H5: Knowledge of online buying sites has a
positive effect on consumer trust on online
trading sites.
H6: The previous purchas e experience will
have a positive effect on consumer confidence
in online trading sites.
H7: The previous purchas e experience will
have a positive effect on consumer purchase
intentions
H8: Consumer confidence positively affects the
buying intentions of cons umers of online
trading sites
Fig. 1. Determinants of online trust and their impact on online purchase intention
3. Research Methodology
This study uses a quantitative approa ch with the
aim of obtaining the results of h ypothesis proof
through data that has been obtained previousl y
according to theories and concepts that ha ve
been there before. This research is descriptive
which explains the relationship or influence
among the variables studied (Malhotra, 2010).
Survey was carri ed out to coll ect primary data.
The measurement of e-commerce knowl ed ge
was adapted by Li et al., (2008). The
measurement of perceived reputation was
adapted by Lee et al., (2003) and Jarvenpaa et
al., (1999). The measurement of prior online
purchase experience was adapted by Pentina et
al., (2011). The measurement of perceived
technolog y was adapted by Gefen et al., (2003)
and Lee et al., (2006). T he measurement of
perceived risk was adapted and developed by
Ling et al., (2011), Kim et al., (1999)., and
Jarvenpaa et al., (1999). The measurement of
online trust was adapted by Mc Knight et al.,
(1998). Online purchase intentio n was
measured by scale adapted by Pavlou, (2003).
Finally, the survey instrument us ed in this study
consisted of a total of 25 items related to the
eight constructs of t he research model. The
items were measured using a 5-point Likert-
type scale for all constructs. T he respondent of
this res earch is the residents in Yogyakarta
which 17 yearsold and above. The eligible
respondent s for thi s research will be adults who
are. The surveys were conducted from
1stDecember2017 to 18th Februar y 2018.The
samples selected of this study are based on the
non-probab ility sampling. In this study,
purposi ve sampling is selected as main
sampling procedure which based on resear ch
objective. This study has gathered 310 potential
respondent s , however after selected based on
the criteria there are about 260 respondents who
can be processed.
The distribution of the respondents is 142 male
respondent s (55 percent of the total
respondent s) and 118 female respondents (45
percent of total respondents). In the other side
the majority of respondents fall into th e age
group of 17-23 years old (62 percent), group of
mor e than 24 years old (38 percent). Based on
mont hl y income analysis indicates that the
majority of respondents fall into the income of
mor e than 3.000.001 (39 percent), followed by
Perceived Technology Perceived Reputation
Online Trust
Online Purchase
Intention
H1 H2
H8
H3
Perceived Risk H4
E-commerce
knowledge
H5
Prior Online
Purchase
Experience
H6
H7
[ Downloaded from ijiepr.iust.ac.ir on 2022-06-27 ]
3 / 9
4 Determinants of Online Trust and Their Impact on Online Purchase Intention In Yogyakarta
International Journal of I ndustrial Engineering & Production Research, September 2022, Vol. 33, No. 3
income of less than 1.000.000 (36 percent), and
income of 1.000.001 3.000.000 (25 percent).
Based on sites that are freq uently visited by
respondents in this question respondents may
fill more than one, and analysis indicates that
the majority of respondents fall into Tokopedia
(53 percent), followed by Lazada (47 percent),
Other sites (38 percent), Bukalapak (32
percent), JD.ID (6 percent), and Blibli (6
percent).
4. Result and Discussion
According to Hair et al. (2017) testing the
model of measurement or out er model is the
thing to do before testing the structural model
or inner model. Therefore, th e authors perform
several tests of convergent validity, inter nal
consistency reliability, and discriminant
validity.
Convergent validity is a method for testing the
accuracy of research tools or items collected
independently of each other in which items in
theory must be related (Carmines, 1979).
Furthermore the authors convergent validity on
the data by using SmartPLS 3.0. In general if
outer loading value> 0.70 or higher. If research
is explorator y then 0.4 is acceptable (Hulland,
1999). The results showedthat the values for all
the constructs were more tha n 0.70 from the
lowest outer loading value of 0.805 (perceived
risk) to the highest of 0.911 (e-com merce
knowledge).All of the items within the
constructs were more than 0.70and were
convergent validity. Then the author also tested
to get the Average Variance Extracted value.
Criteria for convergent validity are AVE
(average variance extracted) proposed by
Fornell and Larcker (1981). A minimum AVE
of 0.5 means at least half of the average
variance of the indicator (Forn ell and Larcker,
1981). The results showedthat the values for all
the constructs were more than 0.5 from the
lowest of 0.691 (perceived risk) to the highest
of 0.795 (e-commerce knowl edge). All of the
items within the constru cts were more than
0.50and were con vergent validity.
Discriminant validity is a method of testing the
lack of relationships between items that are
theoretically unrelated (Carmines, 1979).
Furthermore, the author went on to t est the
discriminant validity of the questionnaire data
by using SmartPLS 3.0.To examine
discriminant validity, the shared variances
between factors is compared to the average
variance extracted of the individual factors
(Fornell et Larker, 1981). This analysis shows
that the shared variance between factors is
lower than the average variance extracted of the
individual factors, confirming dis cri minant
validity. The result shows that the "square root"
of each latent variable itself is higher than the
value of the relationship with other latent
variables then the questionnaire data in this
study is valid and feasible to be processed to
the next stage. According to Hair et al's
recommendations. (2017) to conduct good
validity dis cri minant also using cross loading
analysis then the authors also do cross loadin g
analysis. After analyzing cross-loading, it is
known that all latent variable indicator
relationships with the indicator variables
themselves have a higher cross -load value than
the correlation with other latent variable
indicators (Chin, 2010). T herefore the value of
cross loading is considered valid because it
meets the criteria. Finall y it is necessary to test
the VIF values (variance inflation factor) to test
the multicolinearity potential between items
although not required for the measurement of
the reflectiv e model, therefore only the inner
VIF values used as the evaluation criteria ar e as
suggest ed by Hair et al. (2017). According to
Hair et. al.,(2017), the maximum acceptable
VIF value would be 5.0, thus if VIF value
high erthan 5.0 would indicate a problem with
multicollinearity. the values of Varian ce
Inflation Factor (VIF) for all theconstructs were
less than 5.0 and the range of VIF value was
between 1.902 and 3. 258. The finding indicated
that the probl em of multicollinearit y was
notsignificant in this research.
Internal consistency realibility is a method to
test the extent to which the tools or items are
used for research to assist researchers in
interpreting data, determi ning value, and
relationships among variables (Carmines,
1979). To test the internal consistency
realibility of the authors perform tests on data
using SmartPLS 3.0. In general, Cronbach's
alpha coefficients should be above 0.7 at a
minimum in order to be considered as a good
stren gth of association (Eisingerich dan Rubera
,2010). The value of Cronba chs alpha of e-
commerce knowledge is 0.855, perceived
reputation is 0.803, perceived risk is 0.777,
perceived technology is 0.906, prior purchase
experience is 0.863, online purchase intenti on is
0.784 and online trust is 0.862. In conclusion,
the Cronbach Alphas for each variable is more
than 0.70. According to Djikstra dan Hens eler,
(2015) rho_A value> 0.70 or higher shown the
tools or it ems are us ed for research is reliable.
After analyzingrho_A value, in conclusion the
rho_ A for each variable is above 0.70 from the
[ Downloaded from ijiepr.iust.ac.ir on 2022-06-27 ]
4 / 9
5 Determinants of Online Trust and Their Impact on Online Purchase Intention In Yogyakarta
International Journal of I ndustrial Engineering & Production Research, September 2022, Vol. 33, No. 3
lowest of 0.782 (perceived risk) to the highest
of 0.908 (perceived technol ogy). According to
Bag ozzi dan Yi, (2012) composite realibility
must > 0.70 or higher. If resear ch is exploratory
then 0.60 or higher is acceptable to shown that
data have high reliability. After do composite
reliability the results showedthat the values for
all t he constructs were more than 0.70from the
lowest of 0.87 (perceived risk) to the highest of
0.921 (e-commerce knowledge).
In this study structural model test using
bootstrap with 5000 subsamples in accordance
with the recommendations of Chin (2010) and
Hair et al. (2017). The test done on the inner
model is R Square (explained variable), f
Square (effect size), Q Square (predictive
relevance) and path coefficient (testing
hypothesis). In table 1 will be shown from R
Square, Q Square and SRMR from research
data.
Tab. 1. Results of r square, q square and SRMR test
R2 Q2 SRMR
Online Trust 0.59 0.38 0.06
OnlinePurchase Intention
0.6
0.38
0.06
Hasil
Moderate
In this study the endogenous variable OT
received R2 of 0.59 then entered into the
moderate group (Hair et al., 2017), and the OT
variable got Q2> 0 of 0.38 which means having
predictive releva nce and having SRMR <0.08
means the model is good fit (Henseler et al.,
2017). While endogen variable of OPI got R2
equal to 0.6 then enter into moderate group
(Hair et al., 2017), and OT variable got Q2> 0
is 0.38 which means ha vin g predictive
relevance and having SRMR <0.08 means
model is good fit (Henseler et al ., 2017). Then
for the results of the h ypot hesis test can be seen
in table 2.
Tab. 2. Hypothesis test results
Path
Path
Coeffisient
f
2
T Values
P Values
Sig.
Hipotesis
H1 PT > OT 0.167 0.027 2.784 0.005 ** Diterima
H2 PT > OPI 0.452 0.232 7.539 0.000 *** Diterima
H3
PR > OT
0.167
0.031
2.346
0.019
*
H4
PRS > OT
0.408
0.200
5.898
0.000
***
H5
ECK > OT
0.009
0.000
0.182
0.855
Ns
Tidak Diterima
H6 PPE > OT 0.141 0.015 1.806 0.071 ns Tidak Diterima
H7 PPE > OPI 0.271 0.075 4.370 0.000 *** Diterima
H8
OT > OPI
0.142
0.027
2.590
0.010
*
In general, t values > 1.96 indicates that there is
a significant relationship between latent
variablesand p values > 0.05 in dicates wea k
evidence agai nst the null hypothesis, so you fail
to reject the null h ypothesis (Hair et al.,
2017).Accordi ng to table 2, the significant
value for perceived technology is 0.005. This
value is less than p value of 0.05 and the t value
is 2.784 is more than 1.96 thus, H1 is
support ed, whichproves that the perceived
technolog y is significantly affecting
theconsumers trust. Based ontable 2, the
significant value for perceived t echnology is
0.000. This value is less than p value of 0.05
and the t val ue is 7.