The research inherited four factors from the research model of Nor, Nazarie, and Yusoff (2013): (1) Trust Propensity, (2) Testimonials, (3) Experience in Online Purchasing, (4) Financial Perceived Risk and based on the interview, the study developed two more factors: (5) Product Perceived Risk and (6) Time Perceived Risk.
Trang 1FACTORS INFLUENCING CUSTOMER’S TRUST WHEN PURCHASING
ON FACEBOOK SOCIAL NETWORK
Tran Thi My Phuong1
1 An Giang University
Information:
Received: 10/06/2018
Accepted: 14/09/2018
Published: 02/2019
Keywords:
Trust, trust propensity,
testimonials, risk
ABSTRACT
Trust is the key element for the success of trading on Facebook, therefore, finding the factors influencing trust of customers is necessary The research inherited four factors from the research model of Nor, Nazarie, and Yusoff (2013): (1) Trust Propensity, (2) Testimonials, (3) Experience in Online Purchasing, (4) Financial Perceived Risk and based on the interview, the study developed two more factors: (5) Product Perceived Risk and (6) Time Perceived Risk After surveying 200 customers, the research shows that Trust Propensity affects the most positively to Online Trust and Financial Perceived Risk affects negatively to Online Trust The study implicates to help management build their customers’ trust as well as offers a particular reality in an e-commerce research field
1 INTRODUCTION
Purchasing online has been developing
dramatically in all corners of the world due to its
practical benefits such as help customers save
time, compare prices, access new products easily,
and save transportation cost as well as avoid
traffic congestion Recently, a large number of
customers has moved to shop on social networks,
especially Facebook instead of buying on formal
websites such as Vatgia.com or Lazada.vn
Facebook is considered as a public house of
one-seventh of world’s population and one- third of
Vietnamese population (VNNIC, 2013)
Individuals and businesses, therefore, are taking
advantage of this medium to both sell products
and keep in touch with their friends (Ranganathan
& Ganapathy, 2002)
In spite of many advantages mentioned above, the
majority of Internet users are not confident to make a purchasing decision on Facebook According to Gefen (2000), Harn, Khatibi and Ismail (2006), lack of trust is a major barrier in an online purchase This can be explained that Facebook users are unlimited and there has been
no legal organization regulating trading activity
on it It is easy for a person to cheat customers out
of money on Facebook
Nevertheless, most of the studies focused on the traditional e-commerce and there are few types of researches conducted to examine factors that affect customer’s trust on Facebook This study aims to fill in this gap in order to help businesses increase their customer’s trust by some particular actions; besides, its results can reliably contribute
to research of e-commerce field in general
2 LITERATURE REVIEW
Trang 22.1 Trust and Online Trust
Trust has been identified differently among
researchers Koufaris and Hampton-Sosa (2004)
suggest that trust is the willingness to rely on a
third party after the first interaction with it
Another study conducted by Coulter and Coulter
(2002) defined trust as a key factor to build
long-term relationships between service representatives
and their customers
In the Internet shopping platform, online trust is
also the willingness of a consumer to be
defenseless to the actions of an Internet business
and based on the hopes that he or she will be
treated agreeably, regardless of the ability of the
consumer to regulate or control the Internet
business (Lee & Turban, 2001) Meanwhile, Hui
and Kejin (2009) defined online trust as a process
including requirement, attitude, intention, and
purchase behavior
From various definitions, trust shows an
important role in online shopping Without trust,
customers are not confident enough to make
online purchasing decision and businesses can not
establish sustainable relationships with their
customers
2.2 Social Networks (SNs)
SNs is the top of Internet applications which
allows users to create profiles and keep contact
with friends who are already registered on the site
Some of the examples are MySpace, Facebook,
Friendster and Twitter In Vietnam, Facebook has
about 9 million users (Comscore, 2012) and
becomes a phenomenon in the youngsters even
the elderly and the kids One of the special
functions of Facebook is that users can upload
their photos and make their information visible to
their friends This enables them to communicate
more effectively and make friends unlimitedly In
terms of e-commerce, sellers find it easy to
describe the products and services, and convenient
to reply customer’s questions on Facebook, while
buyers think it easy to approach new products
from their favorite Facebook stores, easy to place
orders or write testimonials about some certain products However, everything has its two sides, Facebook has some disadvantages for both sellers and buyers
2.3 Some typical researches on Online Customer’s Trust
Many studies have been conducted to investigate factors that influence customers’ trust in an online purchase One of recent research is by Nor, Nazarie, and Yusoff (2013) They studied the factors influencing individuals’ trust in an online purchase through social network sites and the results discovered that (1) Trust propensity, (2) online purchasing experiences, (3) testimonials, and (4) money risk affect customers’ trust in online purchase The authors surveyed 129 customers who come from Malaysia, China, by sending emails and uploading the questionnaires
on their individual Facebook accounts Especially, the research used the method of multiple regression analysis to test the relation of the factors and discovered that testimonials have the strongest effect on customer’s trust In other words, good comments and feedbacks which come from old customers are advantageous for attracting new customers and building stable relations among existing customers
Moreover, Gefen (2000) also studied the role of familiarity and trust in an e-commerce context He analyzed the relations of five factors: (1) familiarity, (2) trust propensity, (3) trust, (4) inquiring, and (5) purchasing The results show that both familiarity with an Internet vendor and its processes as well as trust in the vendor influenced purchasing intension Therefore, it is necessary to concentrate on building trust by increasing familiarity Moreover, the study proved that familiarity is primarily people's disposition to trust that influenced their trust in the venders
In addition, another important research was conducted by Kim et al (2007) on “A trust-based consumer decision-making model in electronic commerce: The role of trust, perceived risk, and
Trang 3their antecedents” This study focused on
developing a theoretical framework which
describes the trust-based decision-making process
that a consumer uses when purchasing on the
Internet and on examining the model which use
Structural Equation Modeling method The study
results show that online trust and perceived risk
affect customer’s purchasing decisions strongly
In Vietnam, the number of research on online trust
has been increasing Tran Minh (2012) studied
“Factors affecting customer trust in online
shopping in Vietnam” The author proposed a
research model including four factors affecting
trust in online shopping: (1) Privacy Perceptions,
(2) Security Protection, (3) Perceived Risks, and
(4) Perceived Benefits The study used
convenience sampling and purposing sampling to
select 216 Vietnamese students and office
workers who used to shop online by electronic
payments The results proved that Perceived Risks
and Privacy Perceptions have strong impacts on
online trust This is explained that in a developing
country like Vietnam, customers are not familiar
with online shopping Therefore, they tend to care
about risks rather than profits
Furthermore, Nguyen Mai Trang and Nguyen
Ngoc Bich Tram (2014) studied about the factors
affecting actual use of Facebook in Vietnam by
examining six factors: (1) Social Identity, (2)
Altruism, (3) Telepresence, (4) Perceived ease of
use, (5) Perceived usefulness, and (6) Perceived
encouragement There were 363 Facebook users
in Ha Noi, Da Nang, and Ho Chi Minh City
taking part in the survey The results indicated
that Perceived ease of use, Perceived usefulness,
and Perceived encouragement influence directly
on the actual use of Facebook
In conclusion, the above research overview
indicates that in Vietnam most of the studies
concentrated on online purchasing intention or
Facebook-using behavior while many researchers
in the world have investigated online trust
popularly and profoundly in different sides The
factors mentioned and repeated in most of the researches are Trust Propensity, Online Purchasing Experience, Testimonials, and Perceived Risks This inspires the author to implement a study to survey the factors influencing customer’s online trust when purchasing on Facebook
3 Research methodology and hypothesis 3.1 Research methodology
The study was implemented by combining quantitive and qualitative research methodology After reviewing a wide range of national and international papers and researches related to online trust, the author conducted six focus interviews with six customers who had purchased
on Facebook in order to compare the previous results to the reality and to find out new factors
An online questionnaire was prepared and developed based on the previous results and focus interviewed results The study used convenience sampling method which allows reaching the responders who are available conveniently to participate in the study and targeted Facebook users who had purchased on Facebook All of 38 dependent and independent variables were measured by Likert 5 scales
3.2 Research hypothesis
Trust propensity was mentioned in several types
of researches on e-commerce According to McKnight and Chervany (2002), trust propensity
or disposition to trust can be defined as “general willingness based on extended socialization to depend on others” and concluded that those who have a strong belief in human’s good characteristic may have a high level of trust propensity This finding leads to the first hypothesis of the study:
Hypothesis 1: Trust propensity influences positively on customers’ trust when purchasing on Facebook
Testimonials play an important role in online shopping Testimonials are sentences spoken or
Trang 4written about some feelings of customers after
they experienced the products or services with a
hope that they can recommend or prevent other
customers from buying a particular product or
service In social networks, customers can write
their testimonials publicly or send a message
privately (Dwyer, Hiltz & Passerini, 2007) That
is the reason why businesses on Facebook try to
gather more and more positive testimonials about
their product or service to enhance their
customers’ trust Therefore, the second hypothesis
is suggested
Hypothesis 2: Testimonials influence positively
on customers’ trust when purchasing on
Jiang, Chen and Wang (2008) researched on the
relation between knowledge, trust in online
purchasing and online buying intention of
customers The results pointed out that knowledge
are related positively to trust in online purchasing
In the words, those who have awareness of online
shopping tend to trust and be willing to shop
online because they can identify risk in online
transactions The third hypothesis is given
Hypothesis 3: Experience in online purchasing
influences positively on customers’ trust when
purchasing on Facebook
Purchasing online involves a wide range of
risks in which financial risk is the most
concerned Most of the cases, the customers lose
their money by sending money in advance and do
not receive their ordered products or receive
dissatisfied ones Moreover, the customers have to
disclose their personal information to the sellers
such as phone numbers, home address, e-mail address, and debit or credit card numbers which are attractive to financial frauds (March, 2006) Therefore, the fourth hypothesis is proposed Hypothesis 4: Financial risk influences negatively
on customers’ trust when purchasing on Facebook
Product risk is another widespread issue that every customer concerns before making a purchasing decision Product risk involves both quality and quantity In terms of quantity, customers receive more or fewer items than they order In terms of quality, customers receive products that are not matched to the ones they order, namely they may be damaged or defective and they may have different color or size In another word, product risk is the loss that customers get when the provided products are not matched to the ones that they expect (Horton, 1976) Therefore, the fifth hypothesis is proposed Hypothesis 5: Product risk influences negatively
on customers’ trust when purchasing on Facebook
Despite the convenience of online shopping, a loss of time may occur in all stages such as making inquiries, ordering products, paying the bills, receiving products, even asking for refund Roselius (1971) confirmed in his research that time loss was considered as an additional risk of products and services Therefore, the sixth hypothesis is proposed
Hypothesis 6: Time risk influences negatively on customers’ trust when purchasing on Facebook
Trang 5Figure 1 Research Model
4 Research results
4.1 Scale descriptions
The data was collected from 200 respondents who
have been purchasing on Facebook and 38% are
male Most of them are from 22 to 35 years old
The majority have been buying online for a year
(accounting for 39%), for more than one year
(accounting for 26,5%), and for a month
(accounting for 20,5%)
4.2 Cronbach’s Alpha analysis
The Cronbach’s Alpha coefficient is used to
evaluate the reliability of the scale In addition, it
is a measure of internal consistency on the closely level that the variables of the scale correlate with each other in order to eliminate the inappropriate variables According to Nguyen Dinh Tho (2013),
a good reliability scale must have Cronbach’s Alpha coefficient from 0,70 to 0,80 However, the Cronbach’s Alpha coefficient of scale from 0,6 upward are accepted Furthermore, the condition
to carry out Exploratory Factors Analysis (EFA)
is that the variables must have corrected item-total correlation greater than 0,3 (Nunnally and Burnstein, 1994)
Table 1 Results of Cronbach’s Alpha Analysis
Variables Items Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted
Trust Propensity
Testimonials
Trang 6Variables Items Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted
Online Purchasing Experience
Financial Risks
Product Risks
Product Risks
(After deleting Pro2)
Time Risks
Trang 7Variables Items Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted
Time Risks
(After deleting Time8)
Trust
In the process of verifying reliability, Pro2 and
Time8 were deleted because they both have
Corrected Item-Total Correlation less than 0.3
The remaining 36 variables have Cronbach’s
Alpha coefficient of the constituent larger than
0.7; this means that the variables have closely
connected each other in the same factor The
observing variables in the same factors have
item-total correlation greater than 0.3; therefore all of
them were put into Exploratory Factors Analysis
(EFA)
4.3 Exploratory Factor Analysis on
Independent Variables
The analysis result shows that the independent
variables gather in groups and have loading
coefficients larger than 0.5 There are six factors
grouped: Time Risks, Financial Risks, Product
Risks, Online Purchasing Experience, Trust Propensity, Testimonials Among them, Test2, which belongs to Testimonials, had loading coefficient of less than 0.5 so it was removed from the scale Then, another EFA was conducted
in which KMO test coefficient had a value
of 0.840 > 0.5 showing the suitability of the analysis to the data (Nguyen Dinh Tho, 2013) Bartlett's Test of Sphericity is 4006.98 with Sig = 0.00 < 5% This asserts that variables closely connected each other (Nguyen Dinh Tho, 2013) Cumulative Variance is 71.293 > 50% In other words, the first factor can explain 71.29% of the variation of data (Gerbing and Anderson, 1988) Moreover, Eigenvalue coefficient is greater than one
Trang 8Table 2 Rotated Component Matrixa after deleting Test2
Component
time4 849
time3 847
time6 827
time2 824
time7 823
time1 817
time5 786
Trang 94.4 Exploratory Factor Analysis on
Dependent Variables
KMO Measure of Sampling Adequacy is 0.833 >
0.5 This means that the test is essential Bartlett's
Test of Sphericity is 653.509 with Sig = 0.00<
5% (0.05) showing that the variables correlate
closely each other Cumulative of Variance is
71,148 >50%
The matrix does not rotate because only one
dependent factor was extracted
In conclusion, after conducting Exploratory Factor Analysis, six independent variables and one dependent variable remain as before
4.5 Examining the reliability of the scale after removing Test2
This step aims at re-examining to ensure the scale
is still reliable after removing Test2 Fortunately, the result makes certain that Cronbach’s Alpha of the scale get value 0.819 and reliabilities of all items are larger than 0.5
Table 3 Cronbach’s Alpha after Removing Test2
Variable Items Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted
Testimonials
4.6 The result of Pearson Correlation
Coefficient Analysis
After encoding the average value of dependent
and independent variables, the author analyzes the
Pearson correlation in order to find out the
variables that are relevant to each other and to
remove irrelevant or fewer relevant ones
All of the Sig values are smaller than 5% This means that the Pearson Correlation Coefficient Analysis results are significant in statistics The table shows value from -0.310 to +1, that means the variables are relevant to each other and can be used in regression analysis
Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalization
a Rotation converged in 6 iterations
Trang 10Table 4 Correlation
TRUST Pearson Correlation 1 -.250** -.310** 551** -.254** 512** 475**
TIME Pearson Correlation -.250** 1 100 -.249** 184** 088 003
FIN Pearson Correlation -.310** 100 1 -.240** 227** -.159* -.096
PROP Pearson Correlation 551** -.249** -.240** 1 -.086 362** 558**
PRO Pearson Correlation -.254** 184** 227** -.086 1 083 116
EX Pearson Correlation 512** 088 -.159* 362** 083 1 489**
TEST Pearson Correlation 475** 003 -.096 558** 116 489** 1
Sig (2-tailed) 000 966 177 000 101 000
** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.05 level (2-tailed
N = 200
4.7 Regression analysis
Regression analysis was conducted between
TRUST and six independent variables (1) Trust
Propensity (PROP), (2) Testimonials (TEST), (3)
Online Purchasing Experience (EX), (4) Time
Risks (TIME), (5) Financial Risks (FIN), (6)
Product Risks (PRO)
The research’s linear regression equation is as
following:
TRUST =β 0 + β 1 * PROP + β 2 * TEST + β 3 * EX -
β 4 ATIME- β 5 FIN- β 6 PRO
With R2 of 0.518 indicating that the independent
variables in the regression model can explain
51.8% of the variation of the dependent variable Sig value of F test < 0.05 shows that the model is statistically significant at significance level 5% In addition, VIF coefficients of all factors are < 2 implying that multi-collinearity does not influence much on regression results
All regression coefficients (β) are statistically
significant since Sig β < 5% There are three
regression coefficients taking negative sign corresponding to risk factors that affect negatively TRUST All the six hypotheses were therefore accepted