Factors of service quality affecting customer satisfaction of Highlands Coffee shop in Hanoi

Factors of service quality affecting customer satisfaction of Highlands Coffee shop in Hanoi

The case study of FPT University
----------------------------------------------------------------------------------------

GROUP 1

Lecturer: Mrs. Nguyen Hoang Phuong Linh
July, 2021


Table of Contents

CHAPTER 1: INTRODUCTION
1.1 Topic background
Recently, coffee culture and accompanying products and services are increasingly popular in the
lifestyle of Vietnamese people. In this day and age, people come to a coffee shop not only to
drink coffee but also to want a comfortable space, relax, or experience other services to regain
balance in life.
Today, have many chain coffee stores opening across the country famous as Starbucks, The
Coffee House, Highlands Coffee, etc. Unlike other coffee business companies, Highlands coffee
chooses its own direction, because the customers the company targets are people with high and
moderate-income, and at the same time, the business goal of the company. Highlands coffee's
business goal is to lead the way to lead the high-end retail industry in Vietnam. Highlands Coffee
starting with packaged coffee products in Hanoi in 2000. Highlands Coffee has rapidly
developed and expanded into a famous coffee shop brand and has continuously expanded its
operations at home and abroad since 2002.
Highlands Coffee has chosen for itself the form of chain business to apply. The reality shows that
this business form is very suitable to the reality and needs of customers. Highlands Coffee
succeeds in building a large Highlands Coffee brand while providing customers with the best
products with attentive service and reasonable prices in a modern airy space. Highlands Coffee is

currently the most successful brand of the chain business model - a franchise-oriented business
model. After a long journey, Highlands Coffee has constantly improved and developed to bring
delicious, comparable coffee products in a comfortable and polite space. Highlands Coffee's


coffee cups are not only a familiar drink but also represent a part of the modern culture of
Vietnamese people. Highland has not stopped growing to expand its stores across the whole
country of Vietnam. In 2011, Highlands Coffee had only 50 stores, by 2015 there were 75 stores.
By the end of March 2017, there were a total of 180 stores, spread across 14 provinces and cities
in Vietnam. As of September 2019, the Highlands Coffee franchise system has 296 stores in 25
provinces and cities nationwide. Up to now, Highlands Coffee has had more than 300 stores and
has become a familiar brand for every person.
In Hanoi, Highlands is a place where people often enjoy the space, affirm their own style, or
simply meet and chat with friends and colleagues and enjoy together what Highlands Coffee
brings again. Now, Highlands Coffee has about 96 stores across the districts of Hanoi.
In fact, the needs of customers change constantly, and therefore, in order to best meet those
needs, Highlands Coffee needs to regularly conduct marketing research activities aimed at
exploring market needs as well such as surveying the level of customer satisfaction about the
service quality of the chain so that there can be changes. Therefore, the study of satisfaction is
very necessary.
1.2 Practical problems
Currently, Highland Coffee's market share is decreasing significantly, mainly due to the decrease
in customer satisfaction about the quality of beverage service here, especially after the incident:
''There was a mouse in the cake of Highland Coffee '' in 2008 1 and "Reprinting the expiry date of
canned coffee" in 20112.
In addition, in February 2020, customers also boycotted Highland Coffee because its employees
discriminated against people with disabilities and lottery ticket sellers. 3 To strengthen its position
and increase market share, it is necessary to redefine the level of customer satisfaction with
beverage service at Highland Coffee.
1.3 Research objectives

The main purpose of this study is to determine the factors affecting customer satisfaction about
the quality of beverage service at Highlands Coffee Hanoi. After that, the study will give
1 “Nghi vấn về bánh của Highland coffee có chuột chết - VnExpress”.
2 “Cà phê Highlands 'bắn lại hạn sử dụng” 31 nghìn sản phẩm”.
3 VietnamBiz, “Highlands Coffee bị tố đuổi khách vì tưởng là người bán vé số”.


solutions and recommendations to improve the service quality of Highlands Coffee. This study
was achieved by carrying out the following objectives:




Identify customer satisfaction and factors affect.
Clearly determine the degree of influence of those factors on customer satisfaction
Propose some suggestions to improve service quality in order to respond well

1.4 Research questions
The questions used to find out the research objectives are as follows:
 What factors affect customer satisfaction with Highland Coffee?
 What are some useful recommendations for Highlands Coffee to improve its service quality at

Highlands Coffee?
1.5 Research Scope
The purpose of the survey is to find out the level of customer satisfaction with Highlands
Coffee's service quality. The scope of the study was carried out for customers using services of
Highlands Coffee in Hanoi city.
Target sample characteristics:
o
o

o
o
o
o

Survey type: Online survey
Age: 18 – 40
Gender: Male/Female and others
Expected number of respondents: over
Research scope: Hanoi
Occupation: Student, Office worker, Entrepreneur

1.6 Methodology and data overview
In this research, Primary research includes quantitative methods used to collect through
surveying then analyzed by SPSS. The number of responses received is expected to be 385 and
due to the Covid-19 situation, we will be sending the survey forms online. We will share the
survey forms on food groups on facebook forums with a large audience of students, office
workers in Hanoi.
Secondary research was conducted through research articles, newspapers, websites, and internal
data.


1.7 Outline of thesis
Chap 1: Introduction
Chap 1 provides all basic information about the topic background with all the required
information for the research, including research objective, research question, scope, and
methodology.
Chap 2: Literature review and theoretical models
Chap 2 presents relevant theories that underlie the development of research questions.
Chapter 3: Methodology

Chapter 3 explains the method for doing the research: quantitative analysis, data collection
method and data analyzing method. Then it was explained why research had to do it.
Chapter 4 analyzes data and findings
From the theories of chapter 3. Thereby, finding the factors that improve the quality of beverage
service of Highlands Hanoi.
Chapter 5: Recommendations and Conclusion
Based on the analysis of chapter 4, we will make recommendations on factors to improve the
quality of beverage service of Highlands Ha Noi.
CHAPTER 2: LITERATURE REVIEW AND THEORETICAL MODELS
I. CONCEPT:
1. Concepts of services
According to Zeithaml & Britner (2000), service is the behavior, process, and way of doing a
certain job in order to create use value for the customer to satisfy the needs and expectations of
the customer. Besides, according to Kotler & Armstrong (2004), services are activities or benefits
that businesses can offer to customers in order to establish, strengthen and expand long-term
relationships and cooperation with customers.
Based on Gronroos, he show that “service quality is the result of an evaluation process where the
customer is based on a comparison of the actual service received by the customer with his or her


expectations.” Agreeing with Gronroos, Parasuraman et al. (1985) stated that “service quality is a
form of attitude, resulting from a comparison between the actual service that customers perceive
when consuming. service to their expectations. Service quality is related but not synonymous
with customer satisfaction.” Cronin and Taylor (1992) proposed that service quality is one of the
antecedents of consumer satisfaction, but service quality had less of an effect on purchase
intentions than did consumer satisfaction.
Peter Fdrucker (1954), the father of management, defined "Customers of an enterprise are a
collection of individuals, groups of people, businesses, etc., who have a need to use the product,
services of the business and the desire to satisfy that need”. In business, customers play a vital
role in the survival of the business.

2. Customer satisfaction
According to Bachelet (1995), "customer satisfaction is an emotional response by customers to
their experience with a product or service". Oliva, Oliver and Bearden (1995) said that the
difference, or the relationship of the value of products and services that customers receive
compared to previous products and services will show customer satisfaction.
Similarly, Oliver (1997) states that customer satisfaction is the response of consumers to the
fulfillment of their desires. Sharing this view, Zeithaml and Bitner (2000) argue that, "customer
satisfaction is the customer's evaluation through a product or service that meets their wants and
requirements". This concept has materialized in terms of “customer satisfaction” which is an
assessment measured against a product or service.
According to Kotler (2001), "customer satisfaction is the level of a person's sensory state
resulting from a comparison of the results obtained from the consumption of a product/service
and the customer's expectations." Expectations here are considered human wants or expectations,
derived from personal needs, previous experiences and external information such as advertising,
word of mouth information from family, friends and from seller and competitor information.
Similar to the above studies, Cronin and Taylor 1992 also relied on these studies to measure the
"customer satisfaction" factor. In the study "Measuring Service Quality: A Reexamination and
Extension" Cronin and Taylor used 3 questions to measure the variable "customer satisfaction"


and used the Likert scale with 7 levels from very dissatisfied to very satisfied. Therefore, we will
base on the above studies to develop our research direction.
II. Relevant theoretical frameworks:
1.

Liu et al. (2008)

This is an empirical study of online shopping customer satisfaction in China. The authors test the
hypotheses through multiple regression analysis based on a survey of 1,001 online customers.
The analysis suggests that eight constructs – information quality, web site design, merchandise

attributes, transaction capability, security/privacy, payment, delivery, and customer service – are
strongly predictive of online shopping customer satisfaction, while the effect of response time is
not significant.
2.

SERVQUAL Model of Parasuraman (1988):

The SERVQUAL model is built based on the view that perceived service quality is a comparison
between the expected/expected values (expectations) and the perceived customer values
(perception) studied through the scale. The measurement includes five criterias:






Reliability
Responsiveness
Tangibles
Assurance
Empathy

Patient satisfaction assessment study at Sunyani Regional Hospital, Ghana is a study using
SERVQUAL model by Parasuraman et al., (1988). A total of 214 patients were used in the study.
The

factors

of

service

quality

that

affect

patient

satisfaction

were

Reliability,

Communication/Personal Relationship, Assurance and Responsiveness. In contrast, the Tangible
and Empathic aspects scored positively, confirming the patient's impression of the service.
3.

SERVPERF Model of Cronin and Taylor (1992):

On the basis of Parasuraman's SERVQUAL model, Cronin and Taylor (1992) have overcome and
produced the SERVPERF model, a variant of SERVQUAL. According to SERVPERF model:
Service quality = Perceived level. The factors of service quality that affect customer satisfaction
are: Reliability, Responsiveness, Assurance, Empathy, Tangibles.
Research data were gathered from personal interviews conducted in a medium-sized city in the


southeastern United States. A total of 660 usable questionnaires (all questions answered) were

gathered randomly from consumers at their residences by trained interviewers during a two-week
period in the summer of 1988. The sampling frame was the entire population of the city.
Interviewers were instructed to solicit responses randomly and were assigned city areas to
prevent overlap. Responses were gathered on the service quality offered by two firms in each of
four industries: banking, pest control, dry cleaning, and fast food. The firms and industries were
chosen on the basis of the results of a convenience survey suggesting that these were the four
service industries most familiar to the area's consumers.
 Based on the above studies on customer satisfaction, we find that customer satisfaction

Highlands Coffee Hanoi will be affected by the following 6 factors of service quality:
 Reliability
 Empathy
 Tangibles
 Service efficiency
 Assurance
III. Hypothesis and research model
1.

Hypothesis:

Hypotheses on the influence of perceived service quality factors are posed according to
Parasuraman's research model, including 5 components as follows:
 H1: Reliability has a significant direct influence on customer satisfaction, when the

reliability of food safety is assessed by customers, the customer satisfaction level will
increase and vice versa.
 H2: Customer empathy as a common interest has a significant direct influence on

customer satisfaction, when customer empathy about Highlands increases, customer
satisfaction will increase and vice versa. again.

 H3: Tangibles have a significant direct influence on satisfaction of customers such as

beverage products, facilities, and equipment of the restaurant. When the tangible assets of
Highlands are appreciated by customers, the level of customer satisfaction will increase
and vice versa.
 H4: Service efficiency has a significant direct influence on customers, when the service
efficiency of Highlands staff is appreciated by customers, customer satisfaction will
increase and vice versa.


 H5: Assurance has a significant direct influence on customer satisfaction, when the

assurance of Highlands is appreciated by customers increases, the level of customer
satisfaction will increase and vice versa.
2. Research model
On the basis of understanding and analyzing research models on customer satisfaction, which
have been widely applied in many fields, at the same time, stemming from the research
objectives of the topic, and the characteristics of products and services of the company
Highlands Coffee, the research model proposed to be used in the study of customer
satisfaction for the quality of products and services of Highlands in Hanoi is proposed.


CHAPTER 3: METHODOLOGY
A. Data collection method
I. Sample size
Research results are directly affected by the sample size selection stage. Because larger sample
muscles will yield better results reduce the likelihood of errors and appreciate the accuracy of the
research results.
The sample size is calculated based on a 5: 1 ratio that is very important for EFA research (Hair
et al., 2012). Therefore, this study's minimum size should be at least five times depending on the

quantitative of the question. In this case, 24 questions mean the minimum sample size must be at
least: n = 5 * 24 = 120.
Besides, Tabachnick and Fidell (1996) have created sample size assessments for six levels:
•
•
•
•
•
•

Very poor: 50 surveys
Poor: 100 surveys
Fair: 200 surveys
Good: 300 surveys
Very good: 500 surveys
Excellent: 1000 surveys

In order to conduct a study on the factors of service quality of Highlands Coffee in Hanoi city
affecting customer satisfaction., the target audience is those who may use service quality of
Highlands Coffee in Hanoi city.
II. Primary data
1. Sampling methods
Sampling methods are procedures conducted to select the population's elements to apply
necessary tests and compliance for evaluation. According to Hair (2012), there are two basic
sampling designs which are: probability and nonprobability. In our research, we use nonprobability sampling methods specifically convenience sampling and snowball sampling.
Data are collected via an online survey. Firstly, we will use our personal facebook to send
surveys to our friends, relatives and colleagues; then have them send it to their friends and
acquaintances. Second, we sent the survey form to the group "Fu-hoa lac" of FPT University, the
"hoi review do an co tam", “an gi bay gio nhi” and sent it by commenting the survey in the

comment section on the youtube channel of famous food vloggers such as Ninhtito, khoai lang
thang, etc. The survey is administered in two weeks (from June 28 to July 10, 2021).
2. Questionnaire design
According to Joseph F. Hair (2013), there are seven steps to build the questionnaire:
•

Step 1: Confirm research objectives

•

Step 2: Select appropriate data collection method

•

Step 3: Develop question and scaling

•

Step 4: Determine layout and evaluate the questionnaire

•

Step 5: Obtain initial client approval

•

Step 6: Pretest, revise, and finalize the questionnaire

•

Step 7: Implement the survey

All constructs were measured using a multiple items, five-point Likert scales, ranging from
strongly disagree to strongly agree or from very dissatisfied to very satisfied. There are also
scales used to measure control variables: nominal scale, ratio scale. The measurement items of
Customer satisfaction were adapted from Cronin and Taylor (1992) and Liu et al. (2008).
Perceived service quality was evaluated from five dimensions: reliability, empathy, tangibility,
efficiency, and assurance. The items were taken from Cronin and Taylor (1992) and adapted to
the current context in Vietnam.
The survey consists of 5 parts. In the first part, "Introduction", we explain why we are doing this
survey. The second is a "Filter question" to filter survey participants to see if they are a good fit
for our study. In the third part, are "other questions" questions related to demographics: gender,
age, occupation, etc. This section will help us to know our main customer group and whether we
have found the right target group or not? Next, the "Main question" - this is the most important
part, is the series of questions that will give the results to conduct the analysis and draw the final
conclusion in our research. Finally, the "conclusion" section, in which we will end the survey and
thank the survey respondents for taking the time and patience to complete it.
3. Sample characteristic
The survey collected 295 valid responses. We conducted this survey to determine the factors
related to service quality that affect customer satisfaction when using services at Highlands


Coffee. The scope of the research was carried out in Hanoi city and focused on potential
customer groups who often use services at Highlands Coffee to provide the most objective and
accurate data information. The demographic respondents’ characteristics will be presented in the
next chapter.
B. Data analysis methods
In this research, SPSS was chosen for data analysis. SPSS (Statistics Package for Social
Sciences), designed to measure and analyze data after data collection. SPSS was chosen because

it was the most common analytical data (Mujis, 2010). Through data processing by SPSS, this
study conducted the descriptive analysis, test for measurement models and test for hypotheses.
The demographic respondents’ characteristics and checking for missing values are presented in
descriptive analysis.
To test measurement models, the exploratory factor analysis (EFA) and Cronbach’s Alpha are
used to examine scale’s quality. Factor analysis is a statistical technique for identifying which
underlying factors are measured by a (much larger) number of observed variables. EFA seeks to
uncover the underlying factors of a relatively large set of variables. SPSS tries to find groups of
variables that are highly intercorrelated. In this research, we base on recommendation proposed
by Hair et al.2010 to test measurement models.
 Sampling adequacy :
A Kaiser-Meyer-Olkin (KMO) test is used in research to determine the sampling adequacy of
data that are to be used for Factor Analysis. The KMO test allows us to ensure that the data we
are suitable to run a Factor Analysis and therefore determine whether or not we have set out what
we intend to measure. As a rule, the KMO index must be greater than 0.7.
Bartlett's test of sphericity tests the hypothesis that your correlation matrix is an identity matrix,
which would indicate that your variables are unrelated and therefore unsuitable for structure
detection. Small values (less than 0.05) of the significance level indicates that a factor analysis
may be useful with your data.


In a good factor model, most of the off-diagonal elements will be small. The measure of
sampling adequacy for a variable is displayed on the diagonal of the anti-image correlation
matrix. A variable indicator can be worth to be analyzed if the value of MSA (Measure of
Sampling Adequacy) is more than 0.5.
 Communalities
Communalities represent proportions of variance accounted for by selected components.
Variables having low communalities don't contribute much to measuring the underlying factors.
 Total variance explained
Each component has a quality score called an Eigenvalue. Only components with high

eigenvalues are likely to represent a real underlying factor. We select components whose
Eigenvalue is at least 1. Besides, the percentage of variance accounting for by the first n
components must be higher than 50%.
 Factor loading
The rotated component matrix answers to the question: “Which variables measure which
factors?”. A factor (or component) represents whatever its variables have in common. According
to Hair et al. (2010), factor loading must be higher than 0.5 and one variable is strongly
correlated to one component.
 Cronbach’s Alpha

We rely on Cronbach's alpha to examine internal consistency. Cronbach’s alpha is commonly
used for survey using Likert questions. It varies between 0 and 1. A scale is reliable if
Cronbach’s Alpha ranges from 0.7.
 Correlation

Correlation coefficients provide a numerical summary of the direction and strength of the linear
relationship between two variables. The sign of the correlation coefficient indicates the direction
of the correlation: a positive correlation indicates that as one variable increases, so does the


other; a negative correlation indicates that as one variable increases, the other decreases. The
higher the correlation coefficient–the stronger the level of association. Correlation coefficient can
be either positive or negative. The correlation coefficient should always be in the range of -1 to
1.
To test hypotheses, multiple linear regression analysis is conducted.
 Model summary and overall fit statistics

R-Squared: In the linear regression model, the coefficient of determination, R2, summarizes the
proportion of variance in the dependent variable associated with the predictor (independent)
variables, with larger R2 values indicating that more of the variation is explained by the model,

to a maximum of 1. For regression models with a categorical dependent variable, it is not
possible to compute a single R2 statistic that has all of the characteristics of R2 in the linear
regression model, so these approximations are computed instead. The following methods are
used to estimate the coefficient of determination.

•

Cox and Snell's R24 is based on the log likelihood for the model compared to the log
likelihood for a baseline model. However, with categorical outcomes, it has a theoretical

•

maximum value of less than 1, even for a "perfect" model.
Nagelkerke's R25 is an adjusted version of the Cox & Snell R-square that adjusts the scale

•

of the statistic to cover the full range from 0 to 1.
McFadden's R26 is another version, based on the log-likelihood kernels for the interceptonly model and the full estimated model.

The Durbin-Watson d is a test statistic used to detect the presence of autocorrelation at lag 1 in
the residuals (prediction errors) from a regression analysis. According to Hair et al. (2010) 1.5 <
d < 2.5
 ANOVA table
4 “Cox, D.R. and Snell, E.J. (1989) Analysis of Binary Data. 2nd Edition, Chapman and Hall/CRC,
London. - References - Scientific Research Publishing”.
5 Nagelkerke, “A Note on a General Definition of the Coefficient of Determination”.
6 McFadden, “CONDITIONAL LOGIT ANALYSIS OF QUALITATIVE CHOICE BEHAVIOR”.

Analysis of Variance (ANOVA) is a statistical analysis to test the degree of differences between
two or more groups of an experiment. The results of the ANOVA test are displayed in a tabular
form known as an ANOVA table. The ANOVA table displays the statistics that used to test
hypotheses about the population means. The ANOVA table can be either one way or two way
ANOVA table.
The various column headings that are included in the ANOVA table are as follows:
1. “Source” – It means the source which is responsible for the variation in the data.
2. “DF” – degree of freedom of the data.
3. “SS”- the sum of the squares of the data.
4. “MS”- mean sum of the squares of the data.
5. “F” – F-statistic.
6. “P” – P-value.
According to Hair et al. (2010) if the F-test is highly significant (Sig. < 0.05), we assume that the
model explains a significant amount of the variance in Dependent variable
 Multiple linear regression
Multiple linear regression (MLR), also known simply as multiple regression, is a statistical
technique that uses several explanatory variables to predict the outcome of a response variable.
The goal of multiple linear regression (MLR) is to model the linear relationship between the
explanatory (independent) variables and response (dependent) variable.
•

P-value of beta-coefficient: The “Sig.” column in the coefficients table contains the (2tailed) p-value for each beta-coefficient. According to Hair et al. (2010) a b-coefficient is
statistically significant if its p-value < 0.05 (Sig.).

•

The unstandardized beta-coefficients: each b-coefficient indicates the average increase in
dependent variable associated with a 1-unit increase in a predictor.

•

The standardized beta-coefficients (standardized regression coefficients) tell us the
relative strengths of our predictors (independent variables). We could then compare the
relative strengths of our predictors. Standardized beta coefficients are obtained by
standardizing all regression variables into z-scores before computing b-coefficients.


Standardizing variables applies a similar standard (or scale) to them: According to Hair et
al. (2010) the resulting z scores always have mean of 0 and a standard deviation of 1.
 Multicollinearity in the multiple linear regression model
Multicollinearity occurs when the multiple linear regression analysis includes several
variables that are significantly correlated not only with the dependent variable but also to
each other. According to Hair et al. (2010) Tolerance > 0.1 or VIF < 10
CHAPTER 4: ANALYSES AND FINDINGS
I. Descriptive analysis
1.Data Cleaning
We have 295 surveys that are numbered from 1 to 295. The survey included 29 questions that
were coded when included in SPSS. There are 9 questions that have been reverse coded. After
entering the data into SPSS, we have to check if there was no human error in data entry and
codification.
2. Missing Value
Our survey is sent online and all questions must be answered prior to submission so there is no
problem of missing data in our team's dataset.
3. Demographic analysis
Our group's dataset includes 3 demographic variables: Gender, Age, Personal Income.

Giới tính của anh/chị là gì?
Valid
Valid Nam
Nữ

Total
Table 3.1

Frequency Percent
143
48.6
151
51.4
294
100.0

Percent
48.6
51.4
100.0

Cumulative
Percent
48.6
100.0


Table show that male respondents accounted for 48.6%, female respondents accounted for 51.4%
. In Hanoi city, Highland coffee's customers do not have a significant difference between
male and female
Anh/chị bao nhiêu tuổi?
Valid
Frequency Percent
Valid Dưới 20 tuổi
53

18.0
Trên 30 tuổi
38
12.9
Từ 20-30
203
69.0
tuổi
Total
Table 3.2

294

100.0

Percent
18.0
12.9
69.0

Cumulative
Percent
18.0
31.0
100.0

100.0

The table 3.2 shows that customers under the age of 20 account for 18% , customers over 30
years old accounted for 12.9% , customers aged 20 to 30 accounted for 69% .Concluded that

customers in the age range from 20-30 years old is the main customers at Highlands Coffee in
Hanoi
Table 3.3
Thu nhập hàng tháng hiện tại của anh/chị là bao nhiêu?
Valid
Cumulative
Frequency Percent
Valid Dưới 5 triệu
102
34.7
Trên 10
51
17.3
triệu
Từ 5-10
triệu
Total
Table 3.3

Percent
34.7
17.3

Percent
34.7
52.0
100.0

141

48.0

48.0

294

100.0

100.0

The table 3.3 shows that Highlands Coffee's main customers have an income of 5-10
million/month, accounting for 48%. Customers with income below 5 million/month accounted
for 34.7%. Customers with income over 10 million / 1 month account for 17.3 %
Our group's dataset have 1 respondent’s habits variables
Anh/chị thường tới trải nghiệm dịch vụ tại Highlands Coffee bao nhiêu
lần một tuần?


Valid
Valid Dưới 2 lần trên

Frequency Percent
143
48.6

Cumulative

Percent
48.6

Percent
48.6

tuần
Trên 4 lần trên

34

11.6

11.6

60.2

tuần
Từ 2-4 lần trên

117

39.8

39.8

100.0

tuần
Total

294

100.0

100.0

According to the pie chart, we see that the frequency of customers coming to Highlands Coffee
2-4 times a week accounted for 48.6%, the frequency of customers coming to Highlands Coffee
less than 2 times accounted for 39.8%. The frequency of customers coming to Highlands Coffee
more than 4 times a week is at least 11.5%


II. Testing measurement models

1.Exploratory factor analysis (EFA)
•

EFA for Independent variable(s):

Our research model includes 5 independent variables.
a) Sampling adequacy

Table 1.1
The table 1.1 shows that KMO is 0.909 and Bartlett’s test has Sig= 0.00. < 0.05. According to
Hair et al. (2010), the results are acceptable (or satisfied) and confirm the sampling adequacy.


b) Anti-image

Table 1.2
According to Hair et al. (2010), all indicators (items) are worth to be analyzed because MSA is
more than 0.5.

c) Communalities


Table 1.3
The table show that Variables having communalities > 0.4. All variables have high
communalities. It means that they contribute much to measuring the underlying factors.
d) Total variance explained


Table 1.4
The table shows that the first four principal components with eigenvalues greater than 1 form the
extracted solution. These components explain more 72% of the variability in the original 21
variables. There are a 28% loss of information
e) Factor loading


Table 1.5
All factor loadings are higher than 0.7.
The rotated component matrix shows that:
▪ The first component (1) is measured by: RELY_1, RELY_4,RELY_3,ASSUR_1,EFFI_5 => we
interpret component (1) as “Reliability”
▪ The second component (2) is measured by: TANG_1, TANG_2, TANG_3, TANG_4, EFFI_4
=> we interpret component (2) as “Tangible”
▪ The third component (3) is measured by: EMPA_1, EMPA_2, EMPA_3, EMPA_4=> we
interpret component (3) as “Empathy”
▪ The fourth component (4) is measured by: EFFI_2, EFFI_3 => we interpret component (4) as
“Service efficiency”
•

EFA for dependent variable(s)

Our research model has a dependent variable

Table 1.6
The KMO and Bartlett’s Test table shows that KMO is 0.770 and Bartlett’s test has Sig= 0.00. <
0.05. According to Hair et al. (2010), the results are acceptable (or satisfied) and confirm the
sampling adequacy.
The Anti-image show that all indicators (items) are worth to be analyzed because MSA is more
than 0.5.
The Communalities table show that Variables having communalities > 0.4. All variables have
high communalities.
f) Total variance explained


Table 1.7
The table 1.7 shows that the first principal components with eigenvalues greater than 1 form the
extracted solution. This components explain nearly 90% of the variability in the original 3
variables. There are a 10% loss of information
g) Factor loading

Table 1.8
All factor loadings are higher than 0.7.
The rotated component matrix shows that: The component (1) is measured by: CS_3, CS_2,
CS_1 => we interpret component (1) as “Customer Satisfaction”