(TIỂU LUẬN) TOPIC FACTORS INFLUENCING THE FREQUENCY OF USING ONLINE FOOD DELIVERY OF UNIVERSITY STUDENTS IN HO CHI MINH CITY

FOREIGN TRADE UNIVERSITY
IN HO CHI MINH CITY
----------★★★----------

GROUP REPORT
SUBJECT: ECONOMETRICS
INSTRUTOR: MRS. LÊ HẰNG MỸ HẠNH
TOPIC
FACTORS INFLUENCING THE FREQUENCY OF USING ONLINE FOOD
DELIVERY OF UNIVERSITY STUDENTS IN HO CHI MINH CITY

Group: Community

Class: K58CLC5

Ho Chi Minh City – 06/2021

Class code: 203


Table of contents

TABLE OF CONTENTS
TABLE OF CONTENTS

i

LIST OF FIGURES

ii

LIST OF TABLES

ii

GROUP MEMBER LIST

ii

ABSTRACT

1

CHAPTER 1: INTRODUCTION

2

1. WHAT DOES “TOO-BIG-TO-FAIL” MEAN?

2

2. WHEN DO WE CALL A BANK “TOO-BIG-TO-FAIL”?

3

CHAPTER 2: ANALYSIS OF “TOO-BIG-TO-FAIL” PROBLEMS

5

1. HOW THESE BANKS AFFECT ON THE WHOLE ECONOMY

5

2. PROS AND CONS OF “TOO-BIG-TO-FAIL” BANKS

6

CHAPTER 3: BREAKING UP BIG BANKS

8

1. PROS AND CONS OF BREAKING UP BIG BANKS

8

2. OTHER SOLUTIONS TO “TOO-BIG-TO-FAIL” PROBLEM

9

CHAPTER 4: CONCLUSION

11

REFERENCE

12

1


Danh sách thành viên – Danh mục hình ảnh

LIST OF FIGURES
Figure 1. “Too-big-to-fail” problems

1

Figure 2. Too-big-to-fail definition

2

Figure 3. The first too-big-to-tail problem arose with Continential Illinois

3

Figure 4. Dodd-Frank Wall Street Reform and Consumer Protection Act

3

Figure 5. The bankruptcy of Lehman Brothers

5

Figure 6. Less fear of the big impacts of the failure of a group of banks

8

Figure 7. Conclusion

11

LIST OF TABLES
Table 1. 5 out of 100 biggest banks ranked both by reported total assets and total assets when
derivatives are on a gross, not net (U.S. GAAP) in some countries in 2012

4

GROUP MEMBER LIST
STT

MSSV

HỌ VÀ TÊN

1

1911155028

Nguyễn Hồng Gia Hân

2

1911155063

Lê Yến Nhi

3

1911155079

Chung Ngọc Nhã Thi

4

1911155080

Nguyễn Đan Thi

5

1911155087

Nguyễn Quý Thương

GHI CHÚ

2


Chapter 1: Introduction

CHAPTER 1:

INTRODUCTION

Living in an era where technology is advancing at an unprecedented rate, consumers are
becoming smarter and more demanding than ever before. With the existence of many new
products and services, consumers now have plenty of options in terms of price, quality and
especially, convenience. A busy lifestyle is a signature in this modern time, people barely
have time for themselves when most of their time is spent on work. Briefly speaking, they
want to save as much time as possible. That is the main reason why 75% of the total

Vietnamese population have used Food Delivery Service because they find cooking meals, or
even eating out, very time-consuming. What is remarkable is that 25% of them are new users
who have started to use food delivery services for the first time due to Covid-19, which means
safety from the pandemic is the second most significant reason explaining the rise of Food
Delivery Service. Among some ordering methods, Online Food Delivery Service Application
is the most popular one when greater than 90% use this to order food and drinks.
The Vietnam’s Online Food Delivery Service market is driven by increasing internet and
smartphone penetration, along with an increase in the number of restaurants and other food
chains. Aided by the technical advancements in the Vietnam’s Online Food Delivery Service
market, it is expected to witness a healthy growth in the forecast period of 2021-2026,
growing at a CAGR of 24%. At present, this market is facing fierce competition with the
existence of major brands such as Grabfood, Now, Baemin, Gofood, and Loship. However,
recently, the entry of domestic enterprises such as VinID and Tiki is considered to re-divide
the market share of this potential "piece of cake". Although the Online Food Delivery Service
Application market is booming with the participation of many players sharing the pie, having
an available strong customer base will retain the firm’s position and make it harder for other
brands to enter the market. Therefore, besides attempts to raise brand awareness and attract
new customers, brands must invest in advancing the company and the application itself to
keep its current customers and turn them into loyal ones.
The objective of this research is to study the factors that influence the frequency of ordering
food and drinks from Online Food Delivery Service Applications. To be more practical, the
research helps Online Food Delivery Service firms to have a better insight of customers by
providing information about which factors may limit the number of times a customer
demands for Online Food Delivery Service Applications. Thus, firms can create an effective
strategy not only to meet customers’ needs but even exceed their expectations.

1


Chapter 2:Literature review

CHAPTER 2:
1.

LITERATURE REVIEW

OVERVIEW OF FOREIGN RESEARCH

Today there are more than 10,000 studies, essays, large and small articles related to consumer
behavior, purchasing behavior, and frequency of purchase in the world. Among them must be
the quality studies of Arian Oosthoek (2013); See Siew Sin, Khalil Md Nor, and Ameen M
Al-Agaga (2012); Yogi Tri Prasetyo and his partners (2021); Haiyang Liu (2019)... Here, the
authors would like to present a summary of the results of the above works as a premise for the
next steps of our research.
In the research “What is the impact of Facebook tie strength and behavior on purchase
intention?” by Arian Oosthoek, the author used systematic empirical investigation and survey
to collect data, then applying multiple regression model and correlation coefficient to analyze
the data. The results of the study reach four conclusions as follows.
● Hypothesis 1: Strong ties on Facebook affect online purchase intention more than weak
ties on Facebook.
● Hypothesis 2: Social ties on Social Media affect the purchase intention more for alow
involvement luxury good.
● Hypothesis 3: The higher the Facebook activity, the higher the purchase intention and
attitude towards the product with strong ties.
● Hypothesis 4: People with higher income have a more positive attitude and intention
toward the purchase of high involvement products.
In the study “Factors Affecting Customer Satisfaction and Loyalty in Online Food Delivery
Service during the COVID-19 Pandemic: Its Relation with Open Innovation” of Yogi Tri
Prasetyo and partners, the authors applied methods of the questionnaire, model fit, and
structural equation modeling (SEM Model). Initially, there were 11 hypotheses stated on the

causal relationships between variables but after the first time running the SEM Model it
pointed out that several hypotheses were not important. The final SEM model contained 4
exogenous variables, the two hedonic motivations and information quality affecting
“INTENTION TO USE” and the two prices and promotion affecting “ACTUAL USE”.
Haiyang Liu with his research “ Factors positively influencing customer satisfaction of online
food delivery services of customers in Bangkok and its vicinity” is based on the instrument of
Cronbach’s Alpha Coefficient, Multiple Regression Analysis, and Pearson’s Correlation

2


Chapter 2:Literature review

Coefficient to study customer satisfaction. There were two accepted hypotheses, the most
predictive independent variables were 38 hedonic motivations (β = 0.767), new experience (β
=0.163). Hedonic motivations and new experience could positively influence customer
satisfaction of OFD services of customers in Bangkok and its vicinity at 52%, the rest 48%
were affected by any other variables, these variables were not used in this paper. The result of
VIF value was not more than 4 and the tolerance value exceeded 0.2, indicating that there was
not Multicollinearity in the independent variables. Thus, the standard error equaled ±0.492,
the calculation as:
𝑌

=

1. 870

(customer satisfaction)

+

0. 767𝑋1

(hedonic motivations)

+

0. 163𝑋2
(new experience)

This study can offer benefits for restaurants and other food industries in respect of 44
potential new markets of online purchase, online strategy improvement, or investment
decision for online food business in the future regarding factors predicting the customer
satisfaction towards online food delivery services.
In general, foreign research above quite specifically analyzes the factors affecting customers’
online purchase decisions and offers specific solutions for their countries. Here, the authors
would like to use the above quality studies for reference purposes and as a base to build a
model for research in Vietnam, specifically Ho Chi Minh City.

2.

OVERVIEW OF DOMESTIC RESEARCH

In Vietnam, there have been many typical studies such as “Research on online food ordering
behavior of customers aged 18 - 30 in Ho Chi Minh City”, “Research on factors affecting
online shopping decisions of consumers of university students in Ho Chi Minh City” or
“Factors affecting the behavior of using public transport that has technology applications – A
case study with Grab Bike service”.
In the study “Research on factors affecting online shopping decisions of consumers of
university students in Ho Chi Minh City”, the results show that the scales are reliable through

the Cronbach Alpha test. However, EFA analysis explained that the scale of
CHAMSOCKHACHHANG components was excluded. The remaining 6 factors perceived
usefulness, perceived convenience in payment, perceived ease of use, trust, influence from
external factors and price expectations continued to go into the ANOVA analysis. The results

3


Chapter 2:Literature review

show that the factor “Influence from external factors” was not eligible to participate in this
model, so it should be removed. The remaining 5 factors are proportional to the decision on
the purchase price of consumer products of university students in Ho Chi Minh City. It proves
that the proposed theoretical model is consistent with current reality as well as the hypotheses
in the theoretical model are accepted.
Dr. Nguyen Binh Minh and his partners (2017) with the research “Factors affecting the
behavior of using public transport that have technology applications – A case study with Grab
Bike service” applied Cronbach’s alpha and the multiple regression model to analyze the data.
They concluded that the analysis factors all have a positive impact on the usage behavior of
customers. The results provide suggestions for managers and marketers to pay attention to
quality factors, safety, and ease of use. In addition, factors such as "subjective standards' ';
"perceived usefulness" also need attention. As for the "usefulness" factor, especially taking
advantage of technology and saving time to create a difference and advantage in competition.
In general, researchers in Vietnam mainly analyze online buying behavior, but they are
somewhat incomplete. Researches in Vietnam often pay little attention to providing solutions
for enterprises or give them an unsatisfactory, unrealistic, or unfounded way. Therefore, in our
research, we will overcome the mistakes and shortcomings of previous studies to research and
provide effective and appropriate solutions to develop the field of online food ordering in
Vietnam.

4


Chapter 3: Methodology and data

CHAPTER 3:

METHODOLOGY AND DATA

5


Chapter 4: Result

CHAPTER 4:
1.

RESULTS

SCALE RELIABILITY ANALYSIS

Based on the respective internal consistency explained by the value of Cronbach’s Alpha on
the Appendix 1 on Page a, after conducting the scale reliability analysis by Cronbach’s Alpha
on Stata, we receive the result of Cronbach’s Alpha as the tables below indicates:
● With the scale of the variable PROMO, there are 3 smaller encoded variables including
Promo1, Promo2 and Promo3.
Table 1. Cronbach’s Alpha of the Model – Group 1

The overall Cronbach’s Alpha of Group 1 is 0.8381 (0. 8≤ α < 0. 9) which means that the
internal consistency in this variable group is good. Moreover, the item-rest correlation of the

variables Promo1, Promo2 and Promo3 are 0.7083, 0.7489 and 0.6945 respectively (> 0. 4).
⇨ It is concluded that Promo1, Promo2 and Promo3 are well qualified for the PROMO
variable and there would be no variable necessarily excluded from the variable group.
● With the scale of the variable BARRIER, there are 3 smaller encoded variables including
BAR1, BAR2 and BAR3.
Table 2. Cronbach’s Alpha of the Model – Group 2

6


Chapter 4: Result

The overall Cronbach’s Alpha of Group 2 is 0.8438 (0. 8≤ α < 0. 9) which means that the
internal consistency in this variable group is good. Moreover, the item-rest correlation of the
variables BAR1, BAR2 and BAR3 are 0.7250, 0.6860 and 0.7208 respectively (> 0. 4).
⇨ It is concluded that BAR1, BAR2 and BAR3 are well qualified for the BARRIER variable
and there would be no variable necessarily excluded from the variable group.
● With the scale of the variable CS, there are 4 smaller encoded variables including CS1,
CS2, CS3 and CS4.
Table 3. Cronbach’s Alpha of the Model – Group 3

The overall Cronbach’s Alpha of Group 3 is 0.6937 ( 0. 8≤ α < 0. 9) which means that the
internal consistency in this variable group is good. Moreover, the item-rest correlation of the
variables CS1, CS2, CS3 and CS4 are 0.5420, 0.0937, 0.5660 and 0.7306 respectively (
> 0. 4)
⇨ It is concluded that CS1, CS3 and CS4 are well qualified for the CS variable and there
would be variable CS2 necessarily excluded from the variable group.
2.

EXPLORATORY FACTOR ANALYSIS

After conducting the exploratory factor analysis (EFA) on Stata with the minimum value of
eigenvalue to be retained is 1 (which means that we will exclude the factor with the value of
eigenvalue smaller than 1), we receive the result of factor analysis as the figure in the
Appendix 2 on page a indicates:
Result: Exclude the factor with the eigenvalue smaller than 1 and retain the Factor1, Factor2
and Factor3, which can cumulatively altogether explain 74.66 percent of the deviation of the
data.

7


Chapter 4: Result

● Rotation: orthogonal varimax displaying loading as blank when |loading| < 0.5

Result: There will be 3 main
variable

groups

including

PROMO (Promo1, Promo2 and
Promo3),

BARRIER

(BAR1,

BAR2 and BAR 3) and CS (CS1,
CS3 and CS4).

● Kaiser-Meyer-Olkin (KMO) Test
Running on Stata:

Result: KMO = 0.6277 (> 0. 5)
⇨ The sampling adequacy for each variable in the model and for the complete model is
moderate.
3.

CORRELATION ANALYSIS

Between the dependent variable and independent variables, the correlation analysis using
Pearson’s Correlation Coefficient of Payment, Income, Promo, Barrier and CS that affects the
frequency of using online food delivery.

*Correlation is significant at the 0.1 level
From the analysis, it can be clearly seen that 4
independent variables Payment (𝑟 = 0. 2123),
Income (𝑟 = 0. 3452), PROMO ( 𝑟 = 0. 1405)
and BARRIER (𝑟 =− 0. 1826) will have an effect
on the frequency of using online food delivery at
the

significant

level

0.1.

Meanwhile,

the

8


Chapter 4: Result

independent variable CS (𝑟 =− 0. 0139 ) will have no effect due to its insignificance at the
0.1 level.

4.

MULTICOLLINEARITY TEST

In statistics, the definition of Multicollinearity was, among all the independent variables, a

circumstance

of

a

very

positive

relationship

(StatisticSolutions,

2017).

Higher

multicollinearity proved the higher degree of correlation among independent variables which
might cause deviation away from the true value. Equally, multicollinearity should not appear
as it could lead to incorrect interpreting of MRA results.
After having a regression model, we use the variance inflation factor (VIF) to test
multicollinearity. The value for VIF starts at 1 and has no upper limit. A general rule of thumb
for interpreting VIFs is as follows:
● A value of 1 indicates there is no correlation between a given explanatory variable and any
other explanatory variables in the model.
● A value between 1 and 5 indicates moderate correlation between a given explanatory
variable and other explanatory variables in the model, but this is often not severe enough to
require attention.
● A value greater than 5 indicates potentially severe correlation between a given explanatory
variable and other explanatory variables in the model. In this case, the coefficient estimates
and p-values in the regression output are likely unreliable.
From the appendix, we can see that the VIF value of each independent variable value
between 1 and 5 meaning Multicollinearity does not exist in the independent variables.

9


Chapter 4: Result

5.

5.1.

HETEROSKEDASTICITY TEST
Detecting heteroskedasticity

To test heteroskedasticity, our group uses the Park Test.
{𝐻0: ℎ𝑜𝑚𝑜𝑠𝑐𝑒𝑑𝑎𝑠𝑡𝑖𝑐𝑖𝑡𝑦 (𝑁𝑜 ℎ𝑒𝑡𝑒𝑟𝑜𝑠𝑘𝑒𝑑𝑎𝑠𝑡𝑖𝑐𝑖𝑡𝑦) 𝐻1: ℎ𝑒𝑡𝑒𝑟𝑜𝑠𝑘𝑒𝑑𝑎𝑠𝑡𝑖𝑐𝑖𝑡𝑦 𝑒𝑥𝑖𝑡𝑠
The figure illustrates the step-by-step process of performing a Park test in STATA.
In this case, the coefficient for the variable lin (using the natural log of income as specified by
the Park test) is statistically significant with a p-value of 0.000. Therefore, we reject the
hypothesis of homoscedasticity, it means heteroskedasticity in this model

5.2.

Correcting for heteroskedasticity

To correct this problem, we use heteroskedasticity – robust standard error because they are
valid—at least in large samples—whether or not the errors have constant variance, and we do
not need to know which is the case. (process illustrated in the appendix).

10


Chapter 4: Result

With the significant level 0,05 there are 4 statistically significant independent variables:
Payment (sig=0,01), Income(sig=0,000), Promo(sig=0,047), Barrier(sig=0,008). In addition,
CS (sig=0,759) is not a critical predictor of the frequency of using online food delivery.
Therefore, we have a regression model with heteroskedasticity – robust standard error:
𝐹𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 = 0, 374 + 0, 689𝑃𝑎𝑦𝑚𝑒𝑛𝑡 + 0, 761𝐼𝑛𝑐𝑜𝑚𝑒 + 0, 357𝑃𝑟𝑜𝑚𝑜 – 0, 449𝐵𝑎𝑟𝑟𝑖𝑒𝑟

11


Chapter 5: Conclusion

CHAPTER 5:

CONCLUSION

12


Appendix

APPENDIX 1
CRONBACH’S ALPHA AND INTERNAL CONSISTENCY RELATION
According to Cronbach (1951), Cronbach’s Alpha Rule of thumb indicates the fluctuation of
internal consistency ranged from different value of Cronbach’s Alpha as below:
Table 6. Cronbach’s Alpha and internal consistency relation
Cronbach’s Alpha

Internal Consistency

Cronbach’s Alpha

Internal Consistency

α≥0. 9

Excellent

0. 6≤α < 0. 7

Acceptable

0. 8≤α < 0. 9

Good

0. 5≤α < 0. 6

Poor

0. 7≤α < 0. 8

Moderate

α < 0. 5

Unacceptable

APPENDIX

2

FACTOR ANALYSIS

The process of running factor analysis in Stata and the table of result mentioned in Chapter 4.

1


Appendix

2


References

REFERENCE

3