Empirical Project

Analysis of consumer buying behavior towards online shopping in

China

Konni Roselin Yau Liao

IUP2020031
International Business
University of International Business and Economics

Spring, 2023
Beijing, China

1
Introduction

How does online shopping affect Chinese consumers decision?

The emerge of e-commerce has greatly change people´s perspective of shopping and
shopping channels, specially nowadays with the improvement of living standards and
the rapid spread of the internet, online shopping has become a new way of shopping for
the Chinese market which is one of the largest online shopping markets in the world.
This research is analyzing about factors influence consumer decision toward online
shopping.

For this project, I decided to choose this topic about online shopping in China because
personally I love shopping, but the way I shop in my home-country Panama is so
different compare to China. In Panama, I rarely buy things online because is not very
convenient since there is no door to door delivery, we need to go to specific parcels
places to pick up our packages, and also delivery time takes around one or two weeks
so in conclusion is not convenient at all. What I used majority of the time is to shop
offline stores and supermarkets; however, these pasts months that I came to China, I
have realized that the internet has become part and parcel of consumers ‘daily
activities. One can simply buy anything online only with a click of a button and have
products delivered right at their doorstep.

Online shopping involves purchasing products or services over the Internet. All the
products in online stores are described through text, with photos and multimedia files.
Inclusive, many online stores will provide links for extra information about their products
and even make available, safety procedures, instructions, manufacture specifications
and demonstrations. Some will provide advice or how-to guides, and some allow users
to rate and review their products.

The purpose of this research paper is to identify and analyze the factors that influence
consumer decision to shop online. The effective way to do so is identifying the
independent and dependent variables of this research paper.

An independent variable is one that manipulated, changed, or controlled in a research

experiment. Basically, it is what the researcher changes to trigger an effect on the other
variable. In this study, the independent variables are: convenience, amount of
information, price, limitation and security.

On the other side, dependent variable is one that is being tested in a research
experiment. This variable is usually affected by changes made to the independent
variable. In this study, the dependent variable will be the consumers ‘decision.

In order to finding influence of independent variables towards dependent variable,

research method used in this study is multiple regression analysis. Conclusion in this
study is that consumer decision towards online shopping is affected significantly by
convenience, amount of information, price, limitation and security.

2
Table 1. Conceptual Map of the study

Convenience

Amount of
information

Consumers´ decision
Price

Limitation

Security

Data collection

A total of 450 questionnaires were used for analysis. In the Excel format shows the
demographic profile of the participants. Among the participants, 33.3% are from Beijing,
30% of them are from Shenzhen, 26.7% are from Guangzhou, and 10% from Shanghai.
In the variable of gender shows that there are 305 males and 145 females.
According to the marital status, 74.2% are single and 25.8% are married. Additionally,
for age group, 25.6% of the participants are between the age of 10-19, 55.6% are
between 20-29, 11.1% are between 30-39, 7.78% are between 40-49, and none % for
50 above.

The excel also indicates that 21.1% of the participants are in middle school, 29.6% in
high school, 44.7% hold a bachelor degree, 4.22% hold postgraduate and 2% have
PHD.
In the occupation item, 26.7% are students, 4.44% are managers, 54.4% are staff
members, 8.89% are teachers, and 5.56% have other occupations or are unemployed.
Those who have an income of 1000RMB or below belong to the students, staffs have
an income from 1000-3000RMB, and the teachers and managers have an income
above 4000RMB.

3
Additionally, based on the formulation of the questionnaires, the type of study is
explanatory research, which usually describes the causal relationships between
variables through hypothesis testing.

In this study, there are five attributes used which is convenience (X1), amount of
information (X2), price (X3), limitation (X4), and security (X5); therefore, I obtained a
sample of 5x20= 100 which is used as respondents.

Table 2. Variable Description

Dependent Description Measure Source
Variable
Consumer´s Purchase or not Percentage Statista
Decision

Independent Description Measure Source

Variable
Convenience Comfortability Percentage Statista
Amount of Product info, user Percentage Statista
Information rating and reviews
Price Product´s price, Percentage Statista
shipping cost
Limitation Cannot see, smell Percentage Statista
and try the quality
Security Product quality, Percentage Statista
fake information,
product not send
after payment

III. Descriptive Analysis

The scatter plot of the dependent variable (consumer decision) and primary
independent variable (convenience) generated from STATA software is shown in
figure 1.
Figure 1. Scatter plot between consumer decision and convenience

Besides CONVENIENCE, the scatter plot between consumer decision and other
independent variables are shown in figure 2. However, these scatter plots only show the

4
relationship between variables because further examinations are done in the multiple
regression analysis.

Figure 2. Scatter plots between consumer decision and other independent

variables

Table 3 summarizes the descriptive statistics of the six variables including the number
of observations, mean, standard deviation, minimum, maximum, skewness, and
kurtosis. Focusing in the dependent variable, consumer decision, has a mean of
23.13% which is relatively high but not the highest compare to the independent
variables. On the other side, the standard deviation is higher in the convenience
variable because during the questionnaires, majority of the respondents said that the
reason they shop online is due to the convenience with a 26.80%.

Additionally, the statistic shows two extreme values which is the minimum and the
maximum. For the minimum of 2.22% is for the price and the maximum is 87.54% for
the variable of amount of information.

Descriptive Statistics on the six variables

Table 3.

5
IV. Regression Analysis and Hypothesis Testing

Multiple Regression Model

The model analysis of data used in this study is the analysis of multiple regression
analysis. Multiple regression analysis was used to analyze the influence of independent
variables consisting of convenience (X1), amount of information (X2), price (X3),
limitation (X4), and security (X5). Model equations of multiple regression analysis in this
study can be formulated:

Y = β1 Χ1 + β2 Χ2 + β3 Χ3 + β4 Χ4 + β5 Χ5 +е

Which consist of:

Y: Customer decision

X1: Convenience

X2: Amount of information

X3: Price

X4: Limitation

X5: Security

β 1... β 5 = Coefficient regression independent variable

e = Error estimation

Linear or nonlinear regression

The general guideline for choosing between linear and nonlinear regression is to apply
the linear regression first and determine whether it can obtain a correct fit. In this study
it represents linear regression with robust standard errors on the panel data across.
The result shows that the R-squared is 0.2176, while the root mean squared error is
1.3611, which is relatively close to zero.

The interpretation of the regression model indicates the following:

1. Convenience: Indicates that there is a positive effect between variables X1
towards variable Y, which means that if the higher convenience customer more
likely to decide to shop online and vice versa.
2. Amount of information: Regression coefficient indicates that there is a positive
effect between variables X2 Y variables, which means that if the good amount of
information provided to consumers, customer more likely decide to shop online
and vice versa.

6
 3. Price: Regression coefficient indicates that there is a negative effect between
variable X3 toward variable Y, which means the higher the price the customer
more likely decide to not shop online and vice versa.
4. Limitation: Indicates that there is a negative effect between variable X4 toward
variable Y, which means that more limitation found on shopping online, customer
more likely decide to not shop online.
5. Security: Indicates that there is a positive effect between variables X5 toward
variable Y, which means if the higher value of security offered customer more
likely decide to shop online, and if the security is low the customer more likely
decide to not shop online.

Table 4. Regression model

Multicollinearity Test

For detecting if there is any multicollinearity or not, can be seen from the Variance
Inflation Factor (VIF). If the value of VIF>10 showing the indication of multicollinearity
and vice versa, if VIF< 10 non-multicollinearities is indicated as shown in table below.

Table 5. Multicollinearity Test

X1 1,610 Non-multicollinearity
X2 1,954 Non-multicollinearity
X3 1,675 Non-multicollinearity
X4 1,087 Non-multicollinearity
X5 2,241 Non-multicollinearity

7
Based on the table above, each independent variable showing VIF value is less than
10, so the assumption belongs to the non-multicollinearity.

Hausman Test

H0: E (𝛼| x ) = 0 Random Effect

The null hypothesis is that the preferred model is random effects. If Hausman
Test accept H0 or p- value > 0.05, then the method random effect will be chosen.

H1: E (𝛼 | x ) ≠ 0 Fixed Effect

The alternate hypothesis is that the model is fixed effects. If Hausman Test
receives H1 or

P- value < 0.05, then the method we will be choosing is the fixed effect.

Conclusion

Understanding consumer decision towards online shopping can lead online store
marketer to create the best strategy. This study about consumer primary way of
shopping can be conclude as follow:

-Consumer decision towards online shopping is affected significantly by convenience,

amount of information, price, limitation such as cannot see-feel-touch and extra time
needed to get product on hand before purchase product in online shop, and last is that
they were afraid of being deserted when purchasing product online, consumer decision
is highly affected with security.

- Greatest factor that influence consumer decision to shop online is security, second is
price, third is convenience, followed by limitations, and lastly consumer also considering
amount of information.

In conclusion, online store is perfectly used as an alternative for retailer who did not had
full time for managing brick and mortar store or as a support for an existing physical
store.

Sources

United States Census Bureau. 2020. Distribution of users of leading mobile e-

commerce platform in China.

Uka Tarsadia University. 2019. Factors affecting consumer´s online shopping buying
behavior.