Exploratory Analysis of Factors Affecting Levels of Online Shopping

in the COVID-19

ChiHsiang Ting

1,a

, ChinLien Chang

, Chao Shen

3,a

and Xuyuan Zhang

1,a

Faculty of Management Engineering, Huaiyin Institute of Technology, MeiCheng Rd., Huain An, JiangSu, China

Student Affairs Office, Huaiyin Institute of Technology, MeiCheng Rd., Huain An, JiangSu, China

Jiangsu Smart Factory Engineering Research Centre, College of Management and Engineering, Huaiyin Institute of

Technology, Huai’an, 223003, China

Keywords: Online Shopping, E-Commerce, Consumer Behavior, Home Delivery Service, COVID-19.

Abstract:

This study focuses on the factors influencing online shopping consumer behavior during the COVID-19

epidemic. We conducted an online survey of the population of the Huai'an metropolitan area in Jiangsu

Province. Constructing a behavioral model of consumers' online shopping during the COVID-19 epidemic

using the theory of planned behavior. The least squares statistical analysis shows that attitude has a positive

effect on behavior intention. Subjective norms have a positive effect on perceived behavioral control.

Behavioral intentions have a positive effect on behavior. From the survey results, the categories of

commodities purchased by consumers are mostly daily necessities and epidemic prevention products. The

survey found that people's willingness to change their shopping behavior from in-store shopping to online

shopping has increased. Therefore, enterprises need to pay more attention to health and safety issues when

operating online shopping, in order to reduce consumers' doubts and improve the reputation of enterprises.

1 INTRODUCTION

Since the outbreak of the COVID-19 in 2020, various

industries have been affected. Person-to-person

movement is restricted, and the transportation of

goods is also affected. Scholars Unnikrishna and

Figliozzi (Unnikrishna, Figliozzi 2021) surveyed

Portland consumers in the Vancouver-hillsborough

metro area of shopping behavior during the COVID-

19 pandemic. The results of the survey found that the

epidemic affected consumers' shopping behavior.

Consumers’ online shopping and home delivery

behaviors have grown significantly during the

pandemic.

Also, during the pandemic, in terms of ensuring

customer satisfaction. E-commerce providers should

pay attention to issues such as service quality,

customer perceived value and trust, which affect

consumers’ choice of online shopping and home

delivery (Uzir, Halbusi, Thurasamy, Hock, Aljaberi,

Hasan, Hamid, 2021).

COVID-19 has also affected the operations of

food suppliers. Some scholars have investigated the

https://www.hyit.edu.cn/

https://orcid.org/0000-0003-2377-4000

situation of food cold chain supply in Indonesia.

They found features such as logistics monitoring

equipment such as EDI, RFID and blockchain. The

system can track food quality, build a traceability

system, and ensure food quality and safety

(Masudin,

Ramadhani, Restuputri, 2021).

Based on the above research literature, people

have increased their online shopping behaviors

during the Covid-19 pandemic for their lives. The

close integration of the express delivery industry and

online shopping makes shopping more convenient

and safer. Therefore, from the perspective of

consumers, it is more in line with the interests of

enterprises to explore the consumer behavior of the

combination of express delivery and online shopping.

Therefore, this study will use the theory of planned

behavior to explore this point of view.

2 LITERATURE REVIEW

2.1 Theory of Planned Behavior

This study is an exploratory study that attempts to

explore the impact of COVID-19 on online shopping

Ting, C., Chang, C., Shen, C. and Zhang, X.

Exploratory Analysis of Factors Affecting Levels of Online Shopping in the COVID-19.

DOI: 10.5220/0011349900003440

In Proceedings of the International Conference on Big Data Economy and Digital Management (BDEDM 2022), pages 811-817

ISBN: 978-989-758-593-7

811

intention behavior from a consumer perspective.

This willingness also affects the frequency of home

deliveries. Therefore, the planned behavior theory

proposed by Ajzen (Ajzen 1985) was adopted. This

theory was later extended by scholars such as Ajzen,

Driver, and Fishbein (Driver 1991, Fishbein 2000).

Their research found that human behavior is not

entirely their own will, it will be influenced by the

environment, this influence can be observed and

changed. Therefore, he extended the theory of

rational action (TRA), adding a new concept of

"perceived behavioral control" of the self, and

developed a new model of behavioral theory

research: the project theory of planned behavior

(TPB). The theory includes five elements: attitude,

subjective norm, perceived behavioral control,

behavioral intention, and behavior. Below, we try to

combine this research with theory to explore.

2.2 Attitude

Attitude is the initiator of behavior. It refers to the

positive or negative feelings an individual has when

performing a behavior. When an individual's

conceptual evaluation of this particular behavior

forms an idea, the attitude component is often seen

as the presentation of an individual's important

beliefs about a certain behavior. Researcher Tsai

(Tsai 2021) mentioned the following phenomenon in

the findings of the COVID-19 consumer online

purchasing motivation. (1) Consumers are afraid of

contacting people, reduce going out, and increase

their willingness to change their consumption

patterns. (2) There has been an increase in the

behavior of buying daily necessities near home

rather than online. Therefore, the questionnaire

design item of this research is: after the outbreak of

COVID-19, consumers' attention to the following

questions. (1) Afraid to go out and come into contact

with people? (2) Change consumer behavior? (3)

Change your lifestyle?

2.3 Subjective Norm

Subjective norm refers to the social atmosphere or

environmental pressure that an individual feels about

whether to take a certain behavior. The culture or

atmosphere formed by the collective consciousness

and ideas of the society has the effect of restraint or

pressure on the individual. Subjective norms are

predicting the factors that influence individual

behavioral decisions and measuring the degree of

influence of whether an individual takes a particular

behavior. Therefore, the questionnaire design item

of this research is: after the outbreak of COVID-19,

consumers pay attention to the following factors. (1)

Ability to deliver goods on time? (2) The ability to

handle product delivery errors? (3) Standardized

transportation capacity of goods? (4) The ability to

avoid the quality assurance of goods from virus

infection?

2.4 Perceived Behavioral Control

Perceived behavioral control refers to the amount of

personal experience and the expected size of the

barrier. Perceived behavior has more control over

behavior when a person believes they have more

resources and opportunities and fewer expected

barriers. Forms an individual's confidence and

willingness to do something. It affects in two ways.

One is that it has a motivating meaning for

behavioral intentions. Another is that it can also be

directly used to predict the execution strength of an

action. Therefore, the questionnaire design item of

this research is: after the outbreak of COVID-19,

consumers pay attention to the following factors. (1)

Epidemic prevention management of express

delivery workers? (2) Disinfection of

goods delivery?

(3) The ability to deliver the goods in good condition?

2.5 Behavior Intention

Behavioral intention refers to the subjective

probability of an individual to take a specific

behavior, which reflects the degree of willingness of

an individual to take a specific behavior. We can

infer behavioral intentions, such as preferences,

interests, and hobbies, from observations of

behaviors. Behavioral intent can also be detected

through testing. Therefore, the questionnaire design

item of this research is: consumers' intentions to the

following factors after the outbreak of COVID-19.

(1) After the outbreak, do you rely more on online

shopping? (2) After the outbreak, reduce the number

of times you go out? (3) After the outbreak, avoid

shopping at brick-and-mortar stores?

2.6 Behavior

Scholar Ajzen believes that all factors that may

affect behavior are the effects of behavioral

intentions on behavioral performance. Behavioral

intention is influenced by internal psychological and

external environmental factors. Inner psychology

comes from the individual's own attitude, that is, his

attitude towards taking a particular action. The

external environment is from external pressure.

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

812

From the perspective of theoretical framework,

behavior is the construction of a series of internal

psychological mechanisms, and finally transformed

into explicit behavior. This is also the end result of

the theory of planned behavior. Therefore, the

questionnaire design item of this research is:

behavioral performance of consumers after the

outbreak of COVID-19. (1) The number or amount

of consumption in online shopping has increased

compared to the past? (2) The frequency or amount

of consumption in physical stores has relatively

decreased? (3) Will use the "official online store" of

physical stores to meet the epidemic situation

Demand during the period? (4) The number of

purchases of such products increased in online

shopping? (5) What kind of products are purchased

online? (6) How often do you use express delivery?

As far as this study is concerned, consumers' fear

of the spread of the virus (attitude) and the

willingness to use online shopping (behavioral

intention) are stronger. At the same time, the

willingness to reduce going out during the epidemic

has increased (subjective norm). Consumers reduce

shopping behavior in brick-and-mortar stores and

shift to online shopping behavior at home and

increase the frequency of home delivery (perceived

behavioral control). Therefore, we can observe and

measure consumers' shopping behavior in the face of

the epidemic from the above analysis.

3 RESEARCH METHODS

3.1 Sampling Design

The sampling method for this study will be

convenience sampling. The sample objects are

consumers in Huaian City, Jiangsu Province.

Questionnaires are administered to them via the

Internet. The survey will be released in June 2021.

Pre-test questionnaires were collected, and 50 copies

were collected. The revised formal questionnaire

will be distributed between July 2021 and November

2020, with a valid collection of 350 copies. The

recovery rate was 87%. Questions were measured

using a five-point Likert scale ranging from 5 for

"strongly agree", 4 for "agree", 3 for "normal", 2 for

"disagree", and 1 for "strongly disagree" Express.

We propose six hypotheses as the basis for model

construction: H1 attitudes have a positive effect on

subjective norms. H2 subjective norm has a positive

effect on perceived behavioral control. H3 subjective

norm has a positive effect on behavioral intention.

H4 attitude has a positive effect on behavioral

intention. H5 Perceived behavioral control has a

positive effect on behavioral intention. H6

Behavioral intentions have a positive effect on

behavior.

3.2 Questionnaire Design

In this study, a questionnaire on the impact of the

epidemic on e-commerce and express delivery was

compiled based on relevant domestic and foreign

literatures. The planned behavior theory advocated

by Ajzen is used as the questionnaire infrastructure.

The content of the questionnaire is divided into 5

parts. The first part is attitude: it refers to the degree

of participation of consumers in their willingness to

choose online shopping for items purchased during

the epidemic. The second part is the subjective

norm: it refers to the psychological pressure of

consumers who are afraid of contracting the virus

during the epidemic, and reduce the environmental

pressure of their willingness to change their

shopping behavior in brick-and-mortar stores to

online shopping behavior. The third part is perceived

behavioral control: it refers to the perception that

consumers are affected by the epidemic and change

to online shopping instead of physical store

purchases in order to reduce the virus infection of

the crowd. The fourth part is behavioral intention:

the continuation and spread of the epidemic, and the

intention to adopt online shopping behavior. The

fifth part of behavior refers to the replacement of

individual purchases in brick-and-mortar stores by

online shopping.

4 DATA ANALYSIS

4.1 Pre-Test Questionnaire Design

The questionnaire of this study conducted a

reliability analysis on the questionnaire items of

consumers' attitude towards online shopping during

the epidemic. Scholars believe that a Cronbach's

alpha of 0.5 or above is acceptable, with a high

confidence value between 0.6 and 0.9. The

standardized reliability coefficients of each scale in

this study were Cronbach's α of attitude 0.802,

Cronbach's α of subjective norm 0.907, Cronbach's α

of perceived behavioral control 0.799, Cronbach's α

of behavioral intention 0.800, and Cronbach's α of

behavior 0.837. The reliability test shows that the

data is between 0.800 and 0.910, indicating that each

part of the questionnaire has a good reliability value.

Exploratory Analysis of Factors Affecting Levels of Online Shopping in the COVID-19

813

4.2 Analysis of Sampling Data

4.2.1 Consumer's Socioeconomic

Background

This study randomly selected consumers' attitudes

towards the service quality of the express delivery

industry during the epidemic. The distribution of the

interviewees in the socio-economic background

analysis in this study and the relevant narrative

statistical analysis results are as follows: the gender

ratio of the sample data in this survey is 46% males

and 54% females. The age distribution is 3% under

18 years old, 75% between 18 and 26 years old, 10%

between 26 and 30 years old, 6% between 31 and 40

years old, 4% between 41 and 50 years old, and over

50 years old. Accounted for 2%. The marital status

is 78% unmarried and 22% married. The education

level is 3% for the following junior colleges, 85%

for undergraduates, and 12% for masters and above.

The monthly income is approximately 55% below

4,000 RMB, 25% from 4,000 to 5999 RMB, 8%

from 6,000 to 7,999 RMB, 7% to 8,000 to 10,000

RMB, and 5% above 10,000 RMB. The average

monthly consumption amount spent on online

shopping is approximately 84% of RMB 1,000 or

less, 13% of RMB 1,000 to 2,999, 2% of RMB

3,000 to 4,999, and 1% of RMB 5,000 to 6,999.

Occupation types are 75% of students, 5% of civil

servants, 15% of office workers, 3% of self-

employed persons, and 2% of retirees.

4.2.2 Analysis of Consumers' Purchase of

Goods

After the outbreak of the epidemic in COVID-19,

the categories of online shopping products that

increased in purchase quantity or amount were: food

and beverages, clothing and footwear products, and

household items, accounting for 50%. Disinfection

supplies, masks and sanitary supplies accounted for

33%. Purchases of newspapers, magazines and

books, audio-visual entertainment products, 3C

electronics, beauty and maintenance products, home

appliances, online courses, and health foods

accounted for 17%. The frequency of consumers

using express delivery is: 59% once a week, 20%

once every two weeks, 11% once every three weeks,

and 10% once a month.

4.2.3 Regression Empirical Results and

Analysis

a) Measurement model analysis

This study adopts partial least squares (PLS) path

model calculus, which is a nonparametric method.

Its requirements for the sample size are relatively

loose, and the sample size does not need to be

completely normal distribution. And it is convenient

for sample research and investigation. According to

the judgment criteria, the reliability of a single

variable, the composite reliability (CR), the

Cronbach's α of the latent variable, and the average

extraction variance (AVE) and other observed

values can be used as indicators for judging

reliability and convergence. And Bootstraping is

used to solve the problem of small samples and non-

multivariate normality data to obtain the stability of

estimates between various variables (Hair, Ringle,

Sarstedt, 2011, Chin, 2010).

The reliability of a single measured variable

depends primarily on how well each measured

variable can be explained by the latent variable.

Therefore, scholars suggest that the recommended

factor loading value should be higher than 0.7

(Barclay, Higgins, Thompson, 1995). After

analyzing attitude, subjective norm, perceived

behavior control, behavior intention, behavior and

other factors, most of the factor load values were

greater than 0.8. This study used the Smart PLS 3.0

tool for PLS analysis. The relevant verification

criteria are described below.

The first part is the factor load value of attitude:

after the outbreak of COVID-19, the factor load

value of fear of going out and contacting people was

0.826. The factor load value that changes

consumption behavior is 0.872. The load value of

changing lifestyle factors is 0.837.

The second part is the factor load value of

subjective norm: the factor load value of the

importance of the delivery punctual ability of goods

to the express industry is 0.865. The load value of

the factor of importance of commodity error

handling capacity for the express delivery industry is

0.880. The factor load value of standardized

transportation of goods to the express industry is

0.909. The factor load value of service attitude level

to the importance of express industry is 0.885.

The third part is the factor load value of

perceived behavior control: the factor load value of

the importance of the epidemic prevention

management of the express delivery personnel to the

express delivery industry is 0.777. The load value of

the important factor load value of the disinfection

operation for the delivery of goods to the express

industry is 0.899. The undamaged delivery capacity

of the goods has a factor load value of 0.849 for the

importance of the express delivery industry.

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

814

The fourth part is the factor load value of

behavior intention: After the COVID-19 outbreak,

the factor load value of relying more on online

shopping is 0.826. The factor load value for

reducing the number of trips is 0.850. The factor

load value for avoiding consumption in physical

stores is 0.859.

The fifth part is the factor load value of behavior:

after the outbreak of the COVID-19 epidemic, the

quantity or amount of online shopping consumption

increased compared to the previous factor load value

of 0.899. Compared with the past, the factor load

value of the number or amount of consumption in

the physical store is 0.886. The official online store

in the physical store will be used for consumption,

and the factor load value to meet demand during the

epidemic period is 0.818.

b) Composite reliability (CR)

CR refers to the consistency of variables within

the studied dimensions. The latent variable can be

tested when the CR value of the latent variable and

Cronbach'α is high. Scholars suggest that

Cronbach'α must be greater than 0.7 (0.60-0.70 for

exploratory studies), which is sufficient to indicate

that the latent variables have good internal

consistency. After analyzing the combined validity

of each factor, the combined reliability of attitude is

0.882, the combined reliability of subjective norm is

0.935, the combined reliability of perceived

behavior control is 0.880, the combined reliability of

behavior intention is 0.882, and the combined

reliability of behavior is 0.882. The value is 0.902,

and the combined reliability and validity of each

factor are mostly greater than 0.8.

c) Average extracted variance (AVE)

Average variance extraction (AVE) indicates

how many latent variables a variable can test. It can

be used as judgment reliability, or it can represent

discriminant validity. The AVE value must be

greater than 0.5 to indicate a convergent effect on

the observed variables. Also, check for problems

with collinearity for each set of predictors. The

predicted variance inflation factor (VIF) is less than

0.20, indicating that there is a collinearity problem

(Fornell & Larcker.1981). It has been verified that

the values obtained by the questionnaire items in this

study are all greater than 0.20, and there is no

collinearity. After the analysis, the average variance

of each factor is as those. Attitude (AVE) value was

0.715, subjective norm (AVE) value was 0.783,

perceived behavioral control (AVE) value was

0.711, behavioral intention (AVE) value was 0.714,

and behavioral (AVE) value was 0.754. All The

average variance extraction for factors (AVE) were

all greater than 0.7.

d) PLS module path results

This study uses the least squares method to

analyze the causal relationship between the latent

variables of the structural model. The validation is

set to 300 bootstrap parameters to perform the

validation procedure to obtain the stability of each

variable estimate. Secondly, the analysis effect value

can be used to evaluate the influence of external

variables on the internal dependent variables of

explanatory variables. In general, the influence of

external variables on internal latent variables is 0.02

for small, 0.15 for medium, and 0.35 for large. After

the analysis, the f

effect value of each factor was

analyzed: the influence of attitude on subjective

norm was small, and the f

effect value was 0.072.

The f

effect value of attitude on behavior intention

is 0.573, the effect should be significant. The effect

of perceived behavioral control on behavioral

intention was 0.043 and the effect was not

significant. The f

effect value of behavioral

intention is 0.225, which is significant. The f

effect

of subjective norm on perceived behavioral control

is 0.134, which is significant.

R-Square is a judging path significance test,

which can be used as the explanatory power of the

research model. The coefficient of determination of

R2 represents the size of the potential internal

variables in the structural formula to be explained.

The R2 value can be roughly divided into 0.75 for

large, 0.50 for medium, and 0.25 for small. The R2

determination coefficients of this study are:

subjective norm R2 determination coefficient is

0.067, perceived behavioral control R2

determination coefficient is 0.566, behavioral

intention R2 determination coefficient is 0.445, and

behavioral R2 determination coefficient is 0.692.

Except that the R2 determination coefficient of

subjective norm is less than 0.4, which is not

significant, the R2 determination coefficients of all

other factor dimensions are mostly greater than 0.4.

The predictive power of the study model was

expressed as R-squared value. It represents the

percentage of variance explained by exogenous

versus endogenous variables. Its value is between 0

and 1. The larger the value, the better the

explanatory power of the model. According to the

six research hypotheses proposed in this study, the

results of the overall model relational path validation

show that: attitude, subjective norm, perceived

behavioral control, behavioral intention, and

behavior. Each factor has a positive effect. The

Exploratory Analysis of Factors Affecting Levels of Online Shopping in the COVID-19

815

results of the research hypothesis testing of this

study are shown in Table 1, as shown in Figure 1.

Table 1: hypothesis verification of structural model.

Item Hypothesis

Path

Coefficients

Validation

results

Attitude has a positive

impact on subjective norms

0.259** support

Subjective norms have a

positive influence on

perceptual behavior control

0.752*** support

Subjective norms have a

positive influence on

behavior intentions

0.017* support

Attitude has a positive

effect on behavioral

intention

0.585** support

Perceived behavior control

has a positive effect on

behavior intention

0.234* support

Behavioral intention has a

positive effect on behavior

0.832*** support

*P<0.05,** p<0.01,*** p<0.001

Figure 1: TPB of home deliveries in COVID-19.

5 CONCLUSIONS

The COVID-19 pandemic has affected every aspect

of our lives. Through this actual survey, it is found

that from the perspective of consumers' attitudes and

behaviors in online shopping. Due to the impact of

the epidemic environment, consumers are turning to

online shopping behaviors from physical shopping

behaviors, and the trend of home delivery is

increasing. We conducted an online survey of the

population of the Huai'an metropolitan area in

Jiangsu province. Survey items include the number

of home delivery orders, household and

demographic characteristics, e-commerce and

product preferences, and relevant sociodemographic

variables for survey and data collection. According

to the data, people are afraid of being infected by the

epidemic, so they should reduce the habit of going

out to physical stores to buy goods to avoid being

infected by the virus. This phenomenon has

increased the willingness of consumers to change

their shopping behavior from physical store

shopping to online shopping. Consumers buy

products online, with daily necessities and epidemic

prevention products as the main products. The

number of home deliveries increased significantly

before the COVID-19 period compared to the post-

COVID-19 period. From the perspective of planned

behavior theory, it is explored that the increase of

express delivery business and the increase of online

shopping behavior during the COVID-19 period

have a significant positive correlation with the

impact of the epidemic on consumers.

ACKNOWLEDGMENT

We would like to express our gratitude to Huaiyin

Institute of Technology, the team members of the

innovation and entrepreneurship project for college

students. They are Junlan Jia, Xu Kang, Qing Zhang,

Chenggong Li.

REFERENCES

Ajzen, I. (1985). From intentions to actions: A theory of

planned behavior. In Action-control: From cognition

to behavior. Edited by: Kuhl, J and Beckman, J. 11–

39. Heidelberg: Springer.

Ajzen, I. & Fishbein, M. (2000). “Attitudes and the

attitude-behavior relation: Reasoned and automatic

processes”. In European review of social psychology,

Edited by: Stroebe, W and Hewstone, M. 11, 1–33.

Chichester: Wiley.

Ajzen, I. & Driver, B.L. (1991). Prediction of leisure

participation from behavioral, normative, and control

beliefs: An application of the theory of planned

behavior. Leisure Sciences, 13: 185–204.

Barclay, D., Higgins, C., Thompson, R. (1995). The

Partial Least Squares (PLS) Approach to Causal

Modeling: Personal Computer Adoption and Use as an

Illustration. Technology Studies, 2(2), 285-309.

Chin, W. W. (2010). Bootstrap cross-validation indices for

PLS path model assessment, in Vinzi, V. E. et al.

(eds). Handbook of partial least squares: concepts,

methods and applications. Heidelberg: Springer, 83-

97.

Fornell, C. & Larcker, D. F. (1981). Evaluating Structural

Equation Models with Unobservable Variables and

Measurement Error. Journal of Marketing Research,

18(1), 39-50.

Hair, J. F., Ringle, C., Sarstedt, M. (2011). PLS-SEM:

Indeed a silver bullet. Journal of Marketing Theory

and Practice,19,139-151.

Masudin, I., Ramadhani, A., Restuputri, D. P.(2021).

Traceability system model of Indonesian food cold-

chain industry: A Covid-19 pandemic perspective.

Cleaner Engineering and Technology.

https://doi.org/10.1016/j.clet.2021.100238.

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

816

Tsai, Y. C. (2021). Discussion on Consumers' Motivation

to Improve Online Shopping under the New

Coronavirus. Master's Degree in Business School of

Kainan University, Taiwan.

Unnikrishnan, A. & Figliozzi, M. (2021). Exploratory

analysis of factors affecting levels of home deliveries

before, during, and post- COVID-19. Transportation

Research Interdisciplinary Perspectives.

https://doi.org/10.1016/j.trip.2021.100402 .

Uzir, M. H., Halbusi, H. A., Thurasamy, R., Hock, R. L.

Aljaberi, T. M. A., Hasan, N., Hamid, M. (2021). The

effects of service quality, perceived value and trust in

home delivery service personnel on customer

satisfaction: Evidence from a developing country.

Journal of Retailing and Consumer Services.

https://doi.org/10.1016/j.jretconser.2021.102721.

Exploratory Analysis of Factors Affecting Levels of Online Shopping in the COVID-19

817