CHAPTER I: INTRODUCTION...7Define online shopping method?...7Define online shopping behavior?...7Online shopping trends in Vietnam and other countries...8My company i choose...9PROBLEM S
INTRODUCTION
METHODOLOGY
RESEARCH DESIGN OR RESEARCH PROCESS
To ensure the scientific, the research is carried out through two main stage preliminary research and formal research.
Step Research type Method Data collection techniques
Discussion Double hand Ho Chi Minh
2 Official Quantitative Interview direct Ho Chi Minh
Step 1: Identify the research problem, research question and objective research.
Step 2: Refer to documents related to the research topic It proposes a research model and establishes research hypotheses.
Step 3: Prepare the questionnaire and edit the questionnaire A table draft questions with scales based on previous studies that have been establish Then the questionnaires were delivered to a small group of 5 - 10 Finally, a primary survey was conducted with 300 survey samples.
Step 4: The author conducts surveys and collects data and Questionnaires were sent directly to 300 people.
Step 5: Edit, and adjust the data and reliability of the methods The measurement method is analyzed by Cronbach's Alpha and must meet the coefficient requirement above 0.6.
Step 6: Test the research hypothesis and determine the relationship of the factors in the model through multivariate regression analysis.
Preliminary scale will be used as a reference basis for conducting qualitative research to build questionnaires for observed variables used to measure concepts in the based model In this post-COVID period, researchers will use a one-on-one discussion technique with subjects selected according to a convenient method that still reflects the characteristics of the observed sample set.Subjects selected to participate in the qualitative study are 22-40 years old, have experience in using26 the internet, have purchased goods online or have knowledge of shopping on shopee, and have understanding of shopee The survey was conducted by a combination of methods including: direct interviews with customers, distribution of pre-printed questionnaires to respondents and receiving results after completion Those who want to buy electrical appliances by convenient sampling method are graduate students who have families and have the right to decide to buy electrical appliances.
This study's model incorporates concepts and a preliminary scale developed by foreign authors The author analyzed various groups of factors influencing post-COVID-19 online shopping decisions and identified observable variables for each component scale within the model.
The scale is used as a basis for qualitative research to build a questionnaire for quantitative research Qualitative research was carried out by hand-to-hand discussion method In this study, the two terms “post-covid-19 online sale in COVID-19 purchase” have the same meaning The author uses the term purchase through in qualitative and quantitative research to create closeness and easy understanding for all survey participants Scales were needed to measure the variables accurately, so different variables were selected with appropriate scales The variables are applied according to the 5-point Likert scale, with the scale scale according to the score as follows:
1: Totally disagree, 2: Disagree, 3: Neutral (Compassion) usually), 4: Agree, 5: Totally agree.
This study aims to directly collect customer feedback through survey questionnaires The obtained data serves multiple purposes: evaluating the reliability and validity of the measurement scale, testing the scale's effectiveness, and assessing the model's appropriateness.
To choose an appropriate sample size, according to researchers Hair et al (1998), for exploratory factor analysis (EFA) minimum sample size N≥5*x (x: total number of variables) observe) For authors Tabachnick and Fidell (1996) to conduct regression analysis in the best way, the minimum sample size to be achieved is calculated by the formula N≥8m+50 (where N is the sample size, m is the number of variables model independence) In this study, the author chooses a large enough sample size to satisfy both conditions as suggested by the EFA factor research method and the multiple regression method N≥max (sample size according to EFA requirements; sample size required by multiple regression) This study is expected to sample with a size of
300 samples for 25 observed variables This sample size will be the basis for preparing the number of 320 questionnaires to be distributed.
The authors collect data from other scientific articles, articles published in newspapers or specialized scientific journals on the topic.
In order to minimize the cost and time of the survey, the study uses convenience sampling method The target audience is people of the Z gen group (10-25 years old), who are living in Hanoi and have been shopping online.
The survey was conducted from March 2, 2022 to March 11, 2022 After screening and removing 13 invalid responses, the final number of questionnaires included in the analysis was 224. This data is distributed relatively evenly by gender as well as geographical area in 12 districts in Hanoi city.
The survey object is gen Z, so the number of respondents who do not have a university/college degree account for a large proportion (58.5%) and only 16.5% has a university degree The number of respondents with a monthly personal income of less than 5 million dong accounts for 42%, but still more than 24% have earned over 10 million dong per month In terms of time using the internet per day, the majority of respondents spend 5 hours/day on the internet, 31.3% spend 2 - 5 hours and the number of people who spend less than 2 hours/day using the internet accounts for only 20.1%.
28 figure 1 Descriptive statistics of the study sample
The authors analyzed the data using SPSS 20.0 software Chi-square tests and Cramer's V tests were performed to test the relationship between demographic variables including gender,29 education level, monthly personal income, and time internet use every day with the frequency of online shopping of the study subjects.
The study uses the Cronbach Alpha reliability coefficient and removes the variables with the small-sum correlation coefficient, testing the preliminary scale by exploratory factor analysis (EFA) Multivariate regression analysis Then, test the research hypotheses Analysis of variance ANOVA to test Data analysis is based on demographic variables to analyze differences between the following groups: Male and Female; High Income and Low Income; Young and Elderly; Education and Job Qualifications.
Preliminary assessment of the scale
Evaluation of Cronbach's Alpha reliability coefficient
Purpose This method allows analysis to find out which items should be kept and which items should be omitted from the items to be tested (Hoang Trong and Chu Nguyen Mong Ngoc, 2008) or in other words, help to eliminate observed variables, the scales fail Observable variables with correlation coefficient of total variable (Corrected Item-Total Correlation) less than 0.3 will be eliminated and the standard for choosing the scale when Cronbach's Alpha coefficient is from 0.6 or higher Many researchers agree that when Cronbach's Alpha is from 0.8 or more to close to 1, a good scale, from 0.7 to close to 0.8, is usable There are also
Researchers suggest that Cronbach's Alpha of 0.6 or higher is usable in case the concept being measured is new or new to respondents in the research context (Hoang Trong and Chu Nguyen Mong Ngoc) , 2008).
Factor analysis is primarily utilized to assess both convergent and discriminant validity It proves valuable in identifying the variables required for a research study and in uncovering the interrelationships between them In conducting exploratory factor analysis (EFA), researchers prioritize specific criteria, which include:
KMO (Kaiser-Meyer-Olkin measure of sampling adequacy): is an indicator used to consider the adequacy of factor analysis A large KMO value (between 0.5 and 1) is a sufficient condition for factor analysis to be appropriate If the KMO is less than 0.5, then the factor analysis is likely to be inappropriate for the data Bartlett's test (in factor analysis, it is necessary to test the correlation of variables with each other (H0: variables are not correlated with each other in the population).potentially inappropriate If this test is significant (sig < 0.05), then the observed variables are correlated with each other in the population (Hoang Trong and Chu Nguyen Mong Ngoc, 2008).30
ANALYSIS & DISCUSSION
Samples were collected by convenient method in the form of tables survey questions Through actual survey data collection, there are 20 invalid votes (due to important information or younger age than survey conditions) Therefore, the final correctly selected sample size was 300 samples and was imported into Excel software and further imported into SPSS 16.0 for quantitative analysis.
Cons tant number Total squares df Medium way
Multilinear phenomenon: The results of the multilinear phenomenon test of model 1. Thereby, we see that the variance magnification coefficient (VIF) of independent concepts in
1 715 a 512 504 3956995 model 1 is less than 10; Proof: between independent concepts no multilinear phenomenon occurs.
Model 1 demonstrates compliance with the critical assumptions required for regression analysis The absence of a discernible trend in the estimated residuals indicates adherence to the assumption of constant variance Additionally, the Durbin-Watson statistic of 1.96 suggests that the estimated residuals are independent, implying the absence of any linear correlation between them.
Regression model testing after the covid-19 pandemic
Thus, the study model has a revised R2 of 0.504, meaning that 50.4% of the variability of customer decisions is explained by the variability of independent variables including: ease of use perception, usefulness perception, risk awareness, behavioral control awareness with 95% confidence, Thus, 49.6% of the variation of dependent variables is due to variables other than the unrecognized model Specifically, the regression function is written as follows: Decision 1.172 + 0.171 *ease of use + 0.474 *usefulness - 0.096*risk
In particular, the Beta coefficient of ease of use, usefulness and behavior control has a positive coefficient and has a positive impact on customers' buying decisions Particularly, the beta coefficient of risk has a negative coefficient and has a negative influence on customers' decisions to buy shopping online
According to the 5-degree likert scale, under constant other conditions, when the perception of ease of use increases to 1, the level of the customer's decision to buy increases to 0.171 units Similarly, when the perception of usefulness increases by 1, the level of the customer's buying decision increases to 0.474 units If behavior control increases to 1 unit, the level of purchase increases to 0.067 units Meanwhile, the risk perception increased by 1, the level of customer buying decision decreased by 0.096 units.
Levene analysis df1 df2 Itself.
Levene test result for value Sig =.549 (greater than 5%), so the H0 hypothesis – variance of homogeneous sex groups, is accepted; the right dataset for performing ANOVA audits. The results of the ANOVA test with Sig = 241 (greater than 5%) showed that: hypothesis H0 – no difference in buying decisions between sex groups, is accepted This means that gender factors do not influence a customer's buying decision The average level of buying decisions of men compared to women did not differ.
Levene analysis df1 df2 Itse lf.
Based on the Levene test (Sig = 0.770), the assumption of equal variances across age groups is supported, indicating homogeneity Subsequent ANOVA analysis revealed no significant difference in buying decisions between age groups (Sig = 0.393) This suggests that age does not influence customer decisions to shop online The average buying behavior across different age groups does not exhibit significant variations.
Analyze differences by education level
Levene analysis df1 df2 Itself.
Total pitcher s way df Medium way F Itself.
Levene test result for value Sig = 562 (greater than 5%), so the H0 hypothesis – variance of homogeneous academic level groups, is accepted; the right dataset for performing ANOVA audits The results of the ANOVA test with Sig = 791 (greater than 5%) showed that: hypothesis H0 – no difference in purchasing decisions between education groups, accepted This means that education level does not affect the customer's buying decision The average level of purchasing decisions of education groups did not differ.
Pitchers way df Medium way
Levene test result for value Sig = 216 (greater than 5%), so the hypothesis H0 – variance of homogeneous professional groups, is accepted; the right dataset for performing ANOVA audits.
The results of the ANOVA test with Sig = 979 (greater than 5%) showed that: hypothesis H0 – no difference in buying decisions between specialized groups, is accepted This means that expertise does not influence the customer's buying decision.
Analysis of differences by income
Totals square df Average square
Levene test result for value Sig = 636 (greater than 5%), so the H0 hypothesis – variance of homogeneous income groups, is accepted; the right dataset for performing ANOVA audits.
The results of the ANOVA test with Sig = 379 (greater than 5%) showed that: hypothesis H0 – no difference in buying decisions between income groups, is accepted This means that income does not affect the customer's buying decision.
Recognition rate of online sales websites popular in
In terms of popularity, Shopee emerges as the most prominent online shopping platform, claiming a 24.0% market share Chodientu.vn follows closely with 20.0%, while Chotot.vn stands at 15.7% Dienmay.com captures 23.3% of the market, and 5giay.com secures 17.0%.
Average time of using the Internet/ 1 day According to table Shopee
It shows: Out of a total of 300 survey subjects , 42 subjects (14.0%) access less than 0.5 hours; there are 77 objects (25.7%) visited from 0.5 to 2 hours ; there are 64 objects (21.3%) accessing from 2 to 4 hours ; 65 objects (21.7%) visit from 4 to 6 hours ; there are 52 objects (17.3%) visited more than 6 hours.
Time of use Amount Rate (%)
Statistics of Internet use experience shopee
Experience in use Amount Proportion
Of the 300 survey subjects, the highest number of subjects had experience of using the Internet for more than 7 years with 77 subjects (25.7%), followed by 67 subjects from 5 to 7 years (22.3%) And the minimum is less than 1 year of experience There were 41 subjects (13.7%).
Average time/ 1 visit to websites sixty one 20.3 shopee
Time accessed/ 1 time Amount Rate (%)
Of the 300 subjects, the maximum number of subjects using the Internet was less than 10 minutes with 79 subjects (26.3%), followed by the time from 10 to 30 minutes with 71 subjects (23.7%) And the lowest is unused There were 39 subjects (13.0%).
Number of visits/ 1 month to shopee in recent time
According to table 4.5 shows that: Of the 300 survey subjects , the highest number of times accessing the Internet is from 3 to 5 times with 66 subjects ( 22.0.0 %), followed by the number of visits from 6 to 10 times with 66 subjects object (22.0) %) And the lowest was unused , 41 subjects (13.7%).
Sex ratio of the observed sample
According to statistics in the observed sample, the number of female sex is 60% more than the male gender
Age of the observed sample
Subjects over 40 years old accounted for 16.3%, distributed 24.7% were from 22 to 27 years old, from 28 to 30 years old accounted for 22.0%, from 31 to 40 years old accounted for 19.0%, under 22 years old accounted for 18.0%.
Education level of the observed sample
Statistics in the observed sample show that: high school accounted for 12.8%, professional high school accounted for 18.7%, College accounted for 24.7%, University accounted for 23.0%, Postgraduate accounted for 21.0%
CONCLUSIONS AND RECOMMENDATIONS
The results showed that there were 30 root variables of the scale belonging to 5 components Information from the observation sample showed that the survey subjects were young, concentrated between 22-27 years old Most of them have experience using the Internet, have knowledge of online sale in COVID-19 purchase services.
From the test results, it shows the suitability of the theoretical model for the decision to buy shopping online, as well as the hypotheses that are accepted in this study, bringing a practical meaning to businesses trading shopping online in Ho Chi Minh City Help businesses trading shopping online to get the most effective number of customers, create outstanding competitive advantages and develop sustainably in times of economic crisis The author has focused on the factors that influence the decision to buy shopping online.
The results of the study showed 5 factors affecting the decision to buy shopping online with the level from high to low: Usefulness, Ease of use, Behavior control, Social influence, Risk perception.
The study examined the differences between different groups of customers by gender, age, income, education level, and expertise for the decision to buy shopping online.
Utilizing regression analysis, the author proposes actionable solutions encompassing Usefulness, Ease of Use, Behavior Control, Social Influence, and Risk Perception These recommendations aim to enhance business efficiency through increased positivity, proactive initiatives, and improved customer satisfaction By fostering trust and maximizing efficiency, businesses can achieve optimal performance.
As a result of the descriptive statistics, the rate of awareness of online sale in COVID-19 purchases shows that most Internet users have an interest in buying online, only 13.0% of respondents have never visited an e-commerce website, the rest visit with different levels and different access times In today's Shopee, the awareness rate of Shopee.vn website is 23.9% higher than other sites, followed by dienmay.com website with a recognition rate of 23.3% In terms of the business method of websites, the dienmay.com is an e-commerce website that both introduces products and sells online has been known by many people with a widely developed59 system of stores and supermarkets along with promotion policies to motivate consumers to buy online.
For factors affecting the same dimension on the decision include: Ease of use (B = 0.171); Usefulness (B=0.474); Social influence (B=0.084); Behavior control (B=0.067) It shows that in order to make a positive decision, positive factors and a necessary capacity to carry out the behavior are accompanied by support from the family and society Therefore, businesses need to have a market segment with targeted customers and a suitable marketing policy to motivate consumers In contrast, risk perception (B= -0.096) inversely impacts online purchasing decisions That is, when consumers feel insecure in their transactions and see risks, it reduces the purchase decision Therefore, businesses trading shopping online need necessary solutions to reduce this risk.
Limitations and direction of further research
The author made efforts in carrying out this study, however, also not immune to limitations:
Despite an adjusted R-squared value of 0.504, the study model only accounts for 50.4% of the variation in online sales during COVID-19 purchasing decisions This indicates that a significant portion of the variability remains unexplained, suggesting the need to include additional observational variables in the model.
Appropriate solutions will be offered with particular focus on the current situation and present conditions These solutions aim to enhance the factors that influence online shopping decisions, thereby boosting online purchases in the near future.
The topic uses qualitative and quantitative research methods, which require training in skills and experience in answering survey questions Respondents do not have the skills to answer questions, do not understand or do not understand online shopping services, so the answers are still emotional So, the scales only manifest themselves at a relative level of the decision to buy shopping online.
Finally, the author argues that the factors affecting the decision to buy shopping online are constantly changing according to the diverse needs and desires of customers, in today's market conditions Furthermore, there may be many other factors that have not been addressed in this topic.
The factors studied are the basic factors that influence consumers' decisions to buy shopping online, further studies can study other theoretical contents to build a more accurate and60 complete model of consumers' "online sale in COVID-19 purchase decisions" Besides, it is also necessary to survey other items purchased and sold online such as air tickets, books, newspapers.
REFERENCES
1 Ajzen I., Fishbein M (1975), Belief, Attitude, Intention and Behavior:
An Introduction to theory and research, Addition-Wesley, Reading, MA.
2 Bhimani, A (1996), “Management accounting: European Perspectiver”, Oxford, Oxford University Press.
3 Cox, D.F and Rich, S.V (1964) Perceived risk and consumer decision- making - the case of telephone shopping Journal of Marketing Research, 1, 32-39.
4 Chen, L -D., Gillenson, M L and Sherrell, D L., (2005) “Enticing Online Consumers: An Extended Technology Acceptance Perspective” Information & Management Vol 39, No.8: 705-719.
5 Cheung C.M.K., Lee M.K.O (2005), Research Framework for Consumer Satisfaction with Internet Shopping City University of Hong Kong, China Sprouts: Working Papers on Information Systems, 5(26) http://sprouts.aisnet.org/5-26.
6 Davis, D Fred, and Arbor, Ann (1989), Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology MIS Quarterly September 1989.
7 Dan Wang, Liuzi Yang (2010), Customer Buying Behavior – Online Shopping towards electronic product, Umea School of Business.
8 Hair, J.F Anderson, R.E R.L Tatham and William C Black (1998), Multivariate Data Analysis, Fifth Edition Prentice-Hall Internatinal, Inc.
9 Haubl, G and Trifts, V (2000) Consumer decision making in online shopping environments: The effects of interactive decision aids, Marketing Science, 19 (1), pp 4-21.
10 Haubl, G and Murray, K B (2003) Preference construction and Persistence in digital marketplaces: The role of electronic