4. Data analysis
4.1. Expert interview analysis
4.2.2. Shopping habits during pandemic
The first question in this group of questions in the questionnaire was “Thinking about your everyday life, since the COVID-19/coronavirus pandemic, have you made any changes to your general lifestyle?” in order to understand which changes people implemented in their life due to pandemic triggered environmental conditions.
According to the answers of the respondent’s 81 percent mentioned that with the spread of COVID-19, they started wearing masks and stayed more at home. Fifty-six percent mentioned that there were adhering to social distancing, and 53 percent started to wash their hands more frequently in comparison with the pre-pandemic situation. A bit less than a half of the respondents indicated that they reduced the number of their travels, decided to limit the use of public transport and public places in general, 45%, 44%, and 39% respectively. 42% started to visit physical stores less, and 38% of them mentioned that they started to do shopping more via online channels.
26%
9%
46%
5%
14%
Employment Status
Employed Full-time Employed Part-time Student Self-employed Not employed
Also, among the changes in the daily life of people, you can see that people have reduced their load on sports (28%), switched to work from home (28%), stopped constantly using cash (23%). These changes did not affect the lives of most people; nevertheless, a significant part noted precisely these changes in their everyday life (See. Figure 8).
Figure 8: Changes in lifestyle during pandemic
Source: Author
Respondents were asked to mention whether they decreased their day-to-day spending or not. The results show that the answers were divided into two equal parts. Half of the participants responded that they had reduced their daily costs and the other half responded negatively.
Three questions were asked related to the online shopping frequency. The first question (See. Figure 9) is “How often did you start using e-commerce to purchase products which normally bought in-store due to coronavirus?”. Majority of the answers (38%) state that shoppers “Never used” online stores during the pandemic but continued to use physical stores to purchase the products they needed. 30% of the respondents mentioned that they used the online stores once a month. And the rest mentioned that they used online shops “Once in two weeks” (12%), “More than once a week” (10%), and “Once a month” (10%). Overall, the
overwhelming majority of the respondents (62%) have used online shops at least once a month in order to purchase required products.
Figure 9: Frequency of e-commerce use due to COVID-19
Source: Author
The second question related to the online shopping frequency is, “Compared to the amount of online shopping you did prior to the coronavirus (COVID-19) outbreak, would you say, you have been doing more, less, or about the same amount of shopping online?”. As it is illustrated in Figure 10, almost half of the respondents answered that with the start of the pandemic, the frequency of their orders through online stores has not changed. 31% of the respondents indicated that during the coronavirus outbreak, they did more online shopping, compared with the time before the pandemic situation. 22% mentioned that they oppositely used the service of online shops less frequently.
10%
10%
12%
38%
30%
How often did you start using e-commerce to purchase products which normally bought in-store due to coronavirus?
Once a week More than once a week Once in two weeks Never used Once a month
Figure 10: Changes in use frequency of online shopping due to COVID-19
Source: Author
The last question in this group concerns future plans for online shopping “Compared to the amount of online shopping you did prior to the coronavirus (COVID-19) outbreak, would you say, you have been doing more, less, or about the same amount of shopping online? “.
The responses are similar to the ones that respondents provided for the previous question.
Almost 50% of the respondents are not going to change the habit with regard to the frequency of online shopping. 30% stated that they are planning to use online shops more frequently;
however, 22% plan to use online stores less often.
Considering the future plans of the respondents regarding the online shopping (See.
Figure 11), approximately half of them mentioned that they do not intend to make more or fewer purchases through online sales channels, 29.6% of them plan to increase the number of online purchases, and 21.8%, on the contrary, think about using online sales services less often.
Figure 11: Future plans of doing online shopping
Source: Author
Moving on to the part on what human needs, respondents began to spend more or less money with the advent of the pandemic; according to Figure 12, 69% of respondents began to spend more money on maintaining their own hygiene, on medical expenses such as buying medicines/vitamins or on the same hand sanitizer ... In addition, respondents began to spend more money on food, be it food from restaurants or groceries from supermarkets. As many as 64% of survey participants noted this. Along with this, 33% of respondents noted that with the onset of the pandemic, their costs for utilities, which include electricity, house maintenance, taking out the trash, etc., also increased. 23% noted that they started buying more products for home entertainment, for example, different types of games, be they computer or board games, books, and other entertainment. Almost the same proportion of respondents noted that since the beginning of the pandemic, they began to buy more clothes and spend on their hobbies, 16%, and 14%, respectively. As for travel, going to restaurants, cafes, or bars, expenses for home improvement such as buying electronics, furniture, and others, as well as investments, from 5 to 6 percent of survey participants noted that they had increased their budget for these expenses during the pandemic. Relatively, on services such as beauty salons, hairdressers, etc., 8 percent said they spent more money since the beginning of the pandemic. These results support that due
to pandemic, more respondents started to spend more money on their safety and psychological needs that include needs which include the needs to support health, vitality, safety, and other factors.
Figure 12: For what survey respondents spent more money
Source: Author
From Figure 14, it is possible to see for what respondents reduced their spending during the pandemic. The overwhelming majority of survey participants (81%) noted that during the pandemic, they began to spend less on going to public places where respondents are generally socialized, such as bars, restaurants, cinemas, etc. Second, according to the number of answers to which people did not spend a lot of money, is travel regardless of the purpose, 56% noted that the pandemic also affected this spending. 56% of respondents noted that they began to spend less on buying clothes during the pandemic. The next is the spending on various types of services, almost half of the survey participants (49%) spent less money on services related to cosmetic services. Categories of food, hobbies, and investments have the same rate of 17%, which means that the same number of participants have reduced their costs related to the mentioned categories. 33% spend less on housing costs and 23% on home entertainment. And only 3%, according to the graph decreased money spending for health and hygiene. Therefore, according to the results shown in figure 2, during the pandemic, respondents were spending less money on traveling, going out to socialize, buying clothes and etc.
If to compare the results from Figure 13 and Figure 14, it is possible to notice that the same proportion of respondents spent on utilities like electricity, water supply, etc., more and less during the pandemic. From this we can conclude that the demand for products belonging to this category has not generally changed. In addition, almost the same percentage is observed in the hobby category. The pandemic did not affect how much respondents spend on their hobbies and their own hobbies.
Figure 13: For what survey respondents spent less money
Source: Author
The Table 2 illustrates “Fresh and Cupboard Products” had an average of 2.00 (SD = 0.57, SEM = 0.04, Min = 1.00, Max = 3.00, Skewness = 0.20, Kurtosis = -0.73). According to the results, customers continued to buy the same quantity of products that lie under the category of “Fresh and Cupboard Products.”
Table 2: Summary Statistics Table for Fresh and Cupboard Products Variable
“Confectionary and Sweets” had an average of 2.07 (SD = 0.61, SEM = 0.04, Min = 1.00, Max = 3.00, Skewness = -0.05, Kurtosis = -1.04) (Table 3). Similarly, as in the “Fresh
Variable M SD n SEM Min Max Skewness Kurtosis
Fresh and Cupboard Products 2.00 0.57 264 0.04 1.00 3.00 0.20 -0.73
and Cupboard Products” category, customers did not change their habit to buy more or less products that are related to the category of “Confectionary and Sweets.”
Table 3: Summary Statistics Table for Confectionary and Sweets Variable
Variable M SD n SEM Min Max Skewness Kurtosis
Confectionary and Sweets 2.07 0.61 264 0.04 1.00 3.00 -0.05 -1.04
4.2.3. Perception of online shopping
In order to analyse the perception of online shopping by the consumers in Azerbaijan, the section is divided into seven parts which are: “Access convenience,” “Evaluation convenience,” “Transaction convenience,” “Search convenience,” “Post-purchase convenience,” “Possession convenience” and “Behavioural intentions.”
All items included into separate convenience groups and behavioural intentions were run through the reliability tests. In order to measure the convenience of the access to the websites, three items were used, the coefficient alpha of the “Access convenience” variable is 0.8. “Evaluation convenience” is used to measure the availability of reasonable information regarding the offered products/services on the website. The variable consists of three items with a coefficient alpha of 0.79. “Transaction convenience” consists of three items, with the aim to measure the user-friendliness of the payment procedure, coefficient alpha is 0.78. The variable
“Search convenience” consists of six items and aimed to measure the convenience of the layout, information architecture convenience, and ease of navigation through the websites. The coefficient alpha is 0.86. “Post-purchase convenience” and “Possession convenience” variables are used to measure delivery level. Variables consist of three and four items with coefficient alphas 0.65 and 0.79, respectively. The last variable, which is the dependent variable is
“Behavioural intentions,” initially consisted of three items; after conducting the reliability test, the item “I will use websites more often for online purchases” was removed, the result of the reliability test is 0.82. Table 4 presents the results of the reliability analysis.
Table 4: Descriptive Statistics
Note. N = 264. The lower and upper bounds of Cronbach's α were calculated using a 95% confidence interval.
Correlation
A Pearson correlation analysis was conducted between "Access Convenience,"
"Possession convenience," "Search Convenience," "Transaction Convenience," "Evaluation Convenience," "Post-purchase Convenience" and "Behavioural intentions" in other words, intention to use the online shopping method. Cohen's standard was used to evaluate the strength of the relationship, where coefficients between .10 and .29 represent a small effect size, coefficients between .30 and .49 represent a moderate effect size, and coefficients above .50 indicate a large effect size (Cohen, 1988).
The result of the correlation was examined based on an alpha value of 0.05. A significant positive correlation was observed between "Behavioural intentions" and “Access Convenience”
(rp = 0.54, p < .001, 95% CI [0.45, 0.62]), “Transaction Convenience” (rp = 0.53, p < .001, 95% CI [0.44, 0.61]), “Evaluation Convenience” (rp = 0.42, p < .001, 95% CI [0.32, 0.52]),
“Search Convenience” (rp = 0.52, p < .001, 95% CI [0.43, 0.60]), “Possession convenience”
(rp = 0.56, p < .001, 95% CI [0.47, 0.64]) and "Post-purchase Convenience" (rp = 0.47, p <
.001, 95% CI [0.37, 0.56]).
The correlation coefficient between "Behavioural intentions" and "Access Convenience," "Transaction Convenience," "Search Convenience," and "Possession convenience" indicates a large effect size. However, the correlation coefficient between
"Behavioural intentions" and "Evaluation Convenience" and "Post-purchase convenience"
indicate moderate effect size.
According to the results, the most influential aspect that leads to the increase in the intention to do shopping online is "Possession convenience," which means that the delivery level has a significant impact on the overall shopping experience of the consumers and their further intention to purchase online, what supports H9. Transaction Convenience" (r=.53),
"Access Convenience" (r=.54), and "Search Convenience" (r=.52) have minor differences in the impact level to the online purchase intention; nevertheless, all these aspects positively impact "Behavioural intentions". Table 5 presents the results of the correlation among Convenience Dimensions.
Table 5: Correlation Matrix Among Convenience Dimensions
Variable 1 2 3 4 5 6 7
Access convenience -
Evaluation convenience 0.46** -
Transaction convenience 0.56** 0.53** -
Search convenience 0.61** 0.66** 0.60** -
Post-purchase convenience 0.46** 0.44** 0.55** 0.50** -
Possession convenience 0.50** 0.53** 0.60** 0.54** 0.59** - Behavioural intentions 0.54** 0.42** 0.53** 0.52** 0.47** 0.56** -
Note. N = 264. ** Correlation is significant at the 0.01 level (2-tailed).
In addition to that, correlation analysis was conducted between “Overall Convenience”
and “Behavioural Intention” (see Table 3). “Overall convenience” (α=.88) includes all the mentioned convenience variables. A significant positive correlation was observed between
“Overall Convenience” and “Behavioural Intention” (rp = 0.65, p < .001, 95% CI [0.57, 0.71]).
The correlation coefficient between “Overall Convenience” and “Behavioural Intention” was 0.65, indicating a large effect size. This correlation indicates that as “Overall Convenience,”
“Behavioural Intention” tends to increase. Table 6 presents the results of the correlation between Overall Convenience and Behavioural Intentions.
Table 6: Pearson Correlation Results (Overall Convenience and Behavioural Intentions)
Combination rp 95% CI p
Overall convenience-Behavioural intentions 0.65 [0.57, 0.71] < .001
Note. N = 264. 1
Regression analysis
A linear regression analysis was conducted to assess whether “Access convenience,”
“Transaction Convenience,” “Search Convenience,” “Evaluation convenience,” “Post-purchase Convenience,” and “Possession Convenience” significantly predicted “Behavioural Intentions.”
The results of the linear regression model were significant, F(6,257) = 33.51, p < .001, R2 = 0.44, indicating that approximately 44% of the variance in “Behavioural Intentions” is explainable by “Access convenience,” “Post-purchase Convenience,” “Possession Convenience,” “Evaluation convenience,” “Transaction Convenience,” and “Search Convenience.” “Access convenience” significantly predicted “Behavioural Intentions”, B = 0.24, t(257) = 3.65, p < .001. This indicates that on average, a one-unit increase of “Access Convenience” will increase the value of “Behavioural Intentions” by 0.24 units. “Post-purchase Convenience” did not significantly predict “Behavioural Intentions”, B = 0.10, t(257) = 1.35, p
= .177. Based on this sample, a one-unit increase in “Post-purchase Convenience” does not have a significant effect on “Behavioural Intentions.” “Possession Convenience” significantly predicted “Behavioural Intentions”, B= 0.30, t(257) = 3.81, p < .001. This indicates that on average, a one-unit increase of “Possession Convenience” will increase the value of
“Behavioural Intentions” by 0.30 units. “Evaluation Convenience” did not significantly predict
“Behavioural Intentions”, B = -0.01, t(257) = -0.15, p = .884. Based on this sample, a one-unit increase in “Evaluation Convenience” does not have a significant effect on “Behavioural Intentions.” “Transaction Convenience” did not significantly predict “Behavioural Intentions,”
B = 0.14, t(257) = 1.92, p = .056. Based on this sample, a one-unit increase in “Transaction Convenience” does not have a significant effect on “Behavioural intentions.” “Search convenience” did not significantly predict “Behavioural Intentions,” B = 0.16, t(257) = 1.83, p
= .068. Based on this sample, a one-unit increase in “Search Convenience” does not have a significant effect on “Behavioural Intentions.” Table 7 summarizes the results of the regression model.
Table 7: Results for Linear Regression
Variable B SE 95% CI β t p
Access convenience 0.24 0.07 [0.11, 0.38] 0.23 3.65 < .001 Post-purchase convenience 0.10 0.07 [-0.05, 0.25] 0.08 1.35 .177 Possession convenience 0.30 0.08 [0.15, 0.46] 0.25 3.81 < .001 Evaluation convenience -0.01 0.07 [-0.15, 0.13] -0.01 -0.15 .884 Transaction convenience 0.14 0.07 [-0.00, 0.29] 0.13 1.92 .056 Search convenience 0.16 0.09 [-0.01, 0.33] 0.13 1.83 .068
Note. N = 264. 2 Note. Results: F(6,257) = 33.51, p < .001, R2 = 0.44
Unstandardized Regression Equation: Behavioral_intentions = 0.20 + 0.24*Access_convenience + 0.14*Transaction_convenience + 0.16*Search_convenience - 0.01*Evaluation_convenience + 0.10*Post_purchase_convenience + 0.30*Possession_convenience