**4.5 Inferential analysis**


In this study, the regression technique was used to test the relationship between the dependent and independent variables. The analysis is divided into two parts.

**Table 3.** *Reliability statistics (full data).*

*Upsurge of Online Shopping in Malaysia during COVID-19 Pandemic DOI: http://dx.doi.org/10.5772/intechopen.108049*


#### **Table 4.**

*Pearson's correlation coefficient analysis.*

A simple linear regression (SLR) is carried out to examine the relationship between the PU and PEOU, where PU is considered to be the dependent variable in this relationship. Then, multiple linear regression (MLR) is applied to identify the relationship between the independent variables (PU, PEOU, SN, PR and SI) and the dependent variable (BI).

#### **4.6 Multiple linear regression (MLR)**

As shown in **Table 5**, the R2 value is 0.636, implying that the five variables, PU, PEOU, SN, PR and SI, can explain 63.6% of the variance in BI, while the remaining 36.4% is explained by additional factors not included in this study.

As depicted in **Table 6**, the F-value is 68.947 and p-value is 0.000 at 5% significance level. This means that it is significant as the p-value is lower than 0.05. It demonstrates that the dependent variable (BI) has a substantial relationship with the five independent variables (PU, PEOU, SN, PR and SI).

From the **Table 7**, it can be seen that at 5% significance level, PEOU (p-value- = 0.000), PE (p-value = 0.000) and SI (p-value = 0.001) have significant and positive effects on the dependent variables (BI). This is due to their p-value are less than 0.05. However, SN (p-value = 0.376) and PR (p-value = 0.468) are found insignificantly



*b Predictors: (constant), SI, PEOU, SN, PR, PU.*

*c α = 0.05.*

#### **Table 6.**

*Analysis of variance (ANOVA) for all variables.*


#### **Table 7.**

*Multiple linear regression coefficient.*

associated with BI as their p-values are higher than 0.05. As a result, H2, H3 and H6 are statistically supported, but H4 and H5 are not supported.

Thus, the MLR equation is as follows:

BI ¼ 0*:*507 þ 0*:*284 PU þ 0*:*244 PEOU þ 0*:*049 SN þ 0*:*051 PR þ 0*:*257 SI

#### **4.7 Simple linear regression (SLR)**

The R2 value of 0.539, depicts that PEOU can explain 53.9% of the variance in PU, whereas the remaining 46.1% is explained by the other variables not examined in this study, as presented in **Table 8**.

The **Table 9** shows the F-value (1, 201) = 235.024, and p-value = 0.000 at 5% significance level, which indicates that the PEOU plays a significant role in shaping PU. These clearly shows the positive effect of the PEOU.

**Table 10** shows that PEOU (p-value = 0.000) has a significant and favorable influence on PU as the p-value is less than 0.05. Hence, H1 is supported.

To sum up, the SLR equation can be written as follow:

$$\text{PU} = \text{1.324} + \text{0.702 FeOU}$$


#### **Table 8.**

*Model summary for PEOU and PU.*


*Predictors: (constant), PEOU.*

*c α = 0.05.*

#### **Table 9.**

*Analysis of variance (ANOVA) for PEOU and PU.*


#### **Table 10.**

*Single linear regression coefficient.*

A summary of the hypothesis testing has been developed in **Table 11**. H1, H2, H3 and H6 are supported, whereas H4 and H5 are not supported.

#### **5. Discussion**

The findings suggest that consumers believe PEOU to be a key element in determining the PU of online shopping (p < 0.05). This result was confirmed by the prior studies which done by [19–21]. When the online shopping is simple to be accessed by users, they can attain their shopping objectives easily, and thus, increase the perceived usefulness on online shopping among them. Users may not consider e-shopping to be useful if they have to cope with the difficulties in using online shopping. This is because if they are busy handling the difficulties, they could not see


#### **Table 11.**

*Summary for hypotheses testing.*

the benefits of using it. To sum up, there is a positive relationship between PU and PEOU in online shopping during pandemic.

The result (p < 0.05) implied that PU significantly influence BI. Consumers' behavioral intention to utilize online shopping during a pandemic is profoundly influenced by PU. This finding is in line with the findings of a number of prior studies [25, 26]. By comparing the online purchase to traditional purchase from physical stores, online shopping enables users to reach the items faster and also more choices are available compared to physical stores [27]. The consumers are able to shop in their homes while maintaining social distancing during this pandemic. Hence, the perceived usefulness is established. On top of that, consumers have little opportunities to purchase items other than necessities, like clothing, stationery, etc. during the lockdown [53]. So, online shopping can be a good substitute for physical stores. Consequently, PU is a factor for consumers' BI on using online shopping during pandemic.

PEOU is also found to have positive impact on BI of using online shopping during Covid-19 pandemic (p < 0.05). This significance and positive relationship were evidenced by Verweijimeren [41]; Yadav et al. [30]; and Lisdayanti et al. [31] in their study. Youssef et al. [25] explained the ease of use was contributed by the languages available on the platform, the design of the platform and guidance video provided. In Malaysia, Shopee application, one of the top-ranking online shopping in Malaysia, also provided several languages in the application (i.e., Malay, English and Mandarin) and guidance video to use Shopee application was provided on YouTube too. Thus, users may consider online shopping was easy to use (such as adequate search support provided, provide relevant recommendations, etc.). When users found that online shopping was easy to navigate, they may not get frustrated when using online shopping. Instead, users may find it helpful to achieve their purchase objectives easily and thus, developed the intention to purchase online.

The findings (p > 0.05) indicate that SN is not considered a factor that influences the consumers' BI of using online shopping. This result was supported by several prior literatures [26, 53]. The study of Koch et al. [53] showed internal SN has no influence on customers' BI to adopt online shopping, whereas external SN have significant influence. The study defined internal SN as internal sources of social influence, such as family and

#### *Upsurge of Online Shopping in Malaysia during COVID-19 Pandemic DOI: http://dx.doi.org/10.5772/intechopen.108049*

friends, while external SN is described as external sources of social influence, such as mass media. The insignificant relationship is ascribed to the large Generation Z respondents in this study. Generation Z is people that born between 1997 and 2015, who born into new technology. They rely heavily on digital media. In this study, the questionnaire items design for subjective norm was related to internal SN. However, Generation Z is more easily influenced by mass media reports or expert opinions, especially in this epidemic. This is the reason why findings indicated the insignificant relationship between SN and BI in this study, as most of the respondents are from Generation Z. However, there are also some literatures that are not supporting this result [33, 34, 36].

The results (p > 0.05) also show the PR is not a factor that affects the consumers' BI in using online shopping during pandemic. Several studies also demonstrated the same results as this study [27, 26]. Gao et al. [53] found that the insignificant relationship because people may still be concerned about the potential risk of being infected when they purchased online, even though online shopping is the most suitable and safer alternative method for shopping in brick-and-mortar stores. They might worry that people who packaged and delivered their parcels was infected by virus and in this way, PR is not a factor to influence the consumers' BI to adopt online shopping during Covid-19. Therefore, the same result in this study would be due to the same reason as provided by Gao et al., where Malaysian online consumers may find the risk of Covid-19 virus being transmitted through parcel. However, this insignificant relationship result was contrasted by some prior research [31, 37, 38].

The results (p < 0.05) depict that SI is considered a determinant factor that influences consumers' BI of using online shopping. Many previous studies also recognized that SI plays an important role in the behavioral intention of using online shopping during Covid-19 pandemic [40–42]. As stated in Akar [33], Covid-19 has brought a big change across the world and resulted in instabilities in the society. These pandemic concerns can affect the consumers' online shopping intentions. The study stated that Covid-19 had an influence over the customer behavior in terms of frequency, necessity, payment method, price and products availability. During the MCO period in Malaysia, most of the retail stores were forced to stop operations. Thus, the alternative method available for consumers is to shop through online shopping. This has led consumers to switch to e-shopping, and thus, built the BI of using online shopping during Covid-19 crisis.

#### **5.1 Theoretical implications**

This study's main theoretical implication is that it adds value to the literature of technology dissemination, to be more specific, the online shopping research. This research gives a comprehensive overview of Malaysian consumers' perception towards online shopping throughout the pandemic. This study focused on studying the determinant factors of consumers' behavioral intention to use online shopping during the Covid-19 times. The researchers adopted the TAM model and added some variables, such as subjective norms, perceived risk and situational influences, aimed to fit the current situation. This can provide a deeper understanding about the online shopping in Malaysia and it is useful for future research.

#### **5.2 Practical implications**

Covid-19 hits Malaysia economy heavily, and online shopping is the only way to substitute the physical retail stores. Hence, this study is able to bring implication for Malaysian government agencies to further plan to improve the online shopping usage among the citizens. According to this study, consumers' BI to embrace online shopping during a pandemic is influenced by PU, PEOU and SI. From the insight of this finding, Malaysia government can provide supports to those SME companies and online shopping companies to enhance their online shopping platforms. The government can provide IT training especially to those SME owners who run business in traditional business model, who are not familiar with online shopping technology prior to this pandemic. The pandemic has also significantly affected the day-to-day business operations. Those sellers are still struggling with their businesses to minimize their losses during this difficult time. Accordingly, this study could benefit SME sellers to survive through this pandemic and to avoid suffering losses by adopting online shopping business model.

#### **5.3 Limitations**

Firstly, this is a cross-sectional study. The study result is only limited to a single point of time. However, the Covid-19 impact is still continuing and its influence on consumers is still ongoing and changing, so a longitudinal study is able to capture more accurate result. Secondly, due to time and cost constraint and also the pandemic, this study's data collecting technique was decided to be an online questionnaire. Sometimes, the participants may not read and answer the questions carefully. This might result in inaccurate data collection. In addition to this, the quantitative method indeed provides wide scope of data collection but less detailed, compared to qualitative method which provides narrower but more thorough responds. Apart from these, the sampling location for this study is limited to Malaysia, and more specifically only 10 States involved due to time constraint. Hence, this limits the generalizability of this study to other countries. Lastly, the convenience sampling technique was applied in this study on account of the limited time available. Whilst convenience sampling was easy to apply and participants are readily available, this sampling technique is lack of generalizability to the population as a whole and could lead to a biased result.

#### **6. Conclusion and recommendations**

This study presented a comprehensive understanding about the determinants that contributed to the up-surged online shopping in Malaysia during the pandemic. The study integrated TAM theory by using PU, PEOU, SN, PR and SI variables to provide useful insights on the consumers' BI on using online shopping. The results concluded that perceived usefulness, perceived ease of use and situational influences have significant influence on Malaysian consumers' behavioral intention to adopt online shopping during pandemic times. However, subjective norms and perceived risk have showed an insignificant relationship with consumers' behavioral intention. This research adopted the TAM model and added some variables, such as subjective norms, perceived risk and situational influences, aimed to fit the current situation. This can provide a deeper understanding about the online shopping in Malaysia and it is useful for future research. Furthermore, this study is useful for Malaysian governments and SME owners to gain an insight into the online shopping adoption determinants. Limitations and recommendations have been included in order to enhance future studies. Regarding recommendations, the first recommendation would be to carry out the longitudinal perspective for future research. This can analyze the changes in

### *Upsurge of Online Shopping in Malaysia during COVID-19 Pandemic DOI: http://dx.doi.org/10.5772/intechopen.108049*

respondents in the long term. Although investigating a single point in a decision event is allowed, the antecedent factors that lead to that tipping point is also important. Future research can also focus on changes in before and after the event. Future researchers can also consider to adopt multimethod quantitative if there is sufficient time and budget. By applying two or more quantitative methods, the drawbacks can be offset by each other's benefits. The next recommendation is related to the sampling location. Future research can conduct in the whole country, Malaysia. This can cover all the states in Malaysia, as citizen in different states may have different viewpoints and perception due to the different development of online shopping and logistics among the states. To address the limitation on convenience sampling technique, future researchers can gather large samples in order to strengthen the generalizations.
