Consumers’ Perspectives and Behaviors towards Corporate Social Responsibility—A Cross-Cultural Study

: This study measured consumers’ perspectives and behaviors with respect to corporate social responsibility (CSR). Speciﬁcally, we explored the components of CSR, including CSR aimed towards the environment, society, customers, employees, suppliers, and shareholders. We also examined the impact of CSR practices on brand attitude and purchase intention. The study surveyed 616 consumers across three locations, with detailed questionnaires in four languages. A total of 564 samples (186 from Vietnam, 189 from Indonesia, and 189 from Taiwan) qualiﬁed for data analysis. Additionally, statistics software including LISREL 8.8, STATISTICA 10, and MINITAB 19 were utilized to evaluate our hypotheses and construct a structural model. The results indicated that the consumers across the three areas were not concerned about CSR aimed towards shareholders, while all consumers considered CSR aimed towards themselves. Vietnamese customers prioritized a company’s care for its employees, while both Indonesian and Taiwanese consumers concentrated on the environment and society. In addition, suppliers’ beneﬁts and rights attracted Indonesians’ attention. Furthermore, CSR had a positive signiﬁcant impact on brand reputation in all three cultures. However, while CSR had a positive inﬂuence on customer purchase intention in Indonesia and Taiwan, it did not in Vietnam. we the to


Research Background
These days, many companies are attempting to humanize themselves. Specifically, they are endeavoring to behave like humans in order to win customers' hearts. For example, Howard Schultz-the founder of Starbucks-was inspired by his father's adversity. Schultz's father struggled to make ends meet and was underestimated by his employers; moreover, he was seriously injured in the workplace. The hard times experienced by his father prompted Schultz's motivation to genuinely care about his employees' lives and implement corporate social responsibility (CSR). He established a "Transformation Agenda" targeting the firms' employees [1]. Specifically, Starbucks not only remunerates its employees equally but also offers insurance and other benefits to all part time and full time staff. Additionally, Starbucks has implemented numerous initiatives focused on the environment and the community. Regarding the community, Starbucks operates community stores in collaboration with local nonprofit organizations. Through these endeavors, Starbucks has undertaken a host of activities to help the communities in which they are located. With respect to environmental responsibility, Starbucks has tried to reduce its waste by recycling and conserving water and energy. Furthermore, its CSR performance has earned a great deal of support from its customers and, as a result, has contributed to increased profits. After implementing these strategies, Starbucks bounced back financially, and by 2013 had even exceeded its previous revenue.
well [15] due to the experience of SARS in 2003 [16,17]. At the beginning of the pandemic, on 27 April 2020, Taiwan had only 429 cases and 6 deaths [15]. As of 26 December 2020, just 783 cases were confirmed on this island due to its efficient quarantine policies [18]. Although Taiwan encountered a severe wave of infections in May 2021 with about 15,000 cases by 15 July 2021 [19], the number of daily cases decreased to approximately 30 in July 2021 [15]. Updated in 22 December 2021, Taiwan recorded only 16,840 total cases, 850 total deaths, 705 total cases per million people, compared to 35,501 total cases per million people globally, 36 deaths per million people compared to 691 deaths per million people globally. In contrast, in Indonesia and Vietnam, the situation is much worse. Indonesia, a country with the fourth highest population, is likely to be severely affected by the pandemic [20]. In November 2020, the death rate from COVID-19 in Indonesia was the highest in Asia (8% in 2020) [21]. On 22 December 2021, Indonesia recorded 4,261,000 total cases, 144,034 total deaths, and 15,340 total cases per million people, 519 total deaths/million of the population (https://www.worldometers.info/coronavirus/, accessed on 22 December 2021). The explanation for this serious situation is that vaccination programs were rolled out slowly and most Indonesian people follow Islam which limits the consumption of goods-including vaccines-to only those that meet halal certification. Regarding Vietnam, the government controlled the COVID-19 pandemic well during 2020 due to effective quarantine policies [22]. Since April 2021, the most serious COVID-19 outbreak in Vietnam saw 1,588,000 total cases, 30,250 total deaths, and 16,105 total cases per million people, and 307 total deaths per million people, as updated on 22 December 2021 (https://www.worldometers.info/coronavirus/, accessed on 22 December 2021). Many enterprises have been struggling, and some have entered bankruptcy [23]. Obviously, a pandemic-such as COVID-19-is an unexpected event and can magnify the cost of CSR activities [24] and also amplify the sensitivity of customers. In this situation, from a company's perspective, whether it should take on its social responsibility or focus on profit has always been a widely debated issue. In addition, to what extent consumers would support companies that perform CSR remains unknown. According to prior research [25], culture affects CSR disclosures. Particularly, in Western countries, where individualism is embraced and the power gap is not high, CSR reports are prevalent. However, in Asian countries, which are comparatively conservative, the CSR disclosure rate is lower.

Research Purpose
Based on previous studies [26][27][28], this research aims to examine the components and their associated weights in different cultures. Although the concept of CSR components has existed for over a decade, they are still worth researching in the current COVID-19 pandemic-a sensitive time. Furthermore, this research proposes and examines how consumers perceive CSR as well as the influence of CSR on their attitude toward brand and consumer intention. We also collected demographic information for statistical analysis to increase the precision of data in this research. To carry out this study, we collected samples from consumers in Taiwan-a developed economy-as well as two developing economies-Indonesia and Vietnam-during the COVID-19 pandemic. The results analyze CSR components and customers' perception towards CSR, as well as gauge how CSR influences brand attitude and purchase intention. In the research, we make comparisons in three areas. Lastly, this research also makes theoretical and practical contributions.

Research Structure
Section 1 introduces CSR generally and particularly in Taiwan, Indonesia, and Vietnam. It also briefly mentions the purpose of this research and the study structure. Section 2 is the literature review-it defines CSR components, namely, CSR towards the environment, society, customers, employees, suppliers, and shareholders. It also investigates CSR's effects on customers' attitudes toward brands and their purchase intention. In this section, a model including nine hypotheses is also described. In Section 3, the sampling technique and the research methodology are demonstrated. Specifically, we show the way to distribute, collect, and analyze data. Section 4 presents the hypotheses testing results extracted from analyzing the data collected. Moreover, the results of the statistical analyses show the similarities and differences in the three locations. Section 5 includes the conclusion, discussion, limitations, and further studies.

CSR
The concept of CSR dates back to a long time ago. Especially, after World War II, CSR was adopted and gradually became prevalent [29]. Nowadays, CSR has received a great deal of attention from the public. Many companies issue CSR reports as a means of communicating their CSR performance. Public and company awareness leads to the growth of CSR initiatives [30]. There are many standards used as social reporting frameworks. Global Reporting Initiative (GRI) has become commonly used for reporting sustainability [31] since GRI reports are used to judge improvement year by year, and also facilitates making comparisons with their rivals [32]. It provides three standards, including economic, environmental, and social metrics [33]. The ESG metrics are based on three aspects (environmental, social, and governance); ESG disclosures can also help companies become more reliable and transparent and, then, improve their reputation [34]. Furthermore, ISO 26,000 is not certifiable but offers guidance on socially responsible practices to organizations as well as the public sector [35]. One of the most valuable frameworks is the United Nations Global Compact (UNGC), which correlates firms' strategies and activities with human rights, labor, environment, and anticorruption principles [30]. Regarding employees, Social Accountability International's SA8000 standard and FLA Workplace Code guide on maintaining social practices in the workplace [36].
Although many previous theories and models have studied CSR, to date, there is a scarcity of research models that examine how CSR affects brand attitude and purchase intention. In addition, previous research was conducted on just one specific area and did not examine the interrelations in areas that have different legal, economic, cultural, and social conditions. Moreover, previous research was carried out in normal situations. However, this study is conducted in a crisis, the COVID-19 pandemic, in which the cost of CSR activities [24] and the sensitivity of customers are prone to be high. Therefore, this study proposed the influence of CSR on purchase intention in Vietnam and Indonesia-emerging economies-and Taiwan-a developed economy.
Since the 1950s, there have been many definitions of CSR. In the study, CSR is defined as a set of principles that a company adopts voluntarily beyond legal requirements to be accountable toward the environment, society, customers, employees, suppliers, and shareholders [26,27]. Therefore, it improves quality of life and also contributes to sustainable development. According to [28], CSR is a seven-dimensional construct, which includes the environment, society, community, customers, employees, suppliers, and shareholders. However, we realize that there are some overlaps between the two domains-community and society-and we integrate the two domains into one. Therefore, in this study, CSR includes six subdomains.

CSR towards the Environment
Environmental sustainability and green business management have received a great deal of attention. The environment domain plays a crucially important role in CSR and environmental conservation is a duty of corporations and can be a strategic instruction for how enterprises achieve sustainable development [34]. CSR towards the environment is the combination of the awareness of environmental management and the philosophy of corporate operation to alleviate the effect of manufacturing on the environment [37]. CSR towards the environment has been mentioned widely but lacks quantitative data to evaluate the outcomes, while many frameworks, such as ESG, involve environmental criteria to gauge how well a company serves the environment. CSR towards the environment can be considered as enterprises minimizing energy consumption, using ecofriendly and sustainable materials, and having proper waste management. Moreover, these companies engage in constant efforts to preserve the environment, as well as invest in research and development related to environment protection [28,38,39]. This study assumes the existence of the positive relationship of CSR towards the environment and overall CSR. As a result, the following first hypothesis is developed: Hypothesis 1 (H1). CSR towards the environment has a positive impact on overall CSR.

CSR towards Society
Apart from environmental issues, recently, social issues have emerged. Society is described as the surroundings in which an enterprise deploys its CSR practices and activities. In general, society often expects the returns from a company to afford that company acceptance and legitimacy [40]. Instead of avoiding social issues, enterprises nowadays have taken on CSR to address those problems [41]. CSR towards society can be considered as an enterprise's contributions to community development related to job creation, especially for disabled people, economic development, as well as contributing to enhancing the life quality of the community [28,42]. An assumption is that there is a positive correlation between CSR towards society and overall CSR. Therefore, the hypothesis is developed: Hypothesis 2 (H2). CSR towards society has a positive impact on overall CSR.

CSR towards Customers
Customers are a substantial source of the value of a company [43]. If a company does not take responsibility for its customers, it will be unable to attract customers and generate revenue or profits. In the case that customers are unsatisfied with a company's products or behaviors, they can stop supporting or even boycott the company's products or services. The worst case could be that company is likely to lose profits or go bankrupt. Consequently, CSR towards customers is key in CSR strategy. CSR towards customers can be considered as an enterprise making honest claims to customers through advertising or marketing, providing safe products with high quality, fair prices [28,38,39]. This study speculates the positive link between CSR towards customers and overall CSR. Therefore, the following hypothesis is developed: Hypothesis 3 (H3). CSR towards customers has a positive impact on overall CSR.

CSR towards Employees
Employees are a vital resource of a company [44]. Employees' competence is necessary and can be considered a core value of an enterprise. In cases where employees are not satisfied with their jobs, work performance is affected. The quality and quantity of products and services are prone to be reduced. Taking on CSR towards employees is necessary for a company to maintain its business and improve its outcomes [45]. The foundations of CSR should focus on employees' wellbeing [46]. There are many frameworks, such as Social Accountability International's SA8000 standard, FLA Workplace Code, etc., which provide guidance on maintaining social practices in the workplace [36]. CSR towards employees can be considered as an enterprise paying adequate salaries to their employees [47], creating safe, decent working conditions, and providing professional development and promotion opportunities to their employees [48]. It is also worth mentioning that companies must treat their employees fairly (without region or gender discrimination or abuse) as well as offer their employees adequate medical insurance [28,38,39]. The following hypothesis is assumed: Hypothesis 4 (H4). CSR towards employees has a positive impact on overall CSR.

CSR towards Suppliers
There is an important link between suppliers and a company in a distribution chain and that connection is key in a company's business. The reason for this is that suppliers provide a company with materials with reasonable prices and quality. If a company does not fulfill its duties towards suppliers, the materials' quality may be unreliable. Consequently, its products and services can be affected. Therefore, embracing CSR towards suppliers is of importance for a company to run its business [49]. CSR towards suppliers can be considered as an enterprise providing fair terms and conditions for all suppliers. Furthermore, enterprises must communicate openly, honestly, and confidentially, as well as comply with contractual payment terms [28]. The following hypothesis is assumed: Hypothesis 5 (H5). CSR towards suppliers has a positive impact on overall CSR.

CSR towards Shareholders
Shareholders are individuals who own a company's shares. By investing in a company, shareholders gain benefits from improving the company's financial situation with a high share dividend in return. Besides investing in a company, shareholders can engage in the company's activities, such as consulting and other assistance related tasks. Without shareholders' engagement, a company cannot operate well. In fact, CSR towards shareholders is vital in a business-if a company does not take on the responsibility, it may struggle with financial issues [50]. CSR towards shareholders can be considered as an enterprise investing the capital of shareholders properly. Moreover, companies must respect their shareholders by communicating openly and honestly with them. Finally, companies have to guarantee sustainable development and long term success to the shareholders [28,38,47]. The following hypothesis is developed: Hypothesis 6 (H6). CSR towards shareholders has a positive impact on overall CSR.

Brand Attitude
Brand attitude is considered as the consumers' willingness to show a favorable or unfavorable reaction toward a certain brand [51]. An individual's attitude toward any brand is based on the information or knowledge he or she has gained through particular sources, such as their family, friends, networking, cultural, and global aspects [52]. On the other hand, together with engagement, those attitudes are founded through consumers' experiences [53]. Gradually, consumers form their attitude toward the brand: they may trust, like, or be loyal to the brand [54]. In this study, brand attitude refers to consumers' attitudes towards brands that are socially responsible. Based on firms' CSR engagement, customers may have a positive view of companies or brands [55][56][57]. According to previous research [58], CSR can be important for improving brand attitude. The following hypothesis is developed: Hypothesis 7 (H7). CSR has a positive influence on brand attitude.

Purchase Intention
Purchase intention is the inclination of consumers towards products or services. Specifically, after making a certain evaluation, consumers have an intention to buy products or use services [59]. Many factors influence consumers' intentions [60]. The final decision relies on consumers' preferences as well as external aspects [61]. In this study, purchase intentions refer to consumers' intentions to purchase products from companies taking on social responsibility [62]. Based on the theory of reasoned actions [63], personal attitudes and behaviors have a relationship. According to [64], in the Italian banking industry, CSR practices related to the environment, society, and ecosustainable projects are important; at the same time, customers believe that there is a link between CSR practices and their choice of banks. In the hospitality industry, during the COVID-19 pandemic, CSR practices encouraged customers to prepay services [65]. In the wine industry, CSR towards the environment receives a great deal of attention [66]. Therefore, we assume that purchase intention is influenced by CSR and brand attitude. Thus, we form the final two hypotheses:

Hypothesis 9 (H9). Brand attitude has positively impact on purchase intention.
The utility formulation can be conceptualized as: BR: brand attitude PI: purchase intention The conceptual model related to CSR, Brand attitude, and Purchase intention is demonstrated in Figure 1.

Collect Data and Questionnaire
A survey research method was conducted to obtain insights into CSR and its influence on brand attitude and purchase intention as well as test the proposed conceptual model. All question items of the questionnaire were composed to attain the most precise data to accept or reject the hypotheses in the proposed model. Multidimensional analysis results are needed; consequently, demographic information is collected. The research is exempt from the regulations of the Institutional Review Board (IRB) regarding the involvement of human subjects according to the Helsinki declaration. The reason is the research posed almost no risks to human subjects. Ethical principles were abided by including privacy and confidentiality, etc. [67]. The questionnaires contain information about the reason for conducting the survey, instructions on answering, and a statement on assuring their privacy and confidentiality. There are two parts to the questionnaires. The first part contains 40 questions that asks consumers about their perspective on CSR towards the environment, society, customers, employees, and suppliers, and overall CSR of a hypothetical company, as well as their brand attitude and purchase intention. The sec-

Collect Data and Questionnaire
A survey research method was conducted to obtain insights into CSR and its influence on brand attitude and purchase intention as well as test the proposed conceptual model. All question items of the questionnaire were composed to attain the most precise data to accept or reject the hypotheses in the proposed model. Multidimensional analysis results are needed; consequently, demographic information is collected. The research is exempt from the regulations of the Institutional Review Board (IRB) regarding the involvement of human subjects according to the Helsinki declaration. The reason is the research posed almost no risks to human subjects. Ethical principles were abided by including privacy and confidentiality, etc. [67]. The questionnaires contain information about the reason for conducting the survey, instructions on answering, and a statement on assuring their privacy and confidentiality. There are two parts to the questionnaires. The first part contains 40 questions that asks consumers about their perspective on CSR towards the environment, society, customers, employees, and suppliers, and overall CSR of a hypothetical company, as well as their brand attitude and purchase intention. The second part contains 10 questions about personal information. A 5-point Likert scale (totally disagree = 1 to totally agree = 5) is used to evaluate the variables [68]. The questionnaire was first composed in English and then translated into three languages, including Chinese, Indonesian, and Vietnamese. Therefore, respondents across three areas (Taiwan, Indonesia, and Vietnam) can understand the questionnaire holistically without any confusion during the answering process.
The questionnaires were distributed mainly via Google Forms to alleviate the accumulation of defective data that might otherwise impact the analysis results. Through the qualitative research results, as the preliminary survey, the observed variables have been adjusted and supplemented to match the research. To collect the data, a nonprobability convenience sampling was accepted with the purpose of theory testing [69]. The eligibility criteria are consumers over 16 years old residing in the three target countries with acceptable reading comprehension. Initially, 616 participants took part in the survey, however, 11 respondents were not from the three target areas, 9 samples were under 16 years old, and 32 samples were disqualified because survey takers missed some questions or answered questions carelessly (all answers are the same, etc.). After eliminating improper samples, the number of qualified samples was 564, including 186 respondents from Vietnam, 189 from Indonesia, and 189 from Taiwan. The reliability of research constructs can be assessed by using Cronbach's alpha values [70] in each dataset and the pooled data.

Descriptive Statistics
Descriptive statistics are used to demonstrate the basic traits of data gathered from empirical research in numerous ways [71]. It is considered an indicator for depicting quantitative information in a brief structure. This technique is remarkably useful for helping us to understand and demonstrate features or characteristics of a certain dataset by briefly summarizing the samples and their parameters. The most commonly used types are concentrated trend parameters, including mean, median, and weak, which are used in most -if not all-levels of mathematics and statistics. There are two basic types of measurement: centralized trend measurement and volatility or dispersion measurement. The former describes the center of a dataset whereas the latter demonstrates the dispersion of data. The demographic statistic is usually applied in qualitative and demographic variables [72]. In this study, STATISTICA 10 is applied since it offers a variety of integrated statistical information such as age, occupation, income, etc. [73].

Regression Analysis
The most primitive form of regression analysis is the method of least squares, which was first used scientifically in 1885 [74]. Moreover, the method was developed as the method of least squares procedures [75]. Regression analysis is defined as an analytical method to understand the relationship between a dependent variable and independent variables [76]. In addition, it can be used to estimate and predict the expected value of a dependent variable in a conceptual model. After performing regression analysis with the Minitab 19 software, we obtain results including the beta coefficient (β), p-value, t-value, and R 2 . Next, we can use these to evaluate whether or not the hypotheses should be accepted or rejected. Adjusted R 2 indicates how independent variables influence the dependent variable. In theory, if it is good (over 50%), the study is well justified [77]. In addition, p-value is the probability that the null hypothesis is supported. At large, a variable with a p-value not over 0.05 shows that the variable, within the model, is significant; while a p-value exceeding 0.05 implies that the variable should be removed [77]. Regarding the beta coefficient (β), which independent factor has the largest beta coefficient also has the greatest influence on the change in the dependent variable.

Linear Structural Relations (LISREL)
LISREL is a statistics software for modeling structural equations [78] and was used as a research tool in 1972 [79]. For this research, LISREL 8.8 software was used. The reason for applying it is that we can test the hypotheses in the proposed model easily and flexibly. Moreover, we can use this software in various fields thanks to the ability of this software to easily adapt according to the situation, especially in constructing models for estimating relationships between independent and dependent variables. The third reason is providing crucial indicators such as χ 2 /df, P, RMSEA, AGFI, PGFI, GFI, NFI, CFI, IFI, and SRMR, LISREL also helps us determine the goodness of fit test [80]. Importantly, the results from running LISREL 8.8 also show the model's structural validity, and the t-test helps determine any significant relationship between the two variables. In theory, NFI should be over 0.9 [81], GFI and CFI higher than 0.9 [82].

Demographic Statistics
Questionnaires were distributed over two months from 3 September 2021, to 19 November 2021, via an online survey-Google Forms in Vietnamese, Indonesian, Chinese, and English. According to previous research, monetary incentives are beneficial to enhance the response rate and completion rate [83]. Therefore, the authors used a lottery-style approach, which means we provided a USD 25 gift to the luckiest respondent by a random draw. Overall, we attained 616 responses. However, 52 participants were not qualified and 564 complied with the requirements of the survey. Table 1 depicts the demographic profile of the collected data. There are 186 Vietnamese, 189 Indonesian, and 189 Taiwanese respondents, which accounts for about 33% of each region. According to previous studies, urban vs rural differences are prone to obscure sophisticated influences [84]; measure of urban vs rural characteristics is considered in the research. The percentage of urban participants (78.01%; 440 samples) is four times higher than that for rural counterparts-21.99%; 124 samples. Moreover, 60.82% of the total participants are females while 39.18% are males. Only 11.88% are high school students, the rest completed higher education. Maritalwise, the number of single participants is dominant, with 436 samples (77.30%), while married individuals occupy 22.16% (125 samples). Regarding age, the proportion of individuals from 18 to 35 years old constitutes 82.62% (466 samples). The data show that respondents with the lowest income (less than USD 250 per month) take up the highest rate (37.23%; 210 samples), while those making the highest income (over USD 5000 per month) make up the smallest proportion (0.89%; 5 samples). Based on the database collected, Cronbach's alpha values were utilized to test the reliability of research constructs. Table 2 shows that all Cronbach's alpha indicators of the pooled data and each dataset surpass the generally agreed threshold of 0.8 [85]. Therefore, the research is reliable for each location and the cross locations.

Regression Analysis
Based on the database collected, regression analysis was carried out via Minitab 19 software. According to the proposed models, we performed a regression analysis based on Equations (1)-(3) with four datasets. The results demonstrated in Table 3 show that, in three regression models for Equations (1)-(3) with four datasets, R 2 and adjusted R 2 , are over 0.05, which means all models are well justified. Moreover, we observed that indexes such as R 2 and the adjusted R 2 of model 1 are higher than those of models 2 and 3 in the four datasets. Therefore, model 1 is more justified than models 2 and 3.   Table 4 shows the results of hypotheses testing. For the pooled data, all hypotheses are supported except H6 (β = −0.13 < 0). For each regional data, we find that H3, H7, and H9 are supported, while H6 is rejected in all three areas. Both Indonesian and Taiwanese data support H1, H2, and H8 while those hypotheses are rejected in Vietnam. However, regarding H4, Vietnamese data supports it (β = 0.62 > 0 and p-value ≤ 0.05) whereas both Indonesian and Taiwanese information rejected it. In contrast, H5 is supported with Indonesian data (β = 023 > 0 and p-value ≤ 0.05), while it is rejected in both Vietnam and Taiwan.

LISREL Testing
The covariance matrix depicting the linear relationship between constructs is demonstrated in Table 5 for the pooled Vietnamese, Indonesian, and Taiwanese data. It can be seen that all values in the matrix are positive, showing an increasing linear relationship between the constructs.

• Goodness of fit testing
In the study, goodness of fit testing was used to evaluate the covariance between the variables. Indexes are demonstrated in Table 6. The χ 2 /df of the pooled database and that of each region are significant. The explanation for this is that the data are collected from respondents from all walks of life in three locations (different areas, ages, educational background, marital status, jobs, and monthly income). This index is high and p < 0.000, which means that the one-factor model does not fit the data. However, it cannot negate the proposed models, since χ 2 is sensitive to the size and diversity of the samples [86]. Furthermore, the index is nonparametric statistics [87] and the possibility of finding differences is less than the parameter tests [88]. Consequently, the χ 2 test is not as suited as a general independent test [89]. Indices such as RMSEA, AGFI, and PGFI were not fit. However, indices, namely, GFI, NFI, CFI, and IFI are over 0.9, and SRMR is less than 0.08. Therefore, the model can be considered as accepted. • Hypotheses testing Table 7 shows hypotheses testing. With the pooled data, apart from H6, all hypotheses are supported. For each regional data, H3, H7, and H9 are accepted while H6 is rejected in all three areas. Moreover, Indonesian and Taiwanese data support H1, H2, and H8 whereas those hypotheses are refused in Vietnam. Regarding H4, Vietnamese data support it (Est. = 0.20 > 0 and |t-value| ≥ 1.96) while both Indonesian and Taiwanese information rejects it. Nonetheless, H5 is accepted with Indonesian data while it is rejected with both Vietnamese and Taiwanese data. The result of LISREL for the pooled data, Vietnamese data, Indonesian data, and Vietnamese data are depicted in Figures 2-5 respectively.

Comparison of the Regression Analysis and LISREL Model
Regression analysis is a reliable and powerful tool that can solve the relationship between independent and dependent variables as linear models. In regression models, R 2 and adjusted R 2 demonstrate structural validity. In this research, all indices of the pooled database or each area are over 0.5, particularly, some indices are over 0.7. Therefore, we can consider the model as reaching the fit. Regarding LISREL, the advantage is that it is highly flexible for concluding the relationship of variables. In this study, some indices are fitted, however, others, such as RMSEA, AGFI, and PGFI, are not suitable (RMSEA is much higher than the cut-off value = 0.05 and AGFI and PGFI are way lower than the requirement). Therefore, we could not claim that the model is fully fitted.
Comparing the two methods, we found that the regression model is better than the LISREL model due to its validity. However, both methods showed similarities in supporting or rejecting the hypotheses. Especially for the pooled data, we rejected H6 and accepted the rest. We also rejected two hypotheses with Indonesian data and rejected three with Taiwanese data. However, in Vietnam, four hypotheses were accepted while five hypotheses were rejected.

Comparison of the Regression Analysis and LISREL Model
Regression analysis is a reliable and powerful tool that can solve the relationship between independent and dependent variables as linear models. In regression models, R 2 and adjusted R 2 demonstrate structural validity. In this research, all indices of the pooled database or each area are over 0.5, particularly, some indices are over 0.7. Therefore, we can consider the model as reaching the fit. Regarding LISREL, the advantage is that it is highly flexible for concluding the relationship of variables. In this study, some indices are fitted, however, others, such as RMSEA, AGFI, and PGFI, are not suitable (RMSEA is much higher than the cut-off value = 0.05 and AGFI and PGFI are way lower than the requirement). Therefore, we could not claim that the model is fully fitted.
Comparing the two methods, we found that the regression model is better than the LISREL model due to its validity. However, both methods showed similarities in supporting or rejecting the hypotheses. Especially for the pooled data, we rejected H6 and accepted the rest. We also rejected two hypotheses with Indonesian data and rejected three with Taiwanese data. However, in Vietnam, four hypotheses were accepted while five hypotheses were rejected.

Comparison of the Regression Analysis and LISREL Model
Regression analysis is a reliable and powerful tool that can solve the relationship between independent and dependent variables as linear models. In regression models, R 2 and adjusted R 2 demonstrate structural validity. In this research, all indices of the pooled database or each area are over 0.5, particularly, some indices are over 0.7. Therefore, we can consider the model as reaching the fit. Regarding LISREL, the advantage is that it is highly flexible for concluding the relationship of variables. In this study, some indices are fitted, however, others, such as RMSEA, AGFI, and PGFI, are not suitable (RMSEA is much higher than the cut-off value = 0.05 and AGFI and PGFI are way lower than the requirement). Therefore, we could not claim that the model is fully fitted.
Comparing the two methods, we found that the regression model is better than the LISREL model due to its validity. However, both methods showed similarities in supporting or rejecting the hypotheses. Especially for the pooled data, we rejected H6 and accepted the rest. We also rejected two hypotheses with Indonesian data and rejected three with Taiwanese data. However, in Vietnam, four hypotheses were accepted while five hypotheses were rejected.

Summary of Results
Overall, for pooled data, CSR towards shareholders can be ignored while CSR towards other issues positively influences CSR. In particular, CSR towards customers is the most important factor. Furthermore, CSR significantly affects brand attitude and purchase intention. Moreover, the influence of CSR on brand attitude is higher than it is on purchase intention. Brand attitude significantly positively influences purchase intention.
Observing the Vietnamese data, CSR towards employees is the most important factor, followed by CSR towards customers, while other issues can be ignored. The result shows that CSR significantly influences brand attitude but does not affect purchase intention. Brand attitude still has a significant positive impact on purchase intention.
Regarding the Indonesian data, CSR towards customers is the most important factor, followed by the environment and society, while other factors can be neglected. CSR also has a positive impact on brand attitude and purchase intention (the impact of CSR on the former is much higher than on the latter). Furthermore, purchase intention is still influenced by brand attitude and the influence is higher than CSR's influence on purchase intention).
For the Taiwanese data, CSR towards customers is the most significant influence on CSR, followed by CSR towards society and the environment, whereas other aspects related to employees, suppliers, or shareholders can be ignored. Moreover, CSR has a significant impact on brand attitude. The influence of brand attitude is higher than the effect of CSR on purchase intention.

The Correlation across All Areas
Regarding the pooled data across all three areas, the results showed that some aspectsnamely, the environment, society, customers, employees, and suppliers-have a positive impact on CSR. The explanation lies in the fact that, since the COVID-19 pandemic, customer perspectives have changed significantly. Specifically, for societal issues, employees are given a great deal of consideration. Additionally, together with the increasing awareness of the environment, issues such as environmental deterioration or natural disasters are becoming, in general, a pressing concern across the world [90]-particularly in Vietnam, Indonesia, and Taiwan. However, from the customers' standpoint, the conflict between shareholders' benefits and customers still exists [91]. Implementing CSR leads to an increase in company costs and risks, together with a decrease in shareholders' benefits [92]. In this regard, taking on responsibility towards shareholders has no impact on CSR; this is in contrast to prior research. In terms of brand attitude, CSR has a significantly positive impact on it. In other words, customers are likely to exhibit favorable feelings towards brands that engage in social responsibility. Additionally, the results depict that there is a positive correlation between brand attitude and purchase intent.

The Correlation per Area
The effect of the environment on CSR varies across the three locations. Indonesian customers perceive the environment positively influences CSR, and its effect is higher than in Taiwan. However, it has no positive effect on CSR in Vietnam. In the Quran-the equivalent of the Christian Bible-Earth is mentioned over 400 times and Muslims are called to save it [93]. According to Islamic beliefs, environmental issues are seriously considered [94]. In Indonesia-an Islamic country-in the 1960s, the Suharto regime concentrated on economic growth together with environmental efforts [95]. Therefore, Indonesian people are likely to have a much higher awareness of the environment. Previous research shows that the concept of environmental sustainability bears resemblance to Taoist (Daoist) principles [96]. In Taiwan, many people are Taoists [97]; issues relating to the environment, therefore, are pervasive throughout the culture. Since 1992, Taiwan has progressed positively towards environmental sustainability [98]. As a result, Taiwanese people are paying much attention to the environment. Nonetheless, in Vietnam, consumers' awareness of environmental and social issues is still limited [99]. As mentioned in previous studies [100,101], these days, economic activities are in opposition to the socialist ideology declared by the party state, which has failed the environment and society [8]. In contrast, in Taiwan and Indonesia, social issues have caught the public's attention.
From the customer's perspective, their rights and benefits are crucially important. Consequently, fulfilling duties towards customers positively influence CSR in three areas.
A new point of this research that is different from previous CSR research is that CSR towards shareholders does not impact CSR. This is because customers tend to believe that there is a conflict between shareholders' values and customers' benefits [91]. Regarding employees' benefits, in Vietnam, employees are positively affected by CSR while they are not in the other two areas. On the one hand, Vietnamese culture has been affected by Buddhism, Confucianism, and Taoism (Daoism), therefore, the centrality of nhan (humanity, humanness) plays an important role in moral values [8]. On the other hand, in Taiwan, nowadays, the COVID-19 pandemic has been severe, leading to job losses, and many employees continue to struggle from the pandemic's fallout. Therefore, employee related issues are taken into citizens' consideration. Regarding Indonesia, its economy is the largest in Southeast Asia and the country is an emerging global economy. Over the past decade, the Indonesian economy has developed rapidly and, by 2030, is likely to become the seventh largest economy in the world [102]. It is reasonable to postulate that, in Indonesia, issues related to the economy or supply chain are worth considering. CSR towards suppliers, therefore, has a positive impact on CSR.
Regarding the relationship between CSR and brand attitude, in three areas, CSR has a positive influence on brand attitude. In particular, the impact of CSR in Taiwan is the greatest, followed by Indonesia, and Vietnam. This means that Taiwanese consumers display their affinity towards CSR brands more than Indonesian and Vietnamese consumers. Moreover, when it comes to purchase intention, CSR takes effect in Indonesia and Taiwan. Specifically, Indonesian consumers are influenced by CSR more than their Taiwanese counterparts. Nevertheless, Vietnamese customers are not affected. Regarding the link between brand attitude and purchase intention, three areas experience positive relations.

Theoretical Contribution
The academic contributions of this research are as follows. Firstly, based on stakeholder theory, this research assessed customers' perspectives on overall CSR and in each domain. In addition, this work analyzed CSR components and ascertained the influences of the components, including the environment, society, customers, employees, and suppliers, on CSR. Moreover, the study showed that, in the three areas during the COVID-19 pandemic, taking on responsibility towards shareholders does not influence CSR-an obvious departure from prior studies. This could be explained by the fact that, during the COVID-19 pandemic, customers focused more on their benefits and other issues instead of shareholders' interest. Secondly, regarding the applied theory of reasoned actions [63], this study constructed a model of purchase intention based on CSR and brand attitude. The results showed that, with the pooled data, CSR positively influences brand attitude and purchase intention; additionally, brand attitude also affects customers' purchasing intent. Thirdly, the survey was conducted across three different locations with different cultures and economic statuses. This provides a comprehensive picture of customers' perspectives on CSR to brand attitude and purchase intent during the COVID-19 pandemic. In particular, Vietnamese people focus more on human issues, such as CSR towards customers and employees. Furthermore, in Vietnam-a communist country-CSR does not have a positive effect on purchase intention. However, apart from CSR towards customers, Taiwanese and Indonesian citizens pay attention to environmental and social issues.

Managerial Implications
Evaluating customers' perceptions of CSR allows companies to measure customers' awareness level of their CSR practices [28], and the effect of CSR on brand attitude and purchase intention. This research can be applied to launch improved strategies for businesses. Regarding the importance of CSR, in general, companies should pay more attention to CSR practices to enhance brand attitude and purchase intention in crises such as the current COVID-19 pandemic. However, in Vietnam, the COVID-19 pandemic is likely to amplify customers' perspectives on prices, therefore, companies should weigh the costs and benefits associated with taking on CSR practices. Focusing on each area, it can be seen that Vietnamese citizens are less concerned with human related issues. Therefore, businesses should allocate their budgets towards CSR to address these problems in order to enhance their public image. However, in Taiwan and Indonesia, companies should focus on protecting the environment and improving society. These days, the concept of sustainable cities has received a great deal of attention from the public. Sustainable cities were mentioned in The United Nations' 2030 Agenda [103]. Companies can fulfill their responsibility in developing the environment and society via their CSR practices. Therefore, companies can collaborate with governments to establish a CSR scheme contributing to the development of urban infrastructure and sustainable cities [104]

Limitations and Further Research
This study has several limitations. Firstly, the research did not consider industry characteristics. CSR practices vary, depending on the industry to which a company belongs. Therefore, further research should be conducted according to different sectors. Secondly, the sample size for the pooled data (n = 564) is acceptable [105]. However, the research was conducted across three countries, therefore, the validation of the sample size of each region should be considered. It can be seen that the sample size of each location (about 188) was quite small and should be increased. Moreover, the study used cross-cultural data, however, only investigated individuals in three areas across Asia. Moreover, the sample size in each area is quite small and is therefore not representative. As a result, the results may not be used across other cultures, areas, and/or continents. Further research should be carried in different areas on different continents considering different cultural, economic, and legal settings. Furthermore, most respondents were rather young. In the future, sampling procedures should be considered in a strict manner. Third, the investigation took place in a short period during the pandemic, the results cannot be applied in another context (pre-and postpandemic). The difference in different methodology can be used to make comparisons over time [106]. Further research could apply this method to examine the variation in customers' perspectives towards CSR and its effect on brand attitude and purchase intent before, during, and after crises. Finally, when analyzing the effect of CSR on purchase intention, this research did not incorporate other factors related to customers and products. In the future, those factors should be integrated into the model. Institutional Review Board Statement: Ethical review and approval were waived for this study because they were not required by the university requirements.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: Not applicable.