
Skeleton-based action recognition is an important but challenging task in the study of video understanding and human-computer interaction. However, existing methods suffer from two deficiencies. On the one hand, most methods usually involve manually designed convolution kernel which cannot capture spatial-temporal joint dependencies of complex regions. On the other hand, some methods just use the self-attention mechanism, ignoring its theoretical explanation. In this paper, we proposed a unified spatio-temporal graph convolutional network with a self-attention mechanism (SA-GCN) for low-quality motion video data with fixed viewing angle. SA-GCN can extract features efficiently by learning weights between joint points of different scales. Specifically, the proposed self-attention mechanism is end-to-end with mapping strategy for different nodes, which not only characterizes the multi-scale dependencies of joints, but also integrates the structural features of the graph and an ability of self-learning fusion features. Moreover, the attention mechanism proposed in this paper can be theoretically explained by GCN to some extent, which is usually not considered in most existing models. Extensive experiments on two widely used datasets, NTU-60 RGB+D and NTU-120 RGB+D, demonstrated that SA-GCN significantly outperforms a series of existing mainstream approaches in terms of accuracy.
Citation: Min Li, Ke Chen, Yunqing Bai, Jihong Pei. Skeleton action recognition via graph convolutional network with self-attention module[J]. Electronic Research Archive, 2024, 32(4): 2848-2864. doi: 10.3934/era.2024129
[1] | Hakan KURT, Xuhui Peng . Does corporate social performance lead to better financial performance? Evidence from Turkey. Green Finance, 2021, 3(4): 464-482. doi: 10.3934/GF.2021021 |
[2] | Valentinas Navickas, Rima Kontautiene, Jurgita Stravinskiene, Yuriy Bilan . Paradigm shift in the concept of corporate social responsibility: COVID-19. Green Finance, 2021, 3(2): 138-152. doi: 10.3934/GF.2021008 |
[3] | Siming Liu, Xiaoyan Shen, Tianpei Jiang, Pierre Failler . Impacts of the financialization of manufacturing enterprises on total factor productivity: empirical examination from China's listed companies. Green Finance, 2021, 3(1): 59-89. doi: 10.3934/GF.2021005 |
[4] | Giacomo Boesso, Barbara Fryzel, Marco Ghitti . Aesthetics of CSR communication and perception of ethical leadership, impact on purchase intention in high and low rank CSR firms. Green Finance, 2025, 7(1): 175-199. doi: 10.3934/GF.2025007 |
[5] | Karen Paul . Why corporate social responsibility should be recognized as an integral stream of international corporate governance. Green Finance, 2024, 6(2): 348-362. doi: 10.3934/GF.2024013 |
[6] | Abhisek Mahanta, Naresh Chandra Sahu, Pradeep Kumar Behera, Pushp Kumar . Variations in financial performance of firms with ESG integration in business: The mediating role of corporate efficiency using DEA. Green Finance, 2024, 6(3): 518-562. doi: 10.3934/GF.2024020 |
[7] | Min Hong, Benjamin Drakeford, Kexian Zhang . The impact of mandatory CSR disclosure on green innovation: evidence from China. Green Finance, 2020, 2(3): 302-322. doi: 10.3934/GF.2020017 |
[8] | Gonzalo Maldonado-Guzmán . Green innovation mediates between financial innovation and business sustainability? Proof in the mexican manufacturing industry. Green Finance, 2024, 6(3): 563-584. doi: 10.3934/GF.2024021 |
[9] | Manrui Xu, Khaldoon Albitar, Zhenghui Li . Does corporate financialization affect EVA? Early evidence from China. Green Finance, 2020, 2(4): 392-408. doi: 10.3934/GF.2020021 |
[10] | Óscar Suárez-Fernández, José Manuel Maside-Sanfiz, Mª Celia López-Penabad, Mohammad Omar Alzghoul . Do diversity & inclusion of human capital affect ecoefficiency? Evidence for the energy sector. Green Finance, 2024, 6(3): 430-456. doi: 10.3934/GF.2024017 |
Skeleton-based action recognition is an important but challenging task in the study of video understanding and human-computer interaction. However, existing methods suffer from two deficiencies. On the one hand, most methods usually involve manually designed convolution kernel which cannot capture spatial-temporal joint dependencies of complex regions. On the other hand, some methods just use the self-attention mechanism, ignoring its theoretical explanation. In this paper, we proposed a unified spatio-temporal graph convolutional network with a self-attention mechanism (SA-GCN) for low-quality motion video data with fixed viewing angle. SA-GCN can extract features efficiently by learning weights between joint points of different scales. Specifically, the proposed self-attention mechanism is end-to-end with mapping strategy for different nodes, which not only characterizes the multi-scale dependencies of joints, but also integrates the structural features of the graph and an ability of self-learning fusion features. Moreover, the attention mechanism proposed in this paper can be theoretically explained by GCN to some extent, which is usually not considered in most existing models. Extensive experiments on two widely used datasets, NTU-60 RGB+D and NTU-120 RGB+D, demonstrated that SA-GCN significantly outperforms a series of existing mainstream approaches in terms of accuracy.
The concept of "corporate social responsibility" (CSR) refers to the voluntarily adopted steps and activities by businesses to consider and incorporate social and environmental concerns into their business operations and relationships with stakeholders. (Aguilera et al., 2022; Carroll & Shabana, 2021). Beyond merely adhering to the law, it also entails activities like philanthropy, community involvement, sustainability promotion and ethical corporate conduct (Carroll & Shabana, 2021; Brinkmann et al., 2020). In the past few years, people have paid a lot of attention to CSR in both academic research and how companies do things as corporate practices (Zhao et al., 2023; Chih, 2023; Low, 2016). The several facets of CSR and how it affects enterprises, society and the environment have been studied by academics and researchers (McGee, 2021; Kim, 2022; Ghardallou, 2022; Rahim et al., 2011; Amran & Siti-Nabiha, 2009; Abu-Baker & Naser, 2000; Tsang, 1998). Bangladesh still has a general trend toward CSR schemes (Belal, 2001; Imam, 2000). Nevertheless, only a few companies had taken it seriously to encourage Bangladeshi governments to adopt CSR (Hossain et al., 2019; Masud & Hossain, 2019). Another study conducted by Masud and Ferdous (2016) reported that, with particular regard to CSR project implementation, the Bangladeshi companies still have to meet international standards. A study by Dobers and Halme (2009) suggested CSR is very important for enhancing business performance. CSR was recently adopted by publicly listed enterprises and private companies in Bangladesh, which has received great attention from the Government (Masud & Islam, 2018; Masud et al., 2017 and Goi & Yong, 2009).
Unfortunately, this area has not yet been widely covered in Bangladesh, despite this finding (Masud, 2018). The connection between CSR and corporate success has been investigated in earlier research. The research findings revealed various impacts on businesses' and CSR's performances. CSR's performance findings appear inconsistent, as the studies of Barnett (2007) and Dentchev (2004) found that CSR and corporate performance have a positive relationship. The absence of a noteworthy connection between CSR and company performance has been evident in several studies, including the findings from the study conducted by Ahamed et al. (2014) suggest that corporate social responsibility (CSR) does not lead to an improvement in firm financial performance.
To date, the main focus has been on corporate social responsibility (CSR) (McWilliams & Siegel, 2001). According to a finding, 55% of management agrees that sustainable management helps their firms develop a positive reputation, and 76% of management believes CSR positively impacts long-term shareholder value (Barnett et al., 2006; Eberl & Schwaiger, 2005). The implementation of the corporate social responsibility (CSR) strategy has helped businesses combat stakeholder competition and boost their competitive edge (Torugsa et al., 2012) and excellent performance (Lynch-Wood et al., 2009; Torugsa et al., 2012). Various results have been found between corporate performance and CSR: Mishra and Suar (2010), Mugisa (2011), Handayani et al. (2017), Babalola (2012). Cheng et al. (2014) claimed that corporate social responsibility (CSR) and the performance of the corporation are positively related to each other. According to them, corporate social responsibility (CSR) initiatives include obtaining outside perspectives on the company. Julian and Ofori-dankwa (2013) found that enhancing a company's insubstantial assets can play a mediating role to promote the effectiveness of CSR. In contrast, Olowokudejo et al. (2011), Handayani et al. (2017) and Okwemba et al. (2014) emphasized that CSR and performance of the company have no connection. Moreover, Balmer et al. (2011) confirmed that a company, by ignoring CSR, hardly earns CSR performance benefit. In the past decades, there has been a growing interest in business and academic research in corporate social responsibility (CSR). It was found that, concerning the correlation between CSR and firm performance, a large number of studies have almost exclusively focused on developed countries (Rettab et al., 2009) and large firms (Saeidi et al., 2015). Less attention is given to those organizations which are small and medium in terms of size. The limited number of studies on the relationship between CSR and company success from the perspective of developing countries is thus the most noticeable gap in the literature (Moore & Spence, 2006). A literature review reveals that there is hardly any study that has investigated the impact of CSR on the financial performance of manufacturing companies with a mediating function of corporate reputation in Bangladesh, especially in small manufacturing enterprises.
This study explores the link between CSR and corporate performance with a mediating function of corporate credibility. Bangladesh was chosen for this report because of limited research in developing countries. A study from Bangladesh might also help to understand CSR findings in the sense of developing countries. In the analysis of the correlation between CSR and firm performance, we concentrate on small-scale production companies. CSR studies in Bangladesh have focused largely on major companies but not on small and medium-sized companies, and they have not connected CSR to the market. To fulfill the purpose this report, following question of analysis is put forward: How does corporate reputation mediate the effect of CSR on financial performance in small manufacturing companies in Bangladesh over time? This study is important for business people, policymakers and researchers who want to understand CSR-specific practices that affect business performance and reputation. The study also adds another dimension to the limited literature to explore the impact of CSR on corporate performance in Bangladesh.
The following section discusses current research on CSR in small-scale manufacturing and CSR in developing countries. Then, we present literature about the performance relation of CSR in companies and then establish hypotheses for the analysis. Then, we discuss the methods of analysis, data collection processes and measures used in the study and present the results. We conclude the paper by reviewing findings, drawing conclusions and outlining upcoming research topics.
The definition of CSR clarifies the roles and responsibilities of corporations toward society (Maalim et al., 2023; Vilanova et al., 2009). In earlier studies, the definition of CSR has been developed based on various perspectives such as business ethics (Solomon, 1992), social behavior (Carroll, 1979), business people (Waddock, 2000; Hutton et al., 2001), management of stakeholders (Donaldson & Preston, 1995; Murillo & Lozano, 2006) and corporate rules and regulations (Sacconi, 2011). The concept of "doing well by doing well" illustrates a development plan to support CSR. This is because businesses that perform corporate social responsibilities enjoy various incentives, such as favorable financial performance (Orlitzky, 2013), lower financial costs (Godfrey, 2005), higher credibility (Turban & Greening, 1997) and better financial exposure (Cheng et al., 2014). Despite the diversity of opinions, the CSR framework of Carroll (1979) mainly concentrates on academic evidence, which is universally acknowledged. Mitnick et al. (2023), Ortiz-Martínez et al. (2023), Maignan and Ferrell (2000) and Maignan et al. (1999) demonstrated a variety of CSR functions and created a four-dimensional CSR-measurement system based on Carroll's framework:
● Economic Citizenship: The organization, which must meet the needs of consumers, must become competitive and efficient (García-Rosell et al., 2023).
● Legal Citizenship: Organizations must carry out their functions within the legal framework (Silva Junior et al., 2023).
● Ethical Citizenship: Established ethical norms in society ought to be upheld (Frerichs and Teichert, 2023).
● Discretionary Citizenship: An organization performs the functions in a way that supports others and helps society as a whole (Kusakci and Bushera, 2023).
CSR has recently become one of the key trends to create the credibility and image of a company (Palacios-Manzano et al., 2021; Santos-Jaén et al., 2021; Masud and Hossain, 2019). CSR is required as a moral and ethical responsibility of the organization for manufacturing companies whose activities require input from the broader community, be it as the supply of raw materials, work or the target markets (Kiliç et al., 2015; Khan, 2010; Friedman & Miles, 2001). Manufacturing practices are shown to actively contribute to air and water pollution, environmental damage and disruption (Ingram & Frazier, 1980). Today, given that technological improvements have changed the climate, companies are making numerous efforts to enhance competitiveness through change and innovation (Gimeno-Arias et al., 2021; León-Gómez et al., 2022; Becerra-Vicario et al., 2023; Murcia & Souza, 2009). Environmental management includes the role of environmental management in CSR activities of the manufacturing industry (Farooq et al., 2015). With this business environment in mind, CSR is seen as necessary to enable businesses to satisfy the demands of changing times and achieve sustainable growth.
In this analysis, the first impact of CSR is financial performance. Although some studies have negative or neutral results (Li et al., 2023; Adamkaite et al., 2023; AlAjmi et al., 2023; Al-Sfan, 2023), overall, the current literature suggests that CSR has a positive impact on corporate performance. For instance, some authors found that CSR has a positive relationship with business performance (Reisinger, 2023; Chen, 2023; Chen et al., 2023; Ahamed et al., 2014; Islam et al., 2012; Rettab et al., 2009; Saeidi, et al., 2015; Cochran & Wood, 1984; Johnson & Greening, 1999; Preston & O'Bannon, 1997). In contrast, empirical findings by Boyle et al. (1997) and Wright and Ferris (1997) found negative impact of CSR on the performance of the corporations. Moreover, Iqbal et al. (2012) found a neutral impact of CSR on the corporate performance. Pan et al. (2014) observed that corporate financial success has essential positive relationships with workplace engagement and environmental responsibility. Lee et al. (2016) suggested that small and medium-sized enterprises (SMEs) are encouraged to pursue CSR operations when they see a client gain and interest. For instance, Ahamed et al. (2014) pointed out that the significance of the timeframe when researching CSR ties to company performance should be considered. Most studies found that CSR and success have a positive correlation (Orlitzky, 2008; Orlitzky, 2013; Waddock & Graves, 1997; Aupperle & Van Pham, 1989; Zheng et al., 2014; Maignan, 2001; McWilliams & Siegel, 2000). Additionally, CSR significantly contributes to lowering business risks (Fauzi & Idris, 2009; Orlitzky & Benjamin, 2001). Additionally, numerous findings suggest how CSR raises financial performance (Frooman, 1997). As a result, research should summarize that CSR improves businesses' financial performance. The following hypothesis is therefore suggested:
H1: FP of a manufacturing firm is positively influenced by CSR practices.
In this study, the second effect of CSR is the reputation of the firm. Corporate reputation is the "collective representation of the past actions and prospects of an organization describing how key providers of resources interpret the initiation of an organization (Nardella et al., 2023) and evaluate its ability (Mohd Sofian et al., 2023) to deliver valued results" (Fombrun & Rindova, 1996). Hur et al. (2014) stated that corporate reputation is "the overall impression of the group of stakeholders' perception." To sum up, CR encompasses an overall assessment of the perception of the organization formed from the subjective perceptions and experiences of stakeholders (Nasser et al., 2023; Whetten & Mackey 2002). Noteworthy, researchers have conducted a study to explore the relationship between financial performance, management quality, and CSR (Das et al., 2023; Madueno et al., 2016). CSR activities influence the marketing efforts of the company, including corporate communication, branding and the establishment of reputation (Salam and Jahed, 2023; Febra et al., 2023). According to Morimoto et al. (2005), the concept of CR includes people's perception of an organization, which arises due to interpersonal communications and advertising information (Santos-Jaén et al., 2022). Companies with a good reputation can more readily gain access to capital markets, entice potential investors, charge higher prices and boost their creditworthiness (Aggarwal and Saxena, 2023; Dwertmann, et al., 2023; Fombrun & Rindova, 1996). Firms can, therefore, make an effort by using corporate social reporting to influence their reputation (Javed et al., 2020; Liu et al., 2019; Ali et al., 2020; Pérez 2015). Good reputation gives a competitive edge, which enables a company to set up premium pricing for its product and services (Cabrera-Luján et al., 2023; Fombrun & Rindova, 1996). According to Black and Khanna (2007), organizations make efforts to improve their reputation through utilizing resources and expect to perform well. Positive reputation impacts the choice of suppliers (Bashir, 2022; Galbreath & Shum, 2012) through ensuring quality inputs, which eventually leads to maximum return (Roberts and Dowling 2002). According to Little and Little (2000), well known corporations enjoy decent P/E ratios as a result of their CSR involvement. Corporate reputation is a type of competitive advantage which cannot be simulated but can lead to greater performance (Fombrun & Rindova, 1996). In addition, the research conducted by Yan et al. (2022) reveals a noteworthy association between corporate social responsibility (CSR), organizational trust, and corporate reputation, highlighting their combined influence on fostering sustainable performance. Based on the study conducted by Le (2022) that the perception of a positive corporate image, a strong corporate reputation, and enhanced customer loyalty serve as mediators in transforming CSR efforts into improved performance outcomes for small and medium-sized enterprises (SMEs). On the other hand, Raj and Subramani (2022) found that employer branding plays a mediating role in the relationship between corporate social responsibility (CSR) and corporate reputation. For example, Greening and Turban (2000) established that engagement in CSR programs leads to the company's better reputation in society, providing a competitive advantage. Increased credibility ultimately leads to better company performance over the long term (Aranguren Gómez and Maldonado García, 2022; Baldarelli & Gigli, 2014; Bebbington et al., 2008). Additionally, it was noted by Jiang et al. (2022), Munasinghe and Malkumari (2012), Cox et al. (2004), Mankelow & Quazi (2007), and Graafland & Mazereeuw-Van der Duijn Schouten (2012) that SMEs are motivated by CSR enhancing their reputation, staff morale and financial performance. In the conceptual model, CSR is a prediction that affects FP and CR directly. Further, CR will impact company performance. We suggest the following hypothesis based on the above statements:
H2: CSR practices will have a direct impact on the CR of the manufacturing firm.
H3: CR of a manufacturing firm will have a direct impact on financial performance.
H4: CR mediates the relationship between CSR and FP.
Corporate social responsibility (CSR) is assessed utilizing twenty-four objects established initially by Michael Jantzi Research Associates, Inc., and used by Fauzi et al. (2007). Measurements of CSR are taken from Hossain et al. (2019) and Masud and Hossain (2019). Corporate reputation (CR) has been assessed by four items of the questions of Helm (2007) and Trotta and Cavallaro (2012). Four elements of questions given by Fauzi et al. (2007) are used to measure financial success. In this study, the effect of CSR is first observed on FP, a dependent variable, and second, CSR influence is calculated on CR, that is, the second dependent variable. There are two dependent variables. CSR, FP and CR are included in Appendix A for our measurements. This survey focused on companies that have significant contribution in CSR. A self-administered questionnaire technique was used to collect the data: 360 respondents (who were from junior, middle and top management) were given the developed questionaries, and 300 responses were obtained, which represents an 83.3% response rate. The respondents were from a population (manufacturing firms) that had already invested in CSR. The developed questionnaire was divided into four parts. Part A was made to get demographic information from the respondents, and CSR practices were measured with the support of part B. Respondents were asked to rate their company performance in part C, and part D was developed to measure the corporate reputation. The five-point Likert scale, which ranges from "strongly agree" to "strongly disagree, " was used to develop each item. Statistical Package for the Social Sciences (SPSS) and Analysis of Moment Structures (AMOS) were used to analyze the collected data. Through AMOS, structural equation modeling was carried out. This is due to the fact that SEM is considered good enough to develop the theoretical models and test the hypotheses.
The information about demographic characteristics of the respondents can be seen in Table 1. People who hold managerial roles (junior, middle and senior) overseeing the CSR functions in their respective organizations made up the sample for the study. The sample contains 300 respondents, and among them, 94.33% are male and 5.67% are female. Also, 80% of the respondents were relatively young, ranging 21–39 years of age, and the other 20% were 40 years and above. The sample of the study also comprises highly educated people, as 66% of the people have both undergraduate and postgraduate degrees. However, very few (0.67%) of them were Ph.D. holders. The study samples also reflect job tenure with organization in terms of years of experience: 31.33% are in the range of 1–2 years, 62.33% have 3–5 years' job experience, and 6.34% are in the range of 6–10 years' experience. In addition, the questionnaire respondents were 46% junior managers, 42.33% middle managers and 11.67% senior managers.
Demographic Trait | Attributes | Frequency | Percentage (%) |
Gender | Male | 283 | 94.33 |
Female | 17 | 5.67 | |
Age | 21–29 years old | 73 | 24.33 |
30–39 years old | 167 | 55.67 | |
40–49 years old | 45 | 15.00 | |
50–59 years old | 15 | 5.00 | |
Education Level | Bachelor degree | 100 | 33.33 |
Postgraduate degree | 198 | 66.00 | |
Ph.D. | 2 | 0.67 | |
Job tenure with organization | 1–2 years | 94 | 31.33 |
3–5 years | 187 | 62.33 | |
6–10 years | 19 | 6.34 | |
Job position | Junior manager | 138 | 46.00 |
Middle manager | 127 | 42.33 | |
Senior manager | 35 | 11.67 |
The normality testing of the data is presented in Table 2, using one of the statistical techniques known as the skewness and kurtosis parameters. The appropriate range for a certain parameter remains usually between –1.96 and +1.96 (Azzalini et al., 2016). The values obtained from the normality test are below the threshold value with a significance range of 0.05 (p-value = 5%). As a result, it should be inferred that the tested data is normally distributed because the results also imply that all the values of the variables remain within the predicted range.
Constructs | ENV | ECO | SOC | LEG | ETH | PHI | CR | FP |
Skewness | −0.621 | −0.429 | −0.275 | −0.274 | −0.545 | −0.527 | −0.523 | −0.881 |
Kurtosis | 0.690 | 0.653 | 0.608 | 0.635 | 0.798 | 0.480 | 0.386 | 0.133 |
Non-response bias in questionnaire surveys occurs when respondents do not reply to a survey simultaneously (Hill et al., 1997). It is true that there is a possibility of bias even though the data were gathered and reported by the researcher using the same questionnaire (Podsakoff et al., 2003). For the questionnaire survey, the non-response bias issue must be addressed. Following the suggestions of Ooi et al. (2018), this study conducted a t-test to test the non-response bias problem. The independent and dependent variables were from the same respondents; hence, common method bias might become a concern that could weaken the results' validity. Based on their response waves, the 300 questionnaires were divided into two groups: early and late. Dynamically, it was ensured to the respondents that their identities and responses would be kept private. Results indicated no significant differences between these two groups regarding various items at the 5% (p-values > 0.05) significance level. Accordingly, this study does not have a non-response bias problem. Statistically, the major elements of each variable produced seven separate factors which considered 73.7% of the variance. Similar to this, just 11.6% of the dependent, mediating and independent variables' innate factors produced the common factor, which is much less than the required level of 25% (Podsakoff et al. 2003). At last, the correlation matrix of the constructs was analyzed through following Bagozzi et al. (1991), which showed that no correlation was higher than 0.90. As a result, we have made a conclusion that the dataset is CMV free.
Based on the validity and dependability of the constructs, the quality of measurement is assessed for the measuring components. The study used a two-step assessment process to measure and evaluate the hypotheses by looking at how well the structural model fit the data. By assessing convergent validity, construct reliability and discriminant validity, the goodness of fit was determined.
The average variance extracted (AVE) is mostly what determines convergence validity. The prediction ability of the variables with the average fluctuations is frequently measured using the AVE (Alarcón et al., 2015). AVE must typically be more than 0.50 to meet the requirements for strong convergent validity and high dependability (Hair et al., 2012). The values of AVE were found to be more than 0.50. In addition, the study assessed the CRs and discovered that the values of several CRs are also over the cutoff of 0.70. Thus, we can claim that the variables are strongly related (Leong et al., 2012).
Variables | Items | Factor Loading | CR | AVE |
Environmental Protection-CSR | ENV-1 | 0.887 | 0.904 | 0.702 |
ENV-2 | 0.798 | |||
ENV-3 | 0.903 | |||
ENV-4 | 0.754 | |||
Economic-CSR | ECO-1 | 0.907 | 0.926 | 0.760 |
ECO-2 | 0.970 | |||
ECO-3 | 0.708 | |||
ECO-4 | 0.881 | |||
Societal-CSR | SOC-1 | 0.885 | 0.950 | 0.827 |
SOC-2 | 0.879 | |||
SOC-3 | 0.882 | |||
SOC-4 | 0.987 | |||
Legal-CSR | LEG-1 | 0.890 | 0.912 | 0.722 |
LEG-2 | 0.812 | |||
LEG-3 | 0.789 | |||
LEG-4 | 0.902 | |||
Ethical-CSR | ETH-1 | 0.823 | 0.915 | 0.730 |
ETH-2 | 0.856 | |||
ETH-3 | 0.890 | |||
ETH-4 | 0.847 | |||
Philanthropic-CSR | PHI-1 | 0.890 | 0.916 | 0.734 |
PHI-2 | 0.993 | |||
PHI-3 | 0.798 | |||
PHI-4 | 0.721 | |||
Corporate Reputation | CR-1 | 0.890 | 0.926 | 0.759 |
CR-2 | 0.856 | |||
CR-3 | 0.843 | |||
CR-4 | 0.895 | |||
Financial Performance | FP-1 | 0.856 | 0.947 | 0.818 |
FP-2 | 0.908 | |||
FP-3 | 0.954 | |||
FP-4 | 0.898 |
To determine if respondents' responses in the survey were weakly related to other intrinsic factors or not at all, discriminant validity was used (Alarcón et al., 2015). It is significant to begin comparing the square roots of AVE values and other correlations of constructs to analyze the discriminant applicability for each construct. The discriminant validity can be used whenever the variable's AVE values are higher than the correlation coefficients (Fornell & Larcker, 1981). Additionally, we looked at the constructs' discriminant validity using the ratio of Fornell and Larcker. By employing it, we were able to determine that all square roots of AVEs are greater than the corresponding correlation coefficients, which are supported by the Fornell-Larcker ratios being less than one. As a result, all correlation coefficients were significant at the level of 0.01, indicating that the research's discriminant validity was adequate.
Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
ENV | 0.838 | |||||||
ECO | 00.437** | 00.872 | ||||||
SOC | 00.591** | 00.204** | 00.909 | |||||
LEG | 00.204** | 0.324** | 0.324** | 00.850 | ||||
ETH | 00.488** | 0.271** | 0.793** | 0.793** | 00.854 | |||
PHI | 0.398** | 0.217** | 0.629** | 0.842** | 0.842** | 00.857 | ||
CR | 0.117** | -0.109* | 0.398** | 0.377** | 0.117** | 0.547** | 00.871 | |
FP | −0.151** | 0.488** | 0.271** | 0.629** | 0.177** | 591** | 0.177** | 00.904 |
Table 5 shows that CSR and financial success have a positive link because the path coefficient value was found to be significant (β = 0.187, t = 4.073, p < 0.003). Thus, H1 is fully supported.
Path Analysis | Path-Coefficient | C. R. (t-value) | p-value | Results | |||
H1 | CSR | → | FP | 0.187 | 4.073 | 0.003 | Supported |
H2 | CSR | → | CR | 0.212 | 5.906 | 0.000 | Supported |
H3 | CR | → | FP | 0.270 | 6.588 | 0.000 | Supported |
H4 | CSR | → | CR-FP | 0.162 | 2.801 | 0.007 | Supported |
Results indicate that the relationship between CSR and CR is supported with a path coefficient value of β = 0.212, t = 6.906, p = 0.000. Based on the path analysis, H3 (CR → FP, β = 0.270, t = 6.588, p = 0.000) and H4 (CSR → CR-FP, β = 0.162, t = 2.801, p = 0.007) were empirically supported.
This study employed SEM to validate the mechanism of CSR on FP, with mediation of CR, in order to look at how the variables interacted. The structural equation model has been implemented using AMOS-23.00 software. The final model is depicted in Figure 2. The graph addressed factors like CSR, CR and FP, while the circle focused on residual variables. The R2 values were 0.490 and 0.264 for financial performance and corporate reputation, respectively.
We examine the direct, indirect and total effects of all the explanatory variables in addition to the mediation implications. Table 6 demonstrates that significant direct, indirect and total effects on the corresponding endogenous variables were also observed for the explanatory factors. In aspects of indirect effects on FP, PE (0.2289) has the greatest indirect effect, accompanied by EE (0.1883), AC (0.1763), MS (0.1711) and FS (0.1200).
IV | DV | Direct effect | Indirect Effect | Total effect |
ENV | CR | 0.024 | 0.000 | 0.024 |
FP | 0.000 | 0.054 | 0.054 | |
ECO | CR | 0.104 | 0.000 | 0.104 |
FP | 0.000 | 0.025 | 0.025 | |
SOC | CR | 0.111 | 0.000 | 0.111 |
FP | 0.000 | 0.054 | 0.054 | |
LEG | CR | 0.290 | 0.000 | 0.290 |
FP | 0.000 | 0.117 | 0.117 | |
ETH | CR | 0.179 | 0.000 | 0.179 |
FP | 0.000 | 0.033 | 0.033 | |
PHI | CR | 0.300 | 0.000 | 0.300 |
FP | 0.000 | 0.111 | 0.111 | |
CR | FP | 0.270 | 0.000 | 0.270 |
The methodology proposed by Baron and Kenny (1986) has been used to assess the mediation effect of the variables on CR. Sobel's test for mediation significance was carried out in order to assess the importance of the mediation effect. Such a test would demonstrate the potency of CSR's indirect influence on business performance via company reputation. After analyzing the magnitude of the mediation impact and the results of the Sobel test, it can be said that the p-value of 0.001 is lower than that of 0.05 and that the t-count of 2,649 is larger than the t-table value of 1.96. At 0.05 significance levels, we noted that all mediation effects appeared to be significant. This means that there is still a mediation impact of corporate reputation on the relationship between CSR and firm performance.
Variable | Estimate | PBCI (95%) | Sobel's test | ||||||||
IV | MV | DV | lower | upper | t-value | p-value | Sig. | ||||
ENV→ | CR → | FP | 0.006 | −0.036 | −0.036 | 7.24 | 0.000 | Yes | |||
ECO→ | CR → | FP | 0.028 | 0.105 | 0.105 | 6.29 | 0.000 | Yes | |||
SOC→ | CR → | FP | 0.030 | −0.022 | −0.036 | 4.23 | 0.001 | Yes | |||
LEG→ | CR → | FP | 0.078 | −0.036 | 0.105 | 6.17 | 0.000 | Yes | |||
ETH→ | CR → | FP | 0.048 | 0.105 | −0.022 | 4.59 | 0.001 | Yes | |||
PHI→ | CR → | FP | 0.081 | −0.022 | −0.036 | 5.86 | 0.000 | Yes | |||
Note: PBCI = percentile bootstrap confidence interval, significant at p < 0.05 |
The relationship between CSR and financial success is further complicated by the presence of a mediating variable, such as reputation, which raises the possibility that manufacturing organizations' reputations may contribute to some of the explanation for the influence of CSR on financial performance. This highlights the importance of managing and enhancing corporate reputation as a potential mechanism through which CSR efforts can translate into improved financial outcomes. The study also measured the effect of demographic variables on CSR practices to invest further to improve the corporate reputation. Table 8 presents the SPSS results for the one-way ANOVA between attributes of the respondents' demographic variables and CSR practices. The results show that the relationships between gender and CSR practices (F = 5.453, p < 0.05), age and CSR practices (F = 4.184, p < 0.05), education level and CSR practices (F = 3.819, p < 0.05), job tenure with organization and CSR practices (F = 6.310, p < 0.05) and job position in organization and CSR practices (F = 5.003, p < 0.05) were significant. The companies' current statuses of CSR practices and the influence of demographic variables on CSR practices to invest further to improve corporate reputation in manufacturing companies in Bangladesh are presented in Table 9. Table 9 shows that most of the respondents (95.67%) intend to invest in the future more than the previous investment amount only for corporate reputation in the future. In addition, 90.67% of the respondents already invested in local and mass community at least two percent of total earnings for enhancing corporate reputation.
Variables | F-statistics | P-value |
Gender | 5.453 | 0.012 |
Age | 4.184 | 0.008 |
Education Level | 3.819 | 0.002 |
Job tenure with your organization | 6.310 | 0.001 |
Your job position | 5.003 | 0.000 |
Significant at p < 0.05. |
Our company intent to invest in future more than the previous investment amount only for corporate reputation. | Currently, our company invest at least two percent of the total earnings for enhancing the corporate reputation. | ||||
Frequency | Percentage | Frequency | Percentage | ||
Strongly Disagree | 2 | 0.67% | 7 | 2.33% | |
Disagree | 4 | 1.33% | 8 | 2.67% | |
Neutral | 7 | 2.33% | 13 | 4.33% | |
Agree | 182 | 60.67% | 190 | 63.33% | |
Strongly Agree | 105 | 35% | 82 | 27.33% |
To investigate the influence of demographic issues of professionals in CSR practice, this study applied a one-way ANOVA analysis, which is commonly used to compare means across multiple groups. The analysis can help determine if there are significant differences in the variables of interest (CSR, financial performance and corporate reputation) based on the demographic categories (gender, age, education level, job tenure, and job designation). The one-way ANOVA allows this empirical research to assess whether there are any statistically significant differences in the means of CSR, financial performance and corporate reputation among the different demographic groups. By examining the p-values associated with each group comparison, the study can determine if there are significant differences and gain insights into how these demographic variables may impact the variables of interest. For example, the study might find that there are significant differences in CSR, financial performance or corporate reputation across different gender groups. This could suggest that gender plays a role in shaping the perceptions, behaviors or expectations regarding CSR initiatives and financial performance within the manufacturing companies being studied. Similarly, analyzing the effects of age, education level, job tenure and job designation on CSR, financial performance and corporate reputation can provide valuable insights into how these demographic factors might influence these variables.
The findings demonstrated that CR plays a significant and advantageous mediation function in the interaction between CSR and FP (β = 2.801, p = 0.007 ***). These findings suggest that a company can boost its reputation even if it spends a lot of money on environmental responsibility (Masud & Ferdous, 2016), charitable duty, legal liability, ethical responsibility, economic responsibility and social responsibility. The higher the capacity of a firm to promote its social activities is, the better its reputation (Masud, 2018). Polonsky et al. (2005) and Bromley (2000) believed that CSR can also be used to raise and/or improve reputation. Due to the creation of new laws and increasing demands from various stakeholder groups, CSR has greatly increased in popularity in Bangladesh (Masud & Islam, 2018). As a result, Bangladeshi businesses adopted a new perspective on CSR by moving forward on active charitable work (Masud et al., 2017). Due to the legal requirements imposed by the provisions of the Companies Act of 1994, businesses should invest wisely rather than just spend money. It acted as the primary stimulus for this investigation and examination of the CSR-FP link.
The study was able to establish a CR-FP association that was statistically significant. With this type of rise in CR, a company's performance can advance. Superior CR gives a competitive advantage, and companies with a good reputation can maintain high profits over time (Masud & Ferdous, 2016; Masud & Islam, 2018). Reputation is built throughout years; therefore, famous businesses may not turn a profit in the first year but rather in the years that follow. Therefore, a company's reputation tends to improve when its CSR initiatives are in line with its corporate policies. This has a massive effect on a company's reputation, combined with increased environmental contribution and charitable responsibility efforts. Business organizations of Bangladesh appear to assume that their reputations will rise if their business strategies and environmental efforts are more aligned. The Fortune 500 list also shows that the highly profitable businesses have strong reputations. Regardless of the fierce rivalry in the industry, highly reputable companies typically function successfully (Masud & Hossain, 2019; Masud et al., 2013). Thus, it may be claimed that power of the suppliers has little influence on outcomes in reputable organizations. With all performance indicators for manufacturing enterprises, it has been found that CR has a considerable mediating effect on the CSR-FP association.
We can draw a conclusion that the current study examines the mediating effect of CR on the relationship of CSR and business performance. In fact, these results have theoretical and practical consequences. In this study, Carroll's CSR concepts were first used regarding consumer safeguarding and environmental responsibility (economic, legal, ethical and philanthropic duties). The government as well as private manufacturing organization has recently taken responsibility for environmental t and consumer protection related projects for ensuring sustainable development through CSR practices. The study shows that financial performance is significantly affected by CSR, while corporate reputation was the most significant factor. The value of the environmental contribution must be taught to stakeholders. Second, it was found that company reputation has a mediating effect on the relationship between CSR and FP. It is crucial that the more positive CR is, the more positive an effect it has on FP. Additionally, businesses should make sure to effectively and efficiently use their CSR budgets. Companies may think about taking part in CSR programs that assist the community, the environment and consumer relations. Third, businesses have a duty to promote communal welfare programs and uphold ethical business practices that ensure social benefits and maximize stakeholder value as much as possible. This study makes the case that attempting to meet CSR will help businesses create CR. This outcome is consistent with Sacconi and Antoni's (2010) empirical study and with the theoretical analyses of Chung et al. (2015) and Masud et al. (2013). This study has limitations even though the theoretical model presented in this research is supported. In addition, the study was conducted on a limited sample. Further research could even be carried out on larger samples. In addition, to improve the predictive value of the study model, e. g., customer satisfaction may be used as a mediating variable in the relationship between CR and CSR. Despite all the limitations mentioned previously, it is expected that this study could provide additional information on CSR and its benefits in manufacturing firms and contribute to the related field.
Zhang Jing: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Resources, Data curation, Writing - original draft, Writing - review & editing, Visualization, Supervision, Project administration, Funding acquisition. Gazi Md. Shakhawat Hossain: Conceptualization, Methodology, Validation, Formal analysis, Writing - original draft, Writing - review & editing. Professor Badiuzzaman: Conceptualization, Validation, Data curation, Resources, Visualization, Writing – review & editing. Md. Shahinur Rahman: Analyzed and interpreted the data. Najmul Hasan: Conceptualization, Writing - review & editing.
The authors declare that they have not used Artificial Intelligence (AI) tools in the creation of this article.
The authors are pleased to acknowledge financial support from the National Natural Science Foundation of China under Grant No. 72072064. This study was also supported by the National Social Science Foundation of China under Grant No. 22 & ZD110.
The authors would like to thank Tanjana Saiyed Likhon, Independent Researcher & Assistant Teacher, Monohorpur Govt. Primary School, Rajapur, Jhalokathi, Barishal, Bangladesh, for her honorary assistance in technical support and data collection. We are also grateful to the editor in chief and anonymous reviewers for their valuable comments to improve the quality of our manuscript.
The authors declare no conflict of interest.
[1] |
M. Vrigkas, C. Nikou, I. A. Kakadiaris, A review of human activity recognition methods, Front. Rob. AI, 2 (2015), 28. https://doi.org/10.3389/frobt.2015.00028 doi: 10.3389/frobt.2015.00028
![]() |
[2] |
Z. Sun, Q. Ke, H. Rahmani, M. Bennamoun, G. Wang, J. Liu, Human action recognition from various data modalities: A review, IEEE Trans. Pattern Anal. Mach. Intell., 45 (2022), 3200–3225. https://doi.org/10.1109/TPAMI.2022.3183112 doi: 10.1109/TPAMI.2022.3183112
![]() |
[3] | W. Lin, M. T. Sun, R. Poovandran, Human activity recognition for video surveillance, in 2008 IEEE International Symposium on Circuits and Systems (ISCAS), IEEE, (2008), 2737–2740. https://doi.org/10.1109/ISCAS.2008.4542023 |
[4] |
W. Hu, D. Xie, Z. Fu, W. Zeng, S. Maybank, Semantic-based surveillance video retrieval, IEEE Trans. Image Process., 16 (2007), 1168–1181. https://doi.org/10.1109/TIP.2006.891352 doi: 10.1109/TIP.2006.891352
![]() |
[5] | I. Rodomagoulakis, N. Kardaris, V. Pitsikalis, E. Mavroudi, A. Katsamanis, A. Tsiami, et al., Multimodal human action recognition in assistive human-robot interaction, in 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, (2016), 2702–2706. https://doi.org/10.1109/ICASSP.2016.7472168 |
[6] | K. Simonyan, A. Zisserman, Two-stream convolutional networks for action recognition in videos, in Advances in Neural Information Processing Systems 27 (NIPS 2014), 27 (2014). |
[7] | J. Zhu, Z. Zhu, W. Zou, End-to-end video-level representation learning for action recognition, in 2018 24th International Conference on Pattern Recognition (ICPR), IEEE, (2018), 645–650. https://doi.org/10.1109/ICPR.2018.8545710 |
[8] |
M. R. Sudha, K. Sriraghav, S. Manisha, S. G. Jacob, S. Manisha, Approaches and applications of virtual reality and gesture recognition: A review, Int. J. Ambient Comput. Intell., 8 (2017), 1–18. https://doi.org/10.4018/IJACI.2017100101 doi: 10.4018/IJACI.2017100101
![]() |
[9] |
J. Zhu, W. Zou, Z. Zhu, Y. Hu, Convolutional relation network for skeleton-based action recognition, Neurocomputing, 370 (2019), 109–117. https://doi.org/10.1016/j.neucom.2019.08.043 doi: 10.1016/j.neucom.2019.08.043
![]() |
[10] |
L. Shi, Y. Zhang, J. Cheng, H. Lu, Skeleton-based action recognition with multi-stream adaptive graph convolutional networks, IEEE Trans. Image Process., 29 (2020), 9532–9545. https://doi.org/10.1109/TIP.2020.3028207 doi: 10.1109/TIP.2020.3028207
![]() |
[11] |
K. Cheng, Y. Zhang, X. He, J. Cheng, H. Lu, Extremely lightweight skeleton-based action recognition with shiftgcn++, IEEE Trans. Image Process., 30 (2021), 7333–7348. https://doi.org/10.1109/TIP.2021.3104182 doi: 10.1109/TIP.2021.3104182
![]() |
[12] |
M. Wang, X. Li, S. Chen, X. Zhang, L. Ma, Y. Zhang, Learning representations by contrastive spatio-temporal clustering for skeleton-based action recognition, IEEE Trans. Multimedia, 26 (2023), 3207–3220. https://doi.org/10.1109/TMM.2023.3307933 doi: 10.1109/TMM.2023.3307933
![]() |
[13] |
C. Pang, X. Gao, Z. Chen, L. Lyu, Self-adaptive graph with nonlocal attention network for skeleton-based action recognition, IEEE Trans. Neural Networks Learn. Syst., 2023 (2023), 1–13. https://doi.org/10.1109/TNNLS.2023.3298950 doi: 10.1109/TNNLS.2023.3298950
![]() |
[14] |
M. Trascau, M. Nan, A. M. Florea, Spatio-temporal features in action recognition using 3D skeletal joints, Sensors, 19 (2019), 1–15. https://doi.org/10.3390/s19020423 doi: 10.3390/s19020423
![]() |
[15] |
P. Geng, X. Lu, C. Hu, H. Liu, L. Lyu, Focusing fine-grained action by self-attention-enhanced graph neural networks with contrastive learning, IEEE Trans. Circuits Syst. Video Technol., 33 (2023), 4754–4768. https://doi.org/10.1109/TCSVT.2023.3248782 doi: 10.1109/TCSVT.2023.3248782
![]() |
[16] | T. Xu, W. Takano, Graph stacked hourglass networks for 3d human pose estimation, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), (2021), 16105–16114. |
[17] | B. Doosti, S. Naha, M. Mirbagheri, D. J. Crandall, Hope-net: A graph-based model for hand-object pose estimation, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (2020), 6608–6617. |
[18] | K. Cheng, Y. Zhang, X. He, W. Chen, J. Cheng, H. Lu, Skeleton-based action recognition with shift graph convolutional network, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (2020), 183–192. |
[19] | M. Li, S. Chen, Y. Zhao, Y. Zhang, Y. Wang, Q. Tian, Dynamic multi-scale graph neural networks for 3D skeleton based human motion prediction, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (2020), 214–223. |
[20] | S. Zhang, W. Zhao, Z. Guan, X. Peng, J. Peng, Keypoint-Graph-Driven Learning Framework for Object Pose Estimation, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (2021), 1065–1073. |
[21] | L. Li, W. Zheng, Z. Zhang, Y. Huang, L. Wang, Skeleton-based relational modeling for action recognition, preprint, arXiv: 1805.02556, 2018. |
[22] | W. Zheng, L. Li, Z. Zhang, Y. Huang, L. Wang, Relational network for skeleton-based action recognition, in 2019 IEEE International Conference on Multimedia and Expo (ICME), IEEE, (2019), 826–831. https://doi.org/10.1109/ICME.2019.00147 |
[23] | Q. Ke, M. Bennamoun, S. An, F. Sohel, F. Boussaid, A new representation of skeleton sequences for 3D action recognition, in roceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (2017), 3288–3297. |
[24] | T. S. Kim, A. Reiter, Interpretable 3D human action analysis with temporal convolutional networks, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, IEEE, (2017), 20–28. |
[25] | S. Yan, Y. J. Xiong, D. Lin, Spatial temporal graph convolutional networks for skeleton-based action recognition, in Thirty-Second AAAI Conference on Artificial Intelligence, 32 (2018). https://doi.org/10.1609/aaai.v32i1.12328 |
[26] | A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, et al., Attention is all you need, in Advances in Neural Information Processing Systems 30 (NIPS 2017), (2017), 30. |
[27] | L. Shi, Y. Zhang, J. Cheng, H. Lu, Two-stream adaptive graph convolutional networks for skeleton-based action recognition, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), (2019), 12026–12035. |
[28] | C. Wang, C. Deng, On the global self-attention mechanism for graph convolutional networks, in 2020 25th International Conference on Pattern Recognition (ICPR), IEEE, (2021), 8531–8538. https://doi.org/10.1109/ICPR48806.2021.9412456 |
[29] | A. Shahroudy, J. Liu, T. Ng, G. Wang, NTU RGB+D: A large scale dataset for 3D human activity analysis, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (2016), 1010–1019. |
[30] | J. Liu, A. Shahroudy, M. Perez, G. Wang, L. Duan, A. C. Kot, NTU RGB+D 120: A large-scale benchmark for 3D human activity understanding, in IEEE Transactions on Pattern Analysis and Machine Intelligence, 42 (2019), 2684–2701. https://doi.org/10.1109/TPAMI.2019.2916873 |
[31] | M. Defferrard, X. Bresson, P. Vandergheynst, Convolutional neural networks on graphs with fast localized spectral filtering, in Advances in Neural Information Processing Systems 29 (NIPS 2016), (2016), 29. |
[32] | M. Niepert, M. Ahmed, K. Kutzkov, Learning convolutional neural networks for graphs, in Proceedings of The 33rd International Conference on Machine Learning, PMLR, (2016), 2014–2023. |
[33] | B. Li, X. Li, Z. Zhang, F. Wu, Spatio-temporal graph routing for skeleton-based action recognition, in Proceedings of the AAAI Conference on Artificial Intelligence, 33 (2019), 8561–8568. https://doi.org/10.1609/aaai.v33i01.33018561 |
[34] | T. Li, R. Zhang, Q. Li, Multi scale temporal graph networks for skeleton-based action recognition, preprint, arXiv: 2012.02970, 2020. https://doi.org/10.48550/arXiv.2012.02970 |
[35] | H. Touvron, M. Cord, A. Sablayrolles, G. Synnaeve, H. Jégou, Going deeper with image transformers, in Proceedings of the IEEE/CVF International Conference on Computer Vision, (2021), 32–42. |
[36] | H. Zhang, I. Goodfellow, D. Metaxas, A. Odena, Self-attention generative adversarial networks, in Proceedings of the 36th International Conference on Machine Learning, PMLR, (2019), 7354–7363. |
[37] | Y. Rao, J. Lu, J. Zhou, Attention-aware deep reinforcement learning for video face recognition, in Proceedings of the IEEE International Conference on Computer Vision (ICCV), (2017), 3931–3940. |
[38] | H. Larochelle, G. E. Hinton, Learning to combine foveal glimpses with a third-order Boltzmann machine, in Advances in Neural Information Processing Systems 23 (NIPS 2010), (2010), 23. |
[39] | F. Wang, M. Jiang, C. Qian, S. Yang, C. Li, H. Zhang, et al., Residual attention network for image classification, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (2017), 3156–3164. |
[40] | J. Hu, L. Shen, G. Sun, Squeeze-and-excitation networks, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (2018), 7132–7141. |
[41] | K. Xu, J. Ba, R. Kiros, K. Cho, A. Courville, R. Salakhudinov, et al., Show, attend and tell: Neural image caption generation with visual attention, in Proceedings of the 32nd International Conference on Machine Learning, PMLR, (2015), 2048–2057. |
[42] | M. E. Hussein, M. Torki, M. A. Gowayyed, M. El-Saban, Human action recognition using a temporal hierarchy of covariance descriptors on 3D joint locations, in Twenty-Third International Joint Conference on Artificial Intelligence, 2013. |
[43] | J. Liu, A. Shahroudy, D. Xu, G. Wang, Spatio-temporal lstm with trust gates for 3D human action recognition, European Conference on Computer Vision, Springer, Cham, (2016), 816–833. https://doi.org/10.1007/978-3-319-46487-9_50 |
[44] | M. Li, S. Chen, X. Chen, Y. Zhang, Y. Wang, Q. Tian, Actional-structural graph convolutional networks for skeleton-based action recognition, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), (2019), 3595–3603. |
[45] |
C. Chen, X. Zhao, J. Wang, D. Li, Y. Guan, J. Hong, Dynamic graph convolutional network for assembly behavior recognition based on attention mechanism and multi-scale feature fusion, Sci. Rep., 12 (2022), 1–13. https://doi.org/10.1038/s41598-022-11206-8 doi: 10.1038/s41598-022-11206-8
![]() |
[46] | W. Peng, X. Hong, H. Chen, G. Zhao, Learning graph convolutional network for skeleton-based human action recognition by neural searching, in Proceedings of the AAAI Conference on Artificial Intelligence, 34 (2020), 2669–2676. https://doi.org/10.1609/aaai.v34i03.5652 |
[47] | F. Shi, C. Lee, L. Qiu, Y. Zhao, T. Shen, S. Muralidhar, et al., Star: Sparse transformer-based action recognition, preprint, arXiv: 2017.07089, 2021. https://doi.org/10.48550/arXiv.2107.07089 |
[48] | H. Zhang, H. Geng, G. Yang, Two-stream transformer encoders for skeleton-based action recognition, in 6th International Technical Conference on Advances in Computing, Control and Industrial Engineering (CCIE 2021), Springer, 920 (2022), 272–281. https://doi.org/10.1007/978-981-19-3927-3_26 |
[49] | Y. Meng, M. Shi, W. Yang, Skeleton action recognition based on tranformer adaptive graph convolution, in Journal of Physics: Conference Series, 2170 (2022), 012007. https://doi.org/10.1088/1742-6596/2170/1/012007 |
[50] | W. Kay, J. Carreira, K. Simonyan, B. Zhang, C. Hillier, S. Vijayanarasimhan, et al., The kinetics human action video dataset, preprint, arXiv: 1705.06950, 2017. https://doi.org/10.48550/arXiv.1705.06950 |
[51] |
X. Qin, R. Cai, J. Yu, C. He, X. Zhang, An efficient self-attention network for skeleton-based action recognition, Sci. Rep., 12 (2022), 1–10. https://doi.org/10.1038/s41598-022-08157-5 doi: 10.1038/s41598-022-08157-5
![]() |
[52] | T. N. Kipf, M. Welling, Semi-supervised classification with graph convolutional networks, preprint, arXiv: 1609.02907, 2016. https://doi.org/10.48550/arXiv.1609.02907 |
[53] | Z. Chen, S. Li, B. Yang, Q. Li, H. Liu, Multi-scale spatial temporal graph convolutional network for skeleton-based action recognition, in Proceedings of the AAAI Conference on Artificial Intelligence, 35 (2021), 1113–1122. https://doi.org/10.1609/aaai.v35i2.16197 |
[54] | Z. Liu, H. Zhang, Z. Chen, Z. Wang, W. Ouyang, Disentangling and unifying graph convolutions for skeleton-based action recognition, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), (2020), 143–152. |
1. | Qing Wu, Theeralak Satjawathee, The Real Value of CSR Performance in the NEV Industry: Evidence from China, 2024, 12, 2227-7072, 106, 10.3390/ijfs12040106 | |
2. | Abhisek Mahanta, Naresh Chandra Sahu, Pradeep Kumar Behera, Pushp Kumar, Variations in financial performance of firms with ESG integration in business: The mediating role of corporate efficiency using DEA, 2024, 6, 2643-1092, 518, 10.3934/GF.2024020 | |
3. | Junaid Aftab, Nabila Abid, Huma Sarwar, Anam Amin, Mahsa Abedini, Monica Veneziani, Does corporate social responsibility drive financial performance? Exploring the significance of green innovation, green dynamic capabilities, and perceived environmental volatility, 2024, 31, 1535-3958, 1634, 10.1002/csr.2654 | |
4. | Aleksandra Stoiljković, Slavica Tomić, Bojan Leković, Ozren Uzelac, Nikola V. Ćurčić, The Impact of Capital Structure on the Performance of Serbian Manufacturing Companies: Application of Agency Cost Theory, 2024, 16, 2071-1050, 869, 10.3390/su16020869 | |
5. | Shuangyuan Li, Xiaoli Xu, Fuxian Zhu, Can heterogeneous environmental policies mitigate ESG divergence? -Based on corporate green innovation and bleaching green behavioral options, 2024, 8, 26661888, 100351, 10.1016/j.sftr.2024.100351 | |
6. | Ahmed Diab, Aref M. Eissa, ESG Performance, Auditor Choice, and Audit Opinion: Evidence from an Emerging Market, 2023, 16, 2071-1050, 124, 10.3390/su16010124 | |
7. | Jue Feng, Yingdong Wang, Wenzhi Xi, Does green technology transformation alleviate corporate financial constraints? Evidence from Chinese listed firms, 2024, 10, 24058440, e27841, 10.1016/j.heliyon.2024.e27841 | |
8. | Jorge Luis Canta Honores, Luciano Barcellos-Paula, Mediation of brand reputation and image in the relationship between perceived corporate social responsibility and brand equity and loyalty of technology companies among Generation Z: the moderating role of gender, 2024, 11, 2331-1975, 10.1080/23311975.2024.2424484 | |
9. | Gaoke Liao, Yanling Li, Mengxin Wang, Contagion network of idiosyncratic volatility: Does corporate environmental responsibility matter?, 2024, 129, 01409883, 107168, 10.1016/j.eneco.2023.107168 | |
10. | Jing Zou, Liming Yao, Baitao Wang, Yu Zhang, Xiaojun Deng, How does digital inclusive finance promote the journey of common prosperity in China?, 2024, 150, 02642751, 105083, 10.1016/j.cities.2024.105083 | |
11. | Zhi Zhang, Yanhong Feng, Hongwei Zhou, Liming Chen, Yi Liu, The Impact of Climate Policy Uncertainty on the ESG Performance of Enterprises, 2024, 12, 2079-8954, 495, 10.3390/systems12110495 | |
12. | Muttahir Hussain, Wang Xuetong, Liang Hao, Muhammad Malik, The effects of corporate social responsibility on organizational performance in the construction industry: the mediating role of organizational innovation and organizational governance, 2024, 1573-2975, 10.1007/s10668-024-05775-z | |
13. | Mohamed BOULESNAM, Abdelghani KAHELA, Ouissam HOCINI, EXPLORING THE CORRELATION BETWEEN CORPORATE SOCIAL RESPONSIBILITY INDICATORS AND CORPORATE REPUTATION : CASE STUDY OF LEGO GROUP 2012–2023, 2024, 29, 2353-0758, 7, 10.7862/rz.2024.mmr.17 | |
14. | Shashi Kant, Metasebia Adula, Mershaye Birhane, 2025, chapter 9, 9798369397503, 169, 10.4018/979-8-3693-9750-3.ch009 | |
15. | Juan Liu, Applying Ethics of Care-Based Response Strategies to Mitigate AI-Related Corporate Crisis: The Moderating Role of Crisis Involvement, 2025, 1553-118X, 1, 10.1080/1553118X.2025.2459609 |
Demographic Trait | Attributes | Frequency | Percentage (%) |
Gender | Male | 283 | 94.33 |
Female | 17 | 5.67 | |
Age | 21–29 years old | 73 | 24.33 |
30–39 years old | 167 | 55.67 | |
40–49 years old | 45 | 15.00 | |
50–59 years old | 15 | 5.00 | |
Education Level | Bachelor degree | 100 | 33.33 |
Postgraduate degree | 198 | 66.00 | |
Ph.D. | 2 | 0.67 | |
Job tenure with organization | 1–2 years | 94 | 31.33 |
3–5 years | 187 | 62.33 | |
6–10 years | 19 | 6.34 | |
Job position | Junior manager | 138 | 46.00 |
Middle manager | 127 | 42.33 | |
Senior manager | 35 | 11.67 |
Constructs | ENV | ECO | SOC | LEG | ETH | PHI | CR | FP |
Skewness | −0.621 | −0.429 | −0.275 | −0.274 | −0.545 | −0.527 | −0.523 | −0.881 |
Kurtosis | 0.690 | 0.653 | 0.608 | 0.635 | 0.798 | 0.480 | 0.386 | 0.133 |
Variables | Items | Factor Loading | CR | AVE |
Environmental Protection-CSR | ENV-1 | 0.887 | 0.904 | 0.702 |
ENV-2 | 0.798 | |||
ENV-3 | 0.903 | |||
ENV-4 | 0.754 | |||
Economic-CSR | ECO-1 | 0.907 | 0.926 | 0.760 |
ECO-2 | 0.970 | |||
ECO-3 | 0.708 | |||
ECO-4 | 0.881 | |||
Societal-CSR | SOC-1 | 0.885 | 0.950 | 0.827 |
SOC-2 | 0.879 | |||
SOC-3 | 0.882 | |||
SOC-4 | 0.987 | |||
Legal-CSR | LEG-1 | 0.890 | 0.912 | 0.722 |
LEG-2 | 0.812 | |||
LEG-3 | 0.789 | |||
LEG-4 | 0.902 | |||
Ethical-CSR | ETH-1 | 0.823 | 0.915 | 0.730 |
ETH-2 | 0.856 | |||
ETH-3 | 0.890 | |||
ETH-4 | 0.847 | |||
Philanthropic-CSR | PHI-1 | 0.890 | 0.916 | 0.734 |
PHI-2 | 0.993 | |||
PHI-3 | 0.798 | |||
PHI-4 | 0.721 | |||
Corporate Reputation | CR-1 | 0.890 | 0.926 | 0.759 |
CR-2 | 0.856 | |||
CR-3 | 0.843 | |||
CR-4 | 0.895 | |||
Financial Performance | FP-1 | 0.856 | 0.947 | 0.818 |
FP-2 | 0.908 | |||
FP-3 | 0.954 | |||
FP-4 | 0.898 |
Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
ENV | 0.838 | |||||||
ECO | 00.437** | 00.872 | ||||||
SOC | 00.591** | 00.204** | 00.909 | |||||
LEG | 00.204** | 0.324** | 0.324** | 00.850 | ||||
ETH | 00.488** | 0.271** | 0.793** | 0.793** | 00.854 | |||
PHI | 0.398** | 0.217** | 0.629** | 0.842** | 0.842** | 00.857 | ||
CR | 0.117** | -0.109* | 0.398** | 0.377** | 0.117** | 0.547** | 00.871 | |
FP | −0.151** | 0.488** | 0.271** | 0.629** | 0.177** | 591** | 0.177** | 00.904 |
Path Analysis | Path-Coefficient | C. R. (t-value) | p-value | Results | |||
H1 | CSR | → | FP | 0.187 | 4.073 | 0.003 | Supported |
H2 | CSR | → | CR | 0.212 | 5.906 | 0.000 | Supported |
H3 | CR | → | FP | 0.270 | 6.588 | 0.000 | Supported |
H4 | CSR | → | CR-FP | 0.162 | 2.801 | 0.007 | Supported |
IV | DV | Direct effect | Indirect Effect | Total effect |
ENV | CR | 0.024 | 0.000 | 0.024 |
FP | 0.000 | 0.054 | 0.054 | |
ECO | CR | 0.104 | 0.000 | 0.104 |
FP | 0.000 | 0.025 | 0.025 | |
SOC | CR | 0.111 | 0.000 | 0.111 |
FP | 0.000 | 0.054 | 0.054 | |
LEG | CR | 0.290 | 0.000 | 0.290 |
FP | 0.000 | 0.117 | 0.117 | |
ETH | CR | 0.179 | 0.000 | 0.179 |
FP | 0.000 | 0.033 | 0.033 | |
PHI | CR | 0.300 | 0.000 | 0.300 |
FP | 0.000 | 0.111 | 0.111 | |
CR | FP | 0.270 | 0.000 | 0.270 |
Variable | Estimate | PBCI (95%) | Sobel's test | ||||||||
IV | MV | DV | lower | upper | t-value | p-value | Sig. | ||||
ENV→ | CR → | FP | 0.006 | −0.036 | −0.036 | 7.24 | 0.000 | Yes | |||
ECO→ | CR → | FP | 0.028 | 0.105 | 0.105 | 6.29 | 0.000 | Yes | |||
SOC→ | CR → | FP | 0.030 | −0.022 | −0.036 | 4.23 | 0.001 | Yes | |||
LEG→ | CR → | FP | 0.078 | −0.036 | 0.105 | 6.17 | 0.000 | Yes | |||
ETH→ | CR → | FP | 0.048 | 0.105 | −0.022 | 4.59 | 0.001 | Yes | |||
PHI→ | CR → | FP | 0.081 | −0.022 | −0.036 | 5.86 | 0.000 | Yes | |||
Note: PBCI = percentile bootstrap confidence interval, significant at p < 0.05 |
Variables | F-statistics | P-value |
Gender | 5.453 | 0.012 |
Age | 4.184 | 0.008 |
Education Level | 3.819 | 0.002 |
Job tenure with your organization | 6.310 | 0.001 |
Your job position | 5.003 | 0.000 |
Significant at p < 0.05. |
Our company intent to invest in future more than the previous investment amount only for corporate reputation. | Currently, our company invest at least two percent of the total earnings for enhancing the corporate reputation. | ||||
Frequency | Percentage | Frequency | Percentage | ||
Strongly Disagree | 2 | 0.67% | 7 | 2.33% | |
Disagree | 4 | 1.33% | 8 | 2.67% | |
Neutral | 7 | 2.33% | 13 | 4.33% | |
Agree | 182 | 60.67% | 190 | 63.33% | |
Strongly Agree | 105 | 35% | 82 | 27.33% |
Demographic Trait | Attributes | Frequency | Percentage (%) |
Gender | Male | 283 | 94.33 |
Female | 17 | 5.67 | |
Age | 21–29 years old | 73 | 24.33 |
30–39 years old | 167 | 55.67 | |
40–49 years old | 45 | 15.00 | |
50–59 years old | 15 | 5.00 | |
Education Level | Bachelor degree | 100 | 33.33 |
Postgraduate degree | 198 | 66.00 | |
Ph.D. | 2 | 0.67 | |
Job tenure with organization | 1–2 years | 94 | 31.33 |
3–5 years | 187 | 62.33 | |
6–10 years | 19 | 6.34 | |
Job position | Junior manager | 138 | 46.00 |
Middle manager | 127 | 42.33 | |
Senior manager | 35 | 11.67 |
Constructs | ENV | ECO | SOC | LEG | ETH | PHI | CR | FP |
Skewness | −0.621 | −0.429 | −0.275 | −0.274 | −0.545 | −0.527 | −0.523 | −0.881 |
Kurtosis | 0.690 | 0.653 | 0.608 | 0.635 | 0.798 | 0.480 | 0.386 | 0.133 |
Variables | Items | Factor Loading | CR | AVE |
Environmental Protection-CSR | ENV-1 | 0.887 | 0.904 | 0.702 |
ENV-2 | 0.798 | |||
ENV-3 | 0.903 | |||
ENV-4 | 0.754 | |||
Economic-CSR | ECO-1 | 0.907 | 0.926 | 0.760 |
ECO-2 | 0.970 | |||
ECO-3 | 0.708 | |||
ECO-4 | 0.881 | |||
Societal-CSR | SOC-1 | 0.885 | 0.950 | 0.827 |
SOC-2 | 0.879 | |||
SOC-3 | 0.882 | |||
SOC-4 | 0.987 | |||
Legal-CSR | LEG-1 | 0.890 | 0.912 | 0.722 |
LEG-2 | 0.812 | |||
LEG-3 | 0.789 | |||
LEG-4 | 0.902 | |||
Ethical-CSR | ETH-1 | 0.823 | 0.915 | 0.730 |
ETH-2 | 0.856 | |||
ETH-3 | 0.890 | |||
ETH-4 | 0.847 | |||
Philanthropic-CSR | PHI-1 | 0.890 | 0.916 | 0.734 |
PHI-2 | 0.993 | |||
PHI-3 | 0.798 | |||
PHI-4 | 0.721 | |||
Corporate Reputation | CR-1 | 0.890 | 0.926 | 0.759 |
CR-2 | 0.856 | |||
CR-3 | 0.843 | |||
CR-4 | 0.895 | |||
Financial Performance | FP-1 | 0.856 | 0.947 | 0.818 |
FP-2 | 0.908 | |||
FP-3 | 0.954 | |||
FP-4 | 0.898 |
Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
ENV | 0.838 | |||||||
ECO | 00.437** | 00.872 | ||||||
SOC | 00.591** | 00.204** | 00.909 | |||||
LEG | 00.204** | 0.324** | 0.324** | 00.850 | ||||
ETH | 00.488** | 0.271** | 0.793** | 0.793** | 00.854 | |||
PHI | 0.398** | 0.217** | 0.629** | 0.842** | 0.842** | 00.857 | ||
CR | 0.117** | -0.109* | 0.398** | 0.377** | 0.117** | 0.547** | 00.871 | |
FP | −0.151** | 0.488** | 0.271** | 0.629** | 0.177** | 591** | 0.177** | 00.904 |
Path Analysis | Path-Coefficient | C. R. (t-value) | p-value | Results | |||
H1 | CSR | → | FP | 0.187 | 4.073 | 0.003 | Supported |
H2 | CSR | → | CR | 0.212 | 5.906 | 0.000 | Supported |
H3 | CR | → | FP | 0.270 | 6.588 | 0.000 | Supported |
H4 | CSR | → | CR-FP | 0.162 | 2.801 | 0.007 | Supported |
IV | DV | Direct effect | Indirect Effect | Total effect |
ENV | CR | 0.024 | 0.000 | 0.024 |
FP | 0.000 | 0.054 | 0.054 | |
ECO | CR | 0.104 | 0.000 | 0.104 |
FP | 0.000 | 0.025 | 0.025 | |
SOC | CR | 0.111 | 0.000 | 0.111 |
FP | 0.000 | 0.054 | 0.054 | |
LEG | CR | 0.290 | 0.000 | 0.290 |
FP | 0.000 | 0.117 | 0.117 | |
ETH | CR | 0.179 | 0.000 | 0.179 |
FP | 0.000 | 0.033 | 0.033 | |
PHI | CR | 0.300 | 0.000 | 0.300 |
FP | 0.000 | 0.111 | 0.111 | |
CR | FP | 0.270 | 0.000 | 0.270 |
Variable | Estimate | PBCI (95%) | Sobel's test | ||||||||
IV | MV | DV | lower | upper | t-value | p-value | Sig. | ||||
ENV→ | CR → | FP | 0.006 | −0.036 | −0.036 | 7.24 | 0.000 | Yes | |||
ECO→ | CR → | FP | 0.028 | 0.105 | 0.105 | 6.29 | 0.000 | Yes | |||
SOC→ | CR → | FP | 0.030 | −0.022 | −0.036 | 4.23 | 0.001 | Yes | |||
LEG→ | CR → | FP | 0.078 | −0.036 | 0.105 | 6.17 | 0.000 | Yes | |||
ETH→ | CR → | FP | 0.048 | 0.105 | −0.022 | 4.59 | 0.001 | Yes | |||
PHI→ | CR → | FP | 0.081 | −0.022 | −0.036 | 5.86 | 0.000 | Yes | |||
Note: PBCI = percentile bootstrap confidence interval, significant at p < 0.05 |
Variables | F-statistics | P-value |
Gender | 5.453 | 0.012 |
Age | 4.184 | 0.008 |
Education Level | 3.819 | 0.002 |
Job tenure with your organization | 6.310 | 0.001 |
Your job position | 5.003 | 0.000 |
Significant at p < 0.05. |
Our company intent to invest in future more than the previous investment amount only for corporate reputation. | Currently, our company invest at least two percent of the total earnings for enhancing the corporate reputation. | ||||
Frequency | Percentage | Frequency | Percentage | ||
Strongly Disagree | 2 | 0.67% | 7 | 2.33% | |
Disagree | 4 | 1.33% | 8 | 2.67% | |
Neutral | 7 | 2.33% | 13 | 4.33% | |
Agree | 182 | 60.67% | 190 | 63.33% | |
Strongly Agree | 105 | 35% | 82 | 27.33% |