CSR attribution: Is it the cornerstone of CSR success?

Abstract

Purpose

The current study aims to propose and empirically investigate a conceptual model of the most relevant antecedents and consequences of Corporate Social Responsibility (CSR) attribution, thus providing a practical and concise model as well as examining brand attachment as a mediator explaining the relationship between CSR attribution and its consequences.

Design/methodology/approach

A between-subjects experimental design was employed. The study included two experimental conditions; intrinsic and extrinsic CSR attribution and a control condition. An online self-administered survey was utilised for data collection. The sample was a convenience sample of 336 university students. Both one-way between-groups ANOVA and Partial Least Squares-Structural Equation Modelling (PLS-SEM) were utilised for hypotheses testing.

Findings

The most significant antecedents of CSR attribution in order of importance are the firm's approach to CSR communication, past corporate social performance, CSR type and the firm's call for customers' participation in its CSR. CSR attribution exerted a significant direct positive impact on brand attachment and trust. Three significant indirect consequences of CSR attribution were PWOM intention, purchase intention and brand loyalty intention. Whereas trust played a significant mediating role between CSR attribution and its three indirect consequences, brand attachment exerted significant mediation only between CSR attribution and brand loyalty intention. Brand attachment might mediate the relationship between CSR attribution and purchase intention. However, brand attachment failed to play a mediating role between CSR attribution and PWOM intention.

Originality/value

Several studies marginally investigated CSR attribution. Despite the vital role of CSR attribution in how consumers receive firms' CSR engagement, the availability of CSR attribution-centric studies is limited. By introducing a model of the most relevant antecedents and consequences of CSR attribution, this study aids in understanding the psychological mechanism underlying consumers' CSR attribution and provides valuable implications.

Citation

Hassan Hosny, S.O. and AbdelAziz, G.S. (2024), "CSR attribution: Is it the cornerstone of CSR success?", Journal of Humanities and Applied Social Sciences, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/JHASS-09-2023-0114

1. Introduction

Corporate Social Responsibility (CSR) has become a reality in the operating lives of most successful firms. Numerous studies have verified the positive impact of CSR, in general, on consumers' attitudes and behaviours. For example, Liu et al. (2014) examined the impact of CSR performance in the three domains (environment, society and stakeholders) on brand preference as mediated by perceived brand quality. They found that CSR in general could exercise an effect on brand preference. Nevertheless, it was found that CSR in the stakeholders’ domain had the most significant impact on brand preference. The mediation effect was corroborated. In a related respect, Hasudungan and Saragih (2023) investigated the moderating role of CSR dimensions (economic, environmental and social) on the relationship between brand awareness and relationship intention. Both brand awareness and perceived CSR dimensions were found to have a positive impact on consumer relationship intention. The moderating effect of perceived CSR dimensions was confirmed as well. Nevertheless, CSR as a holistic concept has several related intricacies that play a role in how CSR is received and perceived by the consumer; one of which is CSR attribution.

Consumers' response to CSR messages is usually accompanied by some degree of scepticism. Such scepticism often leads to negative consumer outcomes unlike what is initially desired by the sponsoring firm when it resolves to engage in CSR. A concept closely relevant to CSR scepticism is CSR attribution. CSR attribution is the “causal reasoning consumers engage in when trying to understand a company's CSR activities” (Bhattacharya and Sen, 2004, p. 14). Attribution theory represents the origin of the occurrence of the concept of CSR attribution (Heider, 1958; Weiner, 1985; Gond et al., 2017).

When exposed to a CSR message, consumers could attribute a firm's CSR engagement to intrinsic and altruistic motives such as the society's well-being (intrinsic CSR attribution) or to extrinsic and egoistic motives such as the firm's own well-being, e.g. strategic and financial gains (extrinsic CSR attribution) (Habel et al., 2016). Such attribution affects consumers' consequent reactions to sponsoring firms in terms of their attitudes and behaviours (Ellen et al., 2006; Skarmeas and Leonidou, 2013). Thus, consumers, as rational information processors, engage in a process of attributing a firm's CSR which influences their attitudes and behaviours (Osakwe and Yusuf, 2020). Moreover, it was manifested in many previous studies that intrinsic CSR attribution leads to more positive consumer outcomes than extrinsic CSR attribution (e.g. Ahn and Soeiro, 2021; Choi et al., 2016; Ellen et al., 2006; Hammad et al., 2014; Hur and Kim, 2017; Kim and Choi, 2018; Ogunfowora et al., 2018; Tsai, 2009). Intrinsic CSR attribution was found to have a positive effect on consumers' attitudes towards the sponsoring firm (Wongpitch et al., 2016; Shazly and Mahrous, 2020), the CSR initiative evaluation and purchase intention (Zasuwa, 2016), brand loyalty and positive word of mouth (PWOM) and positive consumers' emotions (García-De los Salmones and Perez, 2018). In contrast, extrinsic CSR attribution was found to have a negative effect on consumers' recommendations intentions and patronage intentions (Vlachos et al., 2009) and could result in negative consumers' emotions (García-De los Salmones and Perez, 2018).

Consequently, since intrinsic CSR attribution enjoys a competitive advantage over extrinsic CSR attribution, it becomes imperative to gain an in-depth understanding of the underlying psychological mechanism that takes place when consumers perceive firms' CSR engagement. This deeper understanding could enable sponsoring firms to manipulate such mechanism, to an extent, to eventually be able to elicit intrinsic CSR attribution on the side of the consumer. Hence, the significance of CSR attribution in CSR-related relationships becomes evident. Therefore, conducting in-depth research to examine how consumers' CSR attribution is shaped – as was called for in several previous studies – becomes a worthwhile research path (e.g. Alhouti et al., 2016; Habel et al., 2016; Hur and Kim, 2017). Also, generally, there is yet a need for studies that explore the factors that impact consumer psychology in the context of green consumption behaviour and sustainability in general (Wang et al., 2023). Thus, the focal interest of this research is to empirically investigate the antecedents and consequences of CSR attribution. In addition, the current study is going to apply the proposed model (Figure 1) in the mobile and telecommunications services context which has relatively been under-examined when CSR attribution was the issue of interest as could be noticed from Table 1.

Many factors were found to affect CSR attribution; duration of commitment to CSR initiatives (Yang et al., 2015), company-cause fit, CSR communication strategy (Austin and Gaither, 2017; Yang et al., 2015), the type of sponsored cause, the type of message appeals (Andreu et al., 2015), cultural orientation, nationality of sponsoring firms (Choi et al., 2016; Lim et al., 2018), corporate ability, interpersonal trust, corporate hypocrite (Marín et al., 2016) and corporate history (Leonidou and Skarmeas, 2017).

CSR attribution-centric studies were observed to be limited especially in the Egyptian context. Insufficient studies were conducted to examine concisely and simultaneously the most relevant antecedents and consequences of CSR attribution. Also, there is a need for understanding in a more profound fashion the psychological mechanism underlying consumers' CSR attribution as requested by previous studies (e.g. Alhouti et al., 2016; Hur and Kim, 2017; Pai et al., 2015). Hence, the current study's main purpose is to propose and empirically investigate a concise model of the most relevant antecedents and consequences of CSR attribution. Based on both: a previous literature review and a previously conducted exploratory study with the aim of exploring the most relevant antecedents and consequences of CSR attribution to the Egyptian context (Hassan Hosny and AbdelAziz, 2022), the following most relevant antecedents of CSR attribution were selected for the purpose of empirical investigation: the duration of commitment to CSR initiatives, the type of CSR, the approach to CSR communication, the amount of invested resources in CSR, past corporate social performance and the call for customers' participation in firms' CSR. In addition, the following positive consumer outcomes were selected for investigation: PWOM intention, purchase intention and brand loyalty intention. Furthermore, the mediating effect of both: brand attachment and trust was selected for examining on the relationship between CSR attribution and the consequent expected positive consumer outcomes.

2. Theoretical framework and hypotheses

2.1 Antecedents of intrinsic CSR attribution

2.1.1 Duration of commitment to CSR

The duration of a firm's commitment to CSR refers to the length of time the firm commits to engaging in its CSR initiatives and supporting the respective causes (Choi et al., 2016; Webb and Mohr, 1998). The duration of a firm's commitment to CSR was found to exert an effect on consumers' perception of CSR initiatives and the formation of their CSR attribution (Choi et al., 2016; Ellen et al., 2006; Groza et al., 2011; Hammad et al., 2014). Many studies demonstrated that a long-term commitment to CSR elicits more intrinsic CSR attribution than a short-term commitment. In contrast, a short-term commitment evokes more extrinsic CSR attribution than a long-term commitment (e.g. Choi et al., 2016; Yang et al., 2015).

H1a.

The duration of commitment to CSR initiatives has a positive impact on intrinsic CSR attribution.

H1b.

Long-term commitment to CSR initiatives leads to higher intrinsic CSR attribution than short-term commitment.

2.1.2 CSR type

One of the different classifications of CSR initiatives is philanthropic vs. business-process CSR. A CSR initiative could be considered philanthropic when it results in a direct benefit to the community. Yet, a business-process CSR initiative usually engenders a direct benefit to the sponsoring firm. It was previously detected that CSR type exerts an influence on consumers' perceptions of CSR initiatives (Habel et al., 2016; Homburg et al., 2013). Business process CSR had a positive effect on trust while philanthropic CSR exercised a positive effect on customer-company identification (Homburg et al., 2013). Also, philanthropic CSR was found to enhance the positive impact of CSR on consumers' perceptions of the firm's CSR costs' level (Habel et al., 2016). Hence, it is proposed that CSR type could exert an influence on customers' CSR attribution.

H2a.

CSR type has a positive impact on intrinsic CSR attribution.

H2a1.

Philanthropic CSR has a positive impact on intrinsic CSR attribution.

H2a2.

Business-process CSR has a positive impact on intrinsic CSR attribution.

H2b.

Philanthropic CSR leads to higher intrinsic CSR attribution than business-process CSR.

2.1.3 Approach to CSR communication

CSR could be communicated to consumers and the public in general using different approaches. For example, it could be communicated extravagantly or conversely, the firm could be silent and not communicate its CSR at all. In addition, a firm could communicate its CSR in a minimalist way that mainly aims to inform the relevant stakeholders of its CSR engagement using instruments that do not involve large extravagance or expenditure. A minimalist approach to CSR communication could be defined as an approach to CSR communication that is characterised by simplicity and humility and does not involve exaggerated expenditure of resources through employing inexpensive channels of communication. In contrast, an ostentatious approach to CSR communication could be defined as an approach to CSR communication that is characterised by extravagance and exaggerated expenditure of resources through employing expensive channels of communication (Dhanesh, 2015; Morsing et al., 2008). The approach to CSR communication usually affects consumers' evaluation of firms' CSR (Du et al., 2010; Kim and Choi, 2018).

Consumers could attribute a firm's CSR to extrinsic motives when the approach to CSR communication is extravagant, ostentatious and deliberate as a result of consumer scepticism especially if the firm is perceived to be spending more on CSR communication than on the supported causes (Dhanesh, 2015; Marín et al., 2016). In contrast, firms' silence about their CSR leads to a loss of many of the opportunities that sponsoring firms usually aim to realise when initially deciding to engage in CSR such as improved consumers' attitudes and behaviours (Du et al., 2010; Van de Ven, 2008). It was even stated before that finding an approach to CSR communication that could generate consumers' intrinsic CSR attribution is a major challenge (Du et al., 2010).

Pérez et al. (2020) conducted a study to explore the impact of information specificity and social topic awareness on consumers' attitude towards CSR messages and sponsoring firms as mediated by message authenticity. Information specificity and social topic awareness were found to improve perceived message authenticity which, in turn, has a positive effect on trust and attitudes towards the CSR message and the sponsoring firm. Perceived message authenticity could be presumed to be one of the byproducts of a minimalist approach to CSR communication and could also be expected to be one of the factors affecting consumers' CSR attribution. Thus, Pérez et al. (2020) showed how the approach to CSR communication could have an impact on perceived CSR authenticity which is closely related to CSR attribution (Alhouti et al., 2016).

Therefore, this research proposes a compromise between ostentatious CSR communication and silence which could be a minimalist approach to CSR communication. Such an approach would be characterised by using a relatively inexpensive communication channel such as social media platforms or firms' websites (Morsing et al., 2008). The purpose of a minimalist approach would be to inform consumers and the public in general of its CSR engagement with neither exaggeration nor extravagance. A minimalist approach to CSR communication is expected to elicit consumers' intrinsic CSR attribution (Chen and Chang, 2013; Dhanesh, 2015).

H3a.

The approach to CSR communication has a positive impact on intrinsic CSR attribution.

H3b.

A minimalist approach to CSR communication leads to higher intrinsic CSR attribution than ostentatious CSR communication.

2.1.4 Amount of invested resources in CSR

The amount of invested resources in CSR refers to the amount of resources the firm dedicates to engaging in its CSR initiatives and supporting the respective causes (Alhouti et al., 2016; Moosmayer and Fuljahn, 2010). Many previous studies investigated and verified the significant effect of the amount of resources invested in CSR on the positive consequences desired out of CSR adoption and on reducing consumers' CSR scepticism (Alhouti et al., 2016; Dahl and Lavack, 1995; Hoeffler and Keller, 2002; Moosmayer and Fuljahn, 2010; Müller et al., 2014; Wilson, 2017).

H4a.

The amount of resources invested in CSR has a positive impact on intrinsic CSR attribution.

H4b.

Larger amounts of resources invested in CSR lead to higher intrinsic attribution than lower amounts of invested resources.

2.1.5 Past corporate social performance

A firm's past corporate social performance could be defined as the length of the firm's history and reputation of CSR engagement. A firm's past corporate social performance or history in CSR engagement could have an effect on how consumers receive a firm's CSR message and thus could play a role in shaping consumers' CSR attribution (Du et al., 2010). That is, the longer the history of the firm in CSR adoption is, the more likely it is that consumers are going to attribute the firm's CSR initiatives to intrinsic motives (intrinsic CSR attribution) and vice versa (García-De los Salmones and Perez, 2018; Hammad et al., 2014; Leonidou and Skarmeas, 2017; Lyu, 2017). In fact, a firm's past corporate social performance is similar to the consistency cue in attribution theory. Hence, it is proposed that when a firm has a long past corporate social performance, consumers are more likely to engage in intrinsic CSR attribution. In contrast, if a firm does not have a history in CSR engagement, consumers are more likely to adopt in extrinsic CSR attribution.

H5a.

Past corporate social performance has a positive impact on intrinsic CSR attribution.

H5b.

The existence of a firm's long past corporate social performance will lead to more intrinsic attribution than if the firm had no past corporate social performance.

2.1.6 Call for customer participation in CSR

Lately, firms tend to include customers in their value-creation process (Arvidsson, 2008). Hence, firms started to allow customers the opportunity to actively participate in their CSR processes in the so-called customer participation in CSR. The moderating role of customers' participation in a firm's CSR was previously probed. It was found that the effect of consumer's sense of personal identification with the sponsoring firm on brand loyalty is stronger when consumers have participated in the firm's CSR. In addition, consumers' personal identification with the sponsoring firm as a result of participation in the firm's CSR was found to contribute to brand attachment which, in turn, results in brand loyalty (Cha et al., 2016). Moreover, the effect of CSR on brand loyalty was ameliorated via customer co-creation (Iglesias et al., 2020). Even though the call for customer participation in firms' CSR is quite different from customer co-creation, nevertheless the collaboration aspect between the firm and the customer is common in both. Therefore, it is expected that the call for customers' participation in firms' CSR is going to have an effect on CSR attribution. That is, it is proposed that the call for customer participation in CSR is going to contribute to customers' perceptions of intrinsic CSR attribution as opposed to a firm that does not allow customer participation in CSR.

H6a.

The call for customers' participation in firms' CSR has a positive impact on intrinsic CSR attribution.

H6b.

Calling for customers' participation in firms' CSR will lead to more intrinsic attribution than if the firm did not call for customers' participation in its CSR.

2.2 Consequences of CSR attribution

2.2.1 Intrinsic CSR attribution and brand attachment

Brand attachment refers to “the bond connecting the consumer to the brand” (Park et al., 2010, p. 1). CSR was found to have some effect on brand attachment. For example, it was previously found that a firm's corporate social performance exerts an influence on consumers' emotional attachment towards the sponsoring firm (Vlachos, 2012). Also, it was found that CSR has an effect on brand attachment (Koh and Fang, 2012). In addition, several studies showed that CSR could have a positive influence on customer affective commitment which could be regarded as corresponding to the emotional aspect of brand attachment (e.g. Chomvilailuk and Butcher, 2014). It is also worth noting that the relationship between CSR attribution, specifically altruistic CSR attribution, and purchase intention was found to be indirect and mediated by attitude towards the firm (Wongpitch et al., 2016). Thus, it could be expected that the relationship between intrinsic CSR attribution and positive consumer outcomes to be indirect.

H7.

Intrinsic CSR attribution has a positive impact on brand attachment.

2.2.2 Intrinsic CSR attribution and trust

Consumer trust refers to “the expectations held by the consumer that the service provider is dependable and can be relied on to deliver on its promises” (Sirdeshmukh et al., 2002, p. 17). Several studies illustrated that CSR attribution and trust are significantly related. For example, trust was found to be mediating the relationship between CSR attribution and consumers' recommendations and patronage intentions (Martínez and Bosque, 2013; Vlachos et al., 2009). Also, it was previously stated that CSR attribution could be a factor affecting the basis of consumers' trust in a firm's CSR (Moscato and Hopp, 2019). In addition, CSR washing – where a firm claims that it is involved in CSR whereas in reality, it is not – was found to have a negative effect on consumers' trust (Chen and Chang, 2013; Ginder et al., 2021). CSR washing is a form of corporate hypocrite which refers to “the discrepancy between CSR talk and CSR action” (Andersen and Høvring, 2020). Furthermore, intrinsic CSR attribution was found to be negatively related to green scepticism (Leonidou and Skarmeas, 2017). Similarly, Liu et al. (2020) investigated the impact of CSR on customer behavioural loyalty in the hotel industry and the impact was validated. Furthermore, trust and brand image were found to mediate the effect of CSR on customer loyalty.

H8.

Intrinsic CSR attribution has a positive impact on consumers' trust in the firm.

2.2.3 Brand attachment and positive word of mouth (PWOM) intention

PWOM refers to the process when consumers transfer their positive attitudes towards a certain brand or firm to other people such as their family and friends in the form of informal conversations about the brand or firm and/or its products or services (Markovic et al., 2018). When consumers experience brand attachment towards a firm, they become ready to exert extra-role behaviours to maintain their relationship with such a firm. For example, it was found that as a result of brand attachment, a consumer becomes more loyal to the brand and could further engage in brand promotion and even brand defence (Park et al., 2010). Moreover, CSR attribution was found to affect consumers' recommendations intentions and PWOM (Karaosmanoglu et al., 2016; Vlachos et al., 2009). Pérez et al. (2019) utilised Du et al.'s (2010) model of CSR communication in investigating the impact of certain aspects of a CSR message such as issue importance, impact, motives (CSR attribution), fit and commitment on consumers' responses such as purchase and advocacy intentions in addition to a few moderating variables. They found that intrinsic CSR attribution has a positive impact on both purchase intention and advocacy intention which almost corresponds to PWOM.

H9.

Brand attachment has a positive impact on PWOM intention.

2.2.4 Brand attachment and purchase intention

Purchase intention could be defined as the consumer's intention to engage in an exchange relationship with a service provider, e.g. sharing business information, maintaining business relationships and conducting business transactions (Pavlou, 2003). Brand attachment was found to be a predictor of difficult-to-enact behaviours that involve resource expenditure such as actual purchase behaviour (Park et al., 2010). Moreover, it was found that brand attachment has an effect on purchase intention in a CSR context (Koh and Fang, 2012).

H10.

Brand attachment has a positive impact on purchase intention.

2.2.5 Brand attachment and brand loyalty intention

Brand loyalty intention could be defined as the consumer's intention to be “strongly committed to repurchase a product or service of a brand, in spite of any contextual influences or marketing efforts of the competing brands” (Markovic et al., 2018, p. 726). As previously mentioned, it was found that brand attachment exercises a significant effect on brand loyalty (Park et al., 2010; Thomson et al., 2005). Also, consumers' brand attachment was found to influence brand loyalty intention in a CSR context (Cha et al., 2016; Vlachos, 2012). In addition, the relationship between customer affective commitment and brand loyalty was previously substantiated (e.g. Markovic et al., 2018).

H11.

Brand attachment has a positive impact on brand loyalty intentions.

2.2.6 Trust and positive word of mouth (PWOM) intention

The relationship between trust and PWOM was extensively studied and verified (e.g. Barreda et al., 2015; Lien and Cao, 2014). In addition, it was found that green scepticism is positively related to negative WOM (Leonidou and Skarmeas, 2017). Also, trust was found to have an effect on positive consumer responses to CSR messages (Miller and Lellis, 2015).

H12.

Trust has a positive impact on PWOM intentions.

2.2.7 Trust and purchase intention

Trust was found to have an effect on purchase intention in a CSR context (e.g. Tian et al., 2011; Semuel and Chandra, 2014). In addition, intrinsic CSR attribution was found to have an effect on purchase intention (Ginder et al., 2021). It was also found that green scepticism has a negative effect on purchase intention (Leonidou and Skarmeas, 2017). In addition, Shazly and Mahrous (2020) conducted a study to determine the main dimensions of cause-related marketing affecting consumer attitudes towards the firm and purchase intention. They found that consumer scepticism has a negative effect on purchase intention. Also, a positive relationship between trust and purchase intention was further verified in the context of ethical fashion (Liu et al., 2021). Moreover, the direct impact of CSR on purchase intention could not be proven, yet, when trust mediated the relationship between CSR and purchase intention, the indirect effect of CSR on purchase intention was corroborated (Semuel and Chandra, 2014). Thus, this study proposes that intrinsic CSR attribution could have a positive impact on trust in the sponsoring firm, which, in turn, could affect purchase intention.

H13.

Trust has a positive impact on purchase intention.

2.2.8 Trust and brand loyalty intention

Many previous studies verified the relationship between trust and brand loyalty (e.g. Sirdeshmukh et al., 2002). In a CSR context, trust was found to mediate the relationship between CSR attribution and brand loyalty intention (Vlachos et al., 2009). Moreover, trust was found to exert an influence on brand loyalty (e.g. Choi and La, 2013; Homburg et al., 2013). Furthermore, despite the fact that the direct effect of CSR on brand loyalty was substantiated. The indirect effect of CSR on loyalty was stronger when it was mediated by trust as well as co-creation (Iglesias et al., 2020).

H14.

Trust has a positive impact on brand loyalty intention.

2.3 Brand attachment and trust as mediators

H15.

Brand Attachment mediates the relationship between intrinsic CSR attribution and PWOM intention.

H16.

Trust mediates the relationship between intrinsic CSR attribution and PWOM intention.

H17.

Brand attachment mediates the relationship between intrinsic CSR attribution and purchase intention.

H18.

Trust mediates the relationship between intrinsic CSR attribution and purchase intention.

H19.

Brand attachment mediates the relationship between intrinsic CSR attribution and brand loyalty intention.

H20.

Trust mediates the relationship between intrinsic CSR attribution and brand loyalty intention.

3. Research method

The study employed a between-subjects experimental design. The sample was composed of 336 university students (54% females). The study depended on a self-administered survey conducted online as a data collection instrument and was launched in Arabic. The study's context was mobile and telecommunications services providers in Egypt. The study comprised two experimental conditions; intrinsic CSR attribution and extrinsic CSR attribution conditions in addition to a control condition.

The participants were first presented with a fictitious written scenario of a mobile and telecommunications services provider (X) that engages in CSR as an experimental stimulus and then they were asked to answer the survey questions regarding the antecedents and consequences of CSR attribution. A fictitious experimental stimulus contributes to a study's internal validity (e.g. Ginder et al., 2021; Ogunfowora et al., 2018).

A pre-test that involved surveying 23 subjects was conducted online to determine the manner of manipulation of the antecedents of CSR attribution in order to reflect a significant difference in the characteristics of the intrinsic vs. the extrinsic CSR attribution conditions. Regarding the duration of commitment to CSR, respondents were asked an open-ended question stating: “What do you think is a firm's long-term commitment to a CSR initiative in terms of years?” The mean score of the responses for the long-term commitment was 4.65 ≈ 5 years. The mode of the responses to the short-term commitment question which was as follows: “What do you think is a firm's short-term commitment to a CSR initiative in terms of months or years?” was one year. As for the amount of invested resources in CSR, two open-ended questions were directed to the responses as follows: “What do you think is a large (small) amount of invested resources in a CSR initiative (in terms of a percentage of profits)?” The mean scores – after excluding the outliers – were 10% for the large amount and 2.43 ≈ 2% for the small amount of invested resources. However, the mode for the small amount was 1%. Hence, in order to provide a more significant difference between the two experimental conditions, the 1% was selected for the small amount. With respect to past corporate social performance, respondents were asked to complete the two following sentences: “A company would be considered one with a long history in CSR if it has been engaging in CSR for …. …. years.” and “A company would be considered one with a short history in CSR if it has been engaging in CSR for …. ….months or years”. The mean score for the long past corporate social performance was 5.5 ≈ 6 years. Nevertheless, it was changed and formulated as follows: “Company X cares about its responsibility towards society and has been engaging in social initiatives almost since its inception” to induce a more significant difference as opposed to the short past corporate social performance which was decided to be one year based on the mode of the responses. As for the CSR-type antecedent, after being presented with 29 possible CSR initiatives selected from the websites of the four mobile and telecommunications services firms operating in Egypt, respondents answered the following 5-point Likert scale regarding each initiative: “The social engagement of Company X focuses on the sustainability of its business” (Business-Process CSR) and “The social engagement of Company X focuses on helping society and people in need” (Philanthropic CSR). Each item of the scale was analysed separately for each initiative. Based on both: mean and mode values, the business-process CSR was decided to be: “Company X for mobile and telecommunications services took on the CSR project of training and qualifying technical employment. This project aimed at training and rehabilitating a number of students and graduates of technical and industrial schools and institutes on the technical services provided by the company and needed by the Egyptian communications market and employing the distinguished trainees in companies working in same field.” The philanthropic CSR initiative was chosen to be “Company X for mobile and telecommunications services cooperated with an NGO specialised in developing and empowering blind people to train blind members on communication and computer skills to prepare them for proper employment and become an integral part in society”.

No pre-test was conducted for the remaining two antecedents; the approach to CSR communication and the call for customer participation in CSR. Regarding the approach to CSR communication, based on a previously conducted exploratory study, it was decided that the minimalist approach to CSR communication be represented as follows: “Company X chose to communicate its CSR activities to its customers and the public through its website and its social media page on Facebook in the form of brief descriptions and reports of this CSR initiative and how the lives of the trainees were improved and positively transformed accompanied by actual photos and videos of the project”. The extravagant or ostentatious approach to CSR communication was portrayed as follows: “Company X chose to communicate its CSR engagement to the public through a TV advertisement using a popular celebrity”. With respect to the call for customer participation in CSR, it was said that “It further offered its young customers and the young public to share in the project by asking them to apply to volunteer in the training process”. In the extrinsic CSR attribution condition, this statement was non-existent.

The manipulation of the antecedents of CSR attribution differed across the treatment conditions. The condition that was designed to elicit more intrinsic CSR attribution was characterised by the firm's long-term commitment to the CSR initiative (5 years), a philanthropic CSR type (developing and empowering blind people through training), a minimalist approach to CSR communication (firm's website and Facebook page), a large amount of resources invested in CSR (10% of profits), a long history of past corporate social performance (since its inception), and the firm's call for customers' participation in its CSR initiatives. In contrast, the extrinsic CSR attribution condition was characterised by a short-term commitment to the CSR initiative (almost a year), a business-process CSR type (training and qualifying technical employment), an ostentatious approach to CSR communication (A TV advertisement), a small amount of invested resources in CSR (1% of profits), a lack of history of past corporate social performance and a lack of the firm's call for customers' participation in its CSR. As for the control condition, the scenario only talked about a mobile and telecommunications services provider (X) that engages in CSR with no reference to any of the antecedents of CSR attribution.

The study depended on previously validated measurement scales except for the measure of the approach to CSR communication and the call for customers' participation in CSR. All measurements depended on 5-point scales.

Firstly, the antecedents of CSR attribution measures are presented as follows: the long-term commitment to CSR and the large amount of invested resources in CSR were adapted from Van den Brink et al. (2006). A two-item Likert scale measured the duration of commitment to CSR. A sample item is “The duration of Company X's CSR initiative is long”. A three-item Likert scale measured the amount of resources invested in CSR. A sample item is “Company X invests a large amount of financial resources in its CSR initiative”. As for the CSR type, the measurement scale developed by Habel et al. (2016) was adapted and utilised. The scale was divided into two; one for the philanthropic CSR type and another for the business-process CSR (two items each). Sample items are “The social engagement of Company X focuses on helping society and people in need” and “The social engagement of Company X focuses on the sustainability of its business”, respectively. Past corporate social performance was measured using a scale developed by Hammad et al. (2014) after adaptation to become a three-item scale. A sample item is “Company X has a long past reputation in social activities”. The approach to CSR communication scale was developed to become a four-item scale and its development was inferred from previously conducted exploratory interviews aimed at exploring the most relevant potential antecedents and consequences of CSR attribution in the Egyptian context. The scale items are: “(1) Company X depends on simple ways in communicating its CSR initiative to its customers and the public, (2) Company X communicates its CSR initiative to its customers and the public in an exaggerated ostentatious way. (Reverse-Coded), (3) Company X spends reasonable resources on communicating its CSR initiative to its customers and the public, and (4) Company X communicates its CSR initiative to its customers and the public in a modest way”. The call for customers' participation in a firm's CSR's scale was developed to become a two-item scale. A sample item is “Company X asks its customers and the public to participate in its CSR initiative”.

Secondly, intrinsic CSR attribution was measured using the three-item scale used in Habel et al. (2016) after adaptation. A sample item is “Company X engages in charitable projects because it is genuinely concerned about being socially responsible”. The same applies to extrinsic CSR attribution. A sample item is “Company X supports good causes to take advantage of the cause”.

Thirdly, the consequences of CSR attribution were measured as follows: a brand attachment measurement scale developed by Park et al. (2010) was utilised after adaptation. It is a four-item scale and a sample item is “To what extent could Company X be a part of you and who you are?” Trust was measured using the scale used by Vlachos et al. (2009). It is composed of three semantic differential items (e.g. Very undependable/Very dependable) in addition to a single Likert scale item “Company X is generally honest and trustworthy”. PWOM intention was measured using a three-item scale used by Markovic et al. (2018) after adaptation. A sample item is “I could say positive things about Company X to other people”. A four-item measurement scale of purchase intention used by Lin et al. (2011) was adapted and adopted. A sample item is “If given the chance, I intend to purchase from company X”. Finally, a three-item measurement scale of brand loyalty intention used by Markovic et al. (2018) was adapted and adopted. A sample item is “If given the chance, I will consider Company X my first choice when I purchase the services they supply”.

4. Results

4.1 Manipulation checks

A manipulation check was conducted to ensure the existence of a statistically significant difference in the manipulation of the antecedents of CSR attribution in the experimental stimuli in the two experimental conditions; the intrinsic CSR attribution and the extrinsic CSR attribution. An independent samples t-test and its non-parametric alternative Mann–Whitney U test were conducted. A statistically significant difference was detected in all of the antecedents of CSR attribution between the intrinsic and the extrinsic conditions except for the amount of resources invested in CSR (Table 2).

4.2 Hypotheses testing

Firstly, a one-way between-groups ANOVA was conducted for two purposes. Firstly, it could serve as a further manipulation check of the statistical difference of the intrinsic CSR attribution variable across the treatment conditions. Secondly, it could contribute to testing the B-hypotheses 1 till 6 so long as a statistically significant difference has already been proven in all the antecedents of CSR attribution between the intrinsic and the extrinsic CSR attribution conditions except for the amount of resources invested in CSR. The results revealed a statistically significant difference in the intrinsic CSR attribution variable at the p < 0.05 level for the three groups: F (2, 333) = 4.5, p = 0.01. However, the difference in mean scores between the groups was small. The effect size was 0.03 as calculated using eta-squared. Post-hoc comparisons using the Tukey HSD test showed that the mean score for Group 1: Intrinsic CSR attribution (M = 3.5, SD = 0.74) was significantly different from Group 2: Extrinsic CSR attribution (M = 3.2, SD = 0.86). Group 3: Control (M = 3.47, SD = 0.90) did not differ significantly from either Group 1 or 2. Hence, as a preliminary conclusion, H1b, H2b, H3b, H5b and H6b could be supported. Further support will be provided after conducting the Partial Least Squares-Structural Equation Modelling PLS-SEM analysis when significant effects of the duration of commitment, CSR-type, approach to CSR communication, past corporate social performance and the call for customers' participation in CSR on intrinsic CSR attribution could be substantiated.

Despite not being hypothesised, extrinsic CSR attribution was examined in each treatment condition to further verify the contrast. The one-way between groups ANOVA results confirmed a statistically significant difference at the p-level < 0.01 in the extrinsic CSR attribution across the three treatment conditions: F (2, 333) = 8.096, p = 0.000. Post-Hoc comparisons using the Tukey HSD test revealed that the mean score for both the extrinsic CSR attribution condition (M = 3.31, SD = 0.71) and the control condition (M = 3.16, SD = 0.87) were significantly different from the intrinsic CSR attribution condition (M = 2.89, SD = 0.86). The effect size was 0.05 which is a small to medium effect. The significant difference was further confirmed between the intrinsic and extrinsic CSR attribution conditions via t-test (where t (235) = −4.088, p = 0.000, two-tailed). The magnitude of the difference in the means (mean difference = 0.42, 95% CI: 0.62 to −0.22) was moderate (eta squared = 0.07).

Secondly, a PLS-SEM analysis was conducted using SmartPLS4 software to test the remaining hypotheses. The proposed model was run three times for each treatment condition: intrinsic CSR attribution, extrinsic CSR attribution and control. As for the measurement model, an adequate level of convergent validity was verified by the Average Variance Extracted (AVE) values for all measurement scales of all the constructs across all three conditions as they were all above the 0.5 threshold except for the Approach to CSR Communication scale in the Intrinsic CSR Attribution condition (0.47 after removing the second item) yet it still could be approximated to 0.5. Nevertheless, in both the extrinsic CSR attribution and the control conditions, the AVE scores were 0.672 and 0.674, respectively. The highest AVE score was 0.89 for the CSR type (Philanthropic) scale in the control condition. Most of the indicators loaded on their respective constructs at the 0.7 threshold or above. Most indicators' loadings ranged from 0.7 (1st item of the Call for Customers' Participation in CSR (intrinsic condition)) till 0.991 (2nd item of the Duration of Commitment to CSR (intrinsic condition)). Only two indicators had loadings below 0.7 in the intrinsic CSR attribution condition, yet, these same items had loadings above 0.7 in the remaining two conditions. Thus, convergent validity has been ensured for all indicators.

Discriminant validity was also assessed by the Heterotrait-Monotrait (HTMT) ratio of correlations, the Fornell-Larcker criterion and cross-loadings of all constructs in each treatment condition separately. The HTMT ratios of all constructs in the intrinsic CSR attribution condition showed a satisfactory level of discriminant validity with values below the 0.85 threshold (Henseler et al., 2015) and ranged from 0.835 (Brand Loyalty Intention to Purchase Intention) till 0.157 (Intrinsic CSR Attribution to Duration of Commitment to CSR) except for the Call for Customers' Participation in CSR to the Approach to CSR Communication (1.002). Yet, the HTMT ratio of these two constructs was below 0.85 in the remaining two conditions. In the extrinsic CSR attribution condition, most constructs had HTMT ratios lower than 0.85 ranging from 0.839 (Brand Loyalty Intention to trust) to 0.155 (Amount of Invested Resources in CSR to CSR Type (Business-Process)). Nevertheless, Purchase Intention to Trust and Brand Loyalty Intention to Purchase Intention had HTMT values of 0.910 and 0.913, respectively. Yet, these ratios were below 0.85 in the intrinsic CSR attribution condition. As for the control condition, most HTMT values were below 0.85 ranging from 0.801 (Brand Loyalty Intention) to 0.149 (Intrinsic CSR Attribution to CSR Type (Business-Process)). However, the HTMT ratio of Purchase Intention to PWOM was 0.919. Nevertheless, the exact ratio was 0.785 in the intrinsic CSR attribution condition.

The Fornell-Larcker criterion manifested a high level of discriminant validity across the three treatment conditions as the square root of the AVE scores for each construct was greater than each construct's correlation with other constructs. In addition, cross-loadings showed that all indicators had higher loadings (larger than 0.5) on the constructs they were intended to measure than on the remaining constructs except for the 1st item of the Duration of Commitment to CSR construct in the intrinsic CSR attribution condition which was 0.395 ≈ 0.4 (less than 0.5). Yet, the loading of the same item in the other two conditions was 0.784 and 0.852.

Reliability was measured using both composite reliability (CR) and Cronbach's Alpha (CA). In PLS-SEM, CR is considered superior to CA in terms of accuracy (Hair et al., 2019). CR values of all constructs in the three treatment conditions confirmed a high level of reliability and ranged from 0.720 (Approach to CSR Communication (intrinsic condition)) to 0.942 (CSR Type-Philanthropic (Control Condition)) except for the Duration of Commitment to CSR construct in the intrinsic condition which had a composite reliability score of 0.69 which could still approximately be considered 0.7. Also, the CR score of the duration of commitment to CSR in the other conditions was high (0.757 and 0.786). CA scores were also above the 0.7 threshold for most constructs further confirming internal consistency reliability. CA scores ranged from 0.733 (CSR Type Philanthropic (intrinsic condition)) to 0.912 (Amount of Invested Resources (intrinsic condition)). Even though a few constructs had a CA score lower than 0.7 in a specific condition, this was compensated for either in the remaining conditions or by their corresponding CR score.

As for the explanatory power of the measurement model, the R-squared value of the study's focal construct which is the intrinsic CSR attribution was satisfactory in the three conditions as follows: 0.368 or 37% (Intrinsic condition), 0.478 or 48% (Extrinsic condition) and 0.571 or 57% (Control condition). As for the remaining endogenous variables' R-squared values, they were adequate and ranged from weak 0.281 (Trust (Intrinsic condition)) till moderate 0.654 (Purchase intention (Extrinsic condition)).

Regarding the structural model, the proposed model was run three times for each treatment condition. Based on the results (kindly, refer to Table 3), H1a was rejected as there was no significant relationship between the duration of commitment to CSR and intrinsic CSR attribution in any of the three conditions; Intrinsic CSR attribution condition (Coefficient = −0.037; t-value = 0.409; p = 0.341), Extrinsic CSR attribution condition (Coefficient = 0.147; t-value = 1.592; p = 0.056) and Control condition (Coefficient = 0.097; t-value = 0.948; p = 0.172). Therefore, H1b was rejected as well. There was no effect in the intrinsic CSR attribution condition (ƒ² = 0.002) and a small effect was detected in the extrinsic CSR attribution condition (ƒ² = 0.028). H2a was partially supported since H2a1 was partially supported while H2a2 was rejected. CSR type – Philanthropic was found to have a significant positive effect on intrinsic CSR attribution in the intrinsic CSR attribution condition (Coefficient = 0.190; t-value = 1.664; p < 0.05) and the control condition (Coefficient = 0.198; t-value = 2.015; p < 0.05). Yet such effect was not found in the extrinsic CSR attribution condition (Coefficient = 0.067; t-value = 0.705; p = 0.240). Business-process CSR did not have a significant relationship with intrinsic CSR attribution in any of the three conditions. Thus, H2b was partially supported as well. The effect size (ƒ²) of Philanthropic CSR on intrinsic CSR attribution was small in the intrinsic condition (0.042) and was non-existent in the extrinsic condition (0.006). Business-process CSR had no effect on intrinsic CSR attribution in both the intrinsic (ƒ² = 0.005) and the extrinsic conditions (ƒ² = 0.000). H3a was supported since the approach to CSR communication exerted a significant positive impact on intrinsic CSR attribution in all conditions; intrinsic CSR attribution (coefficient = 0.387, t-value = 4.108, p < 0.001), extrinsic CSR attribution (coefficient = 0.309, t-value = 3.042, p < 0.01) and control condition (coefficient = 0.384, t-value = 3.544, p < 0.001). Hence, H3b was supported as well especially that the positive effect was stronger in the intrinsic CSR attribution condition (ƒ² = 0.15; medium effect) than in the extrinsic CSR attribution condition (ƒ² = 0.11; small effect). H4a was rejected since the amount of resources invested in CSR had no significant impact on intrinsic CSR attribution in any of the three conditions; intrinsic CSR attribution (Coefficient = 0.085; t-value = 0.784, p = 0.217), extrinsic CSR attribution (Coefficient = −0.085; t-value = 0.830, p = 0.203) and control condition (Coefficient = 0.127; t-value = 1.271, p = 0.102). Therefore, H4b was rejected as well. As a further proof, the effect sizes (ƒ²) in the intrinsic and the extrinsic conditions were 0.009 and 0.008, respectively showing no effect. H5a was supported because past corporate social performance had a significant positive impact on intrinsic CSR attribution in all conditions; intrinsic CSR attribution condition (coefficient = 0.224; t-value = 2.088; p < 0.05), extrinsic CSR attribution condition (coefficient = 0.220; t-value = 1.958; p < 0.05) and control condition (coefficient = 0.223; t-value = 1.958; p < 0.05). Thus, H5b was supported as well. The effect size (ƒ²) values were small in the intrinsic condition (0.06) and the extrinsic condition (0.054). H6a was partially supported as the call for customers' participation in firm's CSR had a significant positive impact on intrinsic CSR attribution in the extrinsic CSR attribution condition. Thus, H6b was partially supported.

H7 was supported since a significant positive impact of intrinsic CSR attribution on brand attachment was found in all three conditions; intrinsic CSR attribution condition (coefficient = 0.533; t-value = 7.690; p < 0.001), extrinsic CSR attribution (coefficient = 0.435; t-value = 5.562; p < 0.001) and control condition (coefficient = 0.657; t-value = 11.320; p < 0.001). The effect size (ƒ²) was large (0.4) in the intrinsic CSR attribution condition and medium (0.234) in the extrinsic CSR attribution condition. H8 was supported as a significant positive impact of intrinsic attribution on trust was proven in all conditions; intrinsic CSR attribution condition (coefficient = 0.530; t-value = 5.727; p < 0.001), extrinsic CSR attribution (coefficient = 0.544; t-value = 7.328; p < 0.001) and control condition (coefficient = 0.666; t-value = 12.555; p < 0.001). The effect size (ƒ²) was large in both the intrinsic and extrinsic CSR attribution conditions; 0.4 and 0.42, respectively.

H9 was rejected because brand attachment did not exert any impact on PWOM intention in any of the three conditions; intrinsic CSR attribution (coefficient = 0.047; t-value = 0.507; p = 0.306), extrinsic CSR attribution (coefficient = 0.041; t-value = 0.504; p = 0.307) and control condition (coefficient = 0.070; t-value = 0.662; p = 0.254). The effect (ƒ²) was also non-existent in the three conditions, the intrinsic CSR attribution (0.002), the extrinsic CSR attribution (0.003) and the control (0.006). H10 was partially supported since a significant positive impact of brand attachment on purchase intention was manifested only in the extrinsic CSR attribution condition (coefficient = 0.179; t-value = 2.577; p < 0.01). However, this impact was not found in either the intrinsic CSR attribution (coefficient = 0.107; t-value = 1.250; p = 0.106) or the control condition (coefficient = 0.018; t-value = 0.253; p = 0.400). The effect size (ƒ²) value was 0.014 in the intrinsic (no effect) and the 0.000 in the control (no effect) conditions, yet the effect was small in the extrinsic condition (0.07). H11 was partially supported since brand attachment had a significant positive effect on brand loyalty intention in both the intrinsic CSR attribution condition (coefficient = 0.218; t-value = 2.188; p < 0.05) and the extrinsic CSR attribution condition (coefficient = 0.186; t-value = 1.960; p < 0.05). The effect size (ƒ²) was 0.044 in the intrinsic condition and 0.06 in the extrinsic condition which are both small. There was no effect in the control condition (0.008). H12 was supported as a significant positive impact of trust on PWOM intention in all conditions; intrinsic CSR attribution (coefficient = 0.496; t-value = 4.371; p < 0.001), extrinsic CSR attribution (coefficient = 0.762; t-value = 9.731; p < 0.001) and control condition (coefficient = 0.762; t-value = 7.637; p < 0.001). The effect size (ƒ²) values in the intrinsic and the extrinsic conditions were 0.3; medium and 0.9; large, respectively. H13 was supported because a significant positive effect of trust on purchase intention was shown in all conditions; intrinsic CSR attribution (coefficient = 0.464; t-value = 5.518; p < 0.001), extrinsic CSR attribution (coefficient = 0.658; t-value = 9.822; p < 0.001) and control condition (coefficient = 0.707; t-value = 9.494; p < 0.001). The effect size (ƒ²) values in the intrinsic and extrinsic conditions were 0.3; medium and 0.8; large, respectively. H14 was supported since a significant positive impact of trust on brand loyalty intention was manifested in all conditions; intrinsic CSR attribution (coefficient = 0.349; t-value = 3.753; p < 0.001), extrinsic CSR attribution (coefficient = 0.556; t-value = 5.891; p < 0.001) and control condition (coefficient = 0.589; t-value = 4.700; p < 0.001). The effect size (ƒ²) values in the intrinsic and extrinsic conditions were 0.112; small and 0.5; large, respectively.

4.2.1 Mediation analysis

H15 was rejected since brand attachment failed to exert a mediating effect on the relationship between intrinsic CSR attribution and PWOM intention in any of the three treatment conditions. In the intrinsic CSR attribution condition, a non-significant indirect effect of intrinsic CSR attribution on PWOM intention through brand attachment was found (β = 0.025, t = 0.496, p = 0.310). Yet, the total effect of intrinsic CSR attribution on PWOM intention was significant (β = 0.502, t = 6.954, p < 0.001). Also, the direct effect of intrinsic CSR attribution on PWOM intention was significant (β = 0.214, t = 2.402, p < 0.01). As for the extrinsic CSR attribution condition, a non-significant indirect effect of intrinsic CSR attribution on PWOM intention through brand attachment was found (β = 0.018, t = 0.482, p = 0.315). Also, the direct effect of intrinsic CSR attribution on PWOM intention was not significant (β = −0.020, t = 0.186, p = 0.426). The results of the control condition showed that the indirect effect of intrinsic CSR attribution on PWOM intention through brand attachment was not significant (β = 0.046, t = 0.646, p = 0.259). Also, the direct effect of intrinsic CSR attribution on PWOM intention was insignificant (β = −0.052, t = 0.553, p = 0.290).

H16 was supported as the mediation effect of trust on the relationship between intrinsic CSR attribution and PWOM intention was significant in the three treatment conditions. In the intrinsic CSR attribution condition, the indirect effect was significant (β = 0.263, t = 3.260, p < 0.01). The total effect was significant (β = 0.502, t = 6.954, p < 0.001). The direct effect was significant (β = 0.214, t = 2.402, p < 0.01). Hence, trust exercises a complementary partial mediating effect on the relationship between intrinsic CSR attribution and PWOM intention. In the extrinsic CSR attribution condition, the indirect effect was significant (β = 0.414, t = 5.391, p < 0.001). The total effect was significant (β = 0.413, t = 4.527, p < 0.001). The direct effect was insignificant (β = −0.020, t = 0.186, p = 0.426). Therefore, a full mediation effect was evident. In the control condition, the indirect effect was significant (β = 0.507, t = 6.004, p < 0.001). The total effect was significant (β = 0.501, t = 6.724, p < 0.001). As for the direct effect was insignificant (β = −0.052, t = 0.553, p = 0.290). Hence, a full mediation effect was evident.

H17 was partially supported because the mediation effect of brand attachment on the relationship between intrinsic CSR attribution and purchase intention was significant only in the extrinsic CSR attribution condition. In the intrinsic CSR attribution condition, the indirect effect was insignificant (β = 0.057, t = 1.225, p = 0.110). The direct effect was significant (β = 0.248, t = 3.234, p < 0.01). In the extrinsic CSR attribution condition, the indirect effect was significant (β = 0.078, t = 2.297, p < 0.05). The total effect was significant (β = 0.512, t = 6.338, p < 0.001). The direct effect was insignificant (β = 0.076, t = 0.936, p = 0.175). Thus, a full mediation effect was evident. The indirect effect in the control condition was insignificant (β = 0.012, t = 0.246, p = 0.403). The direct effect was insignificant (β = 0.078, t = 0.829, p = 0.204) indicating no effect.

H18 was supported since a significant mediating role of trust between intrinsic CSR attribution and purchase intention was evident in all three conditions. In the intrinsic CSR attribution condition, the indirect effect was significant (β = 0.246, t = 3.616, p < 0.001). The total effect was significant (β = 0.551, t = 8.042, p < 0.001). The direct effect was significant (β = 0.248, t = 3.234, p < 0.01). Hence, a complementary partial mediation effect was evident. In the extrinsic CSR attribution condition, the indirect effect is significant (β = 0.358, t = 5.583, p < 0.001). The total effect was significant (β = 0.512, t = 6.338, p < 0.001). The direct effect was insignificant (β = 0.076, t = 0.936, p = 0.175). Hence, a full-mediation effect was evident. In the control condition, indirect effect was significant (β = 0.470, t = 7.051, p < 0.001). The total effect was significant (β = 0.560, t = 7.879, p < 0.001). The direct effect was insignificant (β = 0.078, t = 0.829, p = 0.204). Hence, a full mediation effect was evident.

H19 was partially supported as a significant mediating effect of brand attachment on the relationship between intrinsic CSR attribution and brand loyalty intention was detected in both the intrinsic and extrinsic CSR attribution conditions. In the intrinsic CSR attribution condition, the indirect effect was significant (β = 0.116, t = 2.119, p < 0.05). Also, the total effect was significant (β = 0.419, t = 4.845, p < 0.001). However, the direct effect was insignificant (β = 0.119, t = 1.229, p = 0.109). Hence, a full mediation effect was evident. In the extrinsic CSR attribution condition, the indirect effect was significant (β = 0.081, t = 1.764, p < 0.05). The total effect was significant (β = 0.540, t = 6.538, p < 0.001). The direct effect was insignificant (β = 0.157, t = 1.621, p = 0.053). Hence, a full mediation effect was evident. In the control condition, the indirect effect was insignificant (β = 0.060, t = 0.811, p = 0.209). The total effect was significant (β = 0.536, t = 7.354, p < 0.001). The direct effect was insignificant (β = 0.084, t = 0.733, p = 0.232). Hence, there is no effect.

H20 was supported as a significant mediating effect of trust on the relationship between intrinsic CSR attribution and brand loyalty intention was evident in the three treatment conditions. In the intrinsic CSR attribution condition, the indirect effect was significant (β = 0.185, t = 2.890, p < 0.01). The total effect was significant (β = 0.419, t = 4.845, p < 0.001). The direct effect was insignificant (β = 0.119, t = 1.229, p = 0.109). Hence, a full mediation effect was evident. In the extrinsic CSR attribution, the indirect effect was significant (β = 0.303, t = 4.348, p < 0.001). The total effect was significant (β = 0.540, t = 6.538, p < 0.001). The direct effect was insignificant (β = 0.157, t = 1.621, p = 0.053). Hence, a full mediation effect was evident. In the control condition, the indirect effect was significant (β = 0.392, t = 4.108, p < 0.001). The total effect was significant (β = 0.536, t = 7.354, p < 0.001). The direct effect was insignificant (β = 0.084, t = 0.733, p = 0.232). Hence, a full mediation effect was evident.

5. Discussion

5.1 Antecedents of CSR attribution

The current study proposed six antecedents of CSR attribution. Firstly, the duration of commitment to CSR was found to exert no effect on intrinsic CSR attribution. This result does not comply with most previous studies (e.g. Choi et al., 2016; Ellen et al., 2006; Webb and Mohr, 1998; Yang et al., 2015). For example, long-term commitment to cause-related marketing in Egypt was found to exercise a significant effect on consumers' positive motivational attribution (Hammad et al., 2014). Yet, the commitment to CSR was found to exert no impact on consumers' responses (Pérez et al., 2019). One potential reason for such conflicting results may be that the manipulation of the duration of commitment to CSR was not properly perceived by the subjects. The low loading of the first item of the measure of 0.395 may be a sign of this interpretation. However, this low loading was detected only in the intrinsic CSR attribution condition.

Secondly, it was found that CSR type – philanthropic could exert a significant positive effect on intrinsic CSR attribution. In contrast, CSR type – Business-process failed to exercise any impact on intrinsic CSR attribution. Thus, if a firm adopts philanthropic CSR, there is a probability that consumers are going to attribute the firm's CSR engagement to intrinsic motives (intrinsic CSR attribution). This finding is rather in line with previous research. For example, it was found that philanthropic CSR enhances the positive effect of CSR adoption on customers' perceptions of the level of the firm's CSR costs (Habel et al., 2016).

Thirdly, the approach to CSR communication was found to exercise a significant positive effect on intrinsic CSR attribution. That is, if a firm depends on a minimalist approach in communicating its CSR engagement to customers and the public in general, this could have a positive effect on consumers' intrinsic CSR attribution. Moreover, this positive effect was stronger in the intrinsic CSR attribution condition (medium effect) than in the extrinsic CSR attribution condition (small effect). The reason for this is that the manipulation of the approach to CSR communication in the extrinsic CSR attribution condition was designed to reflect an ostentatious approach rather than a minimalist one. That was a further proof that a minimalist approach contributed to the higher level of intrinsic CSR attribution in the intrinsic CSR attribution condition than in the extrinsic CSR attribution condition. These results are in accordance with many previous studies. It was previously highlighted that an ostentatious approach to CSR communication could lead to consumers' extrinsic CSR attribution (Dhanesh, 2015). Also, Ginder et al. (2021) advised firms to adopt discreet CSR – in which a firm does not communicate its CSR to the public – rather than uniform CSR positioning. It was also found that firms prefer green-hushing to avoid consumer scepticism. In addition, it was found that information specificity – which could be considered an instrumental aspect of the approach to CSR communication – has a positive impact on message authenticity (Pérez et al., 2020) which is closely related to CSR attribution (Alhouti et al., 2016). Hence, the current study recommends a minimalist approach to CSR communication as a compromise between a silent approach and an ostentatious one. The advantage of a minimalist approach is two-fold; it gets to inform the public in general and consumers in particular of a firm's CSR engagement, yet, through simple and non-extravagant methods, thus avoiding consumers' scepticism.

Fourthly, contrary to most previous studies, the amount of resources invested in CSR was found to exercise no impact on intrinsic CSR attribution. Many studies identified the amount of invested resources in CSR as an essential variable affecting CSR attribution (e.g. Hoeffler and Keller, 2002; Irmak et al., 2015). Such contradiction of findings may be due to the lack of significant difference in the manipulation of the amount of resources invested in CSR between the intrinsic and the extrinsic CSR attribution conditions as shown in Table 2. However, a contrast between the two conditions in the level of the amount of invested resources in CSR was deliberately inflicted as shown in the research method section.

Fifthly, a firm's past corporate social performance was found to have a significant positive impact on intrinsic CSR attribution. That is, when a sponsoring firm has a long history in CSR engagement, consumers could attribute such firm's CSR initiatives to intrinsic motives. This finding is in accordance with most previous studies' findings. A firm's history of CSR involvement was shown to impact consumers' perception of such firm's CSR (e.g. Du et al., 2010; Ginder et al., 2021). A firm's past corporate social performance was found to exercise a positive influence on consumers' intrinsic CSR attribution (e.g. García-De los Salmones and Perez, 2018; Hammad et al., 2014; Leonidou and Skarmeas, 2017; Lyu, 2017).

Sixthly, a firm's call for customers' participation in its CSR initiatives might have a positive impact on consumers' intrinsic CSR attribution. This finding is, to an extent, in line with a study that suggested that a firm's call for customers' participation in its CSR could contribute to positive consumer consequences such as personal identification, brand attachment and brand loyalty (Cha et al., 2016).

5.2 Consequences of CSR attribution

Two direct consequences of CSR attribution were examined and verified; brand attachment and trust. Intrinsic CSR attribution was found to have a significant positive impact on brand attachment. Thus, when a firm adopts a CSR initiative that is capable of eliciting consumer's intrinsic CSR attribution, such intrinsic CSR attribution could improve consumers' attachment to the sponsoring firm. A previous study by Vlachos (2012) demonstrated that a firm's corporate social performance exercises a significant influence on consumer's emotional attachment to such sponsoring firm. Intrinsic CSR attribution could also exert a significant positive impact on trust. Therefore, when consumers perceive a firm's CSR engagement as intrinsically-motivated, their trust towards such firm could increase. This finding is in agreement with several previous studies (e.g. Vlachos et al., 2009). Moreover, Miller and Lellis (2015) identified consumers' perception of the intent of an advertiser's campaign as a component of their trust in the advertiser. According to Chen and Chang (2013), there is a negative relationship between CSR washing and trust. It was also found that perceived CSR message authenticity exercises a positive impact on trust (Pérez et al., 2020). Also, intrinsic CSR attribution contributes to consumers’ perception of the sponsoring firm as an institution that holds values such as sincerity and philanthropy.

Three indirect consequences of CSR attribution were investigated; PWOM intention, purchase intention and brand loyalty intention. Brand attachment was found to exert no influence on PWOM intention. This finding contradicts with most previous research. Principally, consumers' brand attachment was found to have an impact on their brand loyalty, PWOM and promoting the brand (Park et al., 2010). CSR attribution was also found to exercise an influence on consumers' recommendation intention (Vlachos et al., 2009), extra-role behaviours such as PWOM (Karaosmanoglu et al., 2016) and brand advocacy (Aljarah et al., 2022; Pérez et al., 2019). It was found that brand attachment might exercise an influence on purchase intention. This comes in congruence with many previous studies. For example, Park et al. (2010) emphasised that consumers' brand attachment could predict their readiness to take on difficult-to-enact behaviours and actual purchasing. Also, Koh and Fang (2012) found that CSR exerts a positive impact on purchase intention through brand attachment. In addition, Pérez et al. (2019) found that intrinsic CSR attribution has a positive effect on purchase intention. Brand attachment was also found to exert a significant positive effect on brand loyalty intention. Thus, if a firm engages in CSR that reflects intrinsic motives, consumers are more likely to experience brand attachment, which in turn, affects their intention to be loyal to this brand. This finding is in line with previous research. For example, brand attachment predicted brand loyalty better than attitude (Park et al., 2010; Thomson et al., 2005). Corporate social performance was found to influence brand loyalty intention through emotional attachment (Vlachos, 2012). Such results were also found in B2B contexts (Homburg et al., 2013).

Trust was found to exert a significant positive effect on PWOM intention which represents accordance with previous research's results. Trust was previously found to have a positive effect on PWOM intention (e.g. Barreda et al., 2015; Lien and Cao, 2014). Also, CSR attribution and trust were found to exercise a positive impact on positive consumer responses to CSR communication (Miller and Lellis, 2015). Moreover, green scepticism was associated with negative word of mouth (Leonidou and Skarmeas, 2017). Trust exercised a significant positive effect on purchase intention. This finding comes as an augmentation of a history of corroboration of such relationship. In a CSR context, trust was found to exercise a significant impact on purchase intention (e.g. Tian et al., 2011; Semuel and Chandra, 2014). In contrast, green scepticism was negatively associated with purchase intention (Leonidou and Skarmeas, 2017). Finally, trust was found to exercise a significant positive impact on brand loyalty intention as in several previous studies (e.g. Sirdeshmukh et al., 2002). Also, CSR perceptions exerted an impact on trust, which in turn, exerted a significant impact on brand loyalty (Choi and La, 2013; Homburg et al., 2013; Liu et al., 2020).

5.3 Mediation of brand attachment and trust

Brand attachment failed to act as a mediator variable between intrinsic CSR attribution and PWOM intention. This finding is not in line with previous studies' findings. CSR attribution was found to exercise an impact on brand advocacy, recommendations intentions and consumer responses to CSR (Aljarah et al., 2022; Karaosmanoglu et al., 2016; Vlachos et al., 2009). Also, brand attachment was previously found to contribute to consumers' proneness to promote the brand (Park et al., 2010). In contrast, trust was found to exert a mediating effect on the relationship between intrinsic CSR attribution and PWOM intention. This finding is congruent with most previous studies. For example, trust was found to act as a mediating variable in the relationship between CSR attribution and consumers' recommendation intentions (Vlachos et al., 2009). CSR attribution was regarded as a component part of trust and trust was found to influence consumers' responses to CSR messages (Miller and Lellis, 2015). Also, CSR attribution was found to have an effect on trust (e.g. Ginder et al., 2021; Moscato and Hopp, 2019) and trust to have an effect on PWOM (e.g. Leonidou and Skarmeas, 2017).

Brand attachment may exert a mediating effect on the relationship between intrinsic CSR attribution and purchase intention. However, this inference should be treated with caution since the effect was evident only in the extrinsic CSR attribution condition. Yet, this result conforms, to an extent, to previous studies. For example, CSR was found to have a positive impact on purchase intention (Sen et al., 2016). More specifically, Koh and Fang (2012) verified the mediating effect of emotional attachment on the relationship between CSR and purchase intention. Also, Park et al. (2010) emphasised that brand attachment predicts consumers' intentions to engage in difficult-to-enact behaviours such as those that require resources expenditure. Trust was found to play a significant mediating role in the relationship between intrinsic CSR attribution and purchase intention. This finding is in congruence with previous studies. For example, in a CSR context, trust was found to exert a positive impact on purchase intention (e.g. Tian et al., 2011; Semuel and Chandra, 2014). CSR washing was found to lead to extrinsic CSR attribution (Ginder et al., 2021). CSR washing was also found to exert a negative impact on trust (Chen and Chang, 2013). Therefore, CSR attribution and trust are closely related. Moreover, intrinsic CSR attribution was found to exercise a positive impact on purchase intention (Ginder et al., 2021) while consumers' scepticism towards green CSR initiatives could have a negative effect on purchase intention (Leonidou and Skarmeas, 2017).

Brand attachment was found to exercise a mediating effect on the relationship between intrinsic CSR attribution and brand loyalty intention. This conforms to previous studies' findings. Emotional attachment was previously found to exert a mediating effect on the relationship between corporate social performance and brand loyalty intention (Vlachos, 2012). Emotional attachment predicted brand loyalty better than attitude (Thomson et al., 2005) and contributed to brand loyalty (Park et al., 2010). Trust was found to play a significant mediating role in the relationship between intrinsic CSR attribution and brand loyalty intention. This is also in line with previous research findings. Intrinsic CSR attribution and trust were found to be related in several studies (e.g. Ginder et al., 2021; Miller and Lellis, 2015; Moscato and Hopp, 2019; Vlachos et al., 2009). Also, consumers' CSR perceptions had an effect on brand loyalty through trust (Choi and La, 2013; Liu et al., 2020). Also, CSR attribution and brand loyalty were found to be indirectly related (Konalingam et al., 2022). In addition, value-driven, strategic and stakeholder-driven CSR attributions were found to affect trust which, in turn, had an influence on customer loyalty (Min et al., 2023). In a B2B context, business-process CSR had a positive effect on customer loyalty through trust (Homburg et al., 2013).

6. Conclusion

The current study aimed at constructing and investigating a practical and concise model of the most relevant antecedents and consequences of CSR attribution. It also aimed to contribute to the CSR attribution literature by including the brand attachment variable as a mediating variable that could further explain the relationship between CSR attribution and the consequent consumer outcomes. Several research gaps were attempted to be filled through the execution of this study; for example, inadequate studies with the aim of developing a concise model of the most relevant antecedents and consequences of CSR attribution, limited CSR attribution-related studies examining multiple antecedents simultaneously, the limited availability of CSR attribution centric studies in Egypt and the need for a deeper understanding of the psychological mechanism underlying consumers' CSR attribution. The impact of six proposed antecedents of CSR attribution; duration of commitment to CSR, CSR type, approach to CSR communication, amount of invested resources in CSR, past corporate social performance and call for customer participation in CSR was investigated on CSR attribution. Then the impact of CSR attribution on three positive consumer outcomes; PWOM, purchase intention and brand loyalty intention was examined as mediated by both brand attachment and trust.

The study employed a between-subjects experimental design with three treatment conditions; intrinsic CSR attribution, extrinsic CSR attribution and control. The collected data from 336 subjects were analysed via both one-way between groups ANOVA and PLS-SEM. After analysing the data, the following could be concluded. Regarding the antecedents of CSR attribution, CSR type – philanthropic may have a significant impact on intrinsic CSR attribution. The approach to CSR communication may exercise a significant impact on intrinsic CSR attribution. A minimalist approach to CSR communication may contribute to a higher level of consumers' intrinsic CSR attribution than an ostentatious approach to CSR communication. A firm's past corporate social performance may have a significant effect on intrinsic CSR attribution. A sponsoring firm that has a long history in CSR engagement has a higher probability of being perceived as more intrinsically motivated in its CSR endeavours than a firm with no history in CSR. Lastly, a firm's call for customers' participation in its CSR might have a significant impact on intrinsic CSR attribution. The order of importance of the significant antecedents of CSR attribution could be presented as follows based on the path coefficients and the effect size (ƒ²) values: first, the approach to CSR communication, second, the firm's past corporate social performance, third, CSR type – philanthropic, and finally, the call for customers' participation in CSR.

Two significant direct consequences of intrinsic CSR attribution were brand attachment and trust. Three indirect consequences were PWOM intention, purchase intention and brand loyalty intention. Brand attachment might act as a mediating variable in the relationship between intrinsic CSR attribution and purchase intention. Also, brand attachment may exert a mediating effect on the relationship between intrinsic CSR attribution and brand loyalty intention. However, it could have no effect on PWOM intention. As for trust, it succeeded in playing a mediating role in the relationship between intrinsic CSR attribution and the three indirect consequences. In contrast, the duration of commitment to CSR, CSR type – business-process and the amount of invested resources in CSR could not be substantiated as antecedents of intrinsic CSR attribution.

The lack of support that was found for some of the proposed hypotheses such as H1, H4 and H9 could be attributed to the difference of the context of research. That is, the current study is conducted in Egypt which could be considered an emerging market. Thus, maybe the unsupported hypotheses indicate that they don't apply to the Egyptian context, thus establishing boundary conditions (Burgess and Steenkamp, 2006).

Lastly, the proposed antecedents of intrinsic CSR attribution could provide a moderate explanation of the variance in the intrinsic CSR attribution construct as evidenced by the average of the R-squared values in both the intrinsic and extrinsic CSR conditions (42.5%).

6.1 Theoretical implications

One of the main contributions of this study is its attempt to arrive at a concise conceptual model of the most relevant antecedents and consequences of CSR attribution to the Egyptian context. Even though most of the studied constructs have been previously explored, such exploration has been conducted on a fragmented and sporadic level rather than on a comprehensive level (e.g. Koh and Fang, 2012; Vlachos et al., 2009). Hence, this study aspires to have contributed to the literature via aligning these concepts together simultaneously in a comprehensive model and examining their collective impact using the PLS-SEM data analysis procedure. Nevertheless, a study with a somewhat similar aim was previously conducted by Hammad et al. (2014), yet several differences remain. Moreover, this study contributes to a more profound understanding of the psychological mechanism underlying the process by which consumers form their perceptions of firms' CSR as in Pérez and del Bosque (2013) and specifically how they form their CSR attributions by examining and substantiating the proposed conceptual model which was called for by several studies (e.g. Alhouti et al., 2016; Hur and Kim, 2017).

This study also presents a relatively new antecedent of CSR attribution which was not sufficiently investigated in previous CSR attribution-related studies which is firms' call for customer participation in their CSR and it was verified – to an extent – in the quantitative study. This antecedent was primarily uncovered after previously conducting an exploratory study – the discussion of which is considered outside the scope of the current study. Moreover, this study introduced brand attachment as a mediating variable further explaining the relationship between CSR attribution and its consequences. Even though brand attachment was previously investigated as a mediating variable in the relationship between CSR and purchase intention by Koh and Fang (2012), their study was not conducted with the aim of developing a comprehensive model of the antecedents and consequences of CSR attribution, they only investigated individual relationships and CSR attribution was not included in the study. The current study, though, uncovered a lack of influence of brand attachment on PWOM intention, yet, it was deduced that there might be an effect of brand attachment on purchase intention and a more significant effect on brand loyalty intention was detected. On the other hand, trust was found to be a significant mediating variable elucidating the relationship between intrinsic CSR attribution and the three positive consequences; PWOM intention, purchase intention and brand loyalty intention.

In addition, the study attempted to develop a preliminary scale for the approach to CSR communication construct. It is hoped that future studies could further test and prove the reliability and validity of this scale. Notwithstanding that the reliability and validity of the proposed scale was proved satisfactory in the current study.

A final implication is that this study hopes to be a pioneer in attempting to provide a definition of the concept of a “Minimalist approach to CSR communication” certainly not disregarding that the core concept has been introduced and discussed many years ago (e.g. Dhanesh, 2015; Du et al., 2010; Ginder et al., 2021; Morsing et al., 2008; Ogunfowora et al., 2018).

6.2 Practical implications

The current study presents several practical implications. Firstly, generally, after conducting the experiment, the ANOVA results showed that the manipulation of the antecedents of CSR attribution in the extrinsic CSR attribution condition resulted in a higher level of extrinsic CSR attribution than in the intrinsic CSR attribution condition. Therefore, mobile and telecommunications services providers should not engage in CSR initiatives that are characterised by: short-term duration, business-process CSR, extravagant approach to CSR communication, small amount of invested resources in CSR, lack of long history of past corporate social performance or a lack of a call for customers' participation in the firm's CSR. Thus, a firm avoids consumers' extrinsic CSR attribution and its consequent negative outcomes such negative word of mouth.

Regarding the antecedents of CSR attribution, four antecedents were substantiated and they are as follows in order of importance: the approach to CSR communication, past corporate social performance, CSR type and lastly the firms' call for customers' participation in their CSR. It was found that the approach to CSR communication may have a significant effect on intrinsic CSR attribution where a minimalist approach could elicit a higher level of intrinsic CSR attribution than an ostentatious approach. Past corporate social performance may exert a significant effect on intrinsic CSR attribution where a long history of CSR engagement could contribute to higher intrinsic CSR attribution than a lack of history of CSR engagement. CSR type – philanthropic may exercise a significant effect on intrinsic CSR attribution. The call for customers' participation in CSR might exercise a significant effect on intrinsic CSR attribution. Thus, marketing managers or CSR officials of mobile and telecommunications services providers, when engaging in CSR, should mainly focus on engaging in CSR as a strategic cause along with its lifelong core business operations and not only as a one-time endeavour reflecting a long history of CSR engagement and past corporate social performance. They should select a philanthropic CSR initiative that results in no direct return to the sponsoring firm such as the initiative utilised in the experimental stimuli of the intrinsic CSR attribution condition (developing and empowering blind people to train them on communication and computer skills to prepare them for proper employment and become an integral part in society). When communicating their CSR, they should pay full attention to the approach to CSR communication to consumers and the public in general. A minimalist approach should be adopted that does not involve provocative extravagance for example in terms of large financial resources expenditure in seeking celebrities' representation in TV advertisements. Thus, the communication channel could be the firm's website and its official pages on social media platforms such as Facebook. Finally, a sponsoring firm should call its customers and the public to participate in its CSR initiatives especially that the digitised and connected environment has facilitated the direct interaction between the customer and firms (Iglesias et al., 2020). Hence, the sponsoring firm is going to be perceived as intrinsically motivated in its CSR engagement (consumers' intrinsic CSR attribution) and could realise the consequent positive consumer outcomes such as brand attachment, trust, PWOM intention, purchase intention and brand loyalty intention as previously shown in the study's results. Brand attachment and trust were two significant direct consequences of intrinsic CSR attribution. PWOM, purchase intention and brand loyalty intention were significant indirect consumer outcomes of intrinsic CSR attribution as mediated by brand attachment and trust. It was found that brand attachment might mediate the effect of intrinsic CSR attribution on purchase intention as well as may mediate the effect on brand loyalty intention. As for trust, it mediated the effect of intrinsic CSR attribution on PWOM, purchase intention and brand loyalty intention.

This could be considered a concise prescription for guiding marketing managers or CSR officials on how to customise their CSR initiatives in a manner that would contribute to eliciting intrinsic CSR attribution on the consumers' side. This could as well help in a refined allocation of firms' resources to harvest the desired positive consequences of CSR (Karaosmanoglu et al., 2016).

6.3 Limitations and future research

The study was conducted in Egypt and limited to the mobile and telecommunications services providers' context. A non-probability convenience sample was employed. Hence, the generalisability of the results may be hampered. Fictitious experimental stimuli were utilised which is considered a threat to external validity even though it enhances the internal validity of the study. There is a probability that there are some extraneous variables that could have affected the results of the study and were not controlled for such as participants' previous preference of specific causes. Nevertheless, such impact was hoped to be kept to a minimum through randomisation. The study did not examine some potential moderating variables such as the type of product and the type of consumer decision-making. Yet, the context of the study was limited to a unified service type (mobile and telecommunications) and fictitious brands to minimise the impact of extraneous variables. Moreover, the study aimed to construct a concise model. Thus, the increase in the variables could have resulted in participants' boredom with the survey or dropping. For the same reason of keeping the model concise and comprising the most relevant antecedents, some other potential antecedents of CSR attribution that were revealed from the literature review could not be investigated such as the fit between the sponsored CSR initiative and the firm's core business, type of message appeal in CSR communication (rational or emotional), type of social culture (collectivism vs. individualism) … etc. Finally, the design of the control condition's experimental stimulus may not have been ideal. The stimulus of the control condition was designed to be identical to the other two experimental conditions but without the antecedents of CSR attribution. That is, it was mentioned that the firm engages in CSR but without mentioning any cues of the antecedents of CSR attribution. It might have been more accurate not to mention CSR adoption in the control condition.

Having known the study's limitations, some ideas for future research begin to sprout. A reproduction of the current study could be carried out but with some modifications. Real brands could be utilised instead of fictitious ones to reflect realism and enhance the study's external validity. The study could also be re-applied in a different business context such as the food and beverages context or in a different cultural context, e.g. in an individualistic culture or a developed country. Even more, a comparative study could be conducted between an emerging economy and a developed one, thus unearthing valuable insights and boundary conditions related to how consumers of different cultural or economic backgrounds perceive CSR and form their CSR attributions as suggested in Boseila et al. (2023).A more representative sample could be employed, e.g. a probability or a non-probability quota sample. Potential extraneous variables such as preference for specific causes could be controlled for. The control condition's stimulus could include only the name of the brand and its core business without any reference to CSR adoption. The impact of potential moderating variables could be investigated such as the type of consumer's decision making whether emotional or rational and previous brand preferences. The study could be re-applied in the context of two different product types, e.g. convenience and specialty utilising a 2 × 2 factorial design where there will be two levels of CSR attribution (intrinsic and extrinsic) and two types of products (convenience and specialty). A more comprehensive model could be introduced and investigated with all possible antecedents of CSR attribution using a relatively larger sample size. An approach to CSR communication-centric study could be conducted. That is, a study investigating the impact of the approach to CSR communication on CSR attribution and its consequent positive consumer outcomes could be executed. The study could involve two experimental conditions one representing a minimalist approach to CSR communication and another representing an ostentatious approach to CSR communication. The two approaches could be represented by two TV advertisements; one minimalist (not involving extravagance or a huge expenditure of resources) and another ostentatious (involving extravagance and a huge expenditure of resources). Another alternative is to make the minimalist approach represented by CSR information on a firm's website or on social media platforms and the ostentatious approach be represented by a TV advertisement.

Figures

Figure 1

Proposed conceptual model

Table 1

Literature overview of CSR attribution research

CSR attribution as a dependent variable	CSR attribution as an independent variable	CSR attribution as a mediating variable	CSR attribution as a moderating variable
Duration of commitment to CSR [e.g. Ellen et al. (2006) (Gas Stations/USA); Yang et al. (2015) (Cosmetics/Taiwan); Choi et al. (2016) (Business Supply (office equipment)/USA and Canada; South Korea and India)] Company-Cause fit [e.g. Ellen et al. (2006) (Gas Stations/USA); Yang et al. (2015) (Cosmetics/Taiwan); Marín et al. (2016) (Services (Computer)/Spain); Austin and Gaither (2017) (stigmatised industries) (Beverages (Cola)/USA)] Approach to CSR communication [e.g. Du et al. (2010) (Conceptual paper); Yang et al. (2015) (Cosmetics/Taiwan); Marín et al. (2016) (Services (Computer)/Spain); Lim et al. (2018) (Information Technology/USA and South Korea); Ginder et al. (2021) (Consumer products/USA); Pérez et al. (2020) (Ethical Banking and Ecological Restaurants/Spain)]	Recommendations and Patronage intentions [(e.g. Vlachos et al. (2009) (Telecommunications and Mobile Services/Greece); Pérez et al. (2019) (Ethical banking and Ecological restaurants/Spain)] Consumer responses to CSR [e.g. Tsai (2009), Hammad et al. (2014) (FMCG/Egypt); Miller and Lellis (2015) (Stigmatised Industries) (Qualitative) (Energy/USA); Zasuwa (2016) (Food/Poland)] Purchase Intention [e.g. Zasuwa (2016) (Food/Poland); Pérez et al. (2019) (Ethical banking and Ecological restaurants/Spain)]	(e.g. Groza et al. (2011), Choi et al. (2016), Hur and Kim (2017), Ogunfowora et al. (2018), Kim and Choi (2018)) Between type of CSR strategy and attitude towards the firm and purchase intention [Groza et al. (2011) (Food and Beverage/USA)] Between both cultural orientation and sponsoring firm's nationality and consumer attitudes [Choi et al., 2016) (Business Supply (office equipment)/USA and Canada; South Korea and India)] Between collectivism and CSR perceptions [(Hur and Kim, 2017) (Cross Industry (Major Conglomerates)/South Korea)]	[e.g. Vlachos (2012), Homburg et al. (2013), Pai et al. (2015), Habel et al. (2016)] On the relationship between CSR and emotional attachment [(Vlachos, 2012) (context: NA)] On both business-process CSR and philanthropic CSR engagement and CSR reputation [(Homburg et al., 2013) (Cross Industry B2B)] On CSR perception and brand advocacy and brand equity [(Pai et al., 2015) (Manufacturing and Services/Taiwan)] On CSR engagement and perceived price fairness [(Habel et al., 2016) (Cross Industry B2C/Germany)]
Prior Ethical Reputation [e.g. García-De los Salmones and Perez (2018) (Finance and Banking/Spain)] CSR type [e.g. Andreu et al. (2015) (Services (Restaurants and Banking)/USA)] Cultural Orientation [e.g. Choi et al. (2016) (Business Supply (office equipment)/USA and Canada; South Korea and India); Lim et al. (2018) (Information Technology/USA and South Korea)] Other Factors [e.g. Marín et al. (2016) (Services (Computer)/Spain); Leonidou and Skarmeas (2017) (Eco-friendly products/USA); Ogunfowora et al. (2018) (Energy and Infrastructure/Canada); Moscato and Hopp (2019) (NA/USA)]	Attitude towards sponsoring firm [e.g. Choi et al. (2016) (Business Supply (office equipment)/USA and Canada; South Korea and India); Wongpitch et al. (2016) (Skin care/Thailand); García-De los Salmones and Perez (2018) (Finance and Banking/Spain); (Shazly and Mahrous (2020) (On-demand ride services/Egypt)] Scepticism/Distrust and behavioural intentions [e.g. García-De los Salmones and Perez (2018) (Finance and Banking/Spain)] Other dependent variables [e.g. Zasuwa (2016) (Food/Poland); Alhouti et al. (2016) (NA/USA)]	Between CEO leadership ethics and consumers' intention to volunteer to support the CSR activity, CSR donation behaviour intention and purchase intention [(Ogunfowora et al., 2018) (Energy and Infrastructure/Canada)] Mediates the interaction effects between the crisis issue/post-crisis CSR initiative congruence and crisis type on attitudes toward the corporation and purchase intention Mediates the interaction effects between pre- and post-crisis CSR initiatives' consistency and the crisis issue/pre-crisis CSR initiatives' congruence on attitudes toward the corporation and purchase intention [(Kim and Choi, 2018) (Information Technology Products/USA)]

Source(s): Table by the authors

Table 2

Manipulation checks

	Intrinsic CSR attribution Condition's mean n (119)	Extrinsic CSR attribution Condition's mean n (118)
Duration of commitment to CSR	3.513***	3.09***
Amount of invested resources in CSR	3.08	2.96
Type of CSR (Business-Process)	3.45**	3.75**
Type of CSR (Philanthropic)	4.02***	3.5***
Past corporate social performance	3.58***	3.2***
Approach to CSR communication	3.61***	3.22***
Call for customers' participation in CSR	3.76**	3.47**

Note(s): *p < 0.05 **p < 0.01 ***p < 0.001

Source(s): Table by the authors

Table 3

Hypotheses testing results

Hypothesis	Relationship	Intrinsic CSR attribution condition		Extrinsic CSR attribution condition		Control condition		Decision
		Coefficient	t-value	Coefficient	t-value	Coefficient	t-value
H1a	Duration of Commitment to CSR → Intrinsic CSR Attribution	−0.037	0.409	0.147	1.592	0.097	0.948	Rejected
H2a1	CSR Type (Philanthropic) → Intrinsic CSR Attribution	0.190	1.664*	0.067	0.705	0.198	2.015*	Partially accepted
H2a2	CSR Type (Business-Process) → Intrinsic CSR Attribution	−0.062	0.596	−0.005	0.061	−0.124	1.474	Rejected
H3a	Approach to CSR Communication → Intrinsic CSR Attribution	0.387	4.108**	0.309	3.042**	0.384	3.544**	Accepted
H4a	Amount of Invested Resources in CSR → Intrinsic CSR Attribution	0.085	0.784	−0.085	0.830	0.127	1.271	Rejected
H5a	Past Corporate Social Performance → Intrinsic CSR Attribution	0.224	2.088*	0.220	1.958*	0.223	1.958*	Accepted
H6a	Call for Customers' Participation in CSR → Intrinsic CSR Attribution	−0.061	0.609	0.231	1.915*	0.046	0.505	Partially accepted
H7	Intrinsic CSR Attribution → Brand Attachment	0.533	7.690***	0.435	5.562***	0.657	11.320***	Accepted
H8	Intrinsic CSR Attribution → Trust	0.530	5.727***	0.544	7.328***	0.666	12.555***	Accepted
H9	Brand Attachment → PWOM	0.047	0.507	0.041	0.504	0.070	0.662	Rejected
H10	Brand Attachment → Purchase Intention	0.107	1.250	0.179	2.577**	0.018	0.253	Partially accepted
H11	Brand Attachment → Brand Loyalty Intention	0.218	2.188**	0.186	1.960*	0.091	0.842	Partially accepted
H12	Trust → PWOM	0.496	4.371***	0.762	9.731***	0.762	7.637***	Accepted
H13	Trust → Purchase Intention	0.464	5.518***	0.658	9.822***	0.707	9.494***	Accepted
H14	Trust → Brand Loyalty Intention	0.349	3.753***	0.556	5.891***	0.589	4.700***	Accepted

Note(s): *p-value < 0.05; **p < 0.01; ***p < 0.001

Source(s): Table by the authors