Corporate Social Irresponsibility and Purchase Intention : The Mediating role of Corporate Reputation and Consumer Attitude

An important objective of business research is to understand how organizational practices can influence consumer attitude and behaviors in order to help achieve organizational goals via consumer purchase intention. It was proposed and found in this study that consumers’ perceptions of certain Corporate Social Irresponsible (CSI) practices serves as antecedents of consumer purchase intention (PI) via corporate reputation (CR) and consumer attitude (CA). On the one hand, this finding may help understand the “black box” between CSI practices and PI. On the other hand, while CR refers to customers’ evaluations of the reputation of the organization and CSI represent a significant channel the organization uses to channel its irresponsible behavior to the community, CR and CA literatures have not comprehensively examined the effect of CSI practices on consumers CA and CR beliefs. The findings in the hierarchical regression from a sample of 455 consumers of products in a large corporate organization in Kenya as a study documenting a negative association between CSI practices and CR and CA with PI suggests that consumers draw inferences from the CSI-related treatment they receive in assessing the supportiveness of the organization. By implementing CSI practices that demonstrate the organization does not care about the community and values their contribution, organizations are likely to be perceived as engaging in a high level of irresponsible behaviour. The results of this study add to our knowledge about the antecedents of CR and CA. Moreover, this study bridges the gap in the literature, by combining CSI, CR, CA and PI.


Introduction
Corporate social irresponsibility (CSI) cannot be discussed in isolation of stakeholder theory and corporate social responsibility.Freeman (1984) defines stakeholders as any group or individual who can affect or is affected by the achievement of the firm's objectives; those groups who are vital to the survival and success of the corporation (Freeman, 2004).The theory of corporate social irresponsibility is a contraposition of corporate social responsibility.Positive corporate behaviours are commended by stakeholders, whereas negative business behaviours are condemned by them, and they tend to have serious adverse effects on corporations (Frooman, 1997;Jones, 1995).Corporate social irresponsible behaviours involve a gain by an organization at the expense of society (Armstrong, 1977).According to Armstrong, one key parameter of corporate social irresponsibility is the exploitation of negative externalities.Frooman (1997) adds that socially irresponsible and illegal corporate behaviours cause substantial decreases in shareholders' wealth.Socially irresponsible corporate conducts are reflected in the negative scores of the KLD social ratings database on corporate social performance.In some corporations in the United States, where some CEOs and top executives were involved in unethical financial mismanagement and acted irresponsibly by indulging in either enterprise corruption, creative accounting system, flawed in board governance, jumbo CEO salary or stock market price manipulation caused substantial decreases in shareholders' wealth and employees' compensations in their organization.
Almost every big company has Corporate Social Responsibility (CSR) strategy to meet the requirements of customers and the society as a whole.Consumers are demanding that companies implement social responsibility regardless of their size, shape or origin.Since societies around the world are facing various social problems, they expect from the companies to help in solving them using their resources effectively.Advocates of CSR consider it as a vehicle for development and state that in an increasingly globalized economy competition has become more fierce and traditional differential factors among firms are on the verge of obsolescence, consequently, successful firms will be those that are able to respond to the demands of their stakeholders whatever these demands might be (Doane, D., 2005;Hollender, J., 2004;Idemudia, U.O., 2011).Companies focus shifted from just profit maximization towards more important issues of business survival and the satisfaction of social needs.There is the necessity to balance current stakeholder and societal needs with those of the future (Hildebrand, D., S. Sen and C.B. Bhattacharya, 2011).Therefore, many companies attempt to understand and meet requirements of both their consumers and all stakeholders.According to Lindgreen et al. (2009), CSR must reach out to many different stakeholders; the organization listens and responds to stakeholders that form part of the organization's relationships, networks and interactions.
Unlike CSR, CSI has not been debated on much in the existing literatures of corporate studies.Despite the fact that empirical evidence shows the existence of businesses that behave in a socially irresponsible manner, with its significant consequences, the existing literature has focused little on the concept of CSI.Scholarly interest began with the first study by Armstrong (1977), who put forward that CSI was the dishonest decision made by the directors of a company with the aim of generating shareholder value at the detriment of others, where groups of independent observers played an important role because they were accountable for evaluating the irresponsibility of such behaviour.Thirty years would have to pass before there was decisive return to academic interest in this concept.During the last global economic and financial crisis, and thanks to the broadcasting role of the media, a significant number of business outrages have been reported, which showed an exceptional increase in their occurrence.This new situation has also aroused interest in CSI from a professional point of view; on the one hand, for the important penalties that all irresponsible behaviours can generate at the corporate level and, on the other, because it has become clear that in the wake of the irresponsible behaviour can be perceived an absence of values and ethical principles among the top executives of companies.Business professionals focus on issues such as the implementation of controls to anticipate, monitor and avoid this type of behaviour.For these professionals, it is crucial to have a greater knowledge and better understanding of the mechanisms and tools that can help them to alleviate the possible damage caused by this corporate behaviour.Despite the role of these professionals, consumers' reaction to these companies' irresponsible behavior would have a consequence on their purchase intention.All this, therefore, has generated greater interest and concern in the academic field.It is against this background that the present study gears towards investigating the consequence of the irresponsible behavior of companies on corporate reputation and consumer attitude and their corresponding effect on purchase intention.Some studies have been carried out on the effect of CSI on consumer association but very little has been done on its effect on purchase intention while treating corporate reputation and consumer attitude as mediators.Besides, many studies especially in developing nations have a tendency of establishing relationships between two variables in consumer association research, while not bearing in mind that findings from such kind of studies are subject to estimation errors and biases.This present study gears towards filling the gap by addressing one important aspect of consumer intention, i.e. purchase intention by unfolding the "black box" inherent within corporate social irresponsibility, corporate reputation, consumer attitude and their implication on purchase intention.

Literature Review
The concept of Corporate Social Irresponsibility (CSI) is somewhat new.Even though the expression was invented approximately 40 years back, it has not been explored or received attention in the literature until the last one decade; the exceptional social and business apprehension caused by CSI cases, during the economic crisis, has led researchers again to focus attention on its study (Lin-Hi and Müller, 2013).As with the concept of Corporate Social Responsibility, there is no general agreement on the definition of CSI.In this review, we wanted to bring to the limelight the broad lines and themes of debate, making a suggestion that provide progress in the study of CSI.
Corporate Social Irresponsibility (CSI), defined as behaviours showing a lack of due concern for communities or the environment (Lange and Washburn, 2012), causes punitive actions from stakeholders (Balabanis, 2013;Grappi, Romani, and Bagozzi, 2013;Klein, Smith, and John, 2004;Sweetin et al., 2013;Walsh et al., 2009).The identity of those affected influences to what extent observers will retaliate.Commentators, for example, have bemoaned the lack of stakeholders' protest in the West over serious oil spills in Nigeria (Nossiter, 2010;Vidal, 2010).They have drawn comparisons with the strong response to the Gulf of Mexico oil spill, when BP faced organized boycotts from consumers (Wang, Lee, and Polonsky, 2015).When CSI affects distant others we find it more difficult to identify with the sufferers and feel less emotionally involved in the crisis.This research answers the call for the development of more sophisticated micro-level theories in social responsibility research (Aguinis and Glavas, 2012;Morgerson et al., 2013).A focus on individual level antecedents and mediators contributes to a better understanding of the psychological responses to CSI which complements past focus on strategic or institutional theorizing (McWilliams and Siegel, 2001;Woodward et al., 1996).There is limited evidence on the role played by consumer attitude in reactions to CSI.Russell and Russell (2010) document in-group bias in a CSR context: individuals are more likely to buy from companies that support the local community rather than distant others.However, appraising cases of CSI is psychologically different from interpreting CSR (Murphy and Schlegelmich, 2013).While CSR may benefit the observer, and incentivize reciprocity (Bhattacharya, Korschun, and Sen, 2008), CSI triggers justice evaluations (Carlsmith, Darley, and Robinson, 2002) and desire for revenge (Bechwati and Morrin, 2003).Lange and Washburn (2012) examine conceptually how social identity affects reactions to CSI.Their analysis of irresponsibility attributions suggests that observers are less likely to consider a company irresponsible when 1) the victims of CSI have a different identity and/or 2) the company shares the same identity of the observers.
We draw on social identity theory, corporate reputation and consumer attitude to CSI to develop a conceptual model that explains how the identity of the victims (firm) influences observers' (customers') perceptions of CSI (Figure 1).First, we examine how a firm's irresponsible behaviour influences the customers' perception of corporate reputation.Second, we review literature demonstrating that CSI is a key determinant of consumer attitude.This discussion examines the unique mediating role of these variables in our model.

Corporate Reputation and Corporate Social Irresponsibility
This section examines how the unwholesome activities of firms (i.e.environmental pollution, irresponsible corporate lobbying, tax evasion, and bribery, corruption and contract scandals, etc.) influence stakeholders' perception of the image/reputation of these firms.Some scholars believe reputation and image are synonymous, while others differ.For instance, Bromley (2000) considers reputation as the aggregate of identity and image, while Cornelissen and Thorpe (2002) assert that reputation is the collective representation of past images of an institution through communication or past experience accumulated over time.However, Dowling (2001) distinguishes image from reputation.Dowling argues corporate image is the global evaluation a person has about an organization, while corporate reputation is the attributed values evoked of the company.We argue that perception of corporate reputation is informed by aggregate of information, experiences, observable evidences, including distributed opinions, comments and viewpoints in the public sphere, available to networks of stakeholder groups inside and outside a corporation.
Corporate reputation has been conceived from corporate historical heritage perspective.It is believed that a company's reputation reflects the history of its past actions (Yoon, Guffey, & Kijewski, 1993).Reputation is a historical notion based on the sum of the past behaviours of the entity (Herbig & Mile wicz, 1995).It is a perceptual representation of a corporation's past actions and future prospects and its overall appeal to key stakeholders (Roberts & Dowling, 2002).A review of literature shows scholars (Robertson & Gatignon, 1986;Tauber, 1988) conceive reputation as a set of economic and non-economic attributes ascribed to a corporation.Pruzan (2001) contends that corporate reputation is an attribute vital for stakeholder trust and competitive advantage.Importantly, Fombrun (1996) refers to it as reputational capital.Scholars such as Hanson and Stuart (2001) have established a relationship between corporate misbehaviours and corporate reputation.The damage to the reputation of Broken Hill Propriety Ltd in Australia, resulted in sharp drop in its share price from around AUS$20 to AUS$12 in 1998.Against this backdrop, our first hypothesis is designed to investigate the impact of consumers' perception of corporate social irresponsible behaviours of a large corporation in Kenya on it image/reputation.In order to accomplish this task, the following research hypothesis is made:

Consumer Attitude and Corporate Social Irresponsibility
Attribution theory explained how people infer from self-concept and subsequently affect their attitude and behavior (Heider, 1958).Attributions are what people perceive the reason behind their behavior or the events they observe (Bitner, 1990).Attribution is made via observing a single individual's behavior toward a stimulus object at one point in time.Hence, psychologists inferred that individual's perception would act on the stimulus object rather than the object itself (Sparkman & Locander, 1980).Forehand & Grier (2003) clarify that attribution theory is the process that consumers evaluate marketer motives and how these motives influence consumer subsequent attitude and behavior.Consumers highly concern and evaluate corporate engaging social responsible activities; thus, corporations' effort to increase its reputation, image (Lai, Chiu, Yang, & Pai, 2010), consumer loyalty, trust (Martinez & Rodriguez del Bosque, 2013), consumer behavior (Becker-Olsen et al., 2006), consumer satisfaction (Hsu, 2011) by engaging CSR.Given these facts, the direct opposite will be the reverse.
Responsible manufacturing for instance can encompass a wide range of social and environmental factors, including ensuring fair work practices and minimizing environmental impacts.Costumers punish or reward the companies that obey (or don't obey) the social responsibility initiatives by buying products.Many researchers and managers believe that social responsibility should be applied by costumers.The company presents responsible products as long as customers are demanding such products.The CSI affects customer attitudes and customer attitudes influence their desires and intentions, and desires will lead to behavior (Trudel and Cotte, 2011).The growing number of market survey show, there is a negative relationship between the effects of CSI on consumer behavior (Davids, 1990).Some of evidences indicate that the companies are supporting CSR initiatives like philanthropy, relationship-based marketing, and plan for the protection of minorities, employment and production of social responsibility over the past few years (Sen et al., 2006).Researchers show a positive relationship between CSR initiatives and consumer attitudes toward company (Brown Tom and Peter, 1997).Costumers have a negative attitude toward companies which use child labor and in turn, they have a positive attitude toward companies that employ survivors of natural disasters, this is a fact which is completely confirmed (Wong Szeki, 2012).Therefore, according to the discussion, the following hypothesis can be expressed as: Hypothesis 2: The practice of CSI is negatively related to a high degree of consumer attitude within an organization.

Mediating Effect: Corporate Reputation and Purchase Intention
Corporate reputation has commonly been defined as the combined perception, attitudes, and opinions of various stakeholders including employees, customers, and community members (Fombrun et al. 2000).This perceptual representation of a company is the consequence of a company's past management actions and behavior, and works as a valuable, intangible asset and a competitive advantage for a company (Chun 2005;Fisher-Buttinger and Vallaster 2011;Gibson et al. 2006;Melo and Garrido-Morgado 2012).
To define corporate reputation, there have been deliberations emphasizing several key attributes of it in prior literature.To start with, corporate reputation is developed based on the aggregate perception of all a company's stakeholders (Fombrun et al. 2000;Walker 2010).According to Chun (2005), prior literature has generally classified the major stakeholders into internal and external stakeholders, and marketing literature has focused on customers as internal stakeholders.It triggers us to examine the influence of corporate reputation on the customer evaluation process.Another attribute of corporate reputation is its range from positive to negative (Walker 2010).In prior studies, it has been empirically supported that a positive reputation enhances customer satisfaction and company's performance (Chun 2005) but the critical effects of negative reputation have not been researched on much even though a negative reputation can aggravate the significant effects of positive reputation (Sohn and Lariscy 2012;Walker 2010).Also, corporate reputation has frequently been studied as either a dependent variable (Walker 2010) or as a mediating variable between various independent variables and brand equity (Hur et al. 2013).However, the effect of corporate reputation on consumers' decision processes may be more varied and unique.Wang et al. (2006) noted that corporate reputation might interact with brand equity to enhance corporate performance, which may reinforce or weaken the effect of brand equity.Nevertheless, the relationships among those variables with corporate reputation have seldom been examined, and the moderating effect of corporate reputation, which assesses the interaction between corporate reputation and other variables, has not especially been studied, even though it has been found to affect consumers' attitudes and companies' success (Galbreath 2005;Schwaiger 2004;Wang et al. 2006).Therefore, this study examined the mediating effect of corporate reputation on the relationship between CSI and purchase intention (Fig. 1).

Mediating Effect: Consumer attitude and purchase intention
Psychologists, especially social psychologist believe that the attitude -as a social structure-is so important in establishing and maintaining social order and also believe, attitudes must be changed before behavior (Chaiklin, 2011).With regard to the Social psychology theories, marketing researchers study the relationship between attitude and behavior.This is more like evaluating a process.It means, first attitude then behavior is formed subsequently (Reily et al., 1999).Most of the researches conducted during the 21st century approve that attitude has a huge impact on marketing and consumer behavior.Due to this fact, also, given the fact that costumer behavior is what he/she shows before, during and after buying a product or service.So we can say, costumer attitude to the service or product is the key factor of anticipating and continuing customer behavior, and the attitude is some passivity or a sense of agreement or disagreement about a stimulant.Actually, stimulants are emotional feelings that people have about a phenomenon (Montazeri et al., 2013).Intrinsic factors, external factors and consumer attitudes toward the product are factors affecting consumer's purchase intention (Jaafar et al., 2012).So, the fourth hypothesis is given thus:

Instrument
A questionnaire was administered to the respondents online using a social media platform.The items in the questionnaire were designed with each section addressing a particular variable of the study.Respondents were explained on the essence of the study and assured on the convenience of the data to be collected.Each respondent was to complete the questionnaire under the supervision of the researcher, since the platform offers an opportunity for engagement as they complete the questionnaire.Concepts that were not clear to them were clarified.Collected data was cleaned, coded and entered into SPSS in readiness for analysis.

Data Collection
Data was collected by survey method from 455 customers of a large Kenyan service firm (from 15 th January 2018 to 15 th February 2018).Proportional stratified and random sampling was applied to sample respondents (customers) to the company to give rise to 455 respondents.This method was selected because, each consumer was given an equal opportunity of being selected into the sample and in equal proportion (Cooper et al., 2006).

Measurement Scales
Corporate Social Irresponsibility was measured with four items that were sourced from Jan Kemper et. al., (2013) corporate measurement scale.The consumers perceived CSI scale evaluated consumers' individual experiences of the irresponsible behavior of the focal companies with 5-point Likert scale.The perceptions were measured on the five-point scale of strongly disagree (1) to strongly agree (5).Example of the statement included, 'The focal company does not give back to the communities in which it does business.' The research evaluated corporate reputation using a modified version of the corporate reputation subscale that was adapted from Weiss et al. (1999).We selected three items from the original measure.The corporate reputation scale evaluated the degree of the consumers' perceptions about corporate reputation in the focal companies in Kenya with 5-point Likert scale of strongly disagree (1) to strongly agree (5).Example of the items assessing consumers' perception of the reputation of the focal companies include: 'The focal company is a highly-regarded company.' We evaluated consumers purchase intention using scale that was sourced from Jan Kemper et. al., (2013).We selected the three items from the original measure.The purchase intention scale evaluated the degree of the consumers' perceptions about how the satisfaction they perceived to get from being customers to the focal companies with 5-point Likert scale of strongly disagree (1) to strongly agree (5).Example of the item assessing purchase intention include: 'I will buy most of relevant products/services from the focal company in the future.' In the same way, the consumer attitude scale evaluated the degree of the consumer's perceptions about their general attitude towards the focal company with 5-point Likert scale of strongly disagrees (1) to strongly agree (5).This was actualized by taking four items of consumer attitude as per Kropp, Holden, and Lavack (1999).Example of an item assessing consumer attitude include: "I am willing to pay more for a product if the focal company is donating part of the profits to charity".
In addition to the independent variables, dependent variable and moderating variables, the control variables included: age, gender, education level and the consumer tenure.Age of the consumers was recorded on a continuous scale, while gender of consumers was abbreviated as 1-Female and 0-Male, education level of consumers was coded on a nominal scale with 1-Bachelors, 2-Master, 3-PhD, and 4-Others.The tenure of consumer was abbreviated, 1-regular, 0-Irregular customer.

Results and Discussion
This section presents the research findings including the presentation of the answers given to the questionnaire.The general method used to solicit information from the respondents has been presented in the methodology.However, prior to carrying out the hypothesis testing, we accomplished groundwork scrutiny of the data set to verify for the violations of the underlying assumptions associated with the method of data analysis.A Statistical Package for Social Science (SPSS) version 20.0 was used to analyze the construct validity and reliability and consequently test the research hypotheses.

Descriptive Statistics
The descriptive statistics for the variables under study is as shown below.From the correlation matrix above, all independent and control variables are correlated to the dependent variable PI.There is no multicollinearity amongst the variables too.With respect to the descriptive statistics, the mean education level of respondents is degree and the mean tenure of respondents is 2.94, implying that majority of consumers were regular.

Data Adequacy and Convergent Validity
Here, the Kaiser-Mayer-Olkin Test (KMO), which measures sampling adequacy, was done for each variable and the results showed acceptability.In particular, the results of these statistical analyses showed that; (1) all the study variables exceeded the minimum standard of the KMO value of 0.6 and were significant in Bartlett's test of sphericity (2) all the study variables exceeded the acceptable standard of reliability analysis of 0.70 (Keramati et al., 2011).The statistical results confirmed the measurement scale of this research met the acceptable standard of reliability and validity analyses as given in Table 1b respectively.From the results in table 2, all items loaded well above the rule of thumb of the loading factor of 0.7 (Yong & Pearce, 2013).As suggested by Lee et al. (2009), factor loadings, composite reliability (Venkatesh & Zhang, 2010).Before subjecting data to the analytical model, factor Analysis was carried out on the data.After running severally repeating the data analysis with Varimax rotation and assessing change in Eigen values, all the items loaded well for the given variables.These items had adequate reliability and convergent validity because all factor loadings were greater than 0.7, the composite reliabilities exceeded acceptable criteria of 0.6.After confirming the construct validity and discriminant validity of the constructs, a composite variable was generated for each of the constructs i.e.CSI (for Corporate Social Irresponsibility), CR (for Corporate Reputation), PI (for Purchase Intention, and CA (for Consumer Attitude).

The Effect of Corporate Social Irresponsibility on Purchase Intention
This section attempted to unravel the relationship between CSI, CR, CA and PI so as to address hypothesis I, II and III.The main hypothesis of the study predicted that CR and CA mediated the relationship between CSI and PI.
Before carrying out any meaningful OLS regression in testing the hypothesis, the study utilized Baron and Kenny's (1986)'s recommendations for examining preconditions for mediating effects.The first precondition was used to assess the multicollinearity amongst variables, by carrying out correlations among them.

Condition one:
The independent variables and the proposed mediators must each, be significantly related to the dependent variable when considered separately.
To address this hypothesis, Hierarchical linear regression was carried out with CSI (IV), and the two mediators (CR and CA) and PI as dependent variable.The results are as shown in table 3 below.From the findings in the table above, the separate regression equations between the independent variable (CSI), the two mediators (CR and CA) with the dependent variables, PI were significant, fulfilling the preliminary requirement for variables with a mediation effect.

Condition two: Independent variable(s) to be significantly related to the proposed mediator(s).
This condition was tested by separately regressing the mediators (CR & CA) on the control variables and CSI.The purpose of this step was to establish that zero-order relationships among the variables exist.If one or more of these relationships are non-significant, then we conclude that mediation is not possible or likely.The results of the regression were as shown in table 4 below.4 above, shows that CSI was positively and significantly associated with OS and OI, and hence condition 2 was met in line with Baron and Kenny's (1986).From the results in the table 4 above, CSI is significantly associated to CR (β=-0.581,p<0.01) and to CA (β=-0.576,p<0.01), thus confirming hypothesis one (H1) and hypothesis two (H2) respectively.

Condition three:
The last condition stipulates that the relationship between the independent variable (IV), and the dependent variable (DV), should be weaker or non-significant when the proposed mediator is in the regression equation than when the proposed mediator is not in the equation.To test this, several regressions were carried out with PI as dependent variable and CSI as independent variable, with and without the mediators.Each regression included the control variables and was carried out hierarchically.Since we had only two mediators, so two regressions were carried out to test the mediation effect of each.

Mediation effect of Corporate Reputation in the Relationship between (CSI) and (PI)
To test for the mediation effect of CR in the relationship between CSI and PI, a hierarchical regression was carried out with PI as the DV, CSI as IV.Four control variables, i.e. age, gender, education level and tenure were included.CR was included into the model the last as shown in table 5 below.When consumer purchase intention (CPI) was regressed on corporate social irresponsibility (CSI), gender, age, education, job and tenure were entered at first, then corporate social irresponsibility was entered in step 2, and was found to be significant and negatively related to PI (β=-0.182,p<0.001).In step 3, CR was added to the equation, it was negatively and significantly associated to PI (β=-0.445,p<0.001), while corporate social irresponsibility was non-significance.The IVs explained 13.1% variance in consumer purchase intention and as such the model was slightly of good fit.Taken together, these results indicate that CR fully mediated the relationship between corporate social irresponsibility and consumer purchase intention, which supports Hypothesis three.(H3).

Mediation Effect of Consumer Attitude in CSI and PI Relationship
To test for the mediation effect of CA in the relationship between CSI and PI, a hierarchical regression was carried out with PI as the DV, CSI as IV.Four control variables, i.e. age, gender, education level and tenure were included.CA was included into the model the last, as shown in table 6 below.When PI was regressed on corporate social irresponsibility, gender, age, education, and tenure as control variable entered the equation in step 1, then corporate social irresponsibility entered in step 2 and was found to be significant and negatively related to PI (β=-0.182,p<0.001).Next, when CA was added to the equation, it was found to be significant and negative (β=-0.338,p<0.001), whereas high corporate social irresponsibility was no longer significant.The IV explained about 8.7% of variance in PI, which is slightly, and fairly a fit model.Thus, CA fully mediated the relationship between corporate social irresponsibility and consumer purchase intention.H4 was fully supported.

Direct and Indirect Mediation effects
Having calculated the direct effects (IV-DV relationships, and IV-Mediator-DV relationships), there was need to estimate the sum of the indirect mediation effects.The total mediation effect is what is called indirect effect, which is actually the total mediation effect given by the figure below: Figure 2    The results in table 7 above, shows the coefficients of various IV, DV and Mediators as well as their p-values.
The result shows that the relationship between CSI and PI through is mediated both by CR and CA.In the case of the mediation effect of CA, the relationship between CSI and CR was negative and significant, and that between CR and PI was negative and significant while that of CSI and PI was also significant.Table 7 is a Summary of direct and mediation effects.The Total effect size of each variable can be calculated using the formula below.
Total effect= Direct effect + Indirect Effect.
For example, CSI influences PI through CR as a mediator, whose mediation effect alone is .263.Therefore the total effect on PI is 0.341(.263+.078) in absolute term.Since the scope of this study did not involve calculation of total effect sizes, thus based on the logic above, table 8 was analyzed and offered the summary to the hypothesis testing.The results in table 8 above imply that the total effect of CSI on PI is 0.341, with CR mediating the effect by 0.263 units, while the total effect of CSI on PI is 0.323, with CA mediating the effect by 0.195.This implied that CR is a stronger mediator than CA.Each proposed model seemed to satisfactorily fit to the data.Furthermore, the model explained, over 8% of the variance in the consumer purchase intention, lending strength to the belief that the constructs were suitable to characterize the studied phenomenon.Verification of each hypothesis presented in table 8 was performed by the analysis of magnitude, sign and significance of the standardized path coefficients (Byrne, 2010).The hypotheses, the estimated coefficients and significance levels are shown in table 8. Apparently, CSI negatively influenced CA and CR, while CA and CR negatively mediated the relationship between CSI and PI, hence all the Hypothesis stated earlier were accepted.
In evaluating the mediating role of CR on the relationship between CSI and PI, we used hierarchical regression to provide evidence consistent with a mediating role of CR in the CSI-PI relationship.As illustrated in Table 5, we entered CSI in the first step as a control variable, CR in the second.Because we used multiple regression, statistical significance was assessed with the Wald statistic, which approximates a Z 2 distribution.The third mediation requirement of Kenny et al. (1998), that CR should be associated with PI while controlling for CSI.We found that the increase in the relationship of CSI with CR from Step 2 to Step 3 was statistically significant, (b= -0.445, p< 001).The presumptive mediation effect was complete, with an insignificant relationship between CR and PI still remaining after controlling for CSI.This validates the acceptance of hypothesis 3 (H3).The findings were consistent with CR's mediation of a negative relationship between CSI and PI.This result follows from corporate reputation theory, holding that CSI leads to bad corporate reputation (CR), which in turn, decreases purchase intention by weakening felt obligation of consumers toward the company.Similar procedure was done for CA on the relationship between CSI and PI as illustrated in Table 6 that validates the acceptance of hypothesis 4 (H4) of the study.
From the study result, this hypothesis was confirmed and consistent with Jaafar et al. (2012) study.On the one hand, they stated that the consumers' attitude is an important factor in tendency effect of costumer to buy a product.They also stated that consumer buying behavior is very complex and is influenced by the perceptions and attitudes of consumers.They further argued that the internal and external factor and consumer attitudes toward the product are effective factors on consumer buying intent (Jaafar et al., 2012).On the other hand, Chaiklin pointed out that if you want to change a behavior, you should change the attitude before.He pointed out the social psychologists believe that the change of attitude is a procedure for changing behavior (Chaiklin, 2011).Reily et al (1999) have stated that a behavior emerge after the attitude subsequently.Montazeri et al. (2013) point out that Consumer attitude toward a product or a service is a key factor in predicting consumer behavior (Montazeri et al., 2013).
The results concerning the impact of corporate social irresponsibility provide considerable managerial implications to practitioners who seek ways to effectively manage their company's reputation.The results of the current study ascertain that negative or irresponsible behavior that can aggravate consumers' attitudes and the company should carefully manage a negative reputation to avoid damaging established consumers' attitudes and purchase intentions.The results of this study further suggest that consumers absolutely process the negative corporate reputation given the irresponsible behavior of the company.Any kinds of negativity would damage the relationship with consumers

Conclusion
This research attempts to explore the Impact of corporate social irresponsibility (CSI) on consumers purchase intention in conjunction with corporate reputation and consumer attitude.In conclusion, despite the limitations discussed in the previous section, this study has the prospective to make an important contribution to business research and practice.By connecting consumers' perceptions of corporate social irresponsible (CSI) practices with CR and CA in influencing consumers purchase intention, this study bridges the business research and organizational behavior literatures, and provides guidance on how organizations can foster high levels of corporate reputation and gain positive consumer attitude via the implementation of appropriate corporate practices.The findings on the significant negative effects of CSI on CR and CA demonstrate the importance of CR and CA research and the need for organizations not to engage in corporate irresponsible behaviour that will influence consumers in negating their purchase intention.As mentioned earlier, this research also points to some fascinating directions for future study.

Limitation and future research direction
This study has contributed immensely to business research literature by providing empirical evidence that can enhance the understanding of corporate social irresponsibility, corporate reputation, consumer attitude and purchase intention.However, as in any other research, this study has some limitations.This study based the research on only one company, therefore limiting the generalization of the results.Future researchers may need to investigate the research framework of this study in different companies to generalize the findings across companies.In addition, the current study focused on only corporate reputation and consumer attitude as mediators of negative reputation.Future researchers may want to test the effect of different types of corporate reputation and ensure that the participants' negativity perceptions are measured by an appropriate pretesting procedure.Further to this, examining different types of corporate reputation and consumer attitude and different levels of negative reputation may also lead to interesting results regarding different consumer responses, and it would enrich the empirical research.

Figure 1 :
Figure 1: The model of study below shows the direct and indirect effects relationship between CSI, CR, CA and PI.

Figure 2 :
Figure 2: Direct and indirect effects associated with CR and CA as Mediators.From the figure above, path C is the direct bivariate regression predicting PI from CSI, path a, is a bivariate regression predicting mediator from CSI, and path b and C' are multiple regressions predicting PI from CSI and CR.The total mediation effect is the absolute product of b and C'.The same approach was used for CA and the results for direct and indirect effects (mediation effect) as shown below:

Table 1 :
Descriptive statistics and correlation matrix

Table 1b :
Result of Cronbach's Alpha and KMO

Table 2 :
Results of standardized item loading Source: Authors computation

Table 3 :
Independent regression between independent variables,

Table 4 :
Hierarchical Regressions Predicting the impact of CSI on CR and CA

Table 5 :
Mediation effect of CR in the relationship between CSI and PI

Table 6 :
Mediation effect of CA in the relationship between CSI and PI

Table 7 :
Direct, indirect and Mediation Relationships

Table 8 :
Summary of Relationships and Hypothesis Testing