Perceived CSR and organizational attractiveness : the influence of personal values, age and gender

(1)

1

“

Perceived CSR and organizational attractiveness: the influence

of personal values, age and gender”.

Master Thesis written by: Marlies Schonewille Student number: 10015833

Date: 30-06-2014

MSc. In Business studies- Leadership and Management Track University of Amsterdam

First supervisor: Dr. Nesrien Abu Ghazaleh Second supervisor:

(2)

2

Abstract

In a time where people are regarded as most important assets of organizations, it is of strong importance to attract highly talented people. Besides this, Corporate Social Responsibility (CSR) is a theme that has increased in attention over the past few years. However, the literature that links these two concepts is limited. This study examined the influence of perceived CSR on organizational attractiveness. The research question is: Does perceived CSR influence organizational attractiveness to job seekers? The underlying question is: Does this depend on personal characteristics? A quantitative empirical study is conducted by means of a survey. Results of this study indicate a positive relationship between perceived CSR and organizational attractiveness. However, the influence of the self-transcendence personal value type on this relationship is not proved. Yet this study only found a negative influence of the self-enhancement value type on this relationship. The influence of the other personal characteristics (age, gender) is also not proved.

1. Introduction

Many people acknowledge, that today one of the most valuable assets for organizations are the people who work in it (Tigu, & Tuclea, 2007). Employees who are creative, skilled and focused on customer satisfaction are essential for creating competitive advantage. Therefore, it is of strong importance to attract, motivate and retain highly talented people (Bartlett and Ghoshal, 2002). Today there is a situation with a lot of talented people around and a scarcity of jobs. This makes the competition for job seekers more fierce (Unemployment, 2013). On the other hand, this gives companies the possibility to choose only the people with the most talent. The communication of a company’s Corporate Social Responsibility (CSR) activities can be a useful tool to make an organization more attractive to job seekers (Bhattacharya, Sen and Korschun, 2008). Albinger and Freeman (2000) explain that Corporate Social Performance (CSP) relates positively to employer attractiveness if job seekers have a high level of job choice. They suggest that firms with a high level of Corporate Social Performance have the ability to attract only the most qualified people, because these people usually do have a lot of job choices even in times of scarcity. The importance of CSR for attracting employees is also made clear by Copeland (2003), former CEO of Deloitte and Touche Tohmatsu: “The best professionals in the world want to work in organizations in which they can thrive, and they want to work for companies that exhibit good corporate citizenship”.

(4)

4 CSR is not a new theme, but it is a theme that becomes more and more important (Aguinis and Glavas, 2012). Almost all large companies have a CSR section in their annual report, but there is a lot of dissimilarity in what they write down and how they report their CSR activities (Nielsen and Thomsen, 2007). To clarify what large companies report about CSR, the ministry of economic affairs created the transparency benchmark. This benchmark provides insight into the degree of transparency in social reporting at the 460 largest companies in the Netherlands (Transparency Benchmark, 2013). Despite the fact that the annual report is made public for all stakeholders, increasing organizational attractiveness is only possible if the information in this report is clearly communicated to job seekers trough recruiting tactics. It is the perception of the organizational CSR policy, by job seekers, that will influence organizational attractiveness. This will be the focus of this thesis. The main

research question is: Does perceived Corporate Social Responsibility influence organizational attractiveness to job seekers? The underlying question is: Does this depend on personal characteristics?

Earlier studies on CSR were mainly focused on the relationship between CSR and company performance and especially financial performance (Waddock and Graves, 1997; Barnett and Salomon, 2006; Van der Laan, van Ees and van Witteloostuijn, 2008; van Beurden and Gössling, 2008; Surroca, Tribó and Waddock, 2010). However, besides what CSR could do for company performance there is a stream in the literature that focuses on the relationship between CSR and organizational attractiveness to job seekers (Turban and Greening, 1997; Greening and Turban, 2000; Backhaus, Stone and Heiner, 2002; Albinger and Freeman, 2000). More recent research in this area tried to get a step ahead and investigated why corporate social performance leads to more attraction by job seekers. Results of Jones, Willness and Madey (2010) show that perceived value fit, expected employee treatment and employer prestige were mediating variables in the relationship between perceived CSR and organizational attractiveness. In addition Evans and Davis (2011) acknowledged that CSR is a topic that is studied a lot, but what they were missing were studies that explain the individual level factors that might affect the relationship between perceived CSR and individual reactions. They made a start by investigating whether CSR education and Other-regarding personal value orientation influenced the relationship between perceived CSR and job seekers organizational attraction. Subsequently Aguinis and Glavas (2012) also acknowledge that research must be done on a more micro level (like individual attitudes and personality), so that underlying mechanisms that link CSR with outcomes (such as retention, organizational attractiveness and engagement) can be addressed.

(5)

5 This study wants to extend the current literature about the influence of individual level factors on the relationship between perceived CSR and organizational attractiveness. Besides how and which CSR activities are communicated to influence attractiveness, it is important that organizations acknowledge that reactions to these CSR activities could differ between individuals. The main question of this thesis is positively answered by Turban and Greening (1997), Greening and Turban (2000), Backhaus, Stone and Heiner (2002) and Albinger and Freeman (2000). They all found a positive relationship between (perceived) CSR and organizational attractiveness. However they have done their research in a period with low unemployment, on average 3.5% of the working population (CBS, 2014). Greening and Turban (2000) argue, job seekers have legitimacy, power and urgency when there are plenty of jobs, but today this may no longer be the case. Nowadays there is a situation of high unemployment; 8.6% of the working population in January 2014 (CBS, 2014). It is therefore interesting to study if the relationship between perceived CSR and organizational attractiveness to job seekers still holds in today’s environment. Due to the increasing attention for CSR in today’s environment it is likely that this is still the case, therefore hypothesis 1 is: Job seekers will rate companies as more attractive organizations when they

are perceived as social responsible organizations.

Besides the contribution to the theoretical knowledge that is currently available about the relationship between CSR and organizational attractiveness, results of this study are also practically useful for recruiters. This is because the communication of CSR could lead to more organizational attraction by some people than by others. Recruiters have to take this into account when searching for the right people.

The structure of this thesis is as follows: first a literature review will be conducted in which the main concepts of this study are explained. In addition hypotheses are formed and expressed in a conceptual model. Subsequently the methodology of this study is discussed followed by a results section. Thereafter the discussion, limitations and suggestions for future research are presented. Finally a conclusion will be given.

(6)

6

2. Literature review

This chapter considers the key theoretical literature that is relevant for answering the research question. The first part of this theoretical background explains the importance of people in today’s organizations. Afterwards, organizational attractiveness is discussed, including the main factors that make organizations more attractive. Thereafter, the concept of CSR is explained. Subsequently the role of reputation and the influence of perceived CSR on this reputation is discussed in relation to organizational attractiveness. Finally, the influence of personal characteristics on the relationship between perceived CSR and organizational attractiveness to job seekers is explained.

2.1 People as a sustained competitive advantage

As described in the introduction, people are one of the most important assets for organizations in today’s environment. Without a quality workforce, a sustained competitive advantage can hardly be reached. This is caused by a change in importance of organizational resources; past sources of sustained competitive advantages are less important today (Greening and Turban, 2000). Past sources were focused on access to financial markets, patents and developing economies of scale. These are still important; however the most important factor for organizational success is a quality workforce (Greening and Turban, 2000). Success depends on a firms’ ability to create, manage and transfer knowledge (Greening and Turban, 2000). Barney (1991) defines a sustained competitive advantage as “a value creating strategy not simultaneously being implemented by any current or potential competitors and when these other firms are unable to duplicate the benefits of this strategy”. According to Barney (1991), this depends on the resources an organization employs. This view is called the resource-based view (RBV). A firms’ resources can be divided into three categories: physical capital resources (equipment, technology, access to raw materials), organizational capital resources (planning, reporting and coordination structure) and human capital resources (intelligence, experience, relationships and training of individual people inside the organization). These could provide a sustained competitive advantage if they are valuable, rare, inimitable and non-substitutable. The last category of resources, human capital, is of strong importance and the main focus of this study. Hatch and Dyer (2004) notice that human capital contributes to a sustained competitive advantage due to its inimitability, based on its firm specific, intangible and complex social nature. Therefore it is of strong importance for companies to expand their human capital through attracting highly

(7)

7 qualified employees. This is supported by Greening and Turban (2000, p.257) who argue that “employees are a necessary component to success”.

2.2 Organizational attractiveness

As the previous section makes clear, the contribution of employee skills, knowledge and abilities to organizational performance is more recognized over the past years. Attracting high-performing employees is of strong importance for organizational success and a critical component in personnel selection (Chapman, Uggerslev, Carroll, Piasentin and Jones, 2005). Before organizations can select people, it is important that job seekers are attracted to these organizations. Bratton and Gold (2003, p. 484) define attraction as the "favourable interaction between potential applicants and the images, values and information about an organization." In this way the organization is seen as a desirable place to work. Lievens, Decaesteker, Coetsier and Geirnaert (2001) studied whether objective organizational characteristics influenced organizational attractiveness. Results of their study indicate that prospective employees are more attracted to large and medium sized, multinational and decentralized organizations. Aiman-smith, Bauer and Cable (2001) also studied which factors have an effect on organizational attractiveness. They looked at two important job specific aspects (pay and promotional opportunity) and two important aspects of organizational image (lay-off policy and ecological rating). Overall, ecological rating most strongly influenced prospective employees attractiveness to an organization. Chapman, Uggerslev, Carroll, Piasentin and Jones (2005) performed a meta-analytic review on this subject. Their results indicate that applicant attraction depends on recruiter behaviour, perceived fit, job-organization characteristics (pay, benefits, size, work environment, location and company image) and hiring expectancies. The strongest predictors of organizational attractiveness are the work environment, organizational image (objective organizational characteristics) and perceptions of fit (subjective organizational characteristics). This indicates that if a company has a strong and positive image, it is perceived as having a good work environment. Additionally, if there is a fit between the job seeker and the organization, it is a more attractive organization.

As the previous studies make clear, organizational reputation is of strong importance for the attractiveness of organizations. As Siltaoja (2006) notes, reputation is often seen as one of the most important sources of competitive advantage a company can have. Reputation is defined by Brown and Logsdon (1999, p. 169) as: “the long term combination of outsiders’ assessment about what the organization is, how well it meets its commitments and conforms

(8)

8 to stakeholders’ expectations, and how effectively its overall performance fits with its social-political environment”. Reputation of a company could give stakeholders information about the expected future behaviour of that company. An important factor that could improve this reputation is the CSR activity of a company (Turban and Greening, 1997). The next section will describe CSR.

2.3 Corporate Social Responsibility (CSR)

In a recent article written by Aguinis an Glavas (2012, p.933), CSR is defined as “context-specific organizational actions and policies that take into account stakeholders’ expectations and the triple bottom line of economic, social, and environmental performance”. In the past, four types of organizational responsibilities have been defined: economic, ethical, legal and discretionary responsibilities. Economic responsibilities are the obligations of businesses to meet consumption needs and maintain economic wealth. Ethical responsibilities require businesses to follow the moral rules drawn up by society. Legal responsibilities require businesses to fulfil their economic mission whereby they have to follow the legal obligations. Finally, discretionary responsibilities refer to the philanthropic responsibilities and reflect the desire of society of an active involvement of businesses in creating a better society (Maignan, Ferrel and Hult, 1999). These organizational responsibilities are still perceptible in a study performed by Dahlsrud (2008). He studied past literature that was focused on the concept of CSR. Dahlsrud (2008) came to the conclusion that there are five dimensions that are most related to the concept of CSR: (1) the environmental dimension, which focuses on the natural environment; (2) the social dimension, which focuses on the relationship between business and society; (3) the economic dimension, which focuses on the financial aspects such as describing CSR in terms of business operations; (4) the stakeholder dimension, which focuses on the interaction of the organization with employees, customers, communities and suppliers; and finally (5) the voluntary dimension that describes CSR as behaviour not prescribed by law, behaviour that goes beyond legal obligations.

Corporate social performance (CSP) is introduced as a method to evaluate how well corporate social responsibilities are met by organizations (Albinger and Freeman, 2000). CSP is defined by Wood (1991, p. 693) as “a business organization’s configuration of principles of social responsibility, processes of responsiveness and policies, programs, and observable outcomes as they relate to the firm’s social relationships”. CSP could be divided in 11 dimensions (Backhaus, Stone and Heiner, 2002). Five of these dimensions could have an influence on the attractiveness of an organization to jobseekers. These dimensions are

(9)

9 product issues, community relations, environment, diversity and employee relations. Wherein diversity and environment dimensions have the strongest effect on organizational attractiveness (Backhaus, Stone and Heiner, 2002).

2.4 CSR and organizational attractiveness, the influence of reputation

This section describes the effect of CSR on organizational reputation. In addition, it explains why a positive reputation could lead to more organizational attractiveness to job seekers. This is done according to two theories: Signalling Theory and Social Identity Theory. Finally, the first hypothesis is presented that could give an answer on the main research question: Does perceived corporate social responsibility influence organizational attractiveness to job seekers?

CSR and organizational reputation

Turban and Greening (1997) were one of the first authors that investigated the link between CSP and organizational attractiveness. This is because a new paradigm for business was under development; a paradigm that emphasized the responsibilities of multiple stakeholders instead of only shareholders in the traditional setting. Results of their study indicate that firms with high CSP are more attractive as employers, and firms with high CSP have a better reputation. The fact that CSR activities contribute to the reputation of an organization is also acknowledged by Zyglidopoulos (2001). He argues that the reputation of an organization can be seen as a multi-dimensional construct, wherein performance on CSR is one of the dimensions. The CSR reputation of a firm can be defined as “the knowledge and emotions that various stakeholder groups have concerning the social performance of the particular firm” (Zyglidopoulos, 2001, p. 420). Also Oriesek (2004) acknowledged that CSR has a great influence on organizational reputation. He developed a reputation quotient and CSR contributes up to 55% of this quotient score. Subsequently Coldwell, Billsberry, van Meurs and Marsh (2008) argue that the discretionary and ethical elements that are manifested in a company’s corporate social performance have a large effect on corporate reputation. In addition, results of Arendt and Brettel (2010) show that CSR is a useful concept in corporate identity building. Finally, Aguinis and Glavas (2012) argue: one of the findings that are consistent in the overall literature is that CSR initiatives lead to an improvement of the firm’s reputation.

(10)

10 CSR reputation and organizational attractiveness

Turban and Greening (1997) argue that organizations that score high on CSP have a competitive advantage because they attract more potential applicants than firms that score lower on CSP. Greening and Turban (2000) expand this study by using an individual level of analysis instead of the previous organizational level. Their results indicate that job seekers were more attracted to organizations, will more often request an interview, report greater job pursuit intentions and are more prone to accept job offers if organizations score high on CSP. In addition, results of Behrend, Baker and Thompson (2009) indicate that reputation is a mediating factor between pro-environmental recruiting messages and organizational attractiveness. This implicates that investing in CSR lead to a positive reputation, which could lead to organizational attractiveness. More recently, results of Evans and Davis (2011) show that high perceived CSR leads to more organizational attraction to job seekers. These results implicate that it is not really the objective CSR that is important but the perceived CSR by job seekers, because this will influence the attractiveness perception of an organization. Perceived CSR is “the extent to which and individual perceives an organization as fulfilling the economic, legal, ethical and discretionary responsibilities imposed on the firm by its stakeholders” (Evans and Davis, 2011). It refers to the individual interpretations and evaluations of the fulfilment and importance of CSR activities and processes. Lin, Tsai, Joe and Chiu (2012) studied whether the relationship between perceived CSR and organizational attractiveness is visible in Asian context, namely Taiwan. They find support for the fact that high-perceived CSR leads to more organizational attractiveness to job seekers. The relationship between perceived CSR and organizational attractiveness is caused by the information CSR activities provide to job seekers. A job seeker can use this information to make judgments about job expectations and future work life; it provides information about the appropriateness of that organization for a specific person (Evans and Davis, 2011). There are two theories that further explain why a positive CSR reputation contributes to more organizational attractiveness to job seekers. The two theories are: Signalling Theory and Social Identity Theory.

According to the Signalling Theory, the social policies and programs of an organization can serve as a signal of the working conditions inside that organization and trough this attract job seekers (Turban and Greening, 1997). Applicants do not have complete information about organizations, so the information that they do receive will be interpreted as signals of the working conditions. As Turban and Greening (1997) argue, CSP can signal values and norms that are important to the organization and may influence perceptions of

(11)

11 working conditions that in turn influence organizational attractiveness. Also Srivastava and Lurrie (2001) argue that individuals use signs that are dropped by the firm to draw conclusions about the intentions or actions of that firm. The formal programs and social policies towards stakeholders could attract job seekers because it serves as a signal of the goals and commitments of an organization (Turban and Greening, 2000). In addition Behrend, Baker and Thompson (2009) note that the reputation regarding the environment give signs to prospective employees about how well it treats its employees.

In addition, according to the Social Identity Theory people classify themselves on the basis of various factors into social categories. Being a member of this social category influences an individual’s self-concept (Turban and Greening, 1997). An example of such a category is the organization they work for. Organizational actions on social issues could enhance or damage a firm’s image or reputation, by extension, the images of employees (Turban and Greening, 1997). Backhaus, Stone and Heiner (2002) argue that CSR may be an important factor for job seekers as a way to select a self-enhancing employment setting. In addition, Carmeli, Gilat and Waldman (2007) argue that there is an increased organizational identification when someone works for a socially responsible company.

The previous studies that investigate the link between CSR and organizational attractiveness all ask for a replication of their results among more experienced job seekers in different countries or areas. We are now in a different time period. There is no longer the condition of low unemployment but instead a lot of people are searching for a job (Unemployment, 2013). Lin, Tsai, Joe and Chiu (2011) already looked if perceived CSR could contribute to organizational attractiveness in today’s environment, but their study was done in Asia. There are no studies found that tried to replicate this finding in a European context. This study therefore investigates if the previous relationship between perceived CSR and organizational attractiveness is still evident in today’s Dutch environment. Therefore I hypothesize the following:

Hypothesis 1: Job seekers will rate companies as more attractive organizations when they are perceived as social responsible organizations.

(12)

12 2.4 CSR and organizational attractiveness, the influence of job seekers’ personal characteristics

This section explains the influence of job seekers personal characteristics on the relationship between perceived CSR and organizational attractiveness. The previous sections explained that perceived CSR could lead to more organizational attraction to job seekers mainly due the effect of reputation. However, how CSR is perceived and due to this organizational reputation, depends on individual characteristics. In other words, because people are different from each other, the relationship between perceived CSR and organizational attractiveness will likely differ in strength. The personal characteristics that are included in this study are personal value orientation (based on Schwartz (1994) basic human values), age and gender. Personal value orientation

Bauer and Aiman-Smith (1996) were one of the first authors that looked at the influence of personal characteristics, in this case personal environmental stance, on the relationship between the organizational environmental stance and organizational attractiveness. Personal environmental stance refers to the importance of the environment and the preserving of this environment. The results of Bauer and Aiman-Smith (1996) indicate that a pro-active environment stance of an organization increases organizational attractiveness to job seekers. Furthermore they found that the environmental orientation of an individual moderates the relationship between the environmental stance of the company and job pursuit intentions. However, they did not find this relationship for job attractiveness. Greening and Turban (2000) studied if valuing the environment moderates the relationship between CSP and organizational attractiveness but no significant results were found for this relationship. They explain this by arguing that even if someone does not value the natural environment, an organizations action towards the environment shows the pro-activity of that organization which could in general be attractive.

As argued by Aiman-smith, Bauer and Cable (2001), individuals differ in the needs they have and they try to find jobs in organizations that can best fulfil these needs. This fulfilling of needs will be accomplished if there is a right fit between an organization and the job seeker. For this, not only organizational reputation is important but also the personal values of these job seekers. Judge and Bretz (1992) were one of the first authors who studied the effect of values and then especially applied in a work setting on job choice decisions. Schwartz (1994, p. 20) argues that there are five features included in the definition of values that is widely agreed on: “A value is a (1) belief (2) pertaining to desirable end states or

(13)

13 modes of conduct, that (3) transcends specific situations, (4) guides action or evaluation of behaviour, people and events, and (5) is ordered by importance relative to other values to form a system of value priorities”. This makes values different than attitudes or needs. For example, independence could be a value but thirst is a need. Personal values represent “an enduring belief that a specific mode of conduct or end-state of existence is personally or socially preferable to an opposite or converse mode of conduct or end-state of existence” (Rokeach, 1973, p.5). Personal values can function as an interpretation mechanism, it influences which information receives greater attention or will be seen as more important (Evans and Davis, 2011). As Parks and Guay (2009, p. 676) say: “Values are evaluative; they guide individuals’ judgements about appropriate behaviour both for oneself and for others”. The development of values occurs in the first place through social interaction with role models, like teachers and parents, and secondly through unique learning experiences. Values are relatively stable over time (Parks and Guay, 2009).

The effect of values on the attractiveness of a particular work environment is particularly based on the Person-Organization fit (P-O fit) theory. Kristof (1996, p.4) defines P-O fit as the “compatibility between people and organizations that occurs when at least one entity provides what the other needs or they share similar fundamental characteristics, or both”. This theory explains that individuals search for jobs at organizations that have an organizational environment that fits one’s own personal values. A distinction could be made between supplementary fit and complementary fit. Supplementary fit indicates that the traits of a person are quite similar to traits of employees inside the organization. Complementary fit indicates that a person complements the current organizational environment by providing attributes that are seemingly missing (Kristof, 1996). Complementary fit could be separated in needs-supplies fit and abilities-demands fit (Kristof‐Brown, Zimmerman and Johnson, 2005). The former indicates the needs a person has and the values that are provided by the organization. The later indicates individual skills, knowledge and abilities in comparison to organizational demands.

The work of Judge and Bretz (1992) is later supported by Cable and Judge (1996) who argue that embracing equal values could be important for job seekers to select one organization over the other. It is the perception of this P-O fit that is important to job seekers. The perception of P-O fit depends on the similarity between someone’s own values and the perceived values held by the organization (Albinger and Freeman, 2000). Clary et al. (1998) argue that when people have values that are related to social behaviour they behave pro-socially to express their humanitarian and altruistic concern for others. So people choose

(14)

14 organizations that support CSR as a desire to express their values. Backhaus, Stone and Heiner (2002) also acknowledge that if people perceive congruence between organizational values and personal values the organization is more attractive. In addition, Coldwell et al. (2008) argues that ethically oriented people are attracted to particular organizations by virtue of the CSR aspect of the organizational reputation; this is because of the fit between organizational values and personal values. More recently Evans and Davis (2011) studied which underlying individual factors influence the relationship between perceived CSR and organizational attractiveness. Their research focuses on the role of CSR education and Other-regarding value orientation in relation to job seekers organizational attraction. They argue that congruence between personal and organizational values is of strong importance in the decision to join. Results of the study of Evans and Davis (2011) indicate that perceived CSR has a stronger impact on organizational attractiveness for people who value CSR, like those with an Other-regarding value orientation (helpful, compassionate, equality and loving). People with an Other-regarding value orientation and a low condition of perceived CSR were significantly less attracted to the organization than if this Other-regarding value orientation is not included. Furthermore Tsai and Yang (2010) studied if individual difference variables (Need for affiliation, environmental sensitivity and materialism) have an effect on the relationship among the different dimensions of corporate image (corporate product image, corporate service image, corporate citizenship image and corporate credibility image) and organizational attractiveness. Their findings indicate that environmental sensitivity moderates the relationship between a corporate citizenship image and organizational attractiveness. The P-O Fit theory is supported by Schneiders (1987) ASA model (Attraction, Selection, Attrition) which implicates that job seekers search for employment at organizations that have similar characteristics as themselves. In turn, organizations select those people that best fit their organization. Finally if a person joins an organization and it appears to be a miss-fit, a person leaves. As Schneider (1987) suggests: people make organizations what they are, instead of situations. Wherein attraction and selection occurs through matching values.

One of the most recognized value orientations, and also used for this study, are the personal value dimensions developed by Schwartz (1994). He defines personal values as: a “desirable trans-situational goal varying in importance, which serves as a guiding principle in the life of a person or other social entity” (Schwartz, 1994, p.21). Implicit in this definition is the aspect that distinguishes among values, the type of motivational goals they express. He identified 10 distinct types of basic human values that motivate behaviour and are likely to be recognized across and within cultures. These types can be used to organize personal values.

(15)

15 The 10 types are derived from three universal requirements of the human condition: individual needs as biological organisms, welfare and survival needs of groups and the need for coordinated social interaction (Schwartz, 1994). According to Schwartz (1994) these 10 value dimensions can be grouped into four higher order value types: openness to change (self-direction, hedonism and stimulation), conservatism (security, tradition and conformity), self-transcendence (universalism and benevolence) and self-enhancement (power and achievement). The self-transcendence value type is concerned with transcendence of selfish interest and enhancement of others as both universalism and benevolence show. Universalism includes: ‘’understanding, appreciation, tolerance, and protection for the welfare of all people and for nature’’. Elements include social justice, wisdom, equality and environmental concern (Schwartz and Boehnke, 2004, p.239). Benevolence is the ‘’preservation and enhancement of the welfare of people with whom one is in frequent personal contact’’. Elements include honesty, loyalty, responsibility and helpfulness (Schwartz and Boehnke, 2004, p.239). The focus of an organization on CSR also reflects a concern for interests beyond someone’s own personal interest (Evans and Davis, 2011). Therefore is likely that people with dominant self-transcendence values rate organizations as more attractive if they perceive that these organizations focus strongly on CSR. Therefore I hypothesize the following:

Hypothesis 2a: Self-transcendence relates positively to organizational attractiveness, so that job seekers with a high score on self-transcendence are more attracted to the organizations. Hypothesis 2b: Self-transcendence will moderate the relationship between perceived CSR and organizational attractiveness, so that the relationship is stronger for job seekers that score high on this value type than for job seekers that score low.

Hypothesis 2c: A high score on self-transcendence by job seekers has a stronger moderation effect on the relationship between perceived CSR and organizational attractiveness than job seekers who score high on one of the other three value types.

Age

Besides personal value orientation, as a personal characteristic that may have an influence on the relationship between perceived CSR and organizational attractiveness, age is also incorporated in this study. Newburry, Gardberg and Belkin (2006) studied whether a firm’s foreignness (do they have foreign headquarters) and their degree of internationalization have an influence on organizational attractiveness to job seekers. They examined if age directly

(16)

16 influences this. Their results represent a significant positive relationship between age and organizational attractiveness. In contrast, Aiman-Smith, Bauer and Cable (2001) studied the effect of job factors and organizational image factors on organizational attractiveness to job seekers. Ecological rating was used as an organizational image factor. They included age as a control variable. Their results indicate that age had a significant negative effect on organizational attractiveness. In addition Tsai and Yang (2010) also used age as a control variable in their study to investigate the relationship between different images relevant to social and environmental responsibility, product and credibility an organization could have and organizational attractiveness. Results of their study show that there was a significant negative correlation between age and individual level organizational attractiveness. Albinger and Freeman (2000) suggest that future research must focus on an individual level of analysis instead of an organizational level. For example: the age of job seekers. A study must be accomplished to investigate if younger job seekers put greater value on CSR than older ones in job-choice decisions. On the other hand Gavin and Maynard (1975) reported that older workers are more prone to believe that an organization wants to invest in a solution for social and environmental problems. It is therefore likely that the relationship between perceived CSR and organizational attractiveness is stronger for older people. In addition Aguinis and Glavas (2012) acknowledge that research must be done on a more micro level so that underlying mechanisms that link CSR with outcomes can be addressed. They point out that future research may focus on generational differences and CSR interest. Mcglone, Spain and Mcglone (2011) argue that millennials (people born between 1979 and 2001) believe it is their responsibility to make the world a better place to live in. A large part of them believes that organizations have the responsibility to accompany them in this effort. The majority of the millennials (79%) want to work for an organization that cares about how it contributes to society and even 69% would refuse to work for an organization that is not socially responsible. Subsequently, Meister (2012) argues that employees want an employer who can provide them more than a pay check and even 35% would accept a 15% pay cut to work for a company that is committed to CSR. This effect is according to Meister (2012) particularly apparent in younger generations. Therefore I hypothesize the following:

Hypothesis 3a: Age is negatively related to organizational attractiveness, so that younger job seekers are more attracted to the organizations than older job seekers.

(17)

17 Hypothesis 3b: Age will moderate the relationship between perceived CSR and organizational attractiveness, so that the relationship is stronger for younger job seekers than for older job seekers.

Gender

Finally gender is incorporated in this thesis. In general, benefits of recruiting tactics could differ between men and women. Especially role conflict can cause these differences. Organizational and job attributes that reduce conflict with nwork roles (like location, on-site day care, flexible hours and family friendly benefits) may be more attractive to women than to men (Chapman et al. 2005). This study focuses on different attitudes men and women have in the presence of CSR. In the study of Aiman-Smith et al. (2001), who investigate among others the influence of ecological image on organizational attractiveness, gender is included as a control variable. Results of their study represent a positive relationship between gender and organizational attractiveness, which implicates that females were more attracted to organizations. In contrary, Tsai and Yang (2010) who investigate the relationship between organizational image (citizenship, credibility and product) and organizational attractiveness, report that males are more attracted to organizations than females. Martins and Parsons (2007), who studied the effect of gender diversity management on organizational attractiveness, found no significant differences between male and female and organizational attractiveness. These results indicate the need for more research with regard to gender. Backhaus, Stone and Heiner (2002) argue that women have historically faced more entry barriers and career growth difficulties than men, so they will be more sensitive to activities that reflect conscientiousness of societal needs or the enhancement of society. The presence of positive CSR could suggest a more welcoming and better environment and leads in this way to more organizational attractiveness (Backhaus, Stone and Heiner, 2002). In addition, Chapman et al. (2005) argue that because women historically experienced more discrimination they are more sensitive to injustice. The presence of CSR programmes decrease this injustice and actually give a sense of social justice (Rodrigo and Arenas, 2008). Subsequently, results of Greening and Turban indicate (2000) that if there is a weak treatment of women and minorities, women were less attracted to an organization than men. This suggests that low CSR sends a stronger signal to women than to men. Subsequently, results of Ng and Burke (2005) show that women find diversity management (as part of the CSR practices) more important than men. Studies of Ibrahim and Angelidis (1991; 1993; 1994) indicate that men are more economically driven and women more philanthropically driven.

(18)

18 Smith, Wokutch and Harrington (2001) have also acknowledged this. They used the concept of Corporate Social Orientation (CSO). Results of their study indicate that women have a stronger discretionary orientation than men. They also have a significantly higher ethical orientation than men. Therefore I hypothesize the following:

Hypothesis 4a: Gender is positively related to organizational attractiveness, so that female job seekers are more attracted to the organizations than male.

Hypothesis 4b: Gender will moderate the relationship between perceived CSR and organizational attractiveness, so that the relationship is stronger for female job seekers than for male job seekers.

Conceptual model

A conceptual model (see figure 1) is developed based on the previous described literature and the formed hypotheses. This model is developed to give an answer on the research question: Does perceived Corporate Social Responsibility influence organizational attractiveness to job seekers? The underlying question is: Does this depend on individual characteristics?

(19)

19

3. Methodology

This chapter describes the methods and constructs that are used to undertake this quantitative empirical study. First the research design is explained. Secondly, it is explained how the two organizations that are included in the questionnaire are chosen. Afterwards, the respondents that were included in the sample will be described. Thirdly the survey construction process is explained. In addition an overview is given of the measures that are used for this questionnaire to operationalize the variables from the conceptual model. Subsequently a reliability analysis is performed. Finally an exploratory factor analysis is executed to better understand the structure of the variables.

3.1 Research design

The method used for this study is a self-administered questionnaire. The questionnaire could be filled in anonymous. This because an internet-mediated method was used. In addition this reduced socially desirable answering of the respondents (Saunders, 2009). The survey was developed with an online survey tool called Qualtrics (Qualtrics, 2014). The survey was divided into five parts. The first part was a general question to determine the gender of the respondents. The second part measured the personal value orientation of men and women. The third part measured the attractiveness of Friesland Campina and PostNL. The fourth part measured the perceived CSR of the two organizations and finally the fifth part asked the respondent some general questions, which were necessary to make statements about certain groups. In total the survey consists 12 questions, which included in total 76 items. Respondents were not informed why the two organizations Friesland Campina and PostNL were chosen. The survey was administered in Dutch. This increases the likelihood that all respondents were able to understand the questions well, which increases the validity of the study. In addition, the dropout rate is expected to be lower if the questionnaire will be in Dutch.

Before sending the survey to all respondents, a pilot study was conducted to ensure that the respondents encounter no problems when filling in the questionnaire, which will also increase the likelihood of fewer problems with recording the data (Saunders, 2009). Results of this pilot study represent that some questions were unclear due to translation difficulties. They were adapted so that the Dutch translation was more understandable. All questions included in the survey were drawn from existing scales. These questions were already tested, which will increase the validity of this study. See Appendix 4 for the final survey. The

(20)

20 abbreviations in red indicate which variable they measure, for the full names see Appendix 3. The operationalization of the variables is explained in section 3.3.

3.2 Research sample

The questionnaire is based on real organizations with real CSR reputations, instead of fictitious companies used in previous literature (Greening and Turban, 2000; Evans and Davis, 2011). The decision about which companies to include in the survey is based on the CSR reputation index developed by the Reputation Institute. The Reputation Institute is the world’s leading reputation-based advisory firm (Reputation Institute, 2014). The Reputation Institute created a tool to track corporate reputations globally across stakeholder groups, this tool is called the RepTrak® model. The RepTrak® model is a tool that can track 23 key performance indicators alongside 7 reputation dimensions. These dimensions are: (1) Products/Services, (2) Innovation, (3) Workplace, (4) Citizenship, (5) Governance, (6) Leadership, and (7) Performance (Reputation institute, 2014). If an organization delivers on these 7 dimensions this will lead to: the willingness to purchase products from the company, the willingness to work for the company or to invest in company stock. Three of these dimensions form the CSR reputation index. These are: Citizenship, Governance and Workplace. The organizations chosen for this study are derived from the Dutch annual CSR ranking of this reputation institute, which includes the 30 largest Dutch companies (CSR reputation, 2013). The two chosen organizations are Friesland Campina and PostNL. Friesland Campina has the highest CSR reputation ranking and PostNL has the lowest CSR reputation ranking. In this way it is most likely that there will be a difference in perceived CSR rating between these two organizations. In addition, it is likely that people are familiar with these companies because only the 30 largest companies in the Netherlands are rated.

Subsequently to the company sample, the people sample will be explained. The population on which the sample of this study is based are current job seekers in the Netherlands. To examine how big this group currently is, figures of the UWV (Uitvoerings Instituut Werknemersverzekeringen (body implementing employee insurance schemes)) are used. This because people that are registered at the UWV are actively seeking a job for more than 12 hours a week. Currently the total amount of unemployed people in the Netherlands is 680.000 (CBS, 2013). This is 8.5% of the working population. In addition the number of unemployment insurance benefits reached the point of 460.000 in January 2014, this is the highest point since 2002 (UWV, 2014). The final sample that is used for this study is derived from different sources with different sampling techniques. This is because a diverse sample

(21)

21 was needed due to the moderator variables included in this study. The UWV is, here in the Netherlands, the organization with the most connections to current job seekers. Therefore the UWV is approached to participate in this study. In addition, some other options are executed like placing the link of the survey on forums for job seekers and approaching people on the railway station. Finally connections are made with diverse small organizations that were mainly focused on coaching job seekers and on reintegration. This to create a wide distribution of the questionnaire.

Although it is tried to actively involve people to participate in this study, by raffling a gift voucher, a lot of people did not complete the questionnaire. In total 151 persons started the questionnaire, of this group 38 did not complete the questionnaire sufficiently. With sufficiently is meant; the first 20 questions that measure human values are completed, the attractiveness and perceived CSR questions are answered for at least one of the companies (Friesland Campina or PostNL) and the question if someone is currently actively seeking a job is answered. After the exclusion of these 38 respondents another 6 are excluded, these respondents completed all questions sufficiently but were not actively looking for a job. In addition de data is examined to check if there were outliers that must be excluded from the data. Boxplots are used for this. The boxplots indicate a few outliers, these respondents are analysed. None of the outliers were excluded because all the values were realistic and the questions are personal opinions. The final sample still includes some missing values however; enough questions were sufficiently answered to include the respondents in the analyses. Appendix 2 indicates which questions contain missing values. Because each question did not have above 10% missing data, Hotdeck imputation could be used to replace missing values. Replacing missing values with Hotdeck will lead to better results than with list-wise, pair-wise deletion or mean substitution (Myers, 2011).

3.3 Measures

In order to measure the variables that were depicted in the proposed conceptual model, they were operationalized as follows:

Perceived CSR (independent variable)

Perceived CSR is measured according to a construct developed by Lin, Tsai, Joe and Chiu (2012). They measure perceived CSR alongside four dimensions: perceived economic citizenship, perceived legal citizenship, perceived ethical citizenship and finally perceived philanthropic citizenship. Each dimension consists of five statements. For example, a

(22)

22 statement that is included in the economic dimension is: “the firm supports employees who want to acquire additional education’’ and for the philanthropic dimension a statement is: “this firm is concerned about respecting and protecting the natural environment’’. The statements are answered on a 5 point Likert scale from strongly disagree to strongly agree. For the complete list of all the statements see the final survey in Appendix 4. Results of the factor analysis of Lin, Tsai, Joe and Chiu (2012) represent Cronbach’s alpha’s for the four dimensions that al exceed 0.70, this indicates that it a reliable construct to measure perceived CSR (Field, 2013). In addition al factor loadings were significant; this confirms the convergent validity of the construct. The ratings of the 20 statements are summed and averaged so that they form the organizational perceived CSR variable.

Organizational Attractiveness (dependent variable)

The attractiveness of an organization to job seekers is operationalized according to four statements adapted from the study of Ployhart, Ryan, Bennett (1999). They have analysed organizational perceptions of applicants. The four statements are also used in a more recent PHD study at the University of Amsterdam that was focused on employee behaviour, this increases the reliability and validity of this measure. The four statements are: (1) “This employer is, in my eyes, attractive”, (2) “My attitude towards this employer is positive”, (3) “This employer is, in my view, a good organization”, (4) “My attitude towards this employer is favourable”. These question were rated on a 5 point Likert scales from strongly disagree to strongly agree. The ratings of the four statements are summed and averaged so that they form the organizational attractiveness variable.

Self-transcendence (moderator 1)

Self-transcendence is measured according to the European Social Survey (ESS) human value scale (ESS, 2012). The ESS is a biennial multicounty survey. The ESS human value scale is derived from the Portrait Value Questionnaire (PVQ) that consists of 40 items (Davidov, Schmidt and Schwartz, 2008). PVQ is developed because the original Schwartz Value Survey was less suitable for a large public as it was more cognitively demanding and harder to apply to less educated and younger samples (Bilsky, Janik and Schwartz, 2011). Results of the study of Bilsky et al. (2011) indicate that the 21-item PVQ is a conceptually and empirically appropriate construct to measure values and is a sound basis for cross-cultural research. Each of the 21 items describes a person’s aspirations, goals or wishes that implicitly point to the importance of a value. Every value item (benevolence, power etc.) has two statements, except

(23)

23 for universalism, which has three statements. For example, a statement that is included in the self-direction dimension is: “Thinking up new ideas and being creative is important to

him/her. He/she likes to do things in his/her own original way”. For the power dimension a statement that is included is: “It is important to him/her to be rich. He/she wants to have a lot of money and expensive things”. For the complete list of all the statements see the final survey in appendix 4. The 21 statement scores are combined into 4 higher order value types: openness to change (self-direction, hedonism and stimulation), conservatism (security, tradition and conformity), transcendence (universalism and benevolence) and self-enhancement (power and achievement).

Gender (moderator 2)

Gender is measured according to a dummy variable; 0= is male and 1= is female

Age (moderator 3)

Age is measured through an open question.

Education (control variable)

The control variable included in this study is education. This because Albinger and Freeman (2000) argue that people with lower education do not receive the same information about CSR, due to exposure to different kinds of media or literature. In addition Evans and Davis (2011) argue that CSR education increases the effect of perceived corporate citizenship on job attractiveness because it leads to a better knowledge structure that could give meaning and form to the CSR information. So, due to CSR education, CSR information could be better processed. Education is operationalized by 6 answer choices: lower education, high school, MBO, HBO, university bachelor, and university master.

3.4 Reliability of the variables

As the previous paragraphs indicate there are several variables that are measured by multiple items. To analyse if these variables are reliable, a reliability analysis is performed. Reliability means ‘’that a measure consistently reflect the construct that it is measuring’’ (Field, 2013, p. 706). The Cronbach’s alpha is used for this and is the most common measure that could test scale reliability. In general the Cronbach’s alpha must be 0.7 or higher to be considered as acceptable (Peterson, 1994, p. 381). However, as Nunnally (1994) in Field (2013) argues in early stages of research, a level of 0.5 could already be sufficient. The Cronbach’s alpha of

(24)

24 perceived CSR for Friesland Campina is 0.896 and for PostNL the Cronbach’s alpha is 0.907. This indicates that the perceived CSR variable in this study is reliable for both PostNL and Friesland Campina. The Cronbach’s alpha depends on the number of items that are included in the construct. The Cronbach’s alpha is smaller if the construct does not consist of many items and bigger if the construct consists of many items (Cortina, 1993). Therefore the Cronbach’s alphas of the four dimensions that underlie the perceived CSR variable are also measured. These dimensions all individually have Cronbach’s alpha’s that exceed 0.7, so also the individual dimensions are reliable. However, this study will only use the total perceived CSR score. In addition The Cronbach’s alpha of the organizational attractiveness variable is 0.810 for Friesland Campina and 0.895 for PostNL this indicates that the organizational attractiveness variable is also reliable. The final variables that are analysed are Schwartz (1994) higher order value types. The Cronbach’s alpha’s of the higher order values types are 0.590 for self-transcendence, 0.708 for self-enhancement, 0.699 for openness to change and 0.598 for conservation. In addition the total construct that includes the four higher order value types has a Cronbach’s alpha of 0.688 which is close to 0.7 and above 0.5 and is therefore reliable. However, only the four individual higher order value types are used in this study. 3.5 Exploratory factor analysis

Although the reliabilities of the variables are sufficient, a factor analysis is performed to better understand the structure of the set variables. Because several variables are measured by multiple items it is valuable to know if these items reflect a single construct, in other words are the items driven by the same underlying construct (Field, 2013). This is especially important for Schwartz (1994) basic human values because the ten measured value items are expected to present four higher order value types. For perceived CSR it is not valuable to perform exploratory factor analysis because the underlying dimensions of perceived CSR are not separately used in this study. Schwartz (2012) found support for the fact that the ten value items underlie four higher order value types. However the questions in this study are translated into Dutch. In addition the survey also includes other questions that were not originally combined into one survey. Therefore it is important to check is these four higher order value types are still visible in this study. If the wrong factors are combined to measure a particular construct, problems with multicollinearity could exist. This because some factors would load highly on other factors (Field, 2013).

To discriminate between the factors and maximize the results, the factors are rotated. The rotation technique used in this study is orthogonal varimax rotation. Varimax rotation

(25)

25 “tries to load a smaller number of variables highly on each factor, resulting in more interpretable clusters of factors” (Field, 2013, p. 681). The factor loadings are important to analyse the importance of a variable to a given factor. The significance of a factor loading depends on the sample size (Field, 2013). The critical factor loading for a sample that exceeds 100 is 0.512 according to Stevens (2002). The factor loadings are represented in table 1 in boldface.

A principal component analysis (PCA) was conducted on the 21 items that were assumed to measure the personal values types. PCA “is concerned with establishing which linear components exist within the data and how a particular variable might contribute to that component” (Field, 2013, 675). To test the sampling adequacy the Kaiser-Meyer-Olkin measure of sampling adequacy (KMO) is used. According to Kaiser (1974) a KMO value greater than 0.5 is acceptable. The KMO derived from this factor analysis is: 0.654, this indicates a mediocre sampling adequacy (Field, 2013). In addition Bartlett’s test shows that the correlation between the variables is significantly different form zero. Four components were extracted, together these components explained 45.75% of the variance. As table 1 show: component 1 refers to the openness to change value type, component 2 refers to the self-transcendence value type, component 3 to the self-enhancement value type and component 4 refers to conservation value type. However, the security 1 item, which is included in the conservation variable loaded substantially on component 1, which represents the openness to change dimension. An explanation for this is that, according to Schwartz (2012, p. 6) “exploratory factor analysis is not suitable for discovering the theorized set of relations among values because they form a quasi-circumplex, which EFA does not reveal. Factors obtained in an EFA with rotation will only partly overlap with the 10 values and will exploit chance associations”. In addition the reliability of the conservation variable does not increase if the security 1 item is excluded. Therefore the security 1 item is not excluded from the conservation variable. In addition there is one item that does not load substantially on any component. This item is self-direction 2. If self-direction 2 is excluded from the openness to change variable, the reliability of this variable will increase. Therefore this item is removed from the openness to change variable. The new Cronbach’s Alpha for the openness to change variable is 0.706.

(26)

26 Table 1. Factor loadings for exploratory factor analysis with varimax rotation

Rotated factor loadings

Item 1 2 3 4 Self-Direction 1 0,484 -0,045 0,249 -0,077 Power 1 0,300 -0,256 0,626 -0,005 Universalism 1 -0,108 0,653 0,129 -0,016 Achievement 1 0,079 0,123 0,715 0,074 Security 1 -0,554 0,266 0,032 0,294 Stimulation 1 0,698 0,268 0,099 0,086 Conformity 1 -0,073 -0,134 0,124 0,668 Tradition 1 -0,438 0,308 -0,101 0,317 Benevolence 1 -0,001 0,706 -0,114 0,003 Achievement 2 0,063 0,135 0,714 -0,128 Security 2 0,176 0,362 -0,147 0,510 Stimulation 2 0,760 0,105 0,254 0,067 Power 2 0,083 -0,117 0,711 0,191 Benevolence 2 0,214 0,523 0,072 0,133 Universalism 3 -0,114 0,412 -0,057 -0,101 Hedonism 2 0,454 0,230 0,016 0,417 Universalism 2 0,172 0,609 0,095 0,081 Hedonism 1 0,432 0,323 0,263 0,260 Self-Direction 2 0,214 0,273 0,379 -0,300 Conformity 2 -0,493 0,154 0,324 0,426 Tradition 2 -0,054 -0,033 -0,007 0,622 Note. Factor loadings > 0.512 are in boldface

3.6 Data analysis

To answer the research questions and check the proposed hypotheses two kinds of analyses are conducted: a descriptive analysis of the data and an inferential analysis. The descriptive analysis of the data indicates mean scores, standard deviations and correlations. In addition the assumptions for a multiple linear regression are checked.

The next step is the inferential analysis, although descriptive statistics facilitate representations and summaries of the obtained information, conclusions could mainly be drawn from inferential statistics (Keller, 2005). The inferential analyses will actually test the proposed hypotheses. To test the eight hypotheses two hierarchical multiple linear regression analyses are conducted. The following section will discuss the results of these two regression analyses.

(27)

27

4. Results

This chapter will explain the results of this study. First, the characteristics of the data are explained. Subsequently, the assumptions for a linear regression are analysed and finally the results of the two regression analyses are discussed.

4.1 Characteristics of the data

This section pays attention to the characteristics of the dataset. These characteristics are presented in order to make the right interpretations of the data. The characteristics of the dataset are described in three paragraphs. The three paragraphs address the general characteristics of the dataset, the descriptive statistics of the research variables and a check if the assumptions for performing a multiple linear regression are met.

General characteristics of the data

This paragraph will explain the general characteristics of the dataset.

In total 64 women (59.8%) and 43 men (40.2%) participated in this study. Almost half of the respondents have an HBO education level, this represents 45.8% of the sample. 22 respondents have an education level above HBO and 36 have an education level lower as HBO. In addition the respondents have an average age of 36. Subsequently they look on average for almost 10 months for a job, with one month as a minimum and up till 7 years as a maximum. Of the 107 respondents 7 did apply for a job at Friesland Campina and 9 did apply for a job at PostNL. Only 2 respondents did actually get a job at PostNL. At Friesland Campina no one got the job.

Descriptive statistics and correlations

The following table (table 2) shows the descriptive statistics of the research variables. As table 2 indicates, people score on average the highest on the self-transcendence value type and the lowest on the conservation value type. Subsequently Friesland Campina was included in the questionnaire as a company with a high CSR reputation and PostNL as an organization with a low CSR reputation. This could also be derived from table 2; the average perceived CSR score for Friesland Campina is above three which indicates a positive CSR reputation and the average score of PostNL is around three which indicates a neutral reputation. In addition the average attractiveness score of Friesland Campina is above the neutral score of three, which indicates high organizational attractiveness. In contrast the average attractiveness score of PostNL is below the neutral score, which indicates low organizational

(28)

28 attractiveness. Table 2 also indicates the correlations among the variables. From this correlation matrix it is apparent that there is in particular a mutual correlation between attractiveness and perceived CSR, this is the case for as well Friesland Campina as PostNL. This is a first indication that hypothesis 1: Job seekers will rate companies as more attractive

organizations when they are perceived as social responsible organization, could be

supported. In addition there are some significant correlations between the higher order value types. This is not surprising because they form together a circular motivational continuum (Schwartz, 2012). There is no significant correlation between the self-transcendence value type and perceived CSR for both Friesland Campina and PostNL. This may indicate that hypothesis 2b: Self-transcendence will moderate the relationship between perceived CSR and

organizational attractiveness, so that the relationship is stronger for job seekers that score high on this value type than job seekers that score low, is not supported. However a

significant correlation is visible between self-transcendence and age. Which could indicate support for hypothesis 3b: Age will moderate the relationship between perceived CSR and

organizational attractiveness, so that the relationship is stronger for younger job seekers than for older job seekers. This is because both CSR and self-transcendence are concerned

with enhancement of others and the transcendence of selfish interests. Finally, no correlation is found between gender and one of the variables. This could indicate that there is no support for hypothesis 4b: Gender will moderate the relationship between perceived CSR and

organizational attractiveness, so that the relationship is stronger for female job seekers than for male job seekers.

Besides using the correlations to get a first indication if the proposed hypotheses could be supported. The correlation matrix could also be used to check if there is multicollinearity between the independent variables. Multicollinearity means there is a high significant correlation between two predictor variables (r > 0.7) (Field, 2013). If there is multicollinearity in the data the regression analysis may not be performed. This study uses the Pearson’s correlation coefficient to check if the assumption of no multicollinearity between the independent variables is met. The correlation matrix indicates that there is no multicollinearity in the data as there are no correlations that exceed r > 0.7.

(29)

29 Table 2. Descriptive statistics and correlations

Variables M SD 1 2 3 4 5 6 7 8 9 10

Gender 0,600 0,493

Age 35,610 13,053 -,223* Org. Attractiveness Friesland Campina 3,589 0,548 -,050 ,098 Org. Attractiveness PostNL 2,759 0,811 -,044 -,031 ,184 Perceived CSR Friesland Campina 3,482 0,394 -,160 ,225* ,397** ,090 Perceived CSR PostNL 3,015 0,485 ,040 -,051 ,114 ,564** ,243* Self-transcendence 3,898 0,515 ,162 ,227* ,093 ,061 ,126 ,030 Self-enhancement 3,248 0,702 -,173 -,134 ,180 ,002 ,210* ,073 ,037 Conservation 3,090 0,553 ,100 ,005 ,177 ,263** ,267** ,169 ,228* -,006

Openness to change 3,596 0,634 -,017 -,233* ,019 ,241* 0,063 ,368** ,359** ,253** -,097 * Correlation is significant at the 0.05 level (2-tailed).

** Correlation is significant at the 0.01 level (2-tailed).

Multiple linear regression assumptions

Besides the assumption of no multicollinearity between the independent variables, there are a few other assumptions that must be met for performing a regression analysis. First the variables that are used for analysis must be scale variables, if a nominal variable is used it must be recoded as a dummy variable. The dummy variable included in this study are 0=Male and 1=Female.

Second the model must be linear. This can be checked with a scatterplot. If there is no clear pattern of the residuals; all positive and negative residuals are more or less balanced around the horizontal zero line of the graph, linearity may be assumed.

In Addition the model must include homoscedasticity of the variance. This could be analysed with the same scatterplot. If it does not show a funnel, than homoscedasticity could be assumed. Both the scatterplots do not show a clear pattern of residuals or a funnel therefore linearity and homoscedasticity could be assumed.

The fourth assumption is the assumption of normality. This assumption indicates the normal distribution of the dependent variable. There are a few tests to check if the dependent variable is normally distributed: Kolmogorov-Smirnoff, Skewness and Kurtosis, P-P plots and histograms. Because the sample in this study exceeds the 50 respondents, Kolmogorov Smirnoff and Skewness and Kurtosis are not suitable to check if the variable is normally distributed. P-P plots and histograms are more suitable. P-P plots are shown in appendix 1. If the values lay on the diagonal of the figure a normal distribution is indicated (Field, 2013). These P-P plots show that both dependent variables are normally distributed.

Perceived CSR and organizational attractiveness : the influence of personal values, age and gender