- Email: [email protected]
Mentoring in supervisor–subordinate dyads: Antecedents, consequences, and test of a mediation model of mentorship
Journal of Business Research 62 (2009) 1110–1118
Contents lists available at ScienceDirect
Journal of Business Research
Mentoring in supervisor–subordinate dyads: Antecedents, consequences, and test of a mediation model of mentorship ☆ Orlando C. Richard a, Kiran M. Ismail b,⁎, Shahid N. Bhuian c, Edward C. Taylor d a
The University of Texas at Dallas, United States St. John's University, United States Louisiana Tech University, United States d Piedmont College, United States b c
a r t i c l e
i n f o
Article history: Received 17 September 2007 Accepted 6 September 2008 Keywords: Mentoring Social exchange Trust Organizational commitment Dyadic relationships
a b s t r a c t We develop a framework to look at mentoring as a consequence of employees' values and beliefs, as well as to explore its role in determining the employees' attitudes towards their organizations. Based on social exchange theory, we hypothesize that employees' levels of individualism, collectivism, and trust in supervisor influence the level of supervisory mentoring received. Moreover, mentoring influences employees' affective commitment and intention to quit, as well as mediates the relationships between the proposed antecedents and outcomes. Using structural equation modeling to examine these relationships, we find that within supervisor–subordinate dyads, subordinates report more mentoring when they have collectivist personal values and trust their supervisor. Additionally, more mentoring is positively related to subordinates' affective commitment towards the organizations and negatively related to their intention to quit. We also find that mentoring mediate the relationship between both collectivism values and trust in supervisors and both organizational commitment and intention to quit. © 2008 Elsevier Inc. All rights reserved.
1. Introduction Mentoring relationships are recognized by both academicians and practitioners as a valuable critical resource for employees in organizations. Organizational theorists and management scholars have identified mentoring as an exchange relationship whereby both mentor and protégé gain several benefits from each other (Ragins, 1997; Young and Perrewe, 2000; Young and Perrewe, 2004). For example, compared with non-mentored individuals, mentored employees demonstrate higher levels of objective and subjective positive outcomes such as career development, job satisfaction, socialization, organizational commitment, and career advancements (Allen et al., 2004; Eby et al., 2008). Mentors, in return for the time and effort spent in providing support to the protégés, gain positive outcomes such as career rejuvenation, recognition, personal satisfaction, organizational reputation, and increase in knowledge and power (Noe et al., 2002). While scholars have used various theoretical perspectives to explain mentoring relationships (e.g. leadership, justice, power, exchange, motivation; see Noe et al., 2002, for a detailed review), ☆ A previous version of the paper was presented at the 2007 Academy of Management Annual Meeting, Philadelphia, PA. We thank associate editor Dr. Lei-Yu Wu and three anonymous reviewers of Journal of Business Research for valuable feedback on earlier versions of the paper. ⁎ Corresponding author. St. John's University, Peter J. Tobin College of Business, 8000 Utopia Parkway, Queens, NY 11439, United States. Tel.: +1 718 990 7423. E-mail address: [email protected] (K.M. Ismail). 0148-2963/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2008.09.007
the exchange process between a mentor and protégé lies at the heart of mentoring. Thus, management scholars have suggested that social exchange theory (Homans, 1961; Blau, 1964) is an appropriate theoretical lens for describing and studying the processes involved in mentoring relationships (Olian, et al., 1993; Ensher et al., 2001; Tepper and Taylor, 2003). Based on the social exchange perspective, we define mentoring as a reciprocal exchange relationship between a mentor and a protégé (Young and Perrewe, 2000, 2004). While workplace mentoring has been traditionally defined as a hierarchical relationship between a senior and influential organizational member (mentor) and a junior and less experienced organizational member (Kram, 1983), scholars have identified various forms of mentoring relationships, such as lateral or peer mentoring, supervisory mentoring, team mentoring, and mentoring by an external sponsor (see Eby, 1997; Allen and Eby, 2007; Eby, Rhodes and Allen, 2007; Scandura and Pellegreni, 2007). The focus of the current study is on supervisory mentoring. Our basis for drawing this boundary is the notion that employees tend to interact most frequently with their supervisors than with other agents of the organization, and therefore, supervisors are in the best position to serve as organizational representatives (Tepper and Taylor, 2003). Moreover, extant literature also suggests that supervisors are in the most natural position to, and have the responsibility to provide career and psychosocial support to the subordinates (Eby, 1997: pp. 135/6), and employees are likely to obtain mentoring from their immediate supervisors (Ragins and McFarlin, 1990; Tepper, 1995; Tepper and Taylor, 2003).
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
According to social exchange theory, individuals' beliefs about the support they receive from their employing organizations play an important role in effecting their behaviors and attitudes towards their organizations. While individuals' personal values, as well as the quality of the inter-attitudinal relationships that they experience with their supervisors play an important role in effecting these beliefs, research has not systematically explored the roles of these variables in mentoring relationships. Thus, the main purpose of this study is to take insights from social exchange theory in order to study: (a) employees' personal values (i.e. individualism–collectivism) and employees' relationships with the supervisors (i.e. trust in supervisor) as the antecedents of supervisory mentoring antecedents, (b) employees' commitment to the organization and intention to quit as the outcomes of supervisory mentoring, and (c) the mediating role of supervisory mentoring in the relationships between its antecedents and outcomes. The theoretical model that is being proposed in the current study is shown in Fig. 1. Therefore, this study makes the following contributions. First, based on the notions that mentoring is an exchange relationship between a mentor and a protégé, and that employees regard their supervisors as the representatives of their organizations, we provide a theoretical perspective to look at mentoring as a consequence of employees' personal attitudes, as well as to explore its role in determining the employees' attitudes towards their organizations. Second, we go beyond simple linear regression modeling to investigate more complex relationships as posited by extant theory. Specifically, we use structural equation modeling to examine the role of mentoring as a mediating variable in the relationship between individuals' values and relationships, and their performance outcomes. 2. Theoretical background and hypotheses 2.1. Social exchange theory The basic tenet of social exchange theory is that human interactions are based upon exchange of social and material resources and norms of reciprocity (Homans, 1961; Blau, 1964). The theory posits that people enter into relationships in which benefits and rewards (tangible or intangible) are greater than the costs. Unlike economic exchanges that are characterized by legally-sanctioned contractual arrangements, social exchanges are ‘voluntary actions’ (Blau, 1964) that are embedded in trust that the recipient of a benefit will eventually reciprocate with an equitable, but not necessarily similar, favor (Haas and Deseran, 1981; Gould-Williams and Davies, 2005). While social exchange theory originally focused on explaining relationship between individuals, management scholars have established that the theory plays an important role in explaining the relationship between the organization and its employees (Eisenberger et al., 1986; Whitener, 2001). The ‘voluntary’ exchange relationship between an employee and the employing organization is governed by what management theorists have referred to as ‘psychological contracts’,
1111
defined as an individual's beliefs about the terms of his or her reciprocal relationships with the employing organization (Robinson et al., 1994; Robinson, 1996). As we mentioned in the previous section, such beliefs are effected by individual's personal values, as well as the quality of relationship they experience with their co-workers and supervisors. Based upon the insights offered by social exchange theory, each component of the theoretical model that has been proposed earlier is discussed below. 2.2. Mentoring and organizational commitment Organizational commitment is defined as the level of attachment and identification that an individual feels towards the organization in which he or she is employed (Mowday et al., 1982; Bartlett, 2001). A popular conceptualization of the construct has been offered by Meyer and Allen (1991), who have identified commitment as a multidimensional construct, consisting of three components: affective commitment, which refers to individuals' desire to be committed to the organization due to the emotional attachment felt towards the organization as a result of personal liking or preference, continuance commitment, which refers to individuals' need to remain in the organization due to the costs associated with leaving, and normative commitment, which refers to individuals' moral obligation to be committed to the organization due to their personal values (see also Meyer et al., 1993). Using the social exchange perspective, management scholars have argued that the levels of employees' commitment to their organizations depend upon their beliefs regarding how committed their employing organizations are to them, and the support they receive from their employers (Wayne et al., 1997; Whitener, 2001; Gould-Williams and Davies, 2005). Research in the area has suggested that since supervisors are perceived by employees as agents of the organization, the behaviors of managers and supervisors play critical roles in effecting psychological contracts and signaling the level and nature of the organizations' commitment to the employees, and hence, shape the employees' behaviors and attitudes, or more specifically, their level of commitment, towards their organizations (e.g. Guzzo et al., 1994; Rousseau and Greller,1994; Wayne et al.,1997; Bartlett, 2001; Whitener, 2001; Lee and Bruvold, 2003; Gould-Williams and Davies, 2005). Past research has suggested that mentoring is an important determinant of an employee's affective and continuance commitment to the organization (Koberg et al, 1998; Allen et al., 2004; Payne and Huffman, 2005). This argument is based on the notion that a successful exchange relationships increases employees' commitment to their organizations (Flynn, 2005). We seek to confirm this proposition by exploring the relationship between mentoring and two dependent variables: affective commitment, and intention to quit. Mentors offer different types of social and psychological support to their protégés (Scandura, 1992). The various types of support include social support, coaching and role modeling support, career support, and emotional support (Kram, 1985; Ragins and Cotton, 1999; Ensher et al., 2001). From a social exchange perspective, it can be argued that
Fig. 1. Conceptual model.
1112
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
protégés offer increased attachment, identification, and involvement with their organizations in exchange for the support they receive from their mentors. We had mentioned earlier that from the social exchange perspective, employees' commitment to the organization depends upon their ‘perceptions of how committed their organizations are towards them' (Gould-Williams and Davies, 2005). When subordinates receive higher amount of mentoring from their supervisors, they perceive it as a form of fair treatment from their organizations and reciprocate by increasing their level of citizenship behavior towards their organizations (Tepper and Taylor, 2003). Moreover, since social exchange theory also suggests that people tend to enter and remain into relationships in which benefits outweigh the costs, it can also be argued that, as an intangible benefit, mentoring results in a favorable cost to benefit ratio for the recipients and is therefore negatively related to their intention to quit the organization. In summary, we propose that the psychosocial support provided by supervisors in the form of mentoring is an important signal of organizations' commitment towards the employees, and thereby, increases the levels of employees' commitment towards their organizations. H1a. Supervisory mentoring received by an employee is positively related to his or her level of affective commitment to the organization. H1b. Supervisory mentoring received by an employee is negatively related to his or her intention to quit the organization. 2.3. Trust in supervisor and mentoring Defined in the literature as positive expectation regarding favorableness of another's actions, trust is a critical aspect of social exchange relationships (Blau, 1964; Haas and Dseran, 1981; Robinson, 1996; GouldWilliams and Davies, 2005). Trust reduces opportunism in relationships (Williamson, 1979), thus enabling interacting parties to work together effectively and co-operate (McAllister, 1995; Rousseau et al., 1998). Based on the notion that trust is a determining factor of effective social exchange relationships between individuals who work together (Seabright et al., 1992; McAllister, 1995), we propose that the level of trust between an employee and his or her supervisor is an important determinant of the amount of mentoring received by the employee. From a social exchange perspective, it can be argued that employees who exhibit higher levels of trust in their supervisors, in turn, are able to derive greater benefits out of their exchange relationship with their supervisors. As suggested by literature, high-quality leader–member exchange relationships, that are characterized by trust and loyalty, hold benefits for both supervisors, in the form of motivated, committed, competent, and conscientious subordinates, and employees, in the form of higher appraisals, promotions, and mentorship from the supervisor (Liden and Graen, 1980; Deluga, 1994; McManus and Russell, 1997). Thus, it is likely that supervisors are more willing to invest time and effort in mentoring employees who exhibit greater levels of trust in them, instead of spending time and effort on coaching employees with whom they do not have expectations of developing a high-quality and effective exchange relationships. H2. An employee's trust in his or her supervisor is positively related to the amount of mentoring received from the supervisor. In this study, we are also interested in exploring the mediating role of mentoring in the relationship between its antecedents and outcomes. Based on the social exchange theory, we believe that employees' levels of commitment towards their organization are not directly influenced by their collectivistic values or trust in supervisor. Instead, their levels of commitment are influenced by whether or not they receive benefit from the organization in exchange or return for their collectivistic values or trust in supervisor.
In intra-organizational relationships, psychological contract breach and betrayal of trust are not uncommon phenomena (Robinson, 1996; Elangovan and Shapiro, 1998). Nooteboom (1996) suggests that “trust is related to propensity, not a certainty: it may not be resistant to golden opportunities” (1996: p. 994). Thus, from the social exchange perspective, it can be argued that employees who exhibit high levels of trust in their supervisors expect to gain benefits from their high-quality exchange relationships, and they tend to demonstrate greater levels of organizational commitment only when they are able to derive benefits out of the relationships. Therefore, we also propose that employees who have higher levels of trust in their supervisors will receive higher amount of mentoring, and thus, will exhibit higher levels of organizational commitment. H2a. Mentoring fully mediates the positive relationship between an employee's trust in supervisor and the level of affective commitment to the organization. H2b. Mentoring fully mediates the negative relationship between an employee's trust in supervisor and his or her intention to quit the organization. 2.4. Individualism–collectivism and mentoring While individualism–collectivism constructs have been used at societal level, to describe cultural difference, it has been argued that people from individualistic and collectivist cultures hold idiocentric values, and that the constructs are situation-specific. The constructs are often used by management scholars to measure attitudes and values at individual level of analysis, and capture individuals' preference for ether companionship (collectivism) or personal identity (individualism) (Triandis, 1995; Earley and Gibson, 1998; Noordin et al., 2002). In the context of an organizational setting, individualism and collectivism refer to the patterns in which employees relate to their co-workers, teams and workgroups, supervisors, and organizations. Collectivists define self by in-group memberships (Triandis, 1989; Earley and Gibson, 1998) and seek close and long-term relationships. In an organizational setting, they feel morally obliged to serve their company, and value provisions offered by the organization (e.g. training, physical conditions). Collectivists are particularly concerned about the results of their actions on their organizations (Triandis, 1989), and keeping/losing face in front of others, and this may be a major incentive for them to fulfill the duties expected of them (Hofstede, 1980, 1991). They have strong need for harmony in the work environment. Individualists, on the other hand, have calculative involvement with organization and have strong need for freedom and preference for low-context relationships based on emotional detachment (Earley and Gibson, 1998; Hofstede, 1980, 1991; Triandis, 1995). The individualist preferred work environment is competitive. It has been suggested that collectivists are likely to be provided protection and security in return for their loyalty to their in-groups (Noordin et al., 2002). From a social exchange perspective, we propose that employees with high levels of collectivism tend to build close and high-quality relationships with their supervisors, who, in return for their loyalty, and in respond to their collectivistic employees' need for seeking close, interpersonal relationships and provisions from the organization, offer them higher amount of mentoring, than they are likely to provide to the employees having higher levels of individualistic orientation. H3. An employee's collectivism level is positively related to the amount of supervisory mentoring. H4. An employee's individualism level is negatively related to the amount of supervisory mentoring. According to Shamir (1990), individuals are motivated to contribute to collective work efforts either due to their calculative considerations regarding the nature of rewards and sanctions they
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
expect to receive, or because they are morally obliged to do so, or in order to take advantage of opportunities to express a cherished and salient identity. From a social exchange perspective, it can be argued that individuals are likely to feel committed to their organizations to the extent that the organizations offer them provisions and opportunities that clarify and affirm their values, self-concepts and identities. We propose that mentoring is an example of one such provision or opportunity. Thus, based on the arguments discussed earlier regarding the nature of relationship between mentoring and organizational commitment (Hypothesis 1a and 1b), and the insights offered in the preceding discussion regarding the nature of relationship between individualism–collectivism and mentoring, we propose: H3a. Mentoring fully mediates the positive relationship between an employee's collectivism level and the level of affective commitment to the organization. H3b. Mentoring fully mediates the negative relationship between an employee's collectivism level and his or her intention to quit the organization. H4a. Mentoring fully mediates the negative relationship between an employee's individualism level and the level of affective commitment to the organization. H4b. Mentoring fully mediates the positive relationship between an employee's individualism level and his or her intention to quit the organization. 3. Methodology 3.1. Data collection and setting We tested the hypotheses using data collected from employees of numerous organizations in the metro Atlanta area. We selected a research design that would span as many organizations and supervisor–subordinate dyads as possible thereby maximizing the variance on study variables. We elected to distribute surveys through two MBA classes of about twenty students each with instructions that they find respondents who do not share the same boss. In this way, we avoided inadvertent nesting when the survey design was executed. Multiple responses by subordinates of the same boss would require that we account for variance both within groups and between groups. Our survey distribution method produced a data set that may have as many as twenty responses from a single business organization but no more than one response per boss within those businesses. Twohundred fifty surveys were distributed, 220 were returned and 10 were discarded due to incomplete data. Twenty variables were randomly selected as a test for monotonic responses—another form of non-response. These survey items were then scanned for the same response, if all twenty were equal to the same value then the survey was manually scanned for monotonic responses. Ten surveys were discarded as a result of this procedure. This left a data set with an N of 200. Of the respondents 67% were over 25 years old, 90% of the respondents were Caucasian; 54% were male, and the average yearly earnings were $35,500. 3.2. Measures 3.2.1. Dependent measures Affective commitment was measured using five items adapted from Meyer et al. (1993). (Sample item: “My company has a great deal of personal meaning for me”). One item on a five-point Likert type scale was used to capture intention to quit. It stated “If a similar job, offering the same pay and work, came available to me in another company, I would be likely to transfer.”
1113
3.2.2. Independent measures Collectivism and individualism were measured using two different scales as opposed to a uni-dimensional measure and was designed for the individual level of analysis distinct from national culture measures (Clugston et al., 2001). A five items Likert type scale was designed to measure individualism. (Sample item: “If a group is slowing me down, it is better to leave it and work alone”). Collectivism was measured using a three items Likert type scale. (Sample item: “A person should accept the group's decision even when personally he or she has a different opinion”). Trust was operationalized as a belief in the benevolence of the other party, the matter of relevance in interpersonal relationship risk assessment, by adapting Larzelere and Houston's (1980) 8-item unidimensional dyadic trust scale in close interpersonal relationships, e.g., dating couples, to the supervisor–worker relationship. Trust was measured by asking subordinates to describe how much trust they had in their supervisor on a 5-point Likert type scale with responses ranging from 1—strongly disagree to 5—strongly agree. (Sample item: “My supervisor is truly sincere when making promises”). We used Kram's (1985) conceptual framework for mentoring and Noe (1988) as the basis for our survey items measuring mentoring activity. Noe's (1988) scale contained eight psychosocial related items and seven career help items for a total of 15 items. We adapted 10 survey items from these 15 items for our measure of mentoring. Mentoring activity was measured by asking subordinates to provide how much of each type of mentoring activity they had received from their supervisor on a 5-point Likert type scale with responses ranging from 1—strongly disagree to 5—strongly agree. (Sample item: “My supervisor exposes me to upper management”). 3.2.3. Control measures Length of the mentoring relationship was represented by the number of years with the boss. Age was categorized into five year increments ranging from normal age of entry into the workplace to retirement. Education categories were designed to capture high school dropouts, high school graduates, college attendees, college graduates, master's level graduates and doctoral degrees. Sex was dichotomized as male and female. The number of years with the boss served as a control on mentoring, while age, education, sex and years of experience were controls on two outcomes—affective commitment and intention to quit. Further, in order to control for the degree of similarity between the mentor and protégé, age, education and sex of both employees and bosses were included as controls. 3.3. Confirmatory factor analysis (CFA), reliability, validity, non-response bias, and common method bias We used CFA to assess the scale properties. The CFA Model included five multi-item constructs: collectivism (3 items), individualism (5 items), trust in supervisor (8 items), mentoring (10 items) and affective commitment (5 items). The initial run failed to produce a good fit of the model. Based on poor loadings, modification indices, and high residuals, we detected eighteen bad items: 2 of individualism, 4 of trust in supervisor, 6 of mentoring, and 2 of affective commitment. The deletion of these items was based on content considerations to minimize the reduction of the meaning of the constructs. After deleting these items one by one, the model achieved a good fit: χ2 = 146.884, df =109, GFI =0.924, AGFI =0.893, CFI =0.970, RMSEA =0.042, and PCLOSE=0.779. The t-statistics pertaining to the items ranged from 2.467 to 13.940, i.e., all were significant. We assessed scale reliabilities by estimating average variance extracted (AVEs) and composite reliabilities (Bagozzi and Yi 1988) and found all scales to be reliable. The AVEs for the scales of collectivism, individualism, trust in supervisor, mentoring and affective commitment were 0.987, 0.879, 0.922, 0.851, and 0.882 respectively. The composite reliabilities were 0.99, 0.96, 0.97, 0.96, and 0.96
1114
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Table 1 Correlations, Means, Std. Dev., AVEs, and composite reliabilities.
1. Mentoring 2. Collectivism 3. Individualism 4. Trust in Supervisor 5. Affective commitment 6. Intention to quit 7. Age 8. Education 9. Sex 10. Years of experience Means Standard Deviations AVEs Composite Reliabilities
1
2
3
4
5
6
7
8
9
10
1.0 .157⁎ − .182⁎ .649⁎⁎ .555⁎⁎ − .375⁎⁎ − .127 .131 .100 − .081 3.385 .765 .987 .99
1.0 − .018 .133 .136 − .134 .009 .031 .081 .025 3.272 .663 .879 .96
1.0 − .194⁎⁎ − .122 .106 − .109 − .049 .024 − .137 2.700 .809 .922 .97
1.0 .127 − .474⁎⁎ .023 .059 .186⁎⁎ .016 3.868 .851 .851 .96
1.0 .123 .083 .098 .145⁎ .132 3.533 .969 .882 .96
1.0 − .056 − .104 − .209⁎⁎ − .118 2.290 1.12 – –
1.0 − .103 .051 .880⁎⁎ 3.46 2.50 – –
1.0 .041 − .129 3.33 .90 – –
1.0 − .067 1.46 .50 – –
1.0 14.08 11.29 – –
⁎ = correlation is significant at the 0.05 level (2-tailed). ⁎⁎ = correlation is significant at the 0.01 level (2-tailed).
respectively. If the values of AVEs are greater than 0.50, and the values of composite reliabilities are greater than 0.60, they are considered desirable (Fornell and Larcker 1981). The convergent validity of all scales was supported. The t-values pertaining to all items of all scales were each greater than 2.0 (at p b 0.000). When each indicator's estimated path coefficient on its underlying construct factor is significant (greater than twice its standard error) in the measurement model, the scales of the constructs in the model achieve convergent validity (Anderson and Gerbing 1988). Further, the scales achieved discriminant validity
because no confidence intervals of the correlations for the constructs (Ф values) included 1.0 (p b 0.05) (Anderson and Gerbing 1988). We also tested for non-response bias by comparing the responses of the early respondents with the late respondents to all 18 items (Armstrong and Overton 1977). All multivariate tests results were insignificant: Box's Test of Equality of Covariance Matrices (Sig. = 0.475), Pillai's Trace (Sig. = 0.213), Wilks' Lambda (Sig. = 0.213), Hotelling's Trace (Sig. = 0.213) and Roy's Larget Root (Sig.—0.213) indicating that responses between early respondents and late respondents were not significantly different.
Fig. 2. Structural model 1: full mediation by mentoring.
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Further, following Podsakoff and Organ (1986), Harman's one factor test was conducted where all scale items were entered into a factor analysis. Not a single factor emerged nor was there a general factor that accounted for the majority of the covariance in these variables (the first factor explained 24.76% variance, the second factor 14.56%, the third factor 11.11% and the fourth factor 8.75%). Therefore, common method bias is not considered a severe threat in this study. 4. Results Table 1 shows the means, standards deviations, and correlations among the variables of interest. In order to test the hypotheses, we ran two structural models, as shown in Figs. 2 and 3. Additionally, we ran a third structural model 3 for a further model comparison. Structural model 1 is our hypothesized model, where mentoring fully mediates the influence of the three antecedents and one control variable, the number of years with the boss, on two outcomes. The model also includes seven control variables: age of employee, age of boss, education of employee, education of boss, sex of employee, sex of boss, and employee years of experience. The model demonstrated a fair fit: χ2 = 344.684, df = 233, GFI = 0.891, AGFI = 0.836, CFI = 0.940, and RMSEA = 0.049. Collectivism was positively related to mentoring (see Table 2 for the estimates). Individualism was not related to mentoring. Trust in supervisor was positively related to mentoring. Mentoring was positively related to affective commitment and negatively to the intention to quit. The number of years with the boss was unrelated to mentoring. Employee age was slightly related to affective commitment and positively related to intention to quit. Employee education was slightly related to affective commitment and
1115
was negatively related to intention to quit. Employee sex was positively related to affective commitment and negatively to intention to quit. Employee years of experience were slightly positively related to affective commitment and negatively to intention to quit. Boss's age was unrelated to affective commitment, while positively related to intention to quit. Boss's education was unrelated to both affective commitment and intention to quit. Boss's sex was negatively related to affective commitment and unrelated to intention to quit. Structural model 2 is an alternative model, where mentoring partially mediates the influence of four antecedents on the two outcomes. The model included direct paths from each of the four antecedents to each of the two outcomes (8 paths) in addition to all the paths of structural model 1. The model achieved a reasonable fit: χ2 = 340.684, df = 227, GFI = 0.893, AGFI = 0.834, CFI = 0.938, and RMSEA= 0.050. Collectivism was positively related to mentoring and unrelated to both affective commitment and intention to quit. Individualism was unrelated to mentoring, affective commitment, and intention to quit. Trust in supervisor was positively related to mentoring, unrelated to affective commitment, and negatively related to intention to quit. Mentoring was positively related to affective commitment and negatively to intention to quit. The number of years with the boss was unrelated to mentoring. Employee age was slightly negatively related to affective commitment and positively to intention to quit. Employee education was slightly related to affective commitment and negatively related to intention to quit. Employee sex was positively related to affective commitment but negatively related to intention to quit. Employee years of experience was slightly positively related to affective commitment and negatively related to intention to quit. Boss's age was unrelated to affective commitment, while positively related to intention
Fig. 3. Structural model 2: mentoring as a partial mediator.
1116
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Table 2 Results of analyses. Structural model Structural model Comments 1: full mediation 2: partial mediation Fit indices χ2 df GFI AGFI CFI RMSEA Paths: direct effects (H1a) Mentoring to affective commitment (H1b) Mentoring to intention to quit (H2) Trust in supervisor to mentoring (H3) Collectivism to mentoring (H4) Individualism to mentoring
344.684 233 0.891 0.836 0.940 0.049
340.920 227 0.893 0.834 0.938 0.050
4.071(0.000)
2.9080(.002)
Supported
− 4.154(0.000)
− 1.720(0.043)
Supported
4.485(0.000)
4.498(0.000)
Supported
1.775(0.038) n.s.
1.730(0.042) n.s.
Supported Not supported
n.s.
Supported
− 1.662(0.048) n.s.
Not supported Supported
n.s.
Supported
n.s.
Supported
n.s.
Supported
Model 1 better Model 1 better Model 1 better
Fully mediated by mentoring (H2a) Trust in supervisor to affective commit. (H2b) Trust in supervisor to intention to quit (H3a) Collectivism to affective commitment (H3b) Collectivism to intention to quit (H4a) Individualism to affective commitment (H4b) Individualism to intention to quit Model comparison AIC
580.684
588.920
BCC
617.731
627.850
CAIC
1087.886
1121.911
− 1.456(0.073)
− 1.4800(.069)
1.896(0.029) 1.350(0.089)
2.003(0.022) 1.288(0.099)
− 1.869(.031)
− 1.869(0.031)
1.807(0.036)
1.711(0.043)
− 3.073(0.001) 1.329(0.092)
− 2.855(.002) 1.427(.077)
− 3.150(0.001)
− 3.145(.001)
n.s.
n.s.
n.s. 1.820(0.035) n.s.
n.s. 1.839(0.033) n.s.
n.s. − 3.445(.000)
n.s. − 3.456(0.000)
n.s.
n.s.
Controls Employee age to affective commitment Employee age to intention to quit Employee education to affective commitment Employee education to intention to quit Employee sex to affective commitment Employee sex to intention to quit Employee years of experience to affective commitment Employee years of experience to intention to quit Employee years with the boss to mentoring Boss's age to affective commitment Boss's age to intention to quit Boss's education to affective commitment Boss's education to intention to quit Boss's sex to affective commitment Boss's sex to intention to quit
Structural Model 3's AIC (596.852), BCC (638.608) and CAIC (1468.855) are all higher than that of the two other models indicating that Structural Model 3 is not better than the other two models.
to quit. Boss's education was unrelated to both affective commitment and intention to quit. Boss's sex was negatively related to affective commitment and unrelated to intention to quit.
In order to compare model 1 with model 2, we looked at Akaike information criterion (AIC), Browne–Cudeck criterion (BCC), and Consistent AIC (CAIC) as suggested by Arbuckle and Wothke (1995– 1999). The model that has lowest values for AIC, BCC, and CAIC should be the model of choice. Our hypothesized model 1, fully mediated by mentoring, produced the lowest values for AIC (580.684), BCC (617.731) and CAIC (1087.886) compared to alternative model 2 (AIC = 588.920, BCC = 627.850, and CAIC = 1121.911). Moreover, the results for path coefficients demonstrated that mentoring fully mediated the influences of the antecedents. For the purpose of a further model comparison, a third structural model 3 was estimated. In this model, the four antecedents and mentoring were all treated as antecedents. The outcomes remained the same: affective commitment and intention to quit. The seven additional controls on outcomes were included as antecedents too. The AIC (596.852), BCC (638.608), and CAIC (1468.855) values of model 3 were higher than that of both model 1 and model 2. Therefore, our proposed fully mediated model 1 remained the better model compared to two other alternative models. Furthermore, we undertook a supplemental analysis to test if industry type could moderate the structural paths of the proposed model 1. The data was divided into two samples, one was of service firms and the other was of manufacturing firms. The proposed model 1 was simultaneously estimated with the two samples once with unconstraint structural paths and then again with constraining the corresponding structural paths to equal for the two samples. The ChiSquare difference (622.852[df = 395] − 617.211[df = 389] = 5.641 [df = 6]) was insignificant (5.641, df = 6 b12.592, the critical value). Therefore, the model and the structural paths are not different in service firms and manufacturing firms, i.e., industry type does not appear to be a moderator. 5. Discussion This study examined the role of mentoring as a mediator of collectivism, individualism, and trust in supervisor to affective commitment and intention to quit. The results were generally consistent with our predictions. In particular, both collectivist values and trust in supervisor positively impacted mentoring while individualism had no significant effect. Mentoring mediated the relationship between both collectivism values and trust in supervisors and both organizational commitment and intention to quit. These findings support the theory that personal values and the subordinates' relationship with the supervisor have an important role in creating and influencing employee reception of mentoring from their supervisor. In fact, collectivist values and trust in supervisor plays a pivotal role in the determination of mentoring received from a supervisor. Additionally trust in the supervisor was negatively related to intention to quit. These findings support previous theory regarding the effects of mentoring on organizational commitment and intention to quit (e.g. see Allen et al., 2004; Eby et al., 2008). More importantly, we have provided support for a comprehensive picture of the mentoring model. Nonetheless, there are several limitations to our study along with opportunities for future research. First, our study lacks randomization since the sample was limited to the industry contacts of a small number of working MBA students. Future research would benefit from investigating our proposed relationships in both a controlled laboratory context and in an elaborate field setting. Second, we collected all data from the subordinate. Although our study focuses on individual level constructs from the perspective of the subordinate, collecting some data from the supervisor would have allowed us to more forcefully validate our mentoring measure in particular. More specifically, there should be generally agreement between the supervisor who provides the mentoring and the subordinated that receives the mentoring because such consensus provides
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
greater confidence in the mentoring construct. Nevertheless, we believe while not ideal our approach is acceptable and indeed consistent with previous research that recognizes the protégé is the best source of mentoring data. Third, this study is limited to supervisory mentoring dyads when research acknowledges that many individuals have a network of individuals they receive mentorship from at any one time (Higgins and Kram, 2001). Future research might account for the multiple mentorship relationships to investigate whether our predictions are consistent across different types of mentoring dyads. Finally, all data was collected in a static fashion. Although we controlled for the number of years the employee had been under the supervision of their boss, it would be ideal to use a longitudinal design to investigate how our model predictions hold across the several mentorship developmental stages. Future research is still necessary for a complete framework to be determined. For example, Singelis et al. (1995) have distinguished between vertical and horizontal dimensions of collectivism, suggesting that the former describes the tendency to accept hierarchy within the in-group, while the latter dimension captures individuals' emphasis on equality within the in-group. Perhaps, employees who are stronger on vertical collectivism are more responsive towards supervisory mentoring than employees who are stronger in horizontal collectivism. It could be worthwhile to observe how individuals' responses to mentoring differ across the two dimensions. Moreover, it is possible that other personality attributes (e.g., locus of control, extroversion) and demographic attributes (e.g., gender difference, racial difference, age difference) impact the amount of mentoring received from a supervisor. In addition, it would be interesting to investigate organizational outcomes such as organizational citizenship behavior, actual job performance, and promotions. Such research would require data collection from both members of the dyad. Finally, we did not find individualism related to any of the dependent measures. Continued examination is needed to develop a better understanding of the relationship between individualism and organizational outcomes. 5.1. Managerial implications Research has suggested that mentoring is essential for employee mobility and even compensation. As such, mentoring is a critical aspect of organizational behavior. Social exchange theory provides a fruitful framework for managers to understand supervisor–subordinate dyadic relationships particularly mentorship development. For example, while informal mentoring relationships are determined by both the organization and the employee, it is the organization that can influence the formal nature of mentoring. By providing managers with an understanding of antecedents and consequences of successful mentoring they will be better equipped to foster trust to ensure that the mentoring benefits are realized. References Allen T, Eby L. Common bonds: an integrative view of mentoring relationships. In: Allen TD, Eby LT, editors. Handbook of mentoring: a multiple perspective approach. Malden, MA: Blackwell; 2007. p. 397–419. Allen T, Eby L, Poteet M, Lentz E, Lima L. Career benefits associated with mentoring for protégés: a meta-analysis. J Appl Psychol 2004;89:127–36. Anderson J, Gerbing W. Structural equation modeling in practice: a review and recommended two-step approach. Psychol Bull 1988;103(3):411–23. Arbuckle L, Wothke W. Amos 4.0 user's guide. Chicago, IL: SmallWaters Corporation; 1995–1999. Armstrong J, Overton D. Establishing nonresponse bias in mail surveys. J Mark Res 1977;14:396–402. Bagozzi R, Yi Y. On the evaluation of structural equation models. Acad Market Sci 1988;16(1):74–94. Bartlett K. The relationship between training and organizational commitment: a study in the health care field. Hum Resour Dev Q 2001;12(4):335–52. Blau P. Exchange and power in social life. New York: Wiley; 1964. Clugston M, Howell J, Dorfman P. Does cultural socialization predict multiple bases of foci of commitment? J Manage 2001;5:5-30.
1117
Deluga R. Supervisor trust building, leader–member exchange and organizational citizenship behavior. J Occup Organ Psychol 1994;67(4):315–26. Earley P, Gibson C. Taking stock in our progress on individualism–collectivism: 100 years of solidarity and community. J Manage 1998;24:265–304. Eby L. Alternative forms of mentoring in changing organizational environments: a conceptual extension of the mentoring literature. J Vocat Behav 1997;51(1):125–44. Eby L, Rhodes J, Allen T. Definition and evolution of mentoring. In: Allen TD, Eby LT, editors. Handbook of mentoring: a multiple perspective approach. Malden, MA: Blackwell; 2007. p. 7-20. Eby L, Allen T, Evans S, Ng T, DuBois D. Does mentoring matter? A multidisciplinary meta-analysis comparing mentored and non-mentored individuals. J Vocat Behav 2008;72(2):54-267. Eisenberger R, Huntington R, Hutchison S, Sowa D. Perceived organizational support. J Appl Psychol 1986;71:500–7. Elangovan A, Shapiro D. Betrayal of trust in organizations. Acad Manage Rev 1998;23(3):547–68. Ensher E, Thomas C, Murphy S. Comparison of traditional, step-ahead, and peer mentoring on protégés' support, satisfaction, and perceptions of career success: a social exchange perspective. J Bus Psychol 2001;15(3):419–38. Flynn F. Identity orientations and forms of social exchange in organizations. Acad Manage Rev 2005;30(4):737–50. Fornell C, Larcker D. Evaluating structural equation models with unobservable variables and measurement error. J Mark Res 1981;18:39–50. Gould-Williams J, Davies F. Using social exchange theory to predict the effects of HRM practice on employee outcomes. Publ Manage Rev 2005;7(1):1-24. Guzzo R, Noonan K, Elron E. Expatriate managers and psychological contract. J Appl Psychol 1994;79:617–26. Haas D, Deseran F. Trust and symbolic exchange. Soc Psychol Q 1981;44(March):3-13. Higgins M, Kram K. Reconceptualizing mentoring at work. A developmental network perspective. Acad Manage Rev 2001;26(2):264–88. Hofstede G. Culture's consequences: international differences in work related values. Newbury Park, CA: Sage Publications; 1980. Hofstede G. Culture and organizations: software of the mind. London, England: McGraw-Hill; 1991. Homans G. Social behavior: its elementary forms. Orlando: Harcourt Brace; 1961. Koberg C, Boss R, Goodman E. Factors and outcomes associated with mentoring among health-care professionals. J Vocat Behav 1998;53(1):58–72. Kram K. Phases of the mentor relationship. Acad Manage Rev 1983;26:608–25. Kram K. Mentoring at work: developmental relationships in organizational life. 1st ed. Glenview, IL: Scott, Foresman and Company; 1985. Larzelere R, Huston T. The dyadic trust scale: toward understanding interpersonal trust in close relationships. J Marriage Fam 1980;42(3):595. Lee C, Bruvold N. Creating value for employees: investment in employee development. Int J Hum Resour Manag 2003;14(6):981-1000. Liden R, Graen G. Generalizability of the vertical dyad linkage model of leadership. Acad Manage J 1980;23(3):451–65. McAllister D. Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Acad Manage J 1995;38(1):24–59. McManus S, Russell E. New directions for mentoring research: an examination of related constructs. J Vocat Behav 1997;51:145–61. Meyer J, Allen N. A three-component conceptualization of organizational commitment. Hum Resour Manage Rev 1991;1:61–89. Meyer J, Allen N, Smith C. Commitment to organizations and occupations: extension and test of a three-component conceptualization. J Appl Psychol 1993;78(4):538–51. Mowday R, Porter L, Steers R. Organizational linkages: the psychology of commitment, absenteeism, and turnover. San Diego, CA: Academic Press; 1982. Noe R. An investigation of the determinants of successful assigned mentoring relationships. Pers Psychol 1988;41:457–79. Noe R, Greenberger D, Wang S. Mentoring: what we know and where we might go. Res Pers Hum Resour Manage 2002;21:129–73. Noordin F, Williams T, Zimmer C. Career commitment in collectivist and individualist cultures. Int J Hum Resour Manag 2002;13(1):35–54. Nooteboom B. Trust, opportunism and governance: a process and control model. Organ Stud 1996;17(6):985-1010. Olian J, Carroll S, Giannantonio S. Mentor reactions to protégés: an experiment with managers. J Vocat Behav 1993;43:266–78. Payne S, Huffman A. A longitudinal examination of the influence of mentoring on organizational commitment and turnover. Acad Manage J 2005;48:158–68. Podsakoff P, Organ D. Self-reports in organizational research: problems and prospects. J Manage 1986;12(Winter):531–43. Ragins B. Diversified mentoring relationships in organizations: a power perspective. Acad Manage Rev 1997;22:482–521. Ragins B, Cotton J. Mentor functions and outcomes: a comparison of men and women in formal and informal mentoring relationships. J Appl Psychol 1999;84(4):529–50. Ragins B, McFarlin D. Perceptions of mentor roles in cross-gender mentoring relationships. J Vocat Behav 1990;37:321–39. Robinson S. Trust and breach of the psychological contract. Adm Sci Q 1996;41: 574–99. Robinson S, Kraatz M, Rousseau D. Changing obligations and the psychological contract: a longitudinal study. Acad Manage J 1994;37(1):137–52. Rousseau D, Greller M. Human resource practices: administrative contract makers. Hum Resour Manage 1994;33(3):385–401. Rousseau D, Sitkin S, Burt R, Camerer C. Not so different after all: a cross-discipline view of trust. Acad Manage Rev 1998;23(3):393–404. Scandura T. Mentorship and career mobility: an empirical investigation. J Organ Behav 1992;13:169–74.
1118
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Scandura T, Pellegrini E. Workplace mentoring: theoretical approaches and methodological issues. In: Allen TD, Eby LT, editors. Handbook of mentoring: a multiple perspective approach. Malden, MA: Blackwell; 2007. p. 71–92. Seabright M, Levinthal D, Fichman M. Role of individual attachments in the dissolution of interorganizational relationships. Acad Manage J 1992;35:122–60. Shamir B. Calculations, values, and identities: the sources of collectivistic work motivation. Human Relat 1990;43(4):313–32. Singelis T, Triandis H, Bhawuk D, Gelfand M. Horizontal and vertical dimensions of individualism–collectivism: a theoretical and measurement refinement. Cross-Cult Res 1995;29:240–75. Tepper B. Upward maintenance tactics in supervisory mentoring and nonmentoring relationships. Acad Manage J 1995;38:1191–205. Tepper B, Taylor E. Relationships among supervisors' and subordinates' procedural justice perceptions and organizational citizenship behaviors. Acad Manage J 2003;46:97-105. Triandis H. Cross-cultural studies of individualism and collectivism. In: Berman J, editor. Nebraska symposium on motivation. Lincoln: University of Nebraska Press; 1989. p. 41-133.
Triandis H. Individualism and collectivism. Boulder, CO: Westview; 1995. Wayne S, Shore L, Liden R. Perceived organizational support and leader–member exchange: a social exchange perspective. Acad Manage J 1997;40(1):82-111. Whitener E. Do high commitment human resource practices affect employee commitment? A cross level analysis using hierarchical linear modeling. J Manage 2001;27:515–35. Williamson O. Transaction cost economics: the governance of contractual relations. J Law Econ 1979;22:233–61. Young A, Perrewe P. The exchange relationship between mentors and protégés: the development of a framework. Hum Resour Manage Rev 2000;10(2):177–209. Young A, Perrewe P. The role of expectations in the mentoring exchange: an analysis of mentor and protégé expectations in relation to perceived support. J Manag Issue 2004;XVI(1):103–26.
Contents lists available at ScienceDirect
Journal of Business Research
Mentoring in supervisor–subordinate dyads: Antecedents, consequences, and test of a mediation model of mentorship ☆ Orlando C. Richard a, Kiran M. Ismail b,⁎, Shahid N. Bhuian c, Edward C. Taylor d a
The University of Texas at Dallas, United States St. John's University, United States Louisiana Tech University, United States d Piedmont College, United States b c
a r t i c l e
i n f o
Article history: Received 17 September 2007 Accepted 6 September 2008 Keywords: Mentoring Social exchange Trust Organizational commitment Dyadic relationships
a b s t r a c t We develop a framework to look at mentoring as a consequence of employees' values and beliefs, as well as to explore its role in determining the employees' attitudes towards their organizations. Based on social exchange theory, we hypothesize that employees' levels of individualism, collectivism, and trust in supervisor influence the level of supervisory mentoring received. Moreover, mentoring influences employees' affective commitment and intention to quit, as well as mediates the relationships between the proposed antecedents and outcomes. Using structural equation modeling to examine these relationships, we find that within supervisor–subordinate dyads, subordinates report more mentoring when they have collectivist personal values and trust their supervisor. Additionally, more mentoring is positively related to subordinates' affective commitment towards the organizations and negatively related to their intention to quit. We also find that mentoring mediate the relationship between both collectivism values and trust in supervisors and both organizational commitment and intention to quit. © 2008 Elsevier Inc. All rights reserved.
1. Introduction Mentoring relationships are recognized by both academicians and practitioners as a valuable critical resource for employees in organizations. Organizational theorists and management scholars have identified mentoring as an exchange relationship whereby both mentor and protégé gain several benefits from each other (Ragins, 1997; Young and Perrewe, 2000; Young and Perrewe, 2004). For example, compared with non-mentored individuals, mentored employees demonstrate higher levels of objective and subjective positive outcomes such as career development, job satisfaction, socialization, organizational commitment, and career advancements (Allen et al., 2004; Eby et al., 2008). Mentors, in return for the time and effort spent in providing support to the protégés, gain positive outcomes such as career rejuvenation, recognition, personal satisfaction, organizational reputation, and increase in knowledge and power (Noe et al., 2002). While scholars have used various theoretical perspectives to explain mentoring relationships (e.g. leadership, justice, power, exchange, motivation; see Noe et al., 2002, for a detailed review), ☆ A previous version of the paper was presented at the 2007 Academy of Management Annual Meeting, Philadelphia, PA. We thank associate editor Dr. Lei-Yu Wu and three anonymous reviewers of Journal of Business Research for valuable feedback on earlier versions of the paper. ⁎ Corresponding author. St. John's University, Peter J. Tobin College of Business, 8000 Utopia Parkway, Queens, NY 11439, United States. Tel.: +1 718 990 7423. E-mail address: [email protected] (K.M. Ismail). 0148-2963/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2008.09.007
the exchange process between a mentor and protégé lies at the heart of mentoring. Thus, management scholars have suggested that social exchange theory (Homans, 1961; Blau, 1964) is an appropriate theoretical lens for describing and studying the processes involved in mentoring relationships (Olian, et al., 1993; Ensher et al., 2001; Tepper and Taylor, 2003). Based on the social exchange perspective, we define mentoring as a reciprocal exchange relationship between a mentor and a protégé (Young and Perrewe, 2000, 2004). While workplace mentoring has been traditionally defined as a hierarchical relationship between a senior and influential organizational member (mentor) and a junior and less experienced organizational member (Kram, 1983), scholars have identified various forms of mentoring relationships, such as lateral or peer mentoring, supervisory mentoring, team mentoring, and mentoring by an external sponsor (see Eby, 1997; Allen and Eby, 2007; Eby, Rhodes and Allen, 2007; Scandura and Pellegreni, 2007). The focus of the current study is on supervisory mentoring. Our basis for drawing this boundary is the notion that employees tend to interact most frequently with their supervisors than with other agents of the organization, and therefore, supervisors are in the best position to serve as organizational representatives (Tepper and Taylor, 2003). Moreover, extant literature also suggests that supervisors are in the most natural position to, and have the responsibility to provide career and psychosocial support to the subordinates (Eby, 1997: pp. 135/6), and employees are likely to obtain mentoring from their immediate supervisors (Ragins and McFarlin, 1990; Tepper, 1995; Tepper and Taylor, 2003).
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
According to social exchange theory, individuals' beliefs about the support they receive from their employing organizations play an important role in effecting their behaviors and attitudes towards their organizations. While individuals' personal values, as well as the quality of the inter-attitudinal relationships that they experience with their supervisors play an important role in effecting these beliefs, research has not systematically explored the roles of these variables in mentoring relationships. Thus, the main purpose of this study is to take insights from social exchange theory in order to study: (a) employees' personal values (i.e. individualism–collectivism) and employees' relationships with the supervisors (i.e. trust in supervisor) as the antecedents of supervisory mentoring antecedents, (b) employees' commitment to the organization and intention to quit as the outcomes of supervisory mentoring, and (c) the mediating role of supervisory mentoring in the relationships between its antecedents and outcomes. The theoretical model that is being proposed in the current study is shown in Fig. 1. Therefore, this study makes the following contributions. First, based on the notions that mentoring is an exchange relationship between a mentor and a protégé, and that employees regard their supervisors as the representatives of their organizations, we provide a theoretical perspective to look at mentoring as a consequence of employees' personal attitudes, as well as to explore its role in determining the employees' attitudes towards their organizations. Second, we go beyond simple linear regression modeling to investigate more complex relationships as posited by extant theory. Specifically, we use structural equation modeling to examine the role of mentoring as a mediating variable in the relationship between individuals' values and relationships, and their performance outcomes. 2. Theoretical background and hypotheses 2.1. Social exchange theory The basic tenet of social exchange theory is that human interactions are based upon exchange of social and material resources and norms of reciprocity (Homans, 1961; Blau, 1964). The theory posits that people enter into relationships in which benefits and rewards (tangible or intangible) are greater than the costs. Unlike economic exchanges that are characterized by legally-sanctioned contractual arrangements, social exchanges are ‘voluntary actions’ (Blau, 1964) that are embedded in trust that the recipient of a benefit will eventually reciprocate with an equitable, but not necessarily similar, favor (Haas and Deseran, 1981; Gould-Williams and Davies, 2005). While social exchange theory originally focused on explaining relationship between individuals, management scholars have established that the theory plays an important role in explaining the relationship between the organization and its employees (Eisenberger et al., 1986; Whitener, 2001). The ‘voluntary’ exchange relationship between an employee and the employing organization is governed by what management theorists have referred to as ‘psychological contracts’,
1111
defined as an individual's beliefs about the terms of his or her reciprocal relationships with the employing organization (Robinson et al., 1994; Robinson, 1996). As we mentioned in the previous section, such beliefs are effected by individual's personal values, as well as the quality of relationship they experience with their co-workers and supervisors. Based upon the insights offered by social exchange theory, each component of the theoretical model that has been proposed earlier is discussed below. 2.2. Mentoring and organizational commitment Organizational commitment is defined as the level of attachment and identification that an individual feels towards the organization in which he or she is employed (Mowday et al., 1982; Bartlett, 2001). A popular conceptualization of the construct has been offered by Meyer and Allen (1991), who have identified commitment as a multidimensional construct, consisting of three components: affective commitment, which refers to individuals' desire to be committed to the organization due to the emotional attachment felt towards the organization as a result of personal liking or preference, continuance commitment, which refers to individuals' need to remain in the organization due to the costs associated with leaving, and normative commitment, which refers to individuals' moral obligation to be committed to the organization due to their personal values (see also Meyer et al., 1993). Using the social exchange perspective, management scholars have argued that the levels of employees' commitment to their organizations depend upon their beliefs regarding how committed their employing organizations are to them, and the support they receive from their employers (Wayne et al., 1997; Whitener, 2001; Gould-Williams and Davies, 2005). Research in the area has suggested that since supervisors are perceived by employees as agents of the organization, the behaviors of managers and supervisors play critical roles in effecting psychological contracts and signaling the level and nature of the organizations' commitment to the employees, and hence, shape the employees' behaviors and attitudes, or more specifically, their level of commitment, towards their organizations (e.g. Guzzo et al., 1994; Rousseau and Greller,1994; Wayne et al.,1997; Bartlett, 2001; Whitener, 2001; Lee and Bruvold, 2003; Gould-Williams and Davies, 2005). Past research has suggested that mentoring is an important determinant of an employee's affective and continuance commitment to the organization (Koberg et al, 1998; Allen et al., 2004; Payne and Huffman, 2005). This argument is based on the notion that a successful exchange relationships increases employees' commitment to their organizations (Flynn, 2005). We seek to confirm this proposition by exploring the relationship between mentoring and two dependent variables: affective commitment, and intention to quit. Mentors offer different types of social and psychological support to their protégés (Scandura, 1992). The various types of support include social support, coaching and role modeling support, career support, and emotional support (Kram, 1985; Ragins and Cotton, 1999; Ensher et al., 2001). From a social exchange perspective, it can be argued that
Fig. 1. Conceptual model.
1112
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
protégés offer increased attachment, identification, and involvement with their organizations in exchange for the support they receive from their mentors. We had mentioned earlier that from the social exchange perspective, employees' commitment to the organization depends upon their ‘perceptions of how committed their organizations are towards them' (Gould-Williams and Davies, 2005). When subordinates receive higher amount of mentoring from their supervisors, they perceive it as a form of fair treatment from their organizations and reciprocate by increasing their level of citizenship behavior towards their organizations (Tepper and Taylor, 2003). Moreover, since social exchange theory also suggests that people tend to enter and remain into relationships in which benefits outweigh the costs, it can also be argued that, as an intangible benefit, mentoring results in a favorable cost to benefit ratio for the recipients and is therefore negatively related to their intention to quit the organization. In summary, we propose that the psychosocial support provided by supervisors in the form of mentoring is an important signal of organizations' commitment towards the employees, and thereby, increases the levels of employees' commitment towards their organizations. H1a. Supervisory mentoring received by an employee is positively related to his or her level of affective commitment to the organization. H1b. Supervisory mentoring received by an employee is negatively related to his or her intention to quit the organization. 2.3. Trust in supervisor and mentoring Defined in the literature as positive expectation regarding favorableness of another's actions, trust is a critical aspect of social exchange relationships (Blau, 1964; Haas and Dseran, 1981; Robinson, 1996; GouldWilliams and Davies, 2005). Trust reduces opportunism in relationships (Williamson, 1979), thus enabling interacting parties to work together effectively and co-operate (McAllister, 1995; Rousseau et al., 1998). Based on the notion that trust is a determining factor of effective social exchange relationships between individuals who work together (Seabright et al., 1992; McAllister, 1995), we propose that the level of trust between an employee and his or her supervisor is an important determinant of the amount of mentoring received by the employee. From a social exchange perspective, it can be argued that employees who exhibit higher levels of trust in their supervisors, in turn, are able to derive greater benefits out of their exchange relationship with their supervisors. As suggested by literature, high-quality leader–member exchange relationships, that are characterized by trust and loyalty, hold benefits for both supervisors, in the form of motivated, committed, competent, and conscientious subordinates, and employees, in the form of higher appraisals, promotions, and mentorship from the supervisor (Liden and Graen, 1980; Deluga, 1994; McManus and Russell, 1997). Thus, it is likely that supervisors are more willing to invest time and effort in mentoring employees who exhibit greater levels of trust in them, instead of spending time and effort on coaching employees with whom they do not have expectations of developing a high-quality and effective exchange relationships. H2. An employee's trust in his or her supervisor is positively related to the amount of mentoring received from the supervisor. In this study, we are also interested in exploring the mediating role of mentoring in the relationship between its antecedents and outcomes. Based on the social exchange theory, we believe that employees' levels of commitment towards their organization are not directly influenced by their collectivistic values or trust in supervisor. Instead, their levels of commitment are influenced by whether or not they receive benefit from the organization in exchange or return for their collectivistic values or trust in supervisor.
In intra-organizational relationships, psychological contract breach and betrayal of trust are not uncommon phenomena (Robinson, 1996; Elangovan and Shapiro, 1998). Nooteboom (1996) suggests that “trust is related to propensity, not a certainty: it may not be resistant to golden opportunities” (1996: p. 994). Thus, from the social exchange perspective, it can be argued that employees who exhibit high levels of trust in their supervisors expect to gain benefits from their high-quality exchange relationships, and they tend to demonstrate greater levels of organizational commitment only when they are able to derive benefits out of the relationships. Therefore, we also propose that employees who have higher levels of trust in their supervisors will receive higher amount of mentoring, and thus, will exhibit higher levels of organizational commitment. H2a. Mentoring fully mediates the positive relationship between an employee's trust in supervisor and the level of affective commitment to the organization. H2b. Mentoring fully mediates the negative relationship between an employee's trust in supervisor and his or her intention to quit the organization. 2.4. Individualism–collectivism and mentoring While individualism–collectivism constructs have been used at societal level, to describe cultural difference, it has been argued that people from individualistic and collectivist cultures hold idiocentric values, and that the constructs are situation-specific. The constructs are often used by management scholars to measure attitudes and values at individual level of analysis, and capture individuals' preference for ether companionship (collectivism) or personal identity (individualism) (Triandis, 1995; Earley and Gibson, 1998; Noordin et al., 2002). In the context of an organizational setting, individualism and collectivism refer to the patterns in which employees relate to their co-workers, teams and workgroups, supervisors, and organizations. Collectivists define self by in-group memberships (Triandis, 1989; Earley and Gibson, 1998) and seek close and long-term relationships. In an organizational setting, they feel morally obliged to serve their company, and value provisions offered by the organization (e.g. training, physical conditions). Collectivists are particularly concerned about the results of their actions on their organizations (Triandis, 1989), and keeping/losing face in front of others, and this may be a major incentive for them to fulfill the duties expected of them (Hofstede, 1980, 1991). They have strong need for harmony in the work environment. Individualists, on the other hand, have calculative involvement with organization and have strong need for freedom and preference for low-context relationships based on emotional detachment (Earley and Gibson, 1998; Hofstede, 1980, 1991; Triandis, 1995). The individualist preferred work environment is competitive. It has been suggested that collectivists are likely to be provided protection and security in return for their loyalty to their in-groups (Noordin et al., 2002). From a social exchange perspective, we propose that employees with high levels of collectivism tend to build close and high-quality relationships with their supervisors, who, in return for their loyalty, and in respond to their collectivistic employees' need for seeking close, interpersonal relationships and provisions from the organization, offer them higher amount of mentoring, than they are likely to provide to the employees having higher levels of individualistic orientation. H3. An employee's collectivism level is positively related to the amount of supervisory mentoring. H4. An employee's individualism level is negatively related to the amount of supervisory mentoring. According to Shamir (1990), individuals are motivated to contribute to collective work efforts either due to their calculative considerations regarding the nature of rewards and sanctions they
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
expect to receive, or because they are morally obliged to do so, or in order to take advantage of opportunities to express a cherished and salient identity. From a social exchange perspective, it can be argued that individuals are likely to feel committed to their organizations to the extent that the organizations offer them provisions and opportunities that clarify and affirm their values, self-concepts and identities. We propose that mentoring is an example of one such provision or opportunity. Thus, based on the arguments discussed earlier regarding the nature of relationship between mentoring and organizational commitment (Hypothesis 1a and 1b), and the insights offered in the preceding discussion regarding the nature of relationship between individualism–collectivism and mentoring, we propose: H3a. Mentoring fully mediates the positive relationship between an employee's collectivism level and the level of affective commitment to the organization. H3b. Mentoring fully mediates the negative relationship between an employee's collectivism level and his or her intention to quit the organization. H4a. Mentoring fully mediates the negative relationship between an employee's individualism level and the level of affective commitment to the organization. H4b. Mentoring fully mediates the positive relationship between an employee's individualism level and his or her intention to quit the organization. 3. Methodology 3.1. Data collection and setting We tested the hypotheses using data collected from employees of numerous organizations in the metro Atlanta area. We selected a research design that would span as many organizations and supervisor–subordinate dyads as possible thereby maximizing the variance on study variables. We elected to distribute surveys through two MBA classes of about twenty students each with instructions that they find respondents who do not share the same boss. In this way, we avoided inadvertent nesting when the survey design was executed. Multiple responses by subordinates of the same boss would require that we account for variance both within groups and between groups. Our survey distribution method produced a data set that may have as many as twenty responses from a single business organization but no more than one response per boss within those businesses. Twohundred fifty surveys were distributed, 220 were returned and 10 were discarded due to incomplete data. Twenty variables were randomly selected as a test for monotonic responses—another form of non-response. These survey items were then scanned for the same response, if all twenty were equal to the same value then the survey was manually scanned for monotonic responses. Ten surveys were discarded as a result of this procedure. This left a data set with an N of 200. Of the respondents 67% were over 25 years old, 90% of the respondents were Caucasian; 54% were male, and the average yearly earnings were $35,500. 3.2. Measures 3.2.1. Dependent measures Affective commitment was measured using five items adapted from Meyer et al. (1993). (Sample item: “My company has a great deal of personal meaning for me”). One item on a five-point Likert type scale was used to capture intention to quit. It stated “If a similar job, offering the same pay and work, came available to me in another company, I would be likely to transfer.”
1113
3.2.2. Independent measures Collectivism and individualism were measured using two different scales as opposed to a uni-dimensional measure and was designed for the individual level of analysis distinct from national culture measures (Clugston et al., 2001). A five items Likert type scale was designed to measure individualism. (Sample item: “If a group is slowing me down, it is better to leave it and work alone”). Collectivism was measured using a three items Likert type scale. (Sample item: “A person should accept the group's decision even when personally he or she has a different opinion”). Trust was operationalized as a belief in the benevolence of the other party, the matter of relevance in interpersonal relationship risk assessment, by adapting Larzelere and Houston's (1980) 8-item unidimensional dyadic trust scale in close interpersonal relationships, e.g., dating couples, to the supervisor–worker relationship. Trust was measured by asking subordinates to describe how much trust they had in their supervisor on a 5-point Likert type scale with responses ranging from 1—strongly disagree to 5—strongly agree. (Sample item: “My supervisor is truly sincere when making promises”). We used Kram's (1985) conceptual framework for mentoring and Noe (1988) as the basis for our survey items measuring mentoring activity. Noe's (1988) scale contained eight psychosocial related items and seven career help items for a total of 15 items. We adapted 10 survey items from these 15 items for our measure of mentoring. Mentoring activity was measured by asking subordinates to provide how much of each type of mentoring activity they had received from their supervisor on a 5-point Likert type scale with responses ranging from 1—strongly disagree to 5—strongly agree. (Sample item: “My supervisor exposes me to upper management”). 3.2.3. Control measures Length of the mentoring relationship was represented by the number of years with the boss. Age was categorized into five year increments ranging from normal age of entry into the workplace to retirement. Education categories were designed to capture high school dropouts, high school graduates, college attendees, college graduates, master's level graduates and doctoral degrees. Sex was dichotomized as male and female. The number of years with the boss served as a control on mentoring, while age, education, sex and years of experience were controls on two outcomes—affective commitment and intention to quit. Further, in order to control for the degree of similarity between the mentor and protégé, age, education and sex of both employees and bosses were included as controls. 3.3. Confirmatory factor analysis (CFA), reliability, validity, non-response bias, and common method bias We used CFA to assess the scale properties. The CFA Model included five multi-item constructs: collectivism (3 items), individualism (5 items), trust in supervisor (8 items), mentoring (10 items) and affective commitment (5 items). The initial run failed to produce a good fit of the model. Based on poor loadings, modification indices, and high residuals, we detected eighteen bad items: 2 of individualism, 4 of trust in supervisor, 6 of mentoring, and 2 of affective commitment. The deletion of these items was based on content considerations to minimize the reduction of the meaning of the constructs. After deleting these items one by one, the model achieved a good fit: χ2 = 146.884, df =109, GFI =0.924, AGFI =0.893, CFI =0.970, RMSEA =0.042, and PCLOSE=0.779. The t-statistics pertaining to the items ranged from 2.467 to 13.940, i.e., all were significant. We assessed scale reliabilities by estimating average variance extracted (AVEs) and composite reliabilities (Bagozzi and Yi 1988) and found all scales to be reliable. The AVEs for the scales of collectivism, individualism, trust in supervisor, mentoring and affective commitment were 0.987, 0.879, 0.922, 0.851, and 0.882 respectively. The composite reliabilities were 0.99, 0.96, 0.97, 0.96, and 0.96
1114
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Table 1 Correlations, Means, Std. Dev., AVEs, and composite reliabilities.
1. Mentoring 2. Collectivism 3. Individualism 4. Trust in Supervisor 5. Affective commitment 6. Intention to quit 7. Age 8. Education 9. Sex 10. Years of experience Means Standard Deviations AVEs Composite Reliabilities
1
2
3
4
5
6
7
8
9
10
1.0 .157⁎ − .182⁎ .649⁎⁎ .555⁎⁎ − .375⁎⁎ − .127 .131 .100 − .081 3.385 .765 .987 .99
1.0 − .018 .133 .136 − .134 .009 .031 .081 .025 3.272 .663 .879 .96
1.0 − .194⁎⁎ − .122 .106 − .109 − .049 .024 − .137 2.700 .809 .922 .97
1.0 .127 − .474⁎⁎ .023 .059 .186⁎⁎ .016 3.868 .851 .851 .96
1.0 .123 .083 .098 .145⁎ .132 3.533 .969 .882 .96
1.0 − .056 − .104 − .209⁎⁎ − .118 2.290 1.12 – –
1.0 − .103 .051 .880⁎⁎ 3.46 2.50 – –
1.0 .041 − .129 3.33 .90 – –
1.0 − .067 1.46 .50 – –
1.0 14.08 11.29 – –
⁎ = correlation is significant at the 0.05 level (2-tailed). ⁎⁎ = correlation is significant at the 0.01 level (2-tailed).
respectively. If the values of AVEs are greater than 0.50, and the values of composite reliabilities are greater than 0.60, they are considered desirable (Fornell and Larcker 1981). The convergent validity of all scales was supported. The t-values pertaining to all items of all scales were each greater than 2.0 (at p b 0.000). When each indicator's estimated path coefficient on its underlying construct factor is significant (greater than twice its standard error) in the measurement model, the scales of the constructs in the model achieve convergent validity (Anderson and Gerbing 1988). Further, the scales achieved discriminant validity
because no confidence intervals of the correlations for the constructs (Ф values) included 1.0 (p b 0.05) (Anderson and Gerbing 1988). We also tested for non-response bias by comparing the responses of the early respondents with the late respondents to all 18 items (Armstrong and Overton 1977). All multivariate tests results were insignificant: Box's Test of Equality of Covariance Matrices (Sig. = 0.475), Pillai's Trace (Sig. = 0.213), Wilks' Lambda (Sig. = 0.213), Hotelling's Trace (Sig. = 0.213) and Roy's Larget Root (Sig.—0.213) indicating that responses between early respondents and late respondents were not significantly different.
Fig. 2. Structural model 1: full mediation by mentoring.
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Further, following Podsakoff and Organ (1986), Harman's one factor test was conducted where all scale items were entered into a factor analysis. Not a single factor emerged nor was there a general factor that accounted for the majority of the covariance in these variables (the first factor explained 24.76% variance, the second factor 14.56%, the third factor 11.11% and the fourth factor 8.75%). Therefore, common method bias is not considered a severe threat in this study. 4. Results Table 1 shows the means, standards deviations, and correlations among the variables of interest. In order to test the hypotheses, we ran two structural models, as shown in Figs. 2 and 3. Additionally, we ran a third structural model 3 for a further model comparison. Structural model 1 is our hypothesized model, where mentoring fully mediates the influence of the three antecedents and one control variable, the number of years with the boss, on two outcomes. The model also includes seven control variables: age of employee, age of boss, education of employee, education of boss, sex of employee, sex of boss, and employee years of experience. The model demonstrated a fair fit: χ2 = 344.684, df = 233, GFI = 0.891, AGFI = 0.836, CFI = 0.940, and RMSEA = 0.049. Collectivism was positively related to mentoring (see Table 2 for the estimates). Individualism was not related to mentoring. Trust in supervisor was positively related to mentoring. Mentoring was positively related to affective commitment and negatively to the intention to quit. The number of years with the boss was unrelated to mentoring. Employee age was slightly related to affective commitment and positively related to intention to quit. Employee education was slightly related to affective commitment and
1115
was negatively related to intention to quit. Employee sex was positively related to affective commitment and negatively to intention to quit. Employee years of experience were slightly positively related to affective commitment and negatively to intention to quit. Boss's age was unrelated to affective commitment, while positively related to intention to quit. Boss's education was unrelated to both affective commitment and intention to quit. Boss's sex was negatively related to affective commitment and unrelated to intention to quit. Structural model 2 is an alternative model, where mentoring partially mediates the influence of four antecedents on the two outcomes. The model included direct paths from each of the four antecedents to each of the two outcomes (8 paths) in addition to all the paths of structural model 1. The model achieved a reasonable fit: χ2 = 340.684, df = 227, GFI = 0.893, AGFI = 0.834, CFI = 0.938, and RMSEA= 0.050. Collectivism was positively related to mentoring and unrelated to both affective commitment and intention to quit. Individualism was unrelated to mentoring, affective commitment, and intention to quit. Trust in supervisor was positively related to mentoring, unrelated to affective commitment, and negatively related to intention to quit. Mentoring was positively related to affective commitment and negatively to intention to quit. The number of years with the boss was unrelated to mentoring. Employee age was slightly negatively related to affective commitment and positively to intention to quit. Employee education was slightly related to affective commitment and negatively related to intention to quit. Employee sex was positively related to affective commitment but negatively related to intention to quit. Employee years of experience was slightly positively related to affective commitment and negatively related to intention to quit. Boss's age was unrelated to affective commitment, while positively related to intention
Fig. 3. Structural model 2: mentoring as a partial mediator.
1116
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Table 2 Results of analyses. Structural model Structural model Comments 1: full mediation 2: partial mediation Fit indices χ2 df GFI AGFI CFI RMSEA Paths: direct effects (H1a) Mentoring to affective commitment (H1b) Mentoring to intention to quit (H2) Trust in supervisor to mentoring (H3) Collectivism to mentoring (H4) Individualism to mentoring
344.684 233 0.891 0.836 0.940 0.049
340.920 227 0.893 0.834 0.938 0.050
4.071(0.000)
2.9080(.002)
Supported
− 4.154(0.000)
− 1.720(0.043)
Supported
4.485(0.000)
4.498(0.000)
Supported
1.775(0.038) n.s.
1.730(0.042) n.s.
Supported Not supported
n.s.
Supported
− 1.662(0.048) n.s.
Not supported Supported
n.s.
Supported
n.s.
Supported
n.s.
Supported
Model 1 better Model 1 better Model 1 better
Fully mediated by mentoring (H2a) Trust in supervisor to affective commit. (H2b) Trust in supervisor to intention to quit (H3a) Collectivism to affective commitment (H3b) Collectivism to intention to quit (H4a) Individualism to affective commitment (H4b) Individualism to intention to quit Model comparison AIC
580.684
588.920
BCC
617.731
627.850
CAIC
1087.886
1121.911
− 1.456(0.073)
− 1.4800(.069)
1.896(0.029) 1.350(0.089)
2.003(0.022) 1.288(0.099)
− 1.869(.031)
− 1.869(0.031)
1.807(0.036)
1.711(0.043)
− 3.073(0.001) 1.329(0.092)
− 2.855(.002) 1.427(.077)
− 3.150(0.001)
− 3.145(.001)
n.s.
n.s.
n.s. 1.820(0.035) n.s.
n.s. 1.839(0.033) n.s.
n.s. − 3.445(.000)
n.s. − 3.456(0.000)
n.s.
n.s.
Controls Employee age to affective commitment Employee age to intention to quit Employee education to affective commitment Employee education to intention to quit Employee sex to affective commitment Employee sex to intention to quit Employee years of experience to affective commitment Employee years of experience to intention to quit Employee years with the boss to mentoring Boss's age to affective commitment Boss's age to intention to quit Boss's education to affective commitment Boss's education to intention to quit Boss's sex to affective commitment Boss's sex to intention to quit
Structural Model 3's AIC (596.852), BCC (638.608) and CAIC (1468.855) are all higher than that of the two other models indicating that Structural Model 3 is not better than the other two models.
to quit. Boss's education was unrelated to both affective commitment and intention to quit. Boss's sex was negatively related to affective commitment and unrelated to intention to quit.
In order to compare model 1 with model 2, we looked at Akaike information criterion (AIC), Browne–Cudeck criterion (BCC), and Consistent AIC (CAIC) as suggested by Arbuckle and Wothke (1995– 1999). The model that has lowest values for AIC, BCC, and CAIC should be the model of choice. Our hypothesized model 1, fully mediated by mentoring, produced the lowest values for AIC (580.684), BCC (617.731) and CAIC (1087.886) compared to alternative model 2 (AIC = 588.920, BCC = 627.850, and CAIC = 1121.911). Moreover, the results for path coefficients demonstrated that mentoring fully mediated the influences of the antecedents. For the purpose of a further model comparison, a third structural model 3 was estimated. In this model, the four antecedents and mentoring were all treated as antecedents. The outcomes remained the same: affective commitment and intention to quit. The seven additional controls on outcomes were included as antecedents too. The AIC (596.852), BCC (638.608), and CAIC (1468.855) values of model 3 were higher than that of both model 1 and model 2. Therefore, our proposed fully mediated model 1 remained the better model compared to two other alternative models. Furthermore, we undertook a supplemental analysis to test if industry type could moderate the structural paths of the proposed model 1. The data was divided into two samples, one was of service firms and the other was of manufacturing firms. The proposed model 1 was simultaneously estimated with the two samples once with unconstraint structural paths and then again with constraining the corresponding structural paths to equal for the two samples. The ChiSquare difference (622.852[df = 395] − 617.211[df = 389] = 5.641 [df = 6]) was insignificant (5.641, df = 6 b12.592, the critical value). Therefore, the model and the structural paths are not different in service firms and manufacturing firms, i.e., industry type does not appear to be a moderator. 5. Discussion This study examined the role of mentoring as a mediator of collectivism, individualism, and trust in supervisor to affective commitment and intention to quit. The results were generally consistent with our predictions. In particular, both collectivist values and trust in supervisor positively impacted mentoring while individualism had no significant effect. Mentoring mediated the relationship between both collectivism values and trust in supervisors and both organizational commitment and intention to quit. These findings support the theory that personal values and the subordinates' relationship with the supervisor have an important role in creating and influencing employee reception of mentoring from their supervisor. In fact, collectivist values and trust in supervisor plays a pivotal role in the determination of mentoring received from a supervisor. Additionally trust in the supervisor was negatively related to intention to quit. These findings support previous theory regarding the effects of mentoring on organizational commitment and intention to quit (e.g. see Allen et al., 2004; Eby et al., 2008). More importantly, we have provided support for a comprehensive picture of the mentoring model. Nonetheless, there are several limitations to our study along with opportunities for future research. First, our study lacks randomization since the sample was limited to the industry contacts of a small number of working MBA students. Future research would benefit from investigating our proposed relationships in both a controlled laboratory context and in an elaborate field setting. Second, we collected all data from the subordinate. Although our study focuses on individual level constructs from the perspective of the subordinate, collecting some data from the supervisor would have allowed us to more forcefully validate our mentoring measure in particular. More specifically, there should be generally agreement between the supervisor who provides the mentoring and the subordinated that receives the mentoring because such consensus provides
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
greater confidence in the mentoring construct. Nevertheless, we believe while not ideal our approach is acceptable and indeed consistent with previous research that recognizes the protégé is the best source of mentoring data. Third, this study is limited to supervisory mentoring dyads when research acknowledges that many individuals have a network of individuals they receive mentorship from at any one time (Higgins and Kram, 2001). Future research might account for the multiple mentorship relationships to investigate whether our predictions are consistent across different types of mentoring dyads. Finally, all data was collected in a static fashion. Although we controlled for the number of years the employee had been under the supervision of their boss, it would be ideal to use a longitudinal design to investigate how our model predictions hold across the several mentorship developmental stages. Future research is still necessary for a complete framework to be determined. For example, Singelis et al. (1995) have distinguished between vertical and horizontal dimensions of collectivism, suggesting that the former describes the tendency to accept hierarchy within the in-group, while the latter dimension captures individuals' emphasis on equality within the in-group. Perhaps, employees who are stronger on vertical collectivism are more responsive towards supervisory mentoring than employees who are stronger in horizontal collectivism. It could be worthwhile to observe how individuals' responses to mentoring differ across the two dimensions. Moreover, it is possible that other personality attributes (e.g., locus of control, extroversion) and demographic attributes (e.g., gender difference, racial difference, age difference) impact the amount of mentoring received from a supervisor. In addition, it would be interesting to investigate organizational outcomes such as organizational citizenship behavior, actual job performance, and promotions. Such research would require data collection from both members of the dyad. Finally, we did not find individualism related to any of the dependent measures. Continued examination is needed to develop a better understanding of the relationship between individualism and organizational outcomes. 5.1. Managerial implications Research has suggested that mentoring is essential for employee mobility and even compensation. As such, mentoring is a critical aspect of organizational behavior. Social exchange theory provides a fruitful framework for managers to understand supervisor–subordinate dyadic relationships particularly mentorship development. For example, while informal mentoring relationships are determined by both the organization and the employee, it is the organization that can influence the formal nature of mentoring. By providing managers with an understanding of antecedents and consequences of successful mentoring they will be better equipped to foster trust to ensure that the mentoring benefits are realized. References Allen T, Eby L. Common bonds: an integrative view of mentoring relationships. In: Allen TD, Eby LT, editors. Handbook of mentoring: a multiple perspective approach. Malden, MA: Blackwell; 2007. p. 397–419. Allen T, Eby L, Poteet M, Lentz E, Lima L. Career benefits associated with mentoring for protégés: a meta-analysis. J Appl Psychol 2004;89:127–36. Anderson J, Gerbing W. Structural equation modeling in practice: a review and recommended two-step approach. Psychol Bull 1988;103(3):411–23. Arbuckle L, Wothke W. Amos 4.0 user's guide. Chicago, IL: SmallWaters Corporation; 1995–1999. Armstrong J, Overton D. Establishing nonresponse bias in mail surveys. J Mark Res 1977;14:396–402. Bagozzi R, Yi Y. On the evaluation of structural equation models. Acad Market Sci 1988;16(1):74–94. Bartlett K. The relationship between training and organizational commitment: a study in the health care field. Hum Resour Dev Q 2001;12(4):335–52. Blau P. Exchange and power in social life. New York: Wiley; 1964. Clugston M, Howell J, Dorfman P. Does cultural socialization predict multiple bases of foci of commitment? J Manage 2001;5:5-30.
1117
Deluga R. Supervisor trust building, leader–member exchange and organizational citizenship behavior. J Occup Organ Psychol 1994;67(4):315–26. Earley P, Gibson C. Taking stock in our progress on individualism–collectivism: 100 years of solidarity and community. J Manage 1998;24:265–304. Eby L. Alternative forms of mentoring in changing organizational environments: a conceptual extension of the mentoring literature. J Vocat Behav 1997;51(1):125–44. Eby L, Rhodes J, Allen T. Definition and evolution of mentoring. In: Allen TD, Eby LT, editors. Handbook of mentoring: a multiple perspective approach. Malden, MA: Blackwell; 2007. p. 7-20. Eby L, Allen T, Evans S, Ng T, DuBois D. Does mentoring matter? A multidisciplinary meta-analysis comparing mentored and non-mentored individuals. J Vocat Behav 2008;72(2):54-267. Eisenberger R, Huntington R, Hutchison S, Sowa D. Perceived organizational support. J Appl Psychol 1986;71:500–7. Elangovan A, Shapiro D. Betrayal of trust in organizations. Acad Manage Rev 1998;23(3):547–68. Ensher E, Thomas C, Murphy S. Comparison of traditional, step-ahead, and peer mentoring on protégés' support, satisfaction, and perceptions of career success: a social exchange perspective. J Bus Psychol 2001;15(3):419–38. Flynn F. Identity orientations and forms of social exchange in organizations. Acad Manage Rev 2005;30(4):737–50. Fornell C, Larcker D. Evaluating structural equation models with unobservable variables and measurement error. J Mark Res 1981;18:39–50. Gould-Williams J, Davies F. Using social exchange theory to predict the effects of HRM practice on employee outcomes. Publ Manage Rev 2005;7(1):1-24. Guzzo R, Noonan K, Elron E. Expatriate managers and psychological contract. J Appl Psychol 1994;79:617–26. Haas D, Deseran F. Trust and symbolic exchange. Soc Psychol Q 1981;44(March):3-13. Higgins M, Kram K. Reconceptualizing mentoring at work. A developmental network perspective. Acad Manage Rev 2001;26(2):264–88. Hofstede G. Culture's consequences: international differences in work related values. Newbury Park, CA: Sage Publications; 1980. Hofstede G. Culture and organizations: software of the mind. London, England: McGraw-Hill; 1991. Homans G. Social behavior: its elementary forms. Orlando: Harcourt Brace; 1961. Koberg C, Boss R, Goodman E. Factors and outcomes associated with mentoring among health-care professionals. J Vocat Behav 1998;53(1):58–72. Kram K. Phases of the mentor relationship. Acad Manage Rev 1983;26:608–25. Kram K. Mentoring at work: developmental relationships in organizational life. 1st ed. Glenview, IL: Scott, Foresman and Company; 1985. Larzelere R, Huston T. The dyadic trust scale: toward understanding interpersonal trust in close relationships. J Marriage Fam 1980;42(3):595. Lee C, Bruvold N. Creating value for employees: investment in employee development. Int J Hum Resour Manag 2003;14(6):981-1000. Liden R, Graen G. Generalizability of the vertical dyad linkage model of leadership. Acad Manage J 1980;23(3):451–65. McAllister D. Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Acad Manage J 1995;38(1):24–59. McManus S, Russell E. New directions for mentoring research: an examination of related constructs. J Vocat Behav 1997;51:145–61. Meyer J, Allen N. A three-component conceptualization of organizational commitment. Hum Resour Manage Rev 1991;1:61–89. Meyer J, Allen N, Smith C. Commitment to organizations and occupations: extension and test of a three-component conceptualization. J Appl Psychol 1993;78(4):538–51. Mowday R, Porter L, Steers R. Organizational linkages: the psychology of commitment, absenteeism, and turnover. San Diego, CA: Academic Press; 1982. Noe R. An investigation of the determinants of successful assigned mentoring relationships. Pers Psychol 1988;41:457–79. Noe R, Greenberger D, Wang S. Mentoring: what we know and where we might go. Res Pers Hum Resour Manage 2002;21:129–73. Noordin F, Williams T, Zimmer C. Career commitment in collectivist and individualist cultures. Int J Hum Resour Manag 2002;13(1):35–54. Nooteboom B. Trust, opportunism and governance: a process and control model. Organ Stud 1996;17(6):985-1010. Olian J, Carroll S, Giannantonio S. Mentor reactions to protégés: an experiment with managers. J Vocat Behav 1993;43:266–78. Payne S, Huffman A. A longitudinal examination of the influence of mentoring on organizational commitment and turnover. Acad Manage J 2005;48:158–68. Podsakoff P, Organ D. Self-reports in organizational research: problems and prospects. J Manage 1986;12(Winter):531–43. Ragins B. Diversified mentoring relationships in organizations: a power perspective. Acad Manage Rev 1997;22:482–521. Ragins B, Cotton J. Mentor functions and outcomes: a comparison of men and women in formal and informal mentoring relationships. J Appl Psychol 1999;84(4):529–50. Ragins B, McFarlin D. Perceptions of mentor roles in cross-gender mentoring relationships. J Vocat Behav 1990;37:321–39. Robinson S. Trust and breach of the psychological contract. Adm Sci Q 1996;41: 574–99. Robinson S, Kraatz M, Rousseau D. Changing obligations and the psychological contract: a longitudinal study. Acad Manage J 1994;37(1):137–52. Rousseau D, Greller M. Human resource practices: administrative contract makers. Hum Resour Manage 1994;33(3):385–401. Rousseau D, Sitkin S, Burt R, Camerer C. Not so different after all: a cross-discipline view of trust. Acad Manage Rev 1998;23(3):393–404. Scandura T. Mentorship and career mobility: an empirical investigation. J Organ Behav 1992;13:169–74.
1118
O.C. Richard et al. / Journal of Business Research 62 (2009) 1110–1118
Scandura T, Pellegrini E. Workplace mentoring: theoretical approaches and methodological issues. In: Allen TD, Eby LT, editors. Handbook of mentoring: a multiple perspective approach. Malden, MA: Blackwell; 2007. p. 71–92. Seabright M, Levinthal D, Fichman M. Role of individual attachments in the dissolution of interorganizational relationships. Acad Manage J 1992;35:122–60. Shamir B. Calculations, values, and identities: the sources of collectivistic work motivation. Human Relat 1990;43(4):313–32. Singelis T, Triandis H, Bhawuk D, Gelfand M. Horizontal and vertical dimensions of individualism–collectivism: a theoretical and measurement refinement. Cross-Cult Res 1995;29:240–75. Tepper B. Upward maintenance tactics in supervisory mentoring and nonmentoring relationships. Acad Manage J 1995;38:1191–205. Tepper B, Taylor E. Relationships among supervisors' and subordinates' procedural justice perceptions and organizational citizenship behaviors. Acad Manage J 2003;46:97-105. Triandis H. Cross-cultural studies of individualism and collectivism. In: Berman J, editor. Nebraska symposium on motivation. Lincoln: University of Nebraska Press; 1989. p. 41-133.
Triandis H. Individualism and collectivism. Boulder, CO: Westview; 1995. Wayne S, Shore L, Liden R. Perceived organizational support and leader–member exchange: a social exchange perspective. Acad Manage J 1997;40(1):82-111. Whitener E. Do high commitment human resource practices affect employee commitment? A cross level analysis using hierarchical linear modeling. J Manage 2001;27:515–35. Williamson O. Transaction cost economics: the governance of contractual relations. J Law Econ 1979;22:233–61. Young A, Perrewe P. The exchange relationship between mentors and protégés: the development of a framework. Hum Resour Manage Rev 2000;10(2):177–209. Young A, Perrewe P. The role of expectations in the mentoring exchange: an analysis of mentor and protégé expectations in relation to perceived support. J Manag Issue 2004;XVI(1):103–26.