As production and management capacities have become globally standardized, many formerly effective strategies (e.g., cost reduction, speed of technology development, and automation) no longer contribute to competitive advantage [1
]. Recognizing these threatening environmental changes, companies are striving to achieve competitive advantages with sustainability-oriented innovations [3
]. One source of advantage can be obtained by engaging and developing talented human resources through organizational development programs such as coaching and mentoring. These popular approaches have been used as catalysts to develop employee potential in business by building one-to-one relationships. South Korea has followed this trend by applying coaching and mentoring extensively and concurrently in the workplace [5
Coaching is an organizational development strategy that can improve the manager-subordinate relationship [6
]. The academic literature reports diverse types of coaching practices, including coaching leadership [7
], executive coaching [8
], and peer coaching [9
]. Although each coaching type has a specific purpose and different coach-coachee relationship, they hold in common the goal to help individuals, groups, or organizations improve competence and performance. Managerial coaching (i.e., coaching leadership or manager-as-coach) is even more prevalent than the other types of coaching in South Korea [5
]. Managerial coaching means that line managers facilitate their employees’ learning as coach in order to encourage the development of a high-performance work environment [11
]. More than ever, managers can develop their subordinates within the organization by constructing horizontal and cooperative relationships. Thus, coaching has recently received attention as a form of leadership practiced by managers, and it has been widely applied in many companies [13
Mentoring is traditionally defined as a developmental relationship between a mentor and protégé who have different levels of expertise and skills [14
]. Mentors are usually competent senior employees who have rich experience in a particular field, while protégés are considered to have limited skills and experience. Mentoring may be informal or formal, but typically the primary functions associated with mentoring are vocational support, psychosocial support, and role modeling [16
]. The successful mentor provides opportunities for the protégé to acquire new skills and challenging project assignments, while at the same time engendering psychological safety and encouragement for the protégé through friendship [18
]. Furthermore, when he or she identifies with the mentor, the protégé may try to imitate the mentor’s attitude, values, and behaviors [15
Coaching is well known for playing an important role in enhancing employees’ work-related performance [6
], and mentoring impacts on employees’ career development [19
]. Coaching and mentoring are acclaimed ways to not only facilitate and support talented employees, but also to accomplish the strategic goals of the company. Accordingly, many companies have adopted simultaneous coaching and mentoring programs in South Korea [5
]. In general, the employee, as coachee and protégé, receives different benefits from the acting supervisory coach and the non-supervisory mentor. As a result, the employees may confront complex challenges when trying to maintain both coach-coachee relationships and mentor-protégé relationships effectively.
However, most research on coaching and mentoring in organizations has conducted each practice independent of the other, without offering simultaneous or integrated perspectives on both practices in combination. While coaching and mentoring are known to have common effects on employee commitment, turnover intention, performance, and related factors, at the same time, few studies addressed the interaction between coaching and mentoring. Hence, this study focuses on the duality that an employee may experience as a coachee in a coach-coachee dyad and as a protégé in a mentor-protégé dyad. The integrated effects of coaching and mentoring were examined for any synergistic or trade-off effects using the data of 247 triadic relationships in Korean companies that carried out concurrent coaching and mentoring programs. The findings suggest practical implications to maximize the beneficial effects of coaching and mentoring and may contribute to organizations’ strategic renewal by more effectively developing talented individuals.
3. Materials and Methods
We tested our hypotheses using a quantitative cross-sectional survey design. Drawing on Brislin’s process [72
], the survey was first translated from English into Korean and the equivalent meaning was verified through back-translation into English. Three independent bilingual subject-matter experts discussed the entire questionnaire in order to minimize variations attributable to language difference. The subject-matter experts included two practitioners and one professor, each of whom has at least 10 years of experience in human resource management.
The survey was conducted from 15 January to 15 March 2016. Participants were asked to respond to questions on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). The data for this study were collected from 17 South Korean firms, which had each concurrently adopted both a coaching program and a mentoring program, based on a convenience sampling approach. We first contacted eight consulting service firms that had the experience to provide coaching or mentoring programs and to implement coaching or mentoring systems for their client companies. Then we were introduced to the human resource managers of the client companies. The industries of the participating firms were manufacturing, financial services, and information services with five, five, and six organizations within each of these industries, respectively.
The target population was 700 employees in the 17 firms, who had been assigned concurrently as coachee as well as protégé in the two programs. The questionnaire was distributed via the intranet in the firms and 275 completed questionnaire sheets were collected. After excluding 28 due to incomplete or inaccurate Likert scale entries, 247 sheets were analyzed finally. There was little diversity regarding gender and education. Our data of the triadic relationships consisted of 94.8% male coaches, 91.5% male mentors, and 86.3% male employees. Regarding education, the percentage of the three actors with undergraduate degrees were 81.4%, and post-graduates comprised 11.2%.
To measure managerial coaching, we used eight items developed by Ellinger, Ellinger, and Keller [12
]. As a popular metric for managerial coaching, this scale focuses on behavioral experiences that reflect managerial coaching [73
]. The reliability coefficient was 0.64, which was below the generally accepted threshold of 0.70 [74
]. Two items were excluded from the original instrument concerning alpha-if-item-deleted, and the coefficient alpha for the measure became established at 0.72. The two items excluded from original 8 items were ‘my manager uses analogies, scenarios, and examples to help me learn’ and ‘to help me think through issues, my manager asks questions, rather than provide solutions’. We only could guess three reasons for the exclusion. First, all coaching behaviors have a certain difficulty level, so the two items could refer to the more difficult behaviors involved when managers try coaching subordinates on workplace. Second, our data were influenced by specific cultures. We found a Korean thesis [75
] that studied the differences in managers’ coaching behaviors among organizational cultures. The study points out the two items that are most different among culture types. Third, we inferred the effect of multidimensional managerial coaching behaviors (e.g., [24
]). The two items were commonly related with indirect examples or questions.
Mentoring, the other main independent variable, was measured using 15 items developed by Scandura and Ragins [76
], including three sub-functions based on the classification by Kram [40
]. The questions consisted of career mentoring (six items), psychosocial mentoring (five items), and role-modeling (four items). The reliabilities were 0.79, 0.87, and 0.73, and the overall coefficient alpha was 0.79. Considering the recommended level of 0.70 [74
], these values could be considered reliable.
Eighteen items based on Meyer et al.’s research [77
] were used to collect data on organizational commitment. The questions were classified into three dimensions; affective commitment, continuance commitment, and normative commitment [50
]. Three questions, of which alpha-if-item-deleted was high, were excluded through reliability verification. The excluded items were commonly reverse keyed like ‘I do not feel emotionally attached to this organization’, ‘If I had not already put so much of myself into this organization I might consider working elsewhere’, and ‘I do not feel any obligation to remain with my current employer’. The reliability of the sub-dimensions, measured by five items per dimension, was affective commitment (0.78), continuance commitment (0.72), and normative commitment (0.85), and the overall reliability was shown to be 0.73. These values, above the minimal acceptable level of 0.70 [74
], indicated the good reliability of the measure.
Seven sets of control variables were considered on the basis of prior research. We created two ‘dummy’ variables, differentiating among three industries; manufacturing, financial services, and information services. We also considered two relationship lengths that were measured as the period of the mentoring relationship and coaching relationship. The perceived similarity was measured using nine items that Eby et al. [78
] developed to measure the mismatch between mentor and protégé in the aspects of personalities, habits, dispositions, life priorities, and related factors. The items were reverse-coded, such that higher values represent greater similarity [41
]. Participants reported their coach and mentor using these questions. Three matches between their coach and themselves, between their mentor and themselves, and between the coach and the mentor were measured. The coefficient alphas were 0.74, 0.78, and 0.81, respectively. Finally, the organizational tenure of each coach and mentor was included, using the number of years employed in the organization.
We examined non-response bias by comparing the first 25% and the last 25% of returned surveys in terms of all items. Using multivariate analysis of variance, we found no significant differences between the two groups on any measures or overall (p
> 0.05). We therefore concluded that non-response bias was not a significant problem in this study [79
]. Furthermore, we conducted one factor test for checking for common method bias raised by self-reported measures and cross-sectional study design [80
]. An exploratory factor analysis of all items resulted in seven unrotated factors with eigenvalues greater than 1.0. The first factor, moreover, accounted for only 18.3% of total variance. Since a single factor did not emerge as dominant, common method bias is unlikely to be problematic in this study.
Confirmatory factor analysis was performed on four key variables; homogeneity with coach and mentor, managerial coaching behavior, mentoring practice, and organizational commitment. As items showing factor loadings lower than 0.50 were eliminated, and the reliability and validity of the measuring scales became acceptable. The excluded items with lower factor loading were the same items with lower coefficient alphas. The lowest composite reliability (CR) score was 0.71, the lowest average variance explained (AVE) was 0.62, and the lowest coefficient alpha was 0.72. These were satisfied with general cut-off scores (CR = 0.70, AVE = 0.50) [81
As the coachee and protégé respondents were likely nested among their respective coaches and mentors, multi-level structures should also be considered. We checked the influence belonging to each coach or mentor. The unconditional model of hierarchical linear modeling was analyzed in organizational commitment. The results showed random effects to vary, but the higher-level unit was not significant (u0j
= 1.075, p
> 0.05 in coach level; u0j
= 1.087, p
> 0.05 in mentor level). The insignificance suggests that employees’ responses were not affected by the higher-level unit. The intraclass correlations, ICC(1), were 0.041 in the coach level and 0.038 in the mentor level. ICC(2) values were 0.090 in the coach level and 0.097 in the mentor level. If accepting the effect of the higher-level unit, ICC(1) values should vary from 0.05 to 0.20 and ICC(2) values should be over 0.50 [82
]. Both ICC(1) and ICC(2) were low in our data, which suggests that the effect of multi-level structures was not serious.
shows the descriptive statistics for the main variables including mean, standard deviation, and zero-order correlation. The relationship length of the mentor and coach, tenure of coach, managerial coaching behavior, and mentoring practice were significantly related to employees’ organizational commitment.
The results of the hierarchical regression analysis included four models, as depicted in Table 2
. The impact of the control variables on employees’ organizational commitment was significant in Model 1 (R2
= 0.23; F = 8.99, p
< 0.01). In Model 2, which tested the main effects of the theoretical variables, managerial coaching behavior (β = 0.183, p
< 0.01) and mentoring practice (β = 0.131, p
< 0.05) were positively related to employees’ organizational commitment, while the homogeneity of the coach and mentor was nonsignificant (β = 0.80, p
In Models 3 and 4, the interaction terms were added in sequence to test the moderating effects of mentoring practice and the homogeneity of coach and mentor. Following Aiken and West [83
], the variables involved in the product terms were mean-centered to avoid multicollinearity. The variance inflation factor (VIF) scores of all variables in Models 3 and 4 varied from 1.03 to 1.48, indicating little distortion of the regression results.
Model 3 contained two-way interaction terms generated by each combination of managerial coaching behavior, mentoring practice, and the homogeneity of coach and mentor. Hypothesis 1 was supported because mentoring practice moderates the positive relationship between managerial coaching behavior and employees’ organizational commitment (β = 0.160, p
< 0.01). The simple slope analysis suggested by Aiken and West [83
] of this moderating effect is revealed in Figure 2
. The relationship between managerial coaching behavior and employees’ organizational commitment was positive for conditions of high mentoring practice, but neutral for conditions of low mentoring practice. As we anticipated, the other interaction effects were not significant (βC×H
= −0.018, p
> 0.05; βM×H
= −0.043, p
In Model 4, we entered three-way interaction terms to examine Hypothesis 2. The interaction terms consisted of managerial coaching behavior, mentoring practice, and the influence of the homogeneity of coach and mentor on employees’ organizational commitment. The coefficient was both negative and significant (β = −0.120, p
< 0.05) in support of Hypothesis 2. Figure 3
illustrates the relationship between managerial coaching behavior and employees’ organizational commitment under high or low conditions of mentoring practice and homogeneity of coach and mentor. Managerial coaching behavior had the strongest positive relationship with employees’ organizational commitment, if mentoring practice was high but the homogeneity of coach and mentor was low. However, the relationship between managerial coaching behavior and employees’ organizational commitment was negative when both mentoring practice and the homogeneity of coach and mentor are low.
5. Discussion and Limitations
Prior research [18
] has suggested that managerial coaching and mentoring have a critical impact on developing employees’ potential capabilities. As uncertainty makes predicting the future impossible and complexity increases, the sustainability of enterprises has become dependent on finding ways to realize and improve employees’ potential capabilities. Therefore, enterprises striving for new competitive drivers have great interest in managerial coaching and mentoring. However, researchers have shown scarce interest in the effects of simultaneously implementing managerial coaching and mentoring. The present study aims to fill the gaps of previous research by examining the interactions between managerial coaching and mentoring. We not only confirmed the results of previous studies related to managerial coaching and mentoring, but also found the following implications.
This study focused on the viewpoint of each employee as both a coachee and protégé in their organization’s concurrently implemented coaching and mentoring programs. Previous studies were limited to coaches and mentors as main actors but overlooked employees who are both coachee and protégé. However, the present study investigated the effects of an interaction between managerial coaching and mentoring. Thus, we anticipate that it will contribute to a more practical enhancement of employees’ organizational commitment.
The first finding of our study is that the positive impact of managerial coaching behavior on the organizational commitment of employees is moderated positively by mentoring practices. Although managerial coaching behavior is positively associated with employees’ organizational commitment at high levels of mentoring practice, the relationship is not significant at low levels of mentoring practice. The interaction effects of managerial coaching and mentoring offer important implications for organizations that have adopted managerial coaching and mentoring with vague expectations of their effectiveness. In order to maximize the effect of managerial coaching, organizations should carefully consider the integrated dynamics of managerial coaching and offer mentoring to employees within positive managerial coaching relationships.
Next, the positive moderating effects of mentoring practice were extended to incorporate the homogeneity of coach and mentor in the research model. The positive moderating effect of mentoring practice is significantly strong if the homogeneity of coach and mentor is low. Conversely, the moderating effects of mentoring practice are less negative when both mentoring practice and the homogeneity of coach and mentor are low. When employees come to view coaching and mentoring as complementary resources, a high heterogeneity between coach and mentor could play a role in constructing more diverse social ties according to social support resource theory [71
]. Thus, these complementary effects could extend the explanatory power from the previous research concentrated on the similarity between coach and coachee (e.g., [84
]) or between mentor and protégé (e.g., [41
When the functions provided by a mentor are high, the homogeneity of coach and mentor would not matter in general. By contrast, it would become very critical to ensure a high homogeneity of coach and mentor under the condition that the capabilities of mentor were low. These second findings highlight that companies need to be more careful in selecting mentors and coaches, as well as in optimizing the two practices rather than maximizing their application out of a mistaken assumption that ‘more is better’. Similarly, Ellinger et al. [88
] insisted that the practical effects of empowerment on an employee’s performance could be higher or lower in contexts where the levels of managerial coaching or formal training are substantially higher or lower. Companies should take into account the need to not only understand the functions of coaching and mentoring but also to optimize the homogeneity of coach and mentor.
Despite these findings and implications, this paper has the following limitations that future research can address. First, the effects of managerial coaching and mentoring were measured according to the subjective perception of employees. It is necessary to analyze the opinions of coaches and mentors together in order to enhance the explanatory power of the studies. Second, this study was cross-sectional, so the causality of effects could not be inferred from our findings. The possibility that the organizational commitment of employee leads their interpretation of homogeneity with their coach or mentor, and vice versa, was not ruled out perfectly. Moreover, organizational commitment increased by coaching or mentoring would be developed over time through dynamic interactions. Thus, in order to control the effect of duration in coaching and mentoring, longitudinal or experimental designs need to be employed.
Third, the sample was very male dominated because of a specific characteristic of the workforce structure of Korean companies. It is possible that the research results would be an artifact of the sample. Fourth, it is necessary to apply multi-level analysis in our research model, because the respondents are nested among their respective coachees and mentors. We checked the random effect in the null model and the intraclass correlations as recommended by Bliese [82
], but the effect of the higher-level unit was not statistically significant in our data. The small sample size of each subgroup may cause the insignificance. In our data, the coach had 3 coachees as the maximum and 2.3 coachees on average, and the mentor had 4 protégés as the maximum and 2.7 protégés on average.
Therefore, future research would need to test the proposed relationships by preplanning the designed data to conduct multi-level analysis. Moreover, comparative studies using other outcome variables would be valuable in verifying the differences in the results. For example, organizational satisfaction, promotion, career competencies, turnover intention, and related factors are essential outcomes related to managerial coaching and mentoring in human resource management.