Jennifer K. Giancola ✉ (School for Professional Studies, Saint Louis University) Iwan D. Webster (Department of Psychology, Saint Louis University) Jack C. Friedrich (Department of Psychology, Saint Louis University) Thomas E. Burroughs (Department of Health Management & Policy, Saint Louis University)
Workplace mentoring has typically focused on mentor functions without sound measurement of mentees. This study developed and tested a mentee behaviour scale (MBS) to assess mentors’ and mentees’ perceptions of effective mentee behaviours. The construct and items were derived from the literature and our experience implementing workplace mentoring programmes. An expert panel evaluated content validity before testing the MBS on an employee sample of 295 mentees and 294 mentors. Using a multistep analysis, the MBS was reduced to 24-items with mentee and mentor versions. Results supported MBS reliability and validity for use in workplace mentoring research and practice.
workplace mentoring, mentee assessment, mentor programmes, mentee behaviours, scale development
Accepted for publication: 03 January 2025 Published online: 03 February 2025
© the Author(s) Published by Oxford Brookes University
Workplace mentoring is a tool that aims to drive employee growth and provide learning opportunities intended to further the mentee’s career and professional development (Ragins & Kram, 2007). Studies support the effects of mentoring on positive workplace outcomes for both mentees and mentors (Giacumo, Chen, & Seguinot-Cruz, 2020; Giancola, Guillot, Chatterjee, Bleckman, & Hoyme, 2018; Giancola et al., 2020a; Noe, Clarke, & Klein, 2014; Wen, Chen, Dong, & Shu, 2019). Simply being in a mentoring relationship, however, does not guarantee its success. Research suggests that there are many antecedents influencing partnership outcomes with the mentor and the mentee roles both contributing to the success of mentoring relationships (Eby et al., 2013; Ghosh, 2014; Illies & Reiter-Palmon, 2018; Ivey & Dupré, 2022). A mentoring relationship should be one of mutuality (Kram, 1985) in which both the mentor and mentee are actively engaged and take initiative (Eby & Robertson, 2020).
Research has predominantly measured the characteristics and best practices of effective mentors (Anderson, Chang, Lee, & Baldwin, 2022; Hu & Wang, 2022; Kokt & Dreyer, 2024), partnerships (Straus, Johnson, Marquez, & Feldman, 2013), and formal programmes (Giancola et al., 2020a; Hieker & Rushby, 2020). While there are multiple scales assessing the roles and functions of mentors, there are limited tools for examining mentees (Ragins, 2016). Mentee characteristics and behaviours have been identified, but few measures have been psychometrically validated (Eby, Durley, Evans, & Ragins, 2008). This gap in the literature ignores the interconnected nature of the mentee-mentor relationship, and the contributions of the mentee to partnership success (Ragins, 2016).
The purpose of this study was to develop a psychometrically sound measure of mentee effectiveness behaviours in a workplace mentoring partnership. Creating a valid and reliable assessment will not only provide an instrument for mentoring research, but also enhance the understanding of partner contributions and the implementation of structured mentoring programmes. If organisations have access to an empirically sound measure, they can implement specific mentee training plans to improve mentee skills and, ultimately, improve mentoring relationships and programmes. Furthermore, the assessment could be used as a tool for mentee self-reflection, potentially enhancing the mentoring exchange and positive benefits.
While mentoring spans many disciplines and fields (Eby et al., 2013) including different types and forms (Giacumo et al., 2020), we focused on dyadic mentoring partnerships at work. Based upon the empirical literature, dyadic mentoring was defined as a “one-to-one relationship in which a more experienced person (mentor) provides support, guidance, knowledge and/or opportunity to a less experienced person (mentee) to further the mentee’s career and professional development” (Berk, Berg, Mortimer, Walton-Moss, & Yeo; Johnson & Ridley, 2018).
One approach to examining dyadic workplace mentoring is relational mentoring theory (Janssen, Van Vuuren, & De Jong, 2016). Ragins and Verbos (2007) called for research that examines mentoring from a relational perspective emphasising an interdependent partnership with benefits for both mentees and mentors: essential to understanding high-quality relationships. This contrasts with the one-sided approach of traditional mentoring models and the transactional view of social exchange theory which focus on what the mentor provides as opposed to the active role of the mentee. Ragins (2016) reaffirmed this limitation stating that research, “[…] focused nearly exclusively on the mentor’s behaviours in the mentoring relationship. This approach not only failed to capture the protégé’s behaviours and the behaviours that build the quality of the relationship” (p. 234).
While mentees’ personality characteristics, demographics, job/career history, and actions have been examined (Gisbert-Trejo, Landeta, Albizu, & Fernández-Ferrín, 2019), we focused on the behaviours that the mentee exhibits. Understanding the scope of mentee characteristics is valuable, but these can be harder to identify/develop than overt behaviours. Furthermore, mentee behaviour is one component of the relational mentoring schema (Ragins & Verbos, 2007). Eby and Robertson (2020) provide support for this approach: “Importantly, the relationship science theories examined suggest that the behaviours that occur once the mentoring relationship has been established are just as important, and are likely more predictive of outcomes, than pre-entry characteristics, programme features, and matching” (p. 94). If we are to fully understand high-quality partnerships and test a relational approach to workplace mentoring, then a validated mentee behaviour scale (MBS) is essential.
Criterion validity assesses the degree to which a scale predicts relevant outcomes (DeVellis, 2017), and was tested by examining the relationship between the MBS and mentoring goal attainment and intent to leave the relationship (Eby et al., 2008). Relational mentoring theory underscores the importance of both mentee and mentor contributions in a high-quality partnership (Ragins, 2012). Relationships in which both mentors and mentees exhibit effectual behaviours lead to positive perceptions and increased partnership satisfaction (Berk, Berg, Mortimer, Walton-Moss, & Yeo, 2005). Mentee behaviours like accessibility, initiative, follow-through, and openness to feedback are linked to mentee progress (Giancola et al., 2020a), suggesting that these behaviours support goal attainment. Further, the MBS questions were intentionally derived from research on effective mentee behaviours. We propose that:
Hypothesis 1: Positive perceptions of mentee behaviour will be positively related to mentoring goal attainment.
Mentoring partnerships are subject to interpersonal dynamics similar to other relationships (Ragins, 2012). Separation is an inevitable stage that results from partnership evolution, mentor-mentee fit, and partner interaction (Giancola, Heaney, Metzger, & Whitman, 2016; Kram, 1985). Intentions to leave the mentoring partnership may be an indicator of a negative relationship and/or negative perceptions of one’s partner (Burk & Eby, 2010; Eby et al., 2008). In contrast, when mentees exhibit positive behaviours, mentoring relationships are deemed to be of higher quality, increasing the likelihood that both parties will remain in the relationship (Burk & Eby, 2010). We predict that:
Hypothesis 2: Positive perceptions of mentee behaviour will be negatively related to intent to leave the mentoring partnership.
Convergent validity examines a scale’s relationship to measures of the same or similar constructs (DeVellis, 2017). Given that a psychometrically sound assessment for mentee effectiveness is not currently available, we chose theoretically relevant constructs: relationship quality and mentor effectiveness. It should be noted that perceptions of mentor support/effectiveness are related to, but conceptually distinct from, the mentor/mentee’s evaluation of the relationship and outcomes (Eby et al., 2013). Consistent with a relational perspective, one would assume that positive mentee behaviours are more likely to be reciprocated by the mentor, and vice versa. Consequently, this interdependence leads to higher relationship quality for both parties (Ragins & Kram, 2007). Although few studies have examined the mentee’s role, there is some support for a positive correlation between the mentee’s behaviour and mentoring support, relationship quality and partnership satisfaction (Eby et al., 2013; Giancola et al., 2020a). We hypothesise:
Hypothesis 3: Positive perceptions of mentee behaviour will be positively related to mentoring relationship quality and mentor effectiveness.
Discriminant construct validity demonstrates that the measure is distinct from ‘unrelated’ constructs and unique to the broader nomological net surrounding it (Clark & Watson, 2019; DeVellis, 2017). Similar to Eby et al. (2008), general job satisfaction, social relations at work, and feelings of positive affect were selected to demonstrate that reports of mentee effectiveness behaviours were divergent from those measures. A mentoring relationship is a unique relational exchange and, consequently, perceptions of mentee behaviours should be distinct from general workplace attitudes and mood. We hypothesise that:
Hypothesis 4: Perceptions of mentee behaviour will be distinct from general job satisfaction, social relations at work, and positive affect.
DeVellis (2017) and Clark and Watson (2019) were used as guides throughout the process of scale development and psychometric testing. Based on the literature and our experience implementing mentoring programmes in multiple organisations, we used a deductive approach to develop the operational definition and behavioural categories that represent construct breadth. Mentee effectiveness behaviours were defined as “actions carried out by the mentee that positively contribute to the mentoring relationship process and outcomes.” To ensure that the content domain was covered, 13 categories of effective mentee behaviours were identified from the literature. Next, the researchers, who are experts in the field of workplace mentoring, drew from the literature and their experiences to develop four to eight items per category, resulting in a pool of 71 survey items that spanned the entire content domain of mentee behaviours. The item pool was intentionally redundant with three times the number of desired items (DeVellis, 2017).
A panel of 10 subject matter experts (SMEs) consisting of faculty and professionals evaluated the behavioural categories and items for content-related validity (DeVellis, 2017). The MBS was reduced to 53 items based on both quantitative and qualitative evaluation of the expert review. Per the SME’s recommendation and DeVellis (2017), reverse scored items were eliminated, and some item wording was modified. A 14th category of Maintain Boundaries with three items was added based on the SME feedback. Three to six items per category were retained to ensure the construct domain was covered for the study. The final 14 categories of mentee effectiveness behaviours and references for the Mentee Behaviour Scale (MBS) are in Table 1.
Construct behavioural categories with definitions and supporting references
Participants were recruited from the online research platform Cint (Cint, 2023): a reputable digital survey platform considered a suitable data collection method for applied social science research (Walter, Seibert, Goering, & O’Boyle, 2019). MBS derivative versions (mentor and mentee) were sent in separate participant email requests from Cint. Respondents were eligible to participate if they met the following criteria: at least 22 years of age; worked at least 20 hours per week; resided in the United States; served as a mentee or mentor in a dyadic, career-related mentoring partnership within the last 12 months. To be included in the final participant pool, they had to pass two attention checks and complete at least 75% of the questions. The final usable participant pool included 295 mentees and 294 mentors. This approximates the 300 participants recommended for scale development (Clark & Watson, 2019). Key demographics can be found in Table 2.
Key demographics for both mentees and mentors
The following measures were included in both the mentee and mentor surveys. All items were measured on a 5-point Likert response scale: 1 = strongly disagree to 5 = strongly agree.
Mentee effectiveness behaviours were measured with the 53 item MBS that resulted from the SME panel review. Derivative versions were used for the mentors to evaluate their mentee and for mentees to self-assess. The same questions were used on each of the MBS versions with phrasing slightly modified on the mentor version, where appropriate (Clark & Watson, 2019).
Criterion validity included two measures. Goal attainment was measured with one item that we used in prior mentoring programme development and evaluation (Giancola et al., 2016). Two questions from Burk and Eby (2010) were used to assess intent to leave (α = .88).
Convergent validity was assessed with the following two measures. Five questions from the Allen and Eby (2003) measure of mentoring relationship quality were adapted for this study (α = .85). The Mentor Evaluation Tool (MET; α = .96) was slightly adapted to assess mentor effectiveness from the perspective of both the mentor and mentee. It consisted of 13 questions from Yukawa, Gansky, O’Sullivan, Teherani, and Feldman (2020).
Three measures were used for discriminant validity. Job satisfaction was measured with three questions from the Michigan Organisational Assessment Questionnaire-Job Satisfaction Subscale (MOAQ-JSS; α = .84; Cammann, Fichman, Jenkins, & Klesh, 1979). Social relations at work were assessed with three items from Quinn and Staines (1979); the same items were used by Eby et al. (2008; α = .79). Finally, ten items from the Positive and Negative Affect Schedule (PANAS; α = .84; Watson, Clark, & Tellegen, 1988) measured positive affect.
A muti-step process was used to evaluate the performance and structure of both versions of the 53-item MBS and guide the process of creating a short-form version of the instrument. Multiple imputation using fully conditional specification (FCS) was implemented by the MICE algorithm to address missing data. Predictive mean matching was chosen due to the categorical nature of the data (Van Buuren & Groothuis-Oudshoorn, 2011). Analytic steps included classical test theory methods and modelling from item response theory (IRT), leading to a 24-item MBS. These analyses were repeated for the MBS 24 to evaluate its overall performance.
Item distribution analysis (IDA) was conducted to examine the characteristics of each survey question. The skew for each item was negative but fell within the acceptable range of –2 to 2 (average skew for mentee version = -1.31; average skew for mentor version = -1.26). No ceiling or floor effects or unusual response patterns were observed.
Internal consistency was assessed using Cronbach’s alpha. The mentee version had α = 0.96 and the mentor version had α = 0.97, indicating strong internal consistency.
Factor analysis was conducted to support the unidimensional structure of the MBS prior to IRT (Stevenor & Zickar, 2022). The Kaiser-Meyer-Olkin (KMO) test and Bartlett’s test of sphericity were used to ensure that the requirements were met for adequate common variance, sample size, and inter-item correlation matrix that is significantly different from an identity matrix. These criteria were met for both the mentee and mentor versions of the MBS (mentee version: overall KMO = 0.95, Bartlett’s chi-square = 8578.75, p < 0.001; mentor version: overall KMO = 0.95, Bartlett’s Chi-square = 9504.02, p < 0.001).
Exploratory factory analyses identified 9 factors with eigenvalues > 1.0 but with a steep drop between the first and second eigenvalues for both the mentee version (λ1 = 20.30, λ2 = 1.94) and mentor version (λ1 = 20.09, λ2 = 1.74). This pattern was also observed in the scree and parallel analysis plots. To further support the unidimensional structure of the MBS, three confirmatory factory analyses with Promax rotation were conducted: (1) nine-factor, (2), two-factor, and (3) one-factor (see Table 3 for model fit). These indicators uniformly suggest that a one-factor model provides the best fit to the data which supports the intended theoretical structure.
Exploratory factor analysis of 53-item MBS (mentee and mentor versions)RMSEA = mean squared error of approximation. RMSR = root mean of the residuals. TLI = Tucker Lewis index. BIC = Bayesian information criterion.
Item response theory (IRT) also was used to evaluate and reduce MBS items. Unlike classical test theory, IRT models the individual responses on a set of test questions to their underlying latent trait (θ) and thus provides richer information about item performance. GRM was used to evaluate items with two sets of parameters: 1) threshold parameters, b, also known as location or difficulty parameters, and 2) discrimination parameter, a, also referred to as the slope parameter.
The threshold parameter indicates the point on the latent construct continuum where the probability of selecting a particular response category transitions from one category to the next. Ideal survey items have threshold parameters that are evenly distributed across the entire range of the latent construct. The discrimination parameter indicates how well an item discriminates between individuals with higher and lower levels of the latent trait. This parameter is constant across all response categories on a Likert scale. Items with larger discrimination parameters (i.e., a ≥ 1) are better able to discriminate between high and low scores on the latent MBS trait (Zickar, Russell, Smith, Bohle, & Tilley, 2002). Overall model fit was assessed using AIC, BIC, M2, RMSEA, CFI and the log-likelihood. Overall item fit was assessed using S-X2 (Orlando & Thissen, 2000) and χ2/df (Drasgow, Levine, Tsien, Williams, & Mead, 1995). In addition, character response curves (CRC) and item information curves (IIC) were evaluated for each item.
As shown in Table 4, overall model fit was strong for both the mentee and mentor versions of the 53-item MBS. Similarly, individual item fit was strong for most items in both versions. In the mentee version, S-X2 ranged from 25.16 to 71.56 with all items showing non-significant results, suggesting good fit for all but six items, p < .05. In the mentor version, S-X2 ranged from 25.79 to 73.94 with all except five items showing non-significance. All item discrimination parameters were ≥ 1 indicating that items were effective at differentiating individuals low and high on the underlying trait. Threshold parameters were well-ordered and appropriately spaced.
IRT model fit for 53-item MBS and 24-item MBS (mentee and mentor versions)
Based on the analyses and mentoring theory and research, a multi-faceted approach was used to reduce the MBS to 24 items. Items with stronger factor loadings and moderate inter-item correlations on both versions were given priority. When appropriate, items that did not discriminate well or did not fit both mentor and mentee IRT models were eliminated. Additionally, three items with abnormal IIC’s were removed. Most importantly, the best fitting items for each behavioural category were maintained to ensure that the breadth of the construct was represented. All selected items fit with the theoretical framework and performed as desired on most analyses. The final 24 items for both versions can be found in Table 5.
Final 24 items in the MBS (mentor and mentee versions)
Cronbach’s alpha indicated strong internal consistency for both MBS 24 versions (mentee version, α = 0.94; mentor version, α = 0.95). These results mirror those of the original 53-item versions, suggesting that there was no loss of reliability due to item reduction. Inter-item correlations for both versions of the MBS 24 were strong and positive, but not high enough to suggest redundancy (e.g., no correlations of 0.70 or greater). All were significant at p < .001.
IRT analysis was repeated for the MBS 24 to examine model and item performance (see Table 4). The discrimination factors, a, suggested strong item performance (a ≥ 1.0) for all items. The discrimination parameters for the mentor version ranged from 1.52 to 2.58 and for the mentee version ranged from 1.42 to 2.43. This is a significant improvement over the IRT analyses of original 53-item MBS where 12 items fell short of the criteria for strong item performance. Similarly, the threshold parameters, b, for all MBS 24 items were appropriately ordered and demonstrated good spacing between the response categories. The CRC and IIC also demonstrated good item performance.
The correlation matrix for the MBS 24 (mentor and mentee versions) and validity variables is in Table 6. Criterion validity was supported (Hypothesis 1 and 2) with strong positive associations between perceptions of mentee behaviours and goal attainment from the perspectives of both mentees (r = 0.55, p < .001) and mentors (r = 0.61, p < .001). Significant negative associations were found between mentee behaviours and intentions to leave the mentoring partnership from the perspectives of both mentees (r = - 0.21, p < .05) and mentors (r = - 0.13, p < .05). Correlational analyses also provided strong support for convergent validity (Hypothesis 3). Mentee behaviour was strongly associated with mentoring relationship quality and mentor effectiveness from both mentees (r = 0.67, p < .001, and r = 0.66, p < .001) and mentors (r = 0.72, p < .001 and r = 0.74, p < .001).
Correlations between MBS 24 and mentee (top) and mentor (bottom) outcomes a = p < .05; c = p < .001
To test the theoretical distinction between mentee behaviours and general job satisfaction, social relations at work and positive affect, we used three statistical approaches: correlation, confirmatory factor analysis (CFA), and hierarchical regression (Stevenor & Zickar, 2022). In support of Hypothesis 4, mentee behaviours were strongly positively correlated, but not redundant, with general job satisfaction (mentee version, r = 0.31, p < .05; mentor version, r = 0.33, p < .05), social relations at work (mentee version, r = 0.31, p < .05; mentor version, r = 0.45, p < .05) and positive affect (mentee version, r = 0.36, p < .05; mentor version, r = 0.47, p < .05).
CFA was used to compare the fit of two separate models for each of the discriminant validity measures (Stevenor & Zickar, 2022). The first model was a two-factor structure in which the mentee behaviour items load onto one factor and the items of a discriminant validity measure load onto another factor. This was compared with a second model where a one factor structure in which all items loaded onto the same factor. If mentee behaviours are distinct from the discriminant validity variables, the two-factor structure should fit better than the single factor.
For all variables, a two-factor model fit better, thereby supporting discriminant validity. Comparing mentee behaviours to general job satisfaction, a two-factor model (χ2/df = 2.21, RMSEA = 0.064, CFI = 0.89) fit better than a one-factor model (χ2/df = 2.74, RMSEA = 0.077, CFI = 0.84; Δ χ2 = 173.35, p < .05). In the mentor version, a two-factor model (χ2/df = 2.52, RMSEA = 0.072, CFI = 0.88) fit better than a one-factor model (χ2/df = 3.40, RMSEA = 0.09, CFI = 0.81; Δ χ2 = 286.32, p < .05). Comparing mentee behaviours to social relations at work, a two-factor model (χ2/df = 2.27, RMSEA = 0.07, CFI = 0.89) fit better than a one-factor model (χ2/df = 2.81, RMSEA = 0.08, CFI = 0.84; Δ χ2 = 176.95, p < .05) in the mentee version. In the mentor version, a two-factor model (χ2/df = 2.56, RMSEA = 0.07, CFI = 0.87) fit better than a one-factor model (χ2/df = 3.04, RMSEA = 0.08, CFI = 0.83; Δ χ2 = 159.93, p < .05). For positive affect, a two-factor model (χ2/df = 2.23, RMSEA = 0.07, CFI = 0.87) fit better than a one-factor model (χ2/df = 4.17, RMSEA = 0.10, CFI = 0.66; Δ χ2 = 1021.8, p < .05) in the mentee version. In the mentor version, a two-factor model (χ2/df = 2.19, RMSEA = 0.06, CFI = 0.89) fit better than a one-factor model (χ2/df = 4.33, RMSEA = 0.11, CFI = 0.68; Δ χ2 = 1124.9, p < .05).
The third statistical test for discriminant validity was a series of hierarchical regressions to determine if mentee behaviours accounted for incremental variance in partnership goal attainment beyond each of the discriminant validity variables (Stevenor & Zickar, 2022). For each of the three hierarchical regressions, we added a discriminant validity variable at Step 1 and mentee behaviours at Step 2, examining whether mentee behaviours accounted for an increase in variance explained in partnership goal attainment (i.e., a significant Δ R2). Supporting Hypothesis 4, mentee behaviours accounted for incremental variance beyond general job satisfaction (mentee Δ R2 = 0.03; mentor Δ R2 = 0.02), social relations at work (mentee Δ R2 = 0.01; mentor Δ R2 = 0.07), and positive affect (mentee Δ R2 = 0.03; mentor Δ R2 = 0.06).
This study takes the first step toward developing a reliable and valid measure of mentee effectiveness behaviours. The results support the robust psychometric properties of the MBS mentee and mentor versions. Although inclusion of the initial 53 items was supported, we desired a shorter form. A multifaceted review and set of analyses resulted in 24 items that best reflect the breadth of the construct and behavioural categories.
Hypotheses 1-3 regarding criterion and convergent validity were confirmed. The MBS had strong relationships with mentoring goal attainment and relationship quality suggesting its importance for predicting partnership outcomes. The correlations between the MBS and intent to leave the partnership were significant but weaker than the MBS’ relationships with other criterion/convergent variables. Future research should further examine the variables that determine partnership length and mentee and mentor efforts to sustain the relationship.
Discriminant validity (Hypothesis 4) was supported, though the magnitude of the MBS mentor correlations with work relations and affect were slightly higher than expected. This may be due to overlap with some aspects of relational mentoring. Mentors with stronger work relationships and positive affect may be in higher quality partnerships or, simply, rate their mentee’s behaviour more positively (Huang & Weng, 2012). Nonetheless, the two-factor CFAs and follow-up hierarchical regressions demonstrated that the MBS is distinct from those measures. Mentee effectiveness behaviours overlap with other variables, but collectively they appear to represent a unique construct captured by the MBS.
A notable limitation of the study is that the data was collected from a single sample at one point in time using the same method. This approach fails to capture differences across time or samples and can inflate correlations between variables. To overcome problems with cross-sectional data, a longitudinal and/or experimental study of mentoring pairs is recommended including mentee and mentor partner ratings using the MBS derivative versions; this research could be especially helpful by including unmentored employees to isolate the effects of mentoring from other causes (e.g., spurious effects; Ivey & Dupré, 2022). If the pairs are part of a structured programme, then multiple survey points could be integrated into the programme rollout. Another data collection method, like participant interviews, may be added to counter common method bias (Clark & Watson, 2019).
Cross-validation is an essential next step to investigate the MBS’ performance in other samples and cultures (Clark & Watson, 2019). Using Cint allowed us to more easily and promptly survey a large sample that met our criteria. Online panel survey platforms have been found to yield similar psychometric results to conventional data sources (Walter et al., 2019); and precautions were taken to cull the best data from Cint. Nonetheless, we acknowledge its limitations and plan to test the MBS on diverse populations. It is crucial not to assume that mentee behaviours observed in a predominantly Caucasian U.S. sample will be applicable across different cultures. Factors such as cultural competency, value alignment, and language agreement significantly influence mentoring relationships and must be considered in the assessment of both mentors and mentees (Kinos et al., 2023).
There are ample opportunities to expand MBS’ scope. Derivative versions may be required for other mentoring types (i.e., peer, multilevel, group, cultural), contexts (i.e., education, healthcare, international), and/or fields (i.e., STEM, liberal arts, research; Giacumo et al., 2020; Ivey & Dupré, 2022). In addition to behaviours, subscales can be added to assess mentee effectiveness characteristics like personality, attitudes, professional attributes, etc. (Gisbert-Trejo et al., 2019). Eventually, the MBS could be used in efforts to identify profiles of effective mentees, as has been done for mentors (Anderson et al., 2022), potentially to tailor support to mentee needs throughout structured mentoring programmes, to facilitate effective matching or serve as a metric for programme evaluation (Hieker & Rushby, 2020). Similarities or differences between mentees’ and mentors’ scores on the MBS may have important implications for the mentoring relationship and its outcomes (Illies & Reiter-Palmon, 2018). Furthermore, the MBS should be examined as part of the larger relational mentoring model (Ragins & Verbos, 2007).
Although additional psychometric testing is advised, the MBS is a tool that researchers can use to assess mentees in workplace mentoring studies. Coupled with pre-existing mentor scales, it enables a relational mentoring approach in which the behaviours of both mentors and mentees can be examined to more fully understand the antecedents, processes, and consequences of partnerships (Ragins, 2016). Further, the MBS derivative versions can be used to examine self-other (mentee-mentor) rater agreement of the mentee’s performance (Allen & Eby, 2008). The benefits of this include the ability to identify and mitigate rater and self-assessment bias and blind spots.
The MBS has practical implications for organisations who offer structured mentoring programmes or simply want to enhance mentoring skills. Surprisingly, professionals/employees frequently are unaware of how to successfully give and receive mentorship (Giancola et al., 2016, 2018, & 2020b; Giacumo et al., 2020; Hill et al., 2022). Hence, they can benefit from training programmes that teach best practices for mentees and mentors in high-quality mentoring relationships (Kokt & Dreyer, 2024; Ragins, 2016). The MBS could be used in mentee workshops as a tool for teaching and self-assessing effective behaviours. Similarly, mentors could learn how to use the MBS as part of coaching and providing feedback to their mentees. While the MBS may be used as a teaching, assessment, and coaching tool, it is important to note that mentee and mentor assessments should not be used to evaluate individuals (Giancola et al., 2016). These assessments should be used strictly for self-monitoring, partnership growth, and professional development. The process and outcomes of formal programmes can be evaluated with other methods. While leaders/facilitators/committees may survey individual participants, only aggregate data should be used to evaluate programme-level success factors.
There continues to be a need and desire for mentoring from both employees and employers, but a modified approach is required (Janiak, 2021). The traditional mentoring perspective is one-sided, focusing on what the mentor should do. Relational mentoring provides a comprehensive approach that considers both the mentor’s and mentee’s contributions and outcomes in a high-quality partnership. To this end, the MBS is a tool that can help further contemporary research and practice in workplace mentoring.
Dr. Giancola is a Professor of Leadership and Organisational Development whose consulting experience includes implementing and evaluating workplace mentoring programmes in multiple institutions since 2004.
Mr. Webster is a PhD student in Industrial-Organisational Psychology with research and experience in workplace mentoring, strategic planning, and leadership development.
Mr. Friedrich is a PhD student in Industrial-Organisational Psychology with a background and research experience in occupational health, workplace mentoring, and metascience.
Dr. Burroughs is a Professor of Health Management and Policy whose consulting experience includes analytics, strategy and coaching for healthcare and finance organisations.