Abstract
Mentorship is essential for developing professional competence, enhancing skill acquisition, and supporting identity formation across academic, clinical, and organizational settings. Despite its significance, there is a persistent lack of standardized, psychometrically robust tools for evaluating mentoring effectiveness across disciplines. This paper introduces the Mentoring Quality Index (MQI), a theory driven, modular framework informed by Kram’s mentoring functions, Kirkpatrick’s evaluation model, Working Alliance Theory, and culturally responsive mentoring principles. The MQI comprises 16 key performance indicators across four domains, and incorporates multiperspective triangulation, discrepancy analysis, and structured reporting. Its adaptable design supports varied mentoring contexts through customized weighting, language, and behavioral indicators.
Keywords
Introduction
Mentorship functions as a core developmental mechanism across professional and academic settings, facilitating skill acquisition, career advancement, and the formation of professional identity (Abdelmannan et al., 2025; Ianni et al., 2025). However, as the field of mentorship research has matured, a persistent misalignment has emerged, that is, the rapid expansion of mentoring initiatives has outpaced the development of rigorous, theory-driven methodologies for evaluating their effectiveness. Consequently, existing assessment practices are criticized for being inconsistent, theoretically unsound, and lacking standardized evaluation procedures (Abdelmannan et al., 2025; Ianni et al., 2025). Further, regardless of the calls for psychometrically sound instruments (American Educational Research Association, American Psychological Association, & National Council on Measurement in Education, 2014), existing mentorship assessments often lack documented validity evidence.
This measurement deficit has evolved from a theoretical concern into an operational imperative. Major funding agencies now mandate evidence-based assessments of mentoring quality. The National Institutes of Health (NIH), for example, now requires detailed mentor–trainee assessment plans in grant submissions (effective 2024), compelling institutions to systematically monitor relational dynamics and document measurable changes in mentoring practices (McKlveen, 2024). This requirement marks a significant shift from the provision of mentorship to the empirical demonstration of its impact. In response, contemporary scholarship has intensified efforts to develop and validate instruments capable of capturing the multifaceted nature of mentoring relationships. Researchers have moved beyond satisfaction-based metrics toward more granular constructs. Tools such as the Mentorship Working Alliance Scale assess relational quality and partnership effectiveness (Rogers & Byars-Winston, 2025), while the Mentor Attunement Scale measures a mentor's capacity to meaningfully recognize and respond to a mentee's needs (Pryce & Deane, 2025). Similarly, a systematic review by (Oeben and Grajfoner, 2026) identified five areas influencing coaching effectiveness, including goal alignment and relational bond-constructs directly transferable to mentoring assessment. Additionally, comprehensive evaluation frameworks such as the Context, Input, Process, and Product model offer a structured, multistakeholder approach to assessing program effectiveness (Fatima et al., 2026).
Despite these contributions, a substantial methodological gap persists. Scoping reviews highlight inadequate theoretical grounding in program architectures and widespread reliance on nonvalidated or weakly psychometricized instruments. Many programs struggle to precisely define outcome constructs or align assessments with quantifiable indicators such as competency development, career progression, or leadership capacity. Although mentor-training interventions show positive effects, the mechanisms through which they produce change remain insufficiently understood, rendering the evaluative process a persistent “black box” (Ianni et al., 2025).
To address this methodological shortcoming, this article introduces a quantitative assessment framework developed to advance the rigor and standardization of mentorship evaluation. The framework employs a modular, extensible architecture capable of measuring core constructs across diverse mentoring models while supporting contextual customization. By integrating validated relational and developmental constructs within a coherent evaluation structure, it offers a systematic, scalable, and theoretically grounded mechanism for assessing the full continuum of mentorship processes and outcomes. This article details the conceptual foundations, modular design strategy, and practical applications of the framework, contributing to a more evidence-based and equitable approach to mentoring assessment.
Literature Review
Defining Mentoring and Its Core Functions
Mentoring is broadly defined as a purposeful, collaborative, and dynamic developmental relationship between a more experienced individual (mentor) and a less experienced partner (mentee), characterized by mutual trust and growth (Nuis et al., 2023). Foundational scholarship by Kram (1983) and subsequent work by Ragins and Kram (2007) established an enduring framework for understanding the functions operating within these relationships, distinguishing between two primary forms of support: psychosocial and instrumental. Psychosocial support enhances the mentee's confidence, professional identity, and sense of belonging through role modeling, counseling, and encouragement. Instrumental support, in contrast, advances career development by strengthening professional skills and creating opportunities through coaching, sponsorship, and exposure (Kram, 1983; Ragins & Kram, 2007). This dual-function model has proven remarkably durable, providing a theoretical foundation for decades of subsequent research.
Moreover, Al Hilali et al. (2020) emphasized in their theoretical analysis that coaching and mentoring frameworks foster the development of professional competencies through structured reflective practice, purposeful developmental relationships, and experiential learning processes.
Furthermore, culturally responsive mentorship represents more than a set of techniques for working with diverse populations; it embodies a fundamental reimagining of how relationships support professional growth within contexts of diversity and inequity (Tran et al., 2025). Such an approach moves beyond conventional dyadic support to acknowledge the systemic barriers, including racialized exclusion, institutional neglect, and unclear advancement pathways, that disproportionately affect historically marginalized groups (Iheduru-Anderson et al., 2025; Njelesani et al., 2025). As a mechanism for redressing persistent disparities in educational and professional advancement, culturally responsive mentorship offers benefits that extend to all participants by enriching mentoring practices through the integration of cultural humility, antiracism allyship, and support for navigating workplace discrimination (Njelesani et al., 2025). Empirical evidence demonstrates that structured, culturally focused mentorship programs can enhance cultural safety knowledge, improve professional satisfaction, and foster belonging among underrepresented groups (Fealy et al., 2025; Sagar et al., 2025; Tran et al., 2025). The continued development and implementation of such approaches will therefore be essential for creating truly equitable and inclusive environments where every individual can thrive.
The Expansion of Mentoring Across Disciplinary Contexts
In recent decades, the concepts of coaching and mentoring have expanded well beyond their traditional boundaries. Once confined primarily to education (Nuis et al., 2023) and healthcare (Pfund et al., 2022), their application is now discernible across a diverse range of fields. Empirical investigations have documented the prevalence and impact of mentoring in sports (Bloom et al., 1997, 1998), psychology (Lundgren & Orsillo, 2012), health services (Woolnough & Fielden, 2016), and business administration (Joo et al., 2012). A systematic review by (Janssen et al., 2015) further confirmed that informal mentoring in workplace contexts consistantly predicts career satisfaction, salary progression and organisational committment, though it ofthen repproduces exisiting privilege networks. Across these varied contexts, effective mentoring relationships are consistently linked to critical outcomes. These include fostering clinical competency, shaping professional identity, enhancing psychological well-being (Abdelmannan et al., 2025), and promoting career progression, scholarly productivity, and workforce retention (Abdelmannan et al., 2025; Ianni et al., 2025). The convergence of findings across such diverse settings underscores the fundamental and transferable nature of high-quality mentoring.
Structural Dimensions and Typologies of Mentoring
Beyond its functional and contextual variations, mentoring arrangements can also be classified along primary structural dimensions. Karcher et al. (2006) and Mullen (1994) identify two key axes: participant composition and engagement method. Participant composition refers to the configuration of individuals involved, ranging from the traditional dyad to group or peer-based models. Engagement method describes the mode and setting of interaction, encompassing formal, informal, virtual, and hybrid arrangements. The intersection of these dimensions generates a typology of distinct mentoring models, as presented in Table 1, providing a useful framework for comparing approaches across programs and studies.
Classification of Mentoring Programs by Participation and Engagement.
Furthermore, a core distinction in mentoring is between formal and informal relationships (Mullen, 1994). Early empirical work by (Chao et al., 1992) demonstrated that while both forms provide career and psychosocial support, informal mentoring relationships often yield stronger career outcomes, though formal programmes can be equaly effective when intentionally structured. Formal mentoring involves intentional, structured programs guided by institutional goals. Informal mentoring emerges organically from mutual affinity and voluntary initiation, without external orchestration (Chun & Edwards, 2025). This distinction has significant equity implications. Because informal mentoring depends on existing social networks, it tends to perpetuate homophily, whereby individuals preferentially connect with similar others. This dynamic systematically excludes underrepresented groups from accessing critical developmental relationships.
Mentoring can also be categorized in terms of context or primary focus (Brondyk & Searby, 2013). According to this framework, seven distinct types of mentoring are identified, which are presented in Table 2.
Mentoring Modalities: A Functional Taxonomy.
While effective measurement allows programs to evaluate fidelity, assess outcomes, and target improvements, capturing the nuanced dimensions of mentorship poses significant methodological challenges that researchers must address. Overcoming these challenges through the development of valid and reliable assessment tools therefore constitutes a critical frontier in mentorship research and practice.
The Kirkpatrick model, first introduced in 1979, has become a classic framework in the field of instructional design and training evaluation and remains the most widely used framework for evaluating learning and development interventions (Kirkpatrick, 1996). It assesses program impact across four levels: reaction, learning, behavior, and results. An overview of each level, including its focus, key metrics, and data collection methods, is presented in Table 3.
Summary of the Kirkpatrick Model for Evaluating Training Effectiveness.
The Kirkpatrick model has been extensively applied to evaluate training programs designed for mentors themselves, particularly in educational contexts. Serbanescu and Chircu (2014) found that applying the model to an e-learning platform for mentor-teacher training demonstrated high efficiency, yielding strong learning gains and measurable behavioral and organizational improvements. Further, Smidt et al. (2009) reported that the Kirkpatrick model offers a framework for assessing training programs, helping determine their potential to meet the needs of both the organization and its participating staff.
Similarly, Lin (2011) applied the Kirkpatrick model to evaluate a mentor teacher training program in Taiwan, focusing specifically on the reaction and learning levels. The study, grounded in the framework of the Ministry of Education's Teacher Evaluation for Professional Development initiative, surveyed participants to assess their feedback on the training curriculum and the growth of their mentoring knowledge and skills. Findings revealed that trainees reported positive reactions to the program and demonstrated improvement in their professional abilities as mentor teachers, particularly in the areas of evaluation and instructional supervision. Beyond training mentors, the Kirkpatrick model has been adapted to evaluate the effectiveness of mentorship programs themselves as organizational interventions. Labin (2019) provided a comprehensive framework for applying the four levels to measure mentorship program outcomes, emphasizing the importance of developing metrics aligned with strategic organizational goals. Nsouli et al. (2023) reviewed the literature for the use of the Kirkpatrick model in the assessment of pharmacy students.
Methodology
Despite extensive scholarship establishing mentorship as a key mechanism for professional and academic advancement, the field continues to face a persistent challenge: the lack of standardized, psychometrically robust instruments for assessing mentoring effectiveness. Existing tools remain fragmented across disciplines, are often designed for traditional dyadic structures, and frequently rely on satisfaction-based measures that do not capture the multidimensional complexity of mentoring relationships. Concurrently, increasing accountability demands, such as NIH requirements for Mentor/Trainee Assessment Plans (National Institutes of Health, 2024), underscore the urgent need for evaluation instruments capable of producing credible and actionable evidence of mentoring impact.
To address this gap, the current paper introduces the Mentoring Quality Index (MQI), a structured instrument grounded in established theoretical frameworks, including Kram's dual-function model, contemporary literature on culturally responsive mentorship, and Kirkpatrick's four-level evaluation paradigm. MQI employs a modular and scalable architecture organized around a set of Key Performance Indicators (KPIs) that operationalize mentoring effectiveness across four domains: Relational Foundation, Developmental Support, Cultural Responsiveness, and Measurable Outcomes.
Theoretical Framework
The MQI is underpinned by four interconnected theoretical foundations:
Kirkpatrick's Four-Level Evaluation Model: Kirkpatrick's (1979) model evaluates training across four levels: reaction, learning, behavior, and results. The MQI applies this structure to mentoring, enabling assessment from participant satisfaction to organizational impact and return on investment (Labin, 2019). Kram's Mentoring Functions Framework: The distinction between psychosocial support, such as role modeling, counseling, and affirmation, and instrumental support, including sponsorship, exposure, and coaching, forms the core conceptual framework for understanding mentor behaviors within developmental relationships (Kram, 1983; Ragins & Kram, 2007). The MQI translates these functions into behavioral indicators that capture both the presence and quality of these supportive practices. Working Alliance Theory: Drawing on psychotherapy research, the Mentorship Working Alliance Scale conceptualizes mentoring effectiveness through the relational bond, task agreement, and goal alignment between mentor and mentee. The MQI integrates these dimensions as key predictors of downstream outcomes. Culturally Responsive Mentoring Framework: Recent research emphasizes that effective mentoring in higher education must intentionally consider the cultural identities and lived experiences of both mentors and mentees (Pryce & Deane, 2025). To support policy-relevant evaluation and strengthen institutional accountability, the MQI incorporates indicators that assess cultural attunement, awareness, and adaptive responsiveness. These measures address a long-standing gap in existing assessment tools and enable leaders to more accurately evaluate mentoring quality, monitor equity outcomes, and align mentoring practices with institutional diversity and student-success priorities.
Tool Architecture and KPIs
The MQI is a modular tool containing 16 KPIs across four domains, each rated on a 3-point Likert scale (Likert, 1932) to reduce evaluator bias, with optional behavioral anchors for added reliability. It is administered separately to mentors and mentees to enable multiperspective triangulation.
Domain I: Relational Foundation
This domain evaluates the interpersonal quality and structural coherence of the mentoring relationship. Each KPI is anchored in a distinct theoretical tradition. Trust and Rapport (KPIs 1.1 and 1.2) are informed by Working Alliance Theory, which underscores the relational bond as a foundational determinant of an effective mentoring partnership. Commitment (KPI 1.3) reflects Kram's (1983) framework on mentoring phases and relationship progression. Goal Alignment (KPI 1.4), derived from Goal-Setting Theory, assesses the extent to which mentors and mentees share a clear, mutually agreed-upon understanding of the relationship's aims and intended outcomes. Table 4 depicts the description of these KPIs.
Domain I “Relational Foundation” Key Performance Indicators (KPIs).
Domain II: Developmental Support
Drawing on Kram's dual-function model (1983), this domain assesses the specific supportive behaviors, both psychosocial and instrumental, that characterize the mentoring relationship. Table 5 shows the description of the domain KPIs.
Domain II “Development Support” Key Performance Indicators (KPIs).
Domain III: Cultural Responsiveness
This domain evaluates the mentor's capacity to recognize, respect, and effectively respond to the mentee's cultural identity and lived experiences. Its KPIs are grounded in established theoretical frameworks: Cultural Awareness and Cultural Attunement (KPIs 3.1 and 3.2) draw from the work of Pryce and Deane (2025); Advocacy (KPI 3.3) is informed by the framework proposed by Rogers and Byars-Winston (2025); and Inclusive Environment (KPI 3.4) reflects principles from the broader multicultural mentoring literature. Table 6 describes the domain KPIs.
Domain III “Cultural Responsiveness” Key Performance Indicators (KPIs).
Domain IV: Measurable Outcomes
This domain adapts Kirkpatrick's four-level model to document tangible results emerging from the mentoring relationship. Each KPI corresponds to a distinct Kirkpatrick level: KPI 4.1 (Satisfaction) captures Level 1 (Reaction); KPI 4.2 (Knowledge and Skill Acquisition) assesses Level 2 (Learning); KPI 4.3 (Behavioral Application) measures Level 3 (Behavior); and KPI 4.4 (Career and Organizational Impact) documents Level 4 (Results). Table 7 illustrates the domain KPIs.
Domain IV “Measurable Outcomes” Key Performance Indicators (KPIs).
Multiperspective Administration and Triangulation
The MQI is designed to align with the established standards of reliability and validity. A key methodological strength of the instrument is its structured, multi-informant data collection approach, which gathers parallel assessments from mentors and mentees and, when applicable, incorporates input from program staff or supervisors. This multiperspective structure facilitates several analytic functions. First, discrepancy analysis enables the identification of perceptual differences between mentors and mentees, offering insight into communication gaps or unmet developmental needs. Second, self-report validation is strengthened by triangulating mentee-reported outcomes, such as gains in knowledge or skills, with the corresponding mentor observations, thereby enhancing the robustness of outcome evidence. Third, systematic patterns of discrepancy across respondent groups serve as program-level diagnostic indicators. For example, consistent mentor overestimation of cultural attunement relative to mentee ratings may reveal areas that require targeted training or structural refinement.
Implementation and Scoring Guidelines
Administration Protocol
The MQI assessment protocol should be administered according to a structured, staged schedule. A baseline assessment should be conducted shortly after the initiation of the mentoring relationship, allowing sufficient time for foundational relational dynamics to form. Optional interim pulse checks, consisting of abbreviated subsets of KPIs, may be administered quarterly to support ongoing program monitoring. A comprehensive assessment should be completed annually to capture longitudinal changes in mentoring quality and developmental outcomes. A final administration should occur at the conclusion of the mentoring relationship or at the designated endpoint of the program cycle.
Scoring and Interpretation
Scoring
To reduce the response bias in the evaluation, each of the 16 KPIs across the four domains should be rated using a three-level scale. The response options are “Disagree,” “Neutral,” and “Agree,” with the corresponding weights assigned as follows:
1. Disagree = 1 2. Neutral = 2 3. Agree = 3
Raters should select “Neutral” only when they neither agree nor disagree strongly with a statement. Application of this framework is expected to yield the following scoring outcomes:
The maximum possible score from a single rater, whether mentee or mentor—is 48. The minimum possible score from a single rater is 16. Thus, the scoring range per rater is 32 points (48 − 16). When both raters are considered together, the highest possible combined score is 96 points (48 + 48), and the lowest possible combined score is 32 points (16 + 16). Defining success
The success of the mentoring relationship may be determined by using one of three approaches:
Absolute threshold: a fixed percentage of the maximum combined score is used as the benchmark, for example, 70% of the maximum yields a success threshold of 67.2. Balanced agreement: success is achieved if both the mentee and mentor score above a specified individual threshold. Using the midpoint of 32 as the individual benchmark, a combined score of 64 or higher indicates success. Consensus requirement: both the mentee and mentor must individually score 32 or above, resulting in a combined score of at least 64. This ensures that both parties independently perceive the relationship as successful.
The author strongly recommends adopting the consensus requirement as the most robust method for evaluating the success of the mentoring relationship.
For applications involving different numbers of raters or alternative thresholds, the following formulae can be used to calculate the success mark under any of the three approaches:
Notation
Maximum score per rater = M
Minimum score per rater = m
Number of raters = n
Combined minimum = n × m
Combined maximum = n × M
Range = n × (M − m)
Threshold = T
Success threshold as a percentage = P
Absolute threshold (percentage of maximum) formula: Balanced agreement formula, each rater is required to score ≥T: Consensus requirement: average score per rater ≥
where T could be the midpoint 
Reporting and Visualization
The MQI incorporates a standardized reporting framework that includes individual-level profiles with domain scores and discrepancy analyses, program-level aggregate summaries featuring distributional statistics and benchmarking, longitudinal visualizations that track changes over time, and return on investment calculations linking KPI 4.4 (Career and Organizational Impact) to program expenditures.
Applications Across Mentoring Contexts
The modular architecture of the MQI enables its flexible application across diverse mentoring models and settings. This adaptability allows the framework to accommodate variations in program structure, disciplinary context, and participant needs, while maintaining conceptual rigor and consistency in measurement. By supporting both standardized evaluation and context-specific tailoring, the MQI offers institutions a unified yet versatile tool for assessing mentoring quality and developmental outcomes. Table 8 presents the key considerations for applying the MQI across different mentoring contexts, highlighting how program aims, participant characteristics, disciplinary requirements, and organizational priorities influence the selection and interpretation of relevant indicators. The MQI adapts to each mentoring context by allowing programs to emphasize different domains, weight indicators according to specific developmental goals, and tailor language and behavioral anchors to reflect the cultural, disciplinary, and organizational nuances of the setting.
Applications of the MQI Across Mentoring Contexts.
Note. MQI = Mentoring Quality Index; KPI = key performance indicator; ROI = return on investment.
Conclusion
The MQI offers a comprehensive, theory-driven framework for evaluating mentoring effectiveness across diverse settings. By integrating relational, developmental, cultural, and outcome-oriented dimensions, it addresses key limitations in existing evaluation practices, including the lack of standardized metrics and the limited consideration of cultural responsiveness. Its multi-informant structure enhances the reliability and interpretative depth of assessment data, supporting both individual-level analysis and program-level decision-making. The MQI's modular design enables flexible adaptation while maintaining conceptual rigor, providing a scalable platform for documenting impact and guiding continuous improvement. As accountability expectations intensify, the MQI represents a significant advancement in the systematic measurement of mentoring, strengthening institutional effectiveness and advancing the science and practice of mentoring.
Footnotes
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.