A Commentary on Dhaenens et al. (2025): Centrality as a Conduit - The Critical Role of Recognition and Tie Quality

The Recognition Gap in Structural Advantage

The pivotal theoretical limitation emerges in how Dhaenens et al. (2025) interpret their divergent results. Their result that centrality predicts external marketability (β = 0.22, p < .05, Table 2) but not internal marketability (β = 0.05, p = .52) or career satisfaction (β = 0.05-0.09, p > .05) implies that their model captures prestige signaling but misses relational fulfillment mechanisms. This divergence presents a fundamental theoretical cost: in considering centrality to be the explanatory mechanism, they inadvertently favor visibility-based explanations over experience-based ones, potentially misleading theory and practice about how mentoring actually creates value.

The measurement approach itself contains an important clue. The sociometric question (“who takes an active interest and involvement in your career?”) necessarily embeds recognition into the centrality construct. This creates construct contamination that renders the null internal findings all the more telling: the fact that recognition is at least partially captured while internal outcomes remain non-significant strongly indicates that quality of relationship is the missing mechanism involved. A crucial test would involve examining whether the addition of tie quality measures would then reveal interaction effects for the internal outcome in particular. The theoretical cost in ignoring this is high since it prevents the model from explaining when central positions generate fulfillment versus when they represent mere hollow popularity.

Anchoring the Argument: Tie Quality and the Limits of Eigenvector

The exclusive reliance on eigenvector centrality creates a second theoretical cost by systematically blinding the analysis to relational content. This measure cannot make a distinction between mentors whose centrality is due to deep developmental relationships and others maintaining some level of connection through only superficial interactions. This limitation proves to be particularly damning for mentoring literature because it is the relationship quality characterized by trust and mutual commitment on which benefits hinge (Humberd & Rouse, 2016).

The study’s design inadvertently privileges structural explanations over relational ones. A direct re-analysis of their data could adjudicate between these mechanisms by testing an interaction between centrality and a proxy for relationship intensity, such as the comprehensiveness of mentoring provided (which their 18-item scale captures). Wpredict that centrality’s effect on internal outcomes (satisfaction, internal marketability) will be significant only for those reporting high levels of multifaceted support an indicator of higher-quality, engaged ties. The focal article, which posits centrality as the core driver, predicts no such interaction. If this interpretation were to be supported, it opens up major theoretical implications whereby it could mean that organizations whose focus is solely on situating individuals in central network positions might be able to elevate the visibility of such individuals while compromising the developmental promise of mentoring. Methot et al.’s (2024) work on multiplex relationships illustrates yet again how important it is to distinguish between structural position and relational content, especially nowadays, when the complete mentoring phenomenon occurs in multiple domains.

The Cultural and Spatial Boundaries of Network Value

The theoretical model’s assumption of universal network effects represents a significant limitation, as centrality’s value is conditioned by cultural and spatial contexts through identifiable micro-mechanisms. In cultures where involvement in mentoring is a normative obligation, the attributional value of provision of support diminishes with less likelihood of observers’ attribution of discretionary intent which reduces recognition returns on centrality (Hofstede, 2001). This contrasts with contexts where mentoring is discretionary, making each act a stronger signal of initiative. Spatially, hybrid work introduces a recognition density problem. Remote interactions generate fewer observational witnesses than colocated mentoring, attenuating the conversion of mentoring effort into recognized centrality. Evidence against our account would require demonstrating that the proposed micro-mechanisms are inconsequential. Specifically, evidence against our claims would emerge if: (1) empirical findings showing that the benefits of network centrality are consistent across cultural contexts, indicating that normative expectations (e.g., mentoring as obligation versus discretion) do not alter the value of centrality as we propose, contrary to established principles of cross-cultural organizational behavior (Gelfand et al., 2007) or (2) research results indicating that remote or hybrid work arrangements enhance, or do not diminish, the recognition derived from mentoring interactions, contradicting our premise of reduced “recognition density.” Absence of such findings justifies the necessity for these boundary conditions as they are not minor qualifiers, rather are fundamental components for doing accurate predictions on when and how centrality translates into career value and must be incorporated into any such explicitly stated comprehensive model of mentoring networks.

Practical Imperatives: Fostering Visibility and Equity

The theoretical neglect of relational quality and recognition in Dhaenens et al.’s (2025) framework carries direct practical cost. A narrow focus on fostering centrality risks incentivizing superficial networking over meaningful development, creating the illusion of mentoring infrastructure without substantive benefit. Organizations should instead set up multidimensional recognition systems, separating connectivity’s qualitative impact on significant development. This entails layered tie-strength assessments through relational surveys, peer-nomination algorithms identifying developmentally influential mentors, and HR systems triangulating network metrics with quality metrics. Crucially, these systems need “equity analytics” to detect and correct the systematic under-recognition of mentoring by women and minorities, whose contributions often remain invisible despite their value. Without this integrated approach, rewarding mere centrality risks reinforcing existing inequities, where demographic patterns in networking behavior translate into unequal advancement (Ragins & Verbos, 2017). The practical imperative is to create mentoring scorecards that reflect true developmental impact ensuring what flows through central positions is high-quality, equitably recognized support, not just structural advantage.

Future Research Agenda

Building on this commentary, we propose three testable propositions to advance research:

(1) The relationship between developmental network centrality and career benefits will be strengthened by perceived mentoring tie quality and organizational recognition practices.

Conclusion

Though Dhaenens et al. (2025) map the structure of mentoring, they leave its social processes unexplored. Treating centrality as the core mechanism carries theoretical costs: it cannot explain outcome divergences, obscures cultural and spatial boundaries, and risks cementing inequities. By recasting centrality as a conduit activated by relational quality and recognition, we provide a more complete framework. This temporal gain hinges on differentiating structural opportunity from relational realization. Centrality means potential, but getting the benefits from it necessitates good relationship quality and organizational systems that value development. This paradigm shift moves beyond “who is central?” to “what makes centrality meaningful?” The answer determines whether networks create genuine development or merely reproduce visibility hierarchies. This perspective preserves the focal study’s structural insights while adding the theoretical sophistication comprehend how really mentoring works or does not work in practice to guide research and organizational intervention toward more equitable and effective outcomes.