manager; the present findings provide quantitative support for this claim within a mining context. In game-theoretic terms, supervisors function as reducers of perceived critical mass: their early adoption of cooperative strategies signals to workers that the social cost of change adoption is decreasing (Keser et al., 2012). Groups where supervisors actively promoted change (Groups 1 and 4) achieved cooperative equilibria; the group where supervisory engagement was absent remained trapped in suboptimal mutual-resistance equilibrium. The attitudinal rather than aptitudinal nature of the productivity gap resonates with Herscovitch and Meyer’s (2002) taxonomy of organizational commitment. Workers exhibiting high discipline (M=4.25) but low commitment (M=3.08) display a pattern consistent with continuance commitment—remaining because of perceived costs of leaving rather than affective attachment. This distinction has design implications: adaptive learning systems must address not only knowledge gaps but motivational orientation, targeting the transition from resigned compliance to engaged participation. Klein and Sorra’s (1996) implementation framework suggests this requires simultaneous attention to implementation climate (organizational support) and innovation-values fit (worker perception of benefit). 5.2 Barriers and Facilitators The qualitative analysis revealed an organizational environment fundamentally oriented toward change, with perceived facilitators exceeding barriers at a 4.4:1 ratio and training mentioned 57 times as the primary facilitator. This cooperative substrate (4:1 cooperation ratio in social interaction strategies) constitutes a latent organizational asset—the workforce is not resisting change; it is waiting for structured support to enact it. Three coordination barriers, however, explain why this cooperative potential remains unrealized. First, the “who moves first” dilemma: workers wait for management signals while management waits for worker buy-in, producing mutual inaction despite shared interest in change. Second, trust deficit from historical broken promises: past organizational commitments perceived as unfulfilled have eroded credibility, increasing the perceived risk of early cooperation. Third, early-adopter social costs: workers who cooperate with change before peers do so risk social isolation within their work group—a cost particularly salient in tight-knit mining crews. These three dilemmas explain why uniform training interventions have historically produced heterogeneous outcomes across work groups, and why the identification of multiple Nash equilibria across groups is not merely a theoretical exercise but a practical diagnostic for intervention design. Methodologically, this study addresses what MacMahon et al. (2024) termed the “streetlamp effect” in mining workforce development—the tendency to focus on visible formal training while ignoring contextual factors that determine learning effectiveness. By conducting cultural diagnosis before designing the adaptive system, this approach illuminates the full landscape of worker readiness rather than searching only where it is convenient to look. The productivity model captures Tynjälä’s (2013) presage factors—learner characteristics and organizational context—before designing process interventions, operationalizing the integration gap identified by Lund et al. (2024). 225
RkJQdWJsaXNoZXIy MTM0Mzk2