Humberto M Carvalho, Caio G Miguel, Carlos E Gonçalves
The simulation confirms that, under specified assumptions, the framework successfully recovers embedded theoretical patterns: accelerated growth during scaffolding, maintained gains after withdrawal, successful transfer, differential ZPD width effects, and progression from external to internal regulation. We report this recovery result from a single simulated draw, and separately report whether the same modelling pipeline can discriminate this internalisation pattern from a simulated "compliance" alternative in which scaffolding gains do not persist. A full simulation-based calibration exercise across many replications would strengthen the recovery claim further but is disproportionate to a methodological proof-of-concept and is left for future methodological work.
INTRODUCTION: Sports coaching research remains polarised between direct instruction approaches emphasising feedback and repetition vs. constraints-led approaches favouring athlete discovery through task design. This debate persists partly because existing methods cannot distinguish whether athletes are genuinely internalising tactical understanding or merely following external instructions. We present a methodological simulation study proposing a framework that operationalises Cultural-Historical Activity Theory-particularly Vygotsky's Zone of Proximal Development (ZPD)-through Bayesian multilevel modelling and verbal protocol analysis to track the transition from coach-dependent to self-directed tactical thinking in youth athletes. No human participants were involved, as this study uses exclusively simulated data.
METHODS: We outline a research protocol integrating: (1) a four-phase longitudinal design (baseline → scaffolded coaching → independent test → transfer challenge), (2) verbal protocol instruments assessing spatial awareness, decision-making reasoning, and metacognitive reflection, (3) linguistic coding distinguishing other-regulation from self-regulation, and (4) Bayesian multilevel models capturing individual developmental heterogeneity. Using simulated data ( n = 80 athletes, 25 sessions), we demonstrate-not empirically test-that the framework's statistical model recovers theoretically-predicted developmental patterns embedded in the data-generating process. We additionally simulate a "compliance" data-generating process, in which scaffolding gains do not persist after withdrawal, to test whether the framework can discriminate internalisation from mere compliance.
RESULTS: The simulation confirms that, under specified assumptions, the framework successfully recovers embedded theoretical patterns: accelerated growth during scaffolding, maintained gains after withdrawal, successful transfer, differential ZPD width effects, and progression from external to internal regulation. We report this recovery result from a single simulated draw, and separately report whether the same modelling pipeline can discriminate this internalisation pattern from a simulated "compliance" alternative in which scaffolding gains do not persist. A full simulation-based calibration exercise across many replications would strengthen the recovery claim further but is disproportionate to a methodological proof-of-concept and is left for future methodological work.
DISCUSSION: We provide researchers with a complete methodological package-study design templates, scoring rubrics, coding schemes, statistical specifications, and reproducible code-enabling rigorous empirical tests of Activity Theory in sport. This proof-of-concept bridges Vygotskian theory with computational methods and establishes that the proposed framework has adequate sensitivity to detect theoretically-predicted patterns when they are present in real data.