RESEARCH DOSSIER
Coaching Leadership, Motivational Climate, and Athlete–Coach Relationships
Abstract. Coaching leadership, motivational climate, and athlete–coach relationships form an interconnected psychological system through which sport instruction, evaluation, social influence, and interpersonal experience affect athletic development. This fundamental research monograph reconstructs the emergence of multidimensional sport leadership, achievement-goal approaches to motivational climate, self-determination accounts of psychological need support, and relational models of closeness, commitment, and complementarity. It integrates original scale-development studies, correlational and longitudinal investigations, coach-education experiments, contemporary systematic syntheses, and international safeguarding guidance. The central scientific conclusion is that leadership behavior, athletes’ interpretation of the motivational environment, and the quality of the dyadic relationship are distinguishable but reciprocally connected research objects. Coach behavior can be experimentally modified, although the magnitude, persistence, and transfer of intervention effects vary by behavior, setting, and outcome. Measures of motivational climate and relationship quality have useful evidence of construct validity alongside significant common-method, dyadic, cultural, and causal-inference limitations. The review identifies methodological priorities for multilevel and dyadic designs, independent behavioral observation, developmental differentiation, safeguarding evaluation, and digitally mediated coaching. The method is a critical narrative synthesis of verifiable scientific publications; no new experiment, preregistered systematic review, or independent meta-analysis is claimed.
Keywords: coaching leadership; sport psychology; motivational climate; coach–athlete relationship; achievement goal theory; mastery climate; ego-involving climate; self-determination theory; basic psychological needs; autonomy support; controlling coaching; closeness; commitment; complementarity; sport leadership scale; psychological safety; youth sport; coach education; athlete well-being; safeguarding; dyadic research; Artificial Era.
Introduction and Scientific Object
The coach is simultaneously an instructor of skilled action, an evaluator of progress, an organizer of collective practice, an institutional authority, and a recurring interpersonal partner. Those functions create a distinctive psychological environment. An athlete interprets technical correction as information about competence, interprets selection as an indication of standing, encounters training demands as either intelligible or coercive, and develops expectations about whether the coach will provide help after a mistake. Over seasons, these episodes accumulate into perceived motivational climate and relationship quality. The resulting scientific problem concerns how observable leadership conduct, subjective social interpretation, and repeated dyadic exchanges combine to organize motivation, learning, welfare, and performance.
Coaching leadership denotes patterned influence over the objectives, structure, instruction, coordination, and evaluation of sport participation. Motivational climate denotes the athlete-perceived or independently observed meaning structure of success, failure, improvement, status, choice, and belonging in a training or competition environment. The athlete–coach relationship denotes the continuing interpersonal relation between a particular athlete and a particular coach, including affective connection, intentions to maintain the relationship, cooperative interaction, mutual representations, and disagreement. These definitions refer to distinct levels of analysis. Leadership may be estimated from behavior; climate from communicated criteria and interpreted affordances; relationship quality from the paired individuals’ experience of one another. A supportive team climate may coexist with a difficult particular dyad, and an otherwise close dyad may operate within an institution whose selection rules encourage rivalry or silence.
The monograph addresses five scientific questions. How are leadership, climate, and relational quality defined and measured? Which explanatory mechanisms connect behavior with autonomy, competence, relatedness, self-evaluation, and persistence? What do original interventions establish about changeable coaching behavior and downstream athlete outcomes? Which claims arise chiefly from same-source associations, and which have stronger causal support? Finally, how do age, gender, disability, culture, organizational power, competitive level, and digital technologies change the configuration of these processes? The scope is organized sport, including youth, collegiate, recreational, and elite participation, with attention to both individual and team settings. Clinical treatment of mental disorders and detailed team-cohesion modeling remain neighboring subjects rather than substitutes for this relational and motivational inquiry.
The evidence method is an explicitly selective, critical narrative synthesis. Foundational theories are reconstructed from their original formulations and early operationalizations; primary studies are assessed according to design; recent syntheses are used to evaluate convergence; and 2024 safeguarding guidance situates coach influence within institutional responsibility. A source demonstrating that two questionnaires correlate cannot independently establish a causal pathway. Randomized allocation of coaches to a workshop strengthens inference about coaching behaviors measured after training, but does not automatically establish an effect on injury, performance, or lifetime retention. This hierarchy governs the interpretation of every subsequent chapter.
Historical Formation of Sport Coaching Research
The scientific investigation of sport leadership emerged from organizational psychology, group processes, coaching observation, and educational research. Coaches had long been judged by competitive results and instructional reputation, but such judgments confounded the behavior of the leader with athlete selection, resources, prior competence, and competitive opposition. Research capable of testing leadership required explicit descriptions of conduct and outcomes, together with an account of the circumstances under which a particular style might be effective. The multidimensional leadership model advanced by Chelladurai, P (1980) offered this kind of framework. Its central proposition distinguished behavior demanded by the situation, behavior preferred by participants, and the leader’s actual behavior; congruence was proposed to bear on satisfaction and group functioning. The model turned the question from identifying a universally superior personal style to examining the fit among athlete characteristics, organizational conditions, and coaching action.
Chelladurai, P., & Saleh, S. D (1980) translated dimensions of leader conduct into the Leadership Scale for Sports. Its factors—training and instruction, democratic behavior, autocratic behavior, social support, and positive feedback—brought sport-specific behaviors into a measurable system. The original development reported three samples totaling 485 participants and selected a 40-item, five-factor solution. Some internal consistency estimates were comparatively low, making the publication a foundational instrument-development achievement rather than a permanent guarantee of reliability under every population and version of the scale. The distinction between preferred, perceived, and actual leadership remains crucial when athletes and coaches disagree about what support has been provided.
A second research tradition developed around achievement goals and the environments through which competence is defined. The analysis of mastery-focused educational settings in Ames, C (1992) supplied a vocabulary subsequently applied to coaching: a participant may interpret success as learning, effort, and improvement, or as proving superiority in social comparison. Sport provided a demanding environment in which judgments of success are simultaneously objective, socially ranked, and personally meaningful. Task-involving and ego-involving climates became operational targets; the Newton, M., Duda, J. L., & Yin, Z (2000), helped formalize their assessment.
A third line emerged from motivational psychology. Self-determination theory linked the social organization of action with autonomy, competence, and relatedness, shifting attention from how much motivation athletes express to the quality of its regulation. Mageau, G. A., & Vallerand, R. J (2003) developed a motivational model of coach conduct emphasizing autonomy support, structure, and involvement. This model interpreted motivational behavior as a process influenced by coach beliefs, institutional pressures, and the coach’s reading of athletes. Concurrently, research on the relationship itself expanded beyond generic social support. Jowett and Cockerill (2003)00011-0) explored the significance of relational closeness and coordination in interviews with Olympic medalists; Jowett, S., & Ntoumanis, N (2004) advanced a questionnaire formalizing closeness, commitment, and complementarity.
These traditions have distinctive intellectual histories and operational measures. Their consolidation creates the central contemporary research agenda: leadership describes patterns of conduct, motivational climate describes the social meaning of competence and control, and relational research describes what a particular pair becomes through ongoing interactions. Modern interventions attempt to modify the first and evaluate changes in the second and third; the causal chain must be empirically tested rather than inferred from theoretical compatibility.
Leadership as Multidimensional Behavior
The multidimensional model treats leadership effectiveness as conditional on context, athlete needs, and the alignment between expected and observed behavior. Training and instruction include decisions about demonstration, practice organization, error correction, goal sequencing, and feedback specificity. Democratic practices include eliciting input about achievable choices; autocratic practices centralize decisions; social support includes attention to personal difficulties and team belonging; and positive feedback reinforces progress and effort. The dimensions remain analytically independent: an autocratic decision about a safety-critical drill does not mean the coach lacks instructional competence, while a friendly manner does not ensure adequate technique teaching. The empirical question is whether a specific behavior serves the task and athlete in the relevant circumstances.
Leadership behavior also operates at several time scales. A coach provides momentary technical cues, structures daily training, allocates opportunity across a season, and establishes expectations concerning selection, recovery, and communication. A behavior that appears neutral in one encounter may acquire a different meaning through repetition. For instance, an isolated public correction could be interpreted as ordinary instruction, whereas public correction selectively directed at disfavored athletes may communicate hierarchy and humiliation. Researchers therefore need observation windows long enough to detect stable patterns and variations surrounding wins, losses, mistakes, and injury.
Operational measures should separate coach self-description, athletes’ perceptions, third-party observation, and institutional consequences. The Leadership Scale for Sports is useful for studying athlete preferences and perceived leadership, but its components cannot capture every form of modern coaching governance, nor can questionnaire agreement demonstrate successful performance causally. Studies of adolescent soccer players, including Vella, S. A., Oades, L. G., & Crowe, T. P (2013), illustrate the relevance of leadership and relationship variables to positive developmental experiences. Such research extends the outcome domain beyond win–loss statistics, but observed associations are still susceptible to team selection, reporting style, and the preexisting success of the group.
Leadership models are most useful when they specify moderators. Athlete age, competence, autonomy expectations, current risk, training stage, and team norms may alter the meaning of directiveness or delegated choice. The appropriate design tests a behavior by context interaction, ideally observing change over time. The same model becomes ethically incomplete if it ignores organizational power. In high-stakes systems, a coach may influence scholarships, selection, eligibility, access to care, and future opportunities. Those dependencies shape whether athletes feel able to dissent, report harm, or leave a relationship.
Achievement Goals and Motivational Climate
Achievement goal theory describes how definitions of competence shape achievement behavior. In a task-involving climate, improvement, skill learning, sustained effort, and constructive cooperation are made salient. In an ego-involving climate, normative rank, unequal recognition, comparison, and the visibility of mistakes are prominent. These climate constructs characterize enacted or interpreted features of an environment. They should be distinguished from an athlete’s personal task or ego orientation, which may vary independently of the local coach. A highly comparison-oriented athlete can encounter a mastery-focused coach, and an athlete who values learning can train within a publicly ranking institution.
A task-involving environment may be produced through training design that allows athletes to practice at differentiated levels, feedback focused on controllable strategies, equitable access to instruction, and recognition of improvement. An ego-involving environment may be communicated by selective praise for superior performers, punishment for visible errors, exclusion from instruction, or status signaling that makes peer rank the principal source of worth. Competitive sport inevitably ranks outcomes; the scientific distinction concerns the evaluation practices used to shape competence and belonging throughout participation. Mastery-focused coaching can coexist with rigorous performance standards. Its defining feature is that athletes retain intelligible routes to progress and feedback even when they lose.
Newton, M., Duda, J. L., & Yin, Z (2000) tested revised motivational-climate measurement in female athletes, demonstrating the importance of assessing factor structure rather than treating climate as a generic adjective. Ntoumanis, N (2001) examined links between achievement-goal orientation and motivational regulation in a British university sample. Task orientation was associated with relatively self-determined motives, whereas ego orientation was associated with less self-determined regulation; the authors explicitly noted that explained variance was often modest. The work gives empirical support for theory-relevant associations without making them deterministically predictive of any individual athlete.
Perceived climate is also partly a relational judgment. Olympiou, A., Jowett, S., & Duda, J. L (2008) assessed 591 team-sport athletes and found that task-involving climate perceptions were associated with relational closeness, commitment, and complementarity. The study directly connects domains often measured separately. Its questionnaire-based and cross-sectional design leaves the temporal direction unresolved: climate could contribute to relationship quality, a good relationship could improve interpretations of climate, or stable team characteristics could influence both.
Self-Determination and Basic Psychological Needs
Self-determination theory distinguishes motivation organized around personally endorsed reasons and interest from motivation driven primarily by pressure, threats, surveillance, or contingent approval. Autonomy concerns experienced volition, competence concerns effectiveness in interaction with tasks, and relatedness concerns connection with significant others. These needs represent theoretically distinct pathways through which coaching environments can shape motivated engagement. In Mageau, G. A., & Vallerand, R. J (2003) model, autonomy-supportive practices, appropriate structure, and interpersonal involvement create conditions relevant to all three. The model is a framework for testable mechanisms; it is not itself an experimental demonstration that every link holds in all sporting settings.
Autonomy support includes explaining the rationale for demanding practice, taking athlete perspectives seriously, offering meaningful options within safety and task constraints, and avoiding language that turns every mistake into a threat to identity. Structure includes clear expectations, rational task progression, meaningful corrective feedback, and adequate coaching expertise. These dimensions reinforce one another when choice is accompanied by an understandable path to competence. Permissiveness without instruction can be experienced as abandonment; careful instruction delivered by intimidation can undermine volition and psychological safety. Relatedness support appears in dependable attention, recognition, and connection across both success and failure.
Reinboth, M., Duda, J. L., & Ntoumanis, N (2004) analyzed reports from 265 British adolescent soccer and cricket players. Their structural modeling associated autonomy support, mastery focus, and social support with different dimensions of need satisfaction and athlete welfare. This study offers a coherent mediational account but relied on nonexperimental self-report data; structural equation fit does not identify an intervention effect. Conroy, D. E., & Coatsworth, J. D (2007) studied 165 young swimmers and distinguished coach interest in athletes’ input from praise for autonomous behavior. These are more precise observable strategies than a single global rating of a supportive coach. In a later study of 119 young swimmers, Coatsworth, J. D., & Conroy, D. E (2009) examined sequential associations among autonomy support, need satisfaction, self-perceptions, and indicators of youth development.
Controlling practices are an independently important subject. Bartholomew, K. J., Ntoumanis, N., & Thøgersen-Ntoumani, C (2010) validated the Controlling Coach Behaviors Scale, with dimensions involving controlling rewards, conditional regard, intimidation, and excessive personal control. Such practices can pressure athletes to comply while simultaneously altering their sense of safety, competence, and ownership of goals. Control, directiveness, and legitimate authority require discriminating definitions. A firm safety rule and an act of humiliation are different interpersonal events even when both are expressed in commanding language.
Across these models, the strongest generalizable proposition is that specific patterns of coach behavior are systematically related to athletes’ reports of need satisfaction and motivational quality. Stronger causal claims require interventions with credible control conditions, independent measurement, sufficient follow-up, and attention to clustering within coaches and teams. The distinction between short-term compliance and durable self-regulated engagement is especially important in high-pressure performance settings.
Integrating Climate, Leadership, and Relationship Quality
The three research domains may be organized as interacting levels: coach action; situational and team-level climate; dyadic experience; and individual motivational and developmental outcomes. A provisional causal account begins with the organization of practices, feedback, selection, and interpersonal conduct. These behaviors shape perceived opportunities to improve, exercise agency, and belong; they also communicate whether errors will attract instruction or contempt. Athletes update expectations about the coach and their own position. Over time, those expectations influence disclosure, persistence, help-seeking, attention, and willingness to accept difficult corrective feedback. The system is dynamic because athlete behavior and coach interpretation then feed back into later coaching choices.
This organization clarifies why related measures are not interchangeable. Positive feedback is a leadership behavior, a mastery climate is a pattern of perceived competence-related meaning, and closeness is a property of a particular relationship. If questionnaires use overlapping evaluative language, apparent mediational pathways may reflect shared respondent affect or question wording. A high-quality design therefore measures behavior through independent observation, climate using validated athlete reports or multiple raters, relationship quality from both partners, and outcomes at different time points. It should model athletes nested within coaches and teams and distinguish dyad-level variation from team-wide climate.
This integrative perspective also guards against equating competitive outcomes with leadership merit. Team success is affected by initial skill distribution, financial resources, schedules, opponents, injuries, and stochastic performance. A coach may improve learning and welfare without an immediate change in wins; a winning team may maintain a harmful motivational system. Evidence relating coach behavior to psychological development is therefore necessary on its own terms, with sport performance treated as an important but separately measured endpoint. The next chapters examine the dyadic and institutional mechanisms that determine what coaching actually means to the individual athlete.
The Coach–Athlete Relationship as a Dyadic System
A coach–athlete relationship is a continuing interpersonal system linking two individuals with partly shared and partly unequal goals. Its organization involves histories of cooperation, perceived care, credibility, influence, and the interpretation of one another’s intentions. It is dyadic because neither individual’s report fully describes what the other experiences. A coach may believe that demanding criticism conveys confidence in an athlete’s potential, while the athlete experiences the same exchange as conditional acceptance. Conversely, an athlete may view a coach’s temporary withdrawal as rejection when the coach intends to facilitate independent problem solving. Relationship research makes such asymmetry an empirical question.
Jowett and Cockerill (2003)00011-0) interviewed 12 Olympic medalists, providing qualitative evidence about the significance of relational experiences within elite athletic trajectories. Qualitative accounts reveal how participants understand enduring bonds, loyalty, and support, but retrospective interviews do not establish population prevalence or the causal effect of a specific coaching act. Their distinctive contribution is to define relationship processes from the athlete perspective rather than reduce the relational history to a general leader rating. The athletes’ recollections illuminate how trust may mediate the acceptance of challenging instruction while also revealing the retrospective selectivity of successful careers.
A complete dyadic account separates direct perceptions from meta-perceptions. Direct perception concerns how an athlete regards the coach; meta-perception concerns how the athlete believes the coach regards the athlete. Disagreement between these judgments can become consequential even when neither is objectively correct. Consider an athlete who respects a coach but believes the coach has lost confidence after a mistake. Reduced communication may follow from this belief before the coach has actually changed conduct. The eventual conflict becomes a reciprocal sequence involving inferences, avoidance, and new interactions. Jowett, S (2009) validation work supports the empirical distinction between direct and meta-perspective versions of relationship measurement.
The same dyad can evolve through developmental and institutional transitions. Relationships are initiated through assignment, selection, or invitation, then develop through repeated practice and competition. Changes in role, injury, selection, scholarship dependence, or approaching retirement alter the available choices and the meaning of loyalty. Dyadic quality must therefore be assessed longitudinally; a single survey near the peak of a successful season cannot represent the full relationship. Separate reports from athletes and coaches, confidential collection, observation of interaction, and prospective tracking of ruptures are valuable precisely because a dyad cannot be reduced to one questionnaire respondent.
Closeness, Commitment, and Complementarity
The three-Cs framework organizes relationship quality around closeness, commitment, and complementarity. Closeness refers to affective connection, trust, liking, and respect; commitment refers to the intention and cognitive investment involved in maintaining the relationship; complementarity describes the reciprocal and cooperative fit of interpersonal behaviors. The dimensions are correlated, but their differences matter. A coach and athlete can have high mutual respect while considering a strategic separation; the desire to continue can remain strong in a relationship with limited emotional warmth; and smoothly coordinated instruction can occur within a formally professional relationship. The three-Cs model allows these configurations to be studied without presuming a single uniform relational state.
In two British samples, Jowett, S., & Ntoumanis, N (2004) investigated the Coach–Athlete Relationship Questionnaire (CART-Q) through instrument development, component analysis, and confirmatory modeling. The published findings supported a multidimensional latent structure. This was a measurement study, not an experimental demonstration that higher closeness causes faster skill acquisition. Construct validity is cumulative: factor structure, stability, discriminant correlations, responsiveness to change, and predictions for independently measured outcomes each answer different questions. Jowett, S (2009) further examined the instrument’s nomological network in student athletes, connecting its dimensions with coach support, relationship depth, and conflict.
The same interpersonal relation can be inspected from both ends. Athlete and coach reports can be combined into distinguishable patterns of agreement and disagreement, including high–high, low–low, and asymmetric appraisals. Dyadic analysis can test whether an individual’s perception has effects beyond their partner’s appraisal and whether alignment itself predicts subsequent quality or stability. Such models require linked and identifiable pairs with confidential data collection, which many conventional sport surveys do not possess. Without true dyad identifiers, the word relationship can conceal a fundamentally individual-level measurement.
Relational trust has both instrumental and moral significance. Instrumentally, trust may enable athletes to disclose injury symptoms, concerns about selection, or confusion about training; morally, it concerns the reliability of the person who holds influence over athletic opportunities. Trust should be operationalized through justified expectations and behavior, not equated with unquestioning loyalty. Commitment is similarly contingent. Enduring dedication is compatible with voluntary partnership when it preserves boundaries and exit options; forced loyalty can be a mechanism of vulnerability. The relationship literature becomes stronger when it includes dissent, conflict management, autonomy, and the right to change coaches alongside measures of closeness.
Power, Authority, and Safeguarding
Coaching is an asymmetric relation because the coach often determines training volume, competition selection, evaluation, and access to advancement. In youth sport, age and dependence deepen that asymmetry; in elite sport, sponsorship, reputation, and national-team eligibility may make departure costly. A psychological theory of coaching must examine the consequences of these power differences for voice, consent, protection, and the credibility of negative reports. A coach’s ability to produce immediate compliance is a poor proxy for adaptive influence when the athlete cannot safely refuse.
The distinction between controlling and supportive coaching provides part of the empirical foundation. Bartholomew, K. J., Ntoumanis, N., & Thøgersen-Ntoumani, C (2010) identify controlling rewards, conditional regard, intimidation, and excessive personal control as measurable facets. Conditional regard arises when warmth or approval becomes a reward for obedience or successful outcomes. Intimidation includes threatening conduct that can make reporting, exploration, and error correction psychologically costly. Such practices should be studied with independent behavior measures as well as athlete reports because participants may normalize harmful conduct within a prevailing sport culture.
The Tuakli-Wosornu, Y. A., et al (2024) provides a broad socioecological account of harm. An expert group synthesized evidence and athlete perspectives through a structured process, framing prevention and response as duties shared by individuals, organizations, governing bodies, and the wider system. The consensus distinguishes interpersonal forms of violence and examines contextual conditions that maintain them. This is an expert consensus informed by research and a scoping review; it is not a randomized estimate of the effect of any one safeguarding measure. Its importance for coaching science lies in making organizational accountability an explicit part of interpersonal safety.
The ethical boundaries of leadership cannot depend solely on athletes’ reports of motivation. An intimidated athlete may perform well in the short term, and a successful competitor may report loyalty to a coach whose conduct remains harmful. Safeguarding therefore requires independent evaluation of unacceptable behavior, confidential reporting routes, protection against retaliation, and appropriate referral pathways. Positive relational scores should not be treated as proof that safeguarding arrangements work. Researchers should examine whether prevention policies are implemented, whether disclosures receive timely and fair response, and whether vulnerable athletes can access support without sacrificing selection opportunities.
Autonomy-supportive coaching is compatible with clear boundaries, technical rigor, and nonnegotiable safety practices. The scientifically meaningful distinction is between reasons and procedures that help athletes understand constraints and practices that weaponize dependence. Transparency about criteria, procedural fairness, opportunities to question decisions, and explicit boundaries for one-to-one contact can support both effective instruction and athlete protection. All such measures must be examined in context rather than extrapolated directly from small self-report studies. The neighboring dossier on safeguarding retains the broader institutional and prevalence literature; the present monograph concentrates on how authority changes the meaning and effects of the coach–athlete relationship.
Coach Communication and Feedback
The coach–athlete relationship is enacted through communication. Feedback selects information from athletic performance, expresses evaluation, and shapes what athletes attend to during subsequent practice. Instruction can specify a movement correction, explain the rationale for an adjustment, recognize improvement, or compare an athlete to peers. The same sentence may function as usable performance information or as an evaluation of personal worth, depending on its timing, audience, prior history, and the credibility of the coach. The difference is central to motivational climate because athletes learn from repeated interactions which kinds of effort, mistakes, and progress receive recognition.
Constructive feedback combines informational precision with conditions that preserve agency and dignity. Technical correction is more actionable when it identifies a controllable change, specifies what successful execution would look like, and allows the athlete to test the instruction. Process praise recognizes a specific strategy or improvement, whereas global praise may encourage dependence on approval if its criteria remain opaque. In Conroy, D. E., & Coatsworth, J. D (2007) autonomy-supportive coaching assessment, the two identified components—interest in athlete input and praise for autonomous behavior—illustrate this distinction at the level of observable coaching strategies. The findings support the measurement of such strategies and their association with need satisfaction; they do not supply universal prescriptions for feedback frequency or wording.
Communication is reciprocal. Athletes may withhold symptoms, confusion, and dissatisfaction when their expectations of the coach are negative or when selection incentives punish disclosure. Coaches may misinterpret silence as agreement and respond with even more directive communication. A research design that only measures coach statements cannot reveal whether instructions were understood or whether an athlete felt safe requesting clarification. Recording both sides of decision conversations, together with observed training conditions, can expose misalignment between intended and experienced support.
Team settings create additional complications. Feedback addressed to one athlete is witnessed by teammates, thereby affecting perceived fairness and the climate of the whole group. Unequal allocation of attention may reflect legitimate individualized development, unspoken favoritism, or limited time; researchers should record the rationale and distribution of instruction before labeling it. Coaches also communicate through practices rather than statements: playing time, responsibility, recovery decisions, the willingness to revise plans, and treatment after errors can contradict declared values. This is one reason motivational climate cannot be adequately evaluated from a coach’s self-report of philosophy alone.
Communication technologies add a new layer. Messages, video analysis, performance dashboards, and remote feedback may increase the traceability and availability of information while also increasing surveillance and demands for continual responsiveness. A system that communicates scores constantly might improve specific technical learning while making every action feel evaluated. Questions about perceived agency, privacy, and relationship quality remain empirical, and direct evidence from the traditional face-to-face coaching literature should not be presented as proof of equivalent effects in digitally mediated training.
Coach Education and Intervention Research
Intervention research provides a stronger test of the claim that coaching behavior can be changed. A typical coach education intervention translates motivational and interpersonal theory into workshops, example scenarios, reflection, modeled feedback, and practice plans. Success may be evaluated at several steps: whether coaches learned the content, whether they enacted new behavior, whether athletes perceived the change, whether psychological needs or relationship quality changed, and whether participation, well-being, or performance subsequently improved. Each link requires a separate operational outcome and a suitable interval of measurement.
Randomized assignment of coaches or teams reduces some confounding between training and outcomes, but cluster-level allocation matters. Athletes supervised by the same coach share an intervention exposure and social environment; analyzing them as independent observations can overstate precision. Implementation fidelity, coach attendance, changes in staff, and contamination between groups must also be recorded. Strong claims about enduring transformation require follow-up beyond the immediate workshop period and evidence that observed behavioral changes survive the pressures of competition and organizational demands.
Interventions can target supportive behaviors, disempowering behaviors, or both. These are not simply two ends of a single scale. Reducing intimidation or punitive comparison may follow a different temporal course from learning to elicit athlete input or provide coherent explanatory feedback. Independent behavioral observation is especially valuable when coach education teaches the socially desirable language appearing in later self-report questionnaires. Video coding by trained raters can help identify enacted practices, but the resulting data remain sensitive to observation windows, coder expectations, and the presence of cameras.
A broader outcome framework is needed because training effectiveness, athlete satisfaction, and sporting achievement can diverge. A successful intervention might reduce competitive anxiety without changing a team’s record; it might change observed coach behavior without enough time to change stable relationship quality. The scientifically appropriate inference is domain-specific. Coach education effectiveness is therefore best expressed as a matrix of outcomes, measurement methods, timing, and uncertainty, rather than a single global judgment that a coaching program works.
The Youth Baseball Coach-Training Experiment
An early intervention landmark is Smoll, F. L., Smith, R. E., Barnett, N. P., & Everett, J. J (1993). In youth baseball, eight coaches attended a preseason social support and instructional-effectiveness workshop while ten comparison coaches received no treatment. The research involved 152 boys assessed before and after the season. Players perceived trained coaches as more supportive in theoretically expected ways; they evaluated those coaches more favorably, reported greater enjoyment, and showed stronger team attraction. Win–loss records did not differ across groups. Among boys with relatively low initial self-esteem, the intervention was associated with increased general self-esteem, whereas the corresponding comparison subgroup did not show the same increase.
The experiment is historically important because it changed the coaching environment rather than merely describing existing coaching styles. Its result demonstrates that psychosocial gains can be observed without competitive advantage, an essential challenge to judging every coaching practice by wins. It also highlights heterogeneous treatment effects: participants entering sport with lower self-esteem may have more to gain from a supportive interpersonal environment than those already secure in self-evaluation. Such subgroup findings call for cautious interpretation, especially where sample sizes are modest and interactions require adequate power.
Its limitations matter for modern generalization. The intervention and control conditions contained small numbers of coaches, and the sample reflected a specific period, sport, age range, and gender composition. A contemporary replication would need cluster-appropriate analysis, allocation transparency, intervention fidelity measures, validated multidimensional outcomes, and observation of whether effects persist across seasons. The central lasting contribution is the feasibility of changing coach behavior and the value of independently measuring developmental outcomes.
Motivational Climate Interventions and Anxiety
Performance anxiety is a relevant downstream outcome because evaluative threat is shaped by how coaches respond to errors, recognize participants, and frame competition. An athlete who regards mistakes as occasions for learning may interpret an approaching event differently from an athlete who expects public humiliation or loss of belonging. Achievement goal theory therefore provides a plausible rationale for training coaches to emphasize mastery, individual improvement, and constructive evaluation.
Smith, R. E., Smoll, F. L., & Cumming, S. P (2007) investigated a coach-focused motivational-climate intervention and young athletes’ sport performance anxiety. A related systemic intervention study by Smoll, F. L., Smith, R. E., & Cumming, S. P (2007) involved coach and parent workshops in community basketball and reported lower cognitive and somatic anxiety in the intervention condition relative to a matched comparison. Such evidence supports the feasibility of shifting the psychosocial climate surrounding youth competition. The studies differ in which adults received training and how comparisons were organized; their effects should not be combined as though they represented identical experiments.
Anxiety measures are particularly sensitive to baseline differences, the timing of approaching competitions, and the distinction between worry and somatic arousal. Repeated assessments taken only after wins or losses may confound climate with recent outcomes. A well-designed intervention measures athletes before exposure, specifies its comparison condition, includes meaningful outcome follow-up, and considers the coach or team as the unit of exposure. It should distinguish improvement in athletes’ reported experiences from improvement in independently assessed competition performance.
The practical implications reach beyond simply encouraging positive thinking. Coaches create expectations about how errors will be evaluated; parents and clubs reinforce or contradict those expectations; and athletes differ in temperament and prior experiences. A multilevel intervention that addresses both coaching behavior and organizational incentives may therefore be more durable than a one-time request that coaches use encouraging language. Whether that broader intervention produces lasting changes remains a question to test directly.
Empowering Coaching and the PAPA Program
The Empowering Coaching approach integrates achievement goal theory and self-determination theory to describe supportive and disempowering features of coach-created motivational climates. In Duda, J. L (2013), empowering practices support task involvement, autonomy, and interpersonal connection. Disempowering practices emphasize controlling conduct, social comparison, and thwarted relatedness. This framework is valuable because a coach can display both supportive and harmful practices; a single positive evaluation should not cancel out the presence of intimidation or inconsistent treatment.
A recent direct observational test was published by Tzioumakis, Y. S., Tessier, D., Smith, N., Fabra López, P (2026). Drawing on the Promoting Adolescent Physical Activity project, the investigators randomly assigned 68 grassroots football coaches in England, France, Spain, and Greece to an intervention group (36 coaches) or a comparison group (32 coaches). Their young athletes were approximately 10–15 years old. The intervention comprised about six hours of coach education, delivered through workshops emphasizing the conceptual foundations of empowering coaching, video examples, identification of practices, reflection, and preparation for implementation. Coaching behavior was observed at three points across the season using the Multidimensional Motivational Climate Observation System, reducing dependence on athletes’ global impressions of the coach.
The experiment found a sustained decrease in the observed disempowering climate of intervention coaches, particularly ego-involving practices, compared with controls. Improvements in empowering features, including autonomy and relatedness support, were detectable shortly after intervention and were less persistent by season end. These results matter for both implementation theory and practical interpretation: removing an established harmful practice and acquiring a consistently supportive replacement may require different kinds of learning and reinforcement. The study’s randomization strengthens inference that assignment to training contributed to the observed behavioral differences, but its outcome was the observed coach-created climate, not a demonstrated long-term effect on injuries, lifetime athlete development, or competitive success.
The design has several important qualifications. Only two coaches in the sample were women; nine coaches were lost to follow-up; and the investigators reported that intervention delivery fidelity across countries was not formally assessed. The sample does not warrant an unqualified inference about women’s elite sport, individual sports, para sport, or coaching settings with radically different institutional incentives. A particularly strong future evaluation would replicate the behavioral effect with a broader coach sample, quantify implementation fidelity, and examine whether changes in behavior precede changes in athletes’ motivational regulation, psychological safety, or sustained participation.
A second synthesis extends the motivational mechanism beyond any one trial. Liu, H. J., De Jonge, K. M. M., Den Hartigh, R. J. R., & Van (2026) reviewed basic psychological need support, need satisfaction, and autonomous motivation in coach–athlete relationships. Their meta-analysis reported 102 studies, 339 correlations, 120 independent samples, and 43,675 participants. Reported correlations included coach need support with athlete need satisfaction (r = .47) and athlete need satisfaction with autonomous motivation (r = .37). Those estimates summarize associations across a heterogeneous literature; they do not estimate the causal effect of changing a specific coach practice by one unit. The authors also identified a major structural limitation: the evidence underlying their proposed reciprocal motivational loop was largely individual-level, leaving genuine dyadic feedback in need of direct study.
The intervention and meta-analytic literatures therefore answer complementary questions. The meta-analysis establishes the consistency of theoretical associations over many samples. The randomized observational trial shows that a coach-education program can change identifiable behavior in a specific environment. Neither result alone demonstrates the entire hypothesized causal chain from training through psychological needs to long-run performance. Scientific synthesis requires the distinction to remain visible.
Research Methods and Causal Identification
The field relies on observational surveys, psychometric validation, qualitative interviews, prospective cohorts, intervention trials, and independent behavioral observation. Each approach identifies a different component of the phenomenon. Surveys describe perceptions, covariance, and differences among groups; interviews recover interpretations and contextual histories; observational coding identifies enacted actions; longitudinal designs test temporal ordering; and randomized interventions can estimate the effect of a manipulable program under defined conditions. Evidence quality depends on whether the design matches the claim being made.
A common cross-sectional design asks athletes to rate autonomy support, climate, relational closeness, and well-being during a single session. A positive correlation may reflect a genuine supportive environment, general satisfaction with the season, shared item wording, stable personality differences, or the influence of recent success. Fitting a structural equation model with arrows cannot eliminate these alternatives. Statistical mediation based on simultaneous self-reports is better interpreted as a model of associations. Claims about causal mediation require stronger temporal ordering, adequate covariate adjustment, and ideally experimental manipulation or triangulation with independent behavioral measures.
A causal question should identify its estimand. One possible estimand is the average effect of assigning a coach to an educational workshop on the proportion of training-session minutes independently coded as autonomy supportive. Another is the effect of an intervention on athlete anxiety at season end, measured with baseline adjustment. A third is the effect of coach replacement after harmful conduct on retention over subsequent seasons. Each demands different data and assumptions. Vague outcomes such as coach quality or success are too broad to support reproducible causal estimates.
The nested structure of sport data creates another problem. Athletes are grouped within teams, coaches may train several teams, and athletes can change coaches. Standard errors must reflect such dependence; multilevel models, cluster-robust methods, or appropriate randomization-based inference are necessary when exposure occurs at team or coach level. Spillovers arise because a coach’s new practices affect team norms and peer interactions, so the traditional assumption that one person’s treatment has no effect on another’s outcome can fail. The study design should define whether the unit of inference is the coach, athlete, team, or coach–athlete dyad.
Selection bias is consequential in elite sport. More successful athletes may be assigned to experienced coaches, and coaches judged successful may preferentially recruit highly motivated athletes. Retention may be a collider if athletes with the greatest resources and tolerance for harmful practices remain in a program while others leave. Analyses restricted to surviving elite partnerships can therefore overestimate the benefits or acceptability of coaching styles. Prospective recruitment that includes athletes who discontinue participation, change coaches, or are deselected provides a more complete account of coaching environments.
Reporting should include sampling rules, allocation procedures when used, instrument reliability, attrition, data exclusions, prespecified outcomes, confidence intervals, plausible effect-size interpretation, and conflicts of interest. Where a program contains workshops, coaching materials, and follow-up support, intervention fidelity and the distinct contribution of components should be assessed. A theory-guided intervention may provide a credible developmental rationale, but fidelity, effectiveness, and generalizability remain empirical properties of its implementation.
Psychometric Instruments and Construct Validity
The Leadership Scale for Sports, the Perceived Motivational Climate in Sport Questionnaire-2, the Coach–Athlete Relationship Questionnaire, autonomy-support questionnaires, and controlling-coach measures constitute complementary measurement families. The Chelladurai, P., & Saleh, S. D (1980) describes instructional, democratic, autocratic, social-support, and feedback dimensions of leadership. The Newton, M., Duda, J. L., & Yin, Z (2000) characterizes perceptions of motivational climate. The Jowett, S., & Ntoumanis, N (2004) focuses on closeness, commitment, and complementarity within a dyad. The Conroy, D. E., & Coatsworth, J. D (2007) captures more specific need-supportive strategies, while the Bartholomew, K. J., Ntoumanis, N., & Thøgersen-Ntoumani, C (2010) examines potentially coercive interpersonal practices. Their purpose is not to replace one another.
A psychometric construct must be tied to a measurement theory. Internal consistency indicates how related instrument items are under a model, but says little by itself about accuracy or the absence of response bias. Factorial evidence tests whether proposed latent dimensions are consistent with observed covariation. Measurement invariance asks whether item meanings and score scales are comparable across groups or time points. Predictive validity examines external consequences, while discriminant validity tests whether two named constructs truly provide distinct information. If a questionnaire measures whether the coach feels caring, the response may not establish whether athletes consistently receive fair playing opportunities or experience safety.
The original leadership-scale study reported widely varying internal consistency estimates, illustrating why instrument fidelity cannot be inherited without new testing. Jowett, S., & Ntoumanis, N (2004) established a three-factor measurement foundation, while later validation expanded the network of external correlates. Smith, R. E., Cumming, S. P., & Smoll, F. L (2008) developed the Motivational Climate Scale for Youth Sports for young respondents, with readability suited to children and factor-analytic evidence distinguishing mastery- from ego-involving coaching conduct. Age-appropriate wording is indispensable because adult language about autonomy or status may be understood differently by children.
Construct overlap remains a central risk. Items describing a caring coach may appear in autonomy-support, social-support, and relationship-quality instruments. Strong correlations then partly reflect item content rather than causal linkage between distinct processes. A more discriminating design combines confirmatory factor analysis with multitrait–multimethod data, testing whether behavior ratings from trained observers converge with but remain distinguishable from athlete reports and coach self-ratings. Where English-language measures are translated, semantic and scalar invariance should be tested rather than presumed. Cultural adaptation must preserve the psychological meaning of the construct while allowing the relevant form of supportive behavior to vary by setting.
Sensitivity to change is equally important. An instrument designed to rank athletes by their general perceptions of a coach may be poorly suited to detecting the consequences of one six-hour workshop. Repeated short observation windows can better register changing conduct, while a multi-month relationship inventory may capture cumulative relational quality. Researchers should define the temporal level of the construct before selecting the measure and resist interpreting a stable scale as evidence that no intervention effect occurred in more immediate behavior.
Observation and Longitudinal Assessment
Behavioral observation provides an independent perspective on coach practice. Structured coding can record verbal feedback, athlete choice, opportunities for input, public comparison, responses to errors, expressions of care, and controlling actions. The Multidimensional Motivational Climate Observation System used in Tzioumakis, Y. S., Tessier, D., Smith, N., Fabra López, P (2026) illustrates the value of examining both empowering and disempowering dimensions across successive training sessions. Its use changes the evidence from a judgment about what athletes remember to a coded record of what coaches were observed to do during sampled periods.
Observation has its own measurement limitations. Cameras may alter coach behavior, sampled sessions may differ from high-stakes competitions, coders may disagree about borderline events, and rare but severe misconduct may escape brief sampling periods. Reliability should be assessed at the level of the observed code, with training and calibration procedures described. Blindness to treatment allocation, when possible, reduces expectancy effects. Athletes’ reports remain necessary to establish the interpretation and subjective consequences of the conduct: a video can show that a coach offered a choice but cannot by itself establish that the choice was experienced as meaningful.
Longitudinal analysis examines how coaching environments and relationships change across a season or career. Repeated measurements can distinguish stable between-athlete differences from within-athlete changes. For example, athletes who generally rate coaches positively may also report high well-being, while a decline in perceived coach support for a particular athlete may predict a later change in motivation. Those are different statistical questions, and a correlation pooled across individuals does not answer the second. Longitudinal models are stronger when they include sufficient measurement occasions, plausible lags, and control for stable confounding; they still require caution where unmeasured time-varying events affect both predictors and outcomes.
Dyadic longitudinal models can capture actor and partner effects. A coach’s satisfaction with the relationship may influence future feedback, while an athlete’s satisfaction may influence disclosure and effort; both can evolve in response to injury or competition results. Identifying such feedback requires separately reported data from linked coach–athlete pairs. It also requires special safeguards to protect participant confidentiality, since an athlete must be able to report fear or disagreement without their identifiable response being passed to the coach responsible for selection.
Mixed methods can extend the explanatory reach of quantitative results. Event-contingent diaries document episodes of conflict, praise, or uncertainty close to their occurrence; interviews explain the meaning assigned to those episodes; recorded observations test behavioral concordance. Such triangulation provides a richer test of the proposed causal mechanism than repeatedly administering a global satisfaction scale. For intervention research, independently measured behavioral change should be examined as a candidate mediator only after sufficient evidence of temporal ordering and intervention effects on its relevant components.
Development, Gender, and Cultural Context
Youth sport combines skill acquisition with major developmental processes: identity exploration, peer belonging, expectations of adult authority, growing self-regulation, and decisions about sustained participation. A child’s participation is shaped by parents, schools, coaches, and club rules; their relative influence changes over time. Consequently, leadership arrangements suitable for adult professionals may be unacceptable or developmentally inappropriate when applied to a young athlete. Meaningful autonomy for children commonly involves age-appropriate choices and understandable explanations, combined with safeguards that preserve safety and access to independent adults.
The original youth sport intervention literature has special value because it measures developmental outcomes rather than only performances. Smoll, F. L., Smith, R. E., Barnett, N. P., & Everett, J. J (1993) assessed self-esteem, enjoyment, and team attraction in boys playing baseball; Coatsworth, J. D., & Conroy, D. E (2009) related autonomy-supportive coaching to need satisfaction, self-perceptions, and developmental indicators among youth swimmers. Their conclusions apply most directly to the settings sampled. The broader developmental theory supports extending questions across age groups, but the effects must be evaluated in adolescents, adult recreational athletes, and elite pathways rather than assumed identical.
Gender-related differences are complex because opportunities, coaching structures, cultural expectations, and exposure to harmful conduct vary across sports. An instrument validated primarily in one gender cannot automatically be considered invariant in another. The 2026 PAPA observational trial included only two women coaches, demonstrating the limitations of a frequently cited intervention source for estimating effects among women coaching teams. A more inclusive program would examine women and men coaches, athletes with different gender identities, and interaction patterns affected by organizational representation, maintaining sufficient sample sizes to distinguish measurement from substantive differences.
Para sport and disability require equally substantive adaptation. Autonomy, competence, and relatedness remain relevant research questions, while practical autonomy support may depend on assistive technology, accessibility, classification structures, training environments, and athlete control over accommodations. Coach behavior must be interpreted alongside institutional barriers. Athletes who depend on a coach for travel or care may face distinct constraints on refusing requests or reporting mistreatment; these conditions require careful ethics and appropriate safeguarding rather than simple transfer from studies of able-bodied university teams.
Cross-cultural research should address both language and institutional practice. A democratic coaching item may be understood differently where athletic authority is organized through national teams, schools, clubs, or professional employment. Apparent cultural differences in questionnaire scores may result from translation, response style, sampling, or real institutional variation. Measurement invariance, country-specific contextual data, independent observation, and locally meaningful athlete participation are therefore necessary. The four-country PAPA trial demonstrates the feasibility of coordinated cross-national research but does not exhaust the global cultural range of coaching.
Elite Performance, Teams, and Institutions
Elite sport intensifies selection pressure and makes the coach a gateway to scarce opportunities. Performance reviews may determine contracts, national-team selection, scholarships, and access to specialist services. These conditions can sharpen the motivational value of feedback while making the consequences of relational conflict greater. The scientist must distinguish the technical necessity of demanding preparation from institutional practices that normalize intimidation or concealment of injury. The presence of high performance cannot validate every procedure that preceded it.
Team environments add collective processes. A coach’s treatment of one member signals the basis of fairness and belonging to others. Shared observations can establish team-level climate, while individual athletes may differ substantially in perceived role, trust, and access to feedback. Relationships must therefore be studied at several levels: individual athlete, coach–athlete dyad, training subgroup, team, organization, and governing system. Studies that assess only a team’s average climate can obscure concentrated harm to reserves, new members, or athletes recovering from injury.
Competitive outcomes are especially noisy endpoints. Team wins depend on many determinants outside a coach’s interpersonal practice. Vella, S. A., Oades, L. G., & Crowe, T. P (2013) investigated relations among leadership, relationship quality, team success, and positive development among adolescent soccer players, illustrating the importance of evaluating psychosocial outcomes separately from standings. A rigorous study would measure baseline ability, resources, injury burden, competitive difficulty, and selection decisions before drawing conclusions about the incremental effect of leadership.
The social context can also shape coaches’ own motivation. Financial pressure, unstable employment, professional identity, and parent expectations may reward outcome-focused behavior even when coaches endorse mastery in principle. Mageau, G. A., & Vallerand, R. J (2003) motivational account includes coach orientations and context as potential antecedents of behavior. Liu, H. J., De Jonge, K. M. M., Den Hartigh, R. J. R., & Van (2026) further examined reciprocal associations linking coaches’ need satisfaction and autonomous motivation with the support they provide. The reciprocal interpretation is theoretically attractive, but it calls for linked coach–athlete data and naturalistic longitudinal intervention tests before it can be treated as an established feedback mechanism.
Institutional leadership may amplify or undermine a coach’s interpersonal efforts. An organization that announces a safeguarding policy but rewards coaches only for short-term victory creates conflicting incentives. Transparent selection criteria, staff supervision, independent complaint systems, reasonable workloads, and the separation of psychological care from competitive evaluation are institutional variables deserving measurement. This organizational perspective explains why a training workshop may alter observed practice temporarily yet fail to sustain the effect under unmodified competitive pressures.
Athlete Welfare, Burnout, and Retention
The importance of coaching extends from practice quality to long-term athlete welfare. Participation can provide belonging, mastery, and purpose, while excessive training demands, uncertain selection, and conditional approval can contribute to exhaustion and disengagement. Athlete burnout is typically analyzed through dimensions including physical or emotional exhaustion, reduced sense of accomplishment, and sport devaluation. Its study requires separating demanding training itself from the interpersonal organization of demands, recovery opportunities, and perceived control. A coach can prescribe strenuous work while supporting agency and recovery; another coach can intensify a manageable workload through humiliation, uncertainty, or threats.
Choi, H., Jeong, Y., & Kim, S.-K (2020) investigated associations among perceived autonomy-supportive and controlling coaching, communication, relationship quality, and burnout in 347 Korean collegiate athletes from ten sports. Their modeling found positive links between autonomy-supportive coaching and communication, negative links between controlling coaching and communication, and negative associations of both communication and relationship quality with burnout. Their proposed indirect pathways are consistent with a motivational and relational interpretation. The data remain observational: the study cannot establish that the measured communication variable caused reduced burnout, and negative affect or preexisting burnout may influence how athletes rate their coaches.
Burnout and attrition are different outcomes. Burnout describes a psychological syndrome or pattern of experiences; attrition is a behavioral transition that can follow changing interests, educational demands, injuries, financial circumstances, or deliberate pursuit of other goals. Retention should not automatically be treated as an unquestioned good. An athlete can leave an unsafe sporting environment in a health-protective decision, and athletes can remain in damaging settings when their opportunities are highly constrained. A valid welfare assessment therefore considers voluntary participation, satisfaction, psychological functioning, opportunities to recover, and safe exit alongside continuation rates.
The time structure of welfare matters. Immediate enjoyment after a workshop may differ from stable belonging over a season or a recovery period after injury. Prospective designs should measure workload, sleep, injury, coach exposure, perceived climate, relationship quality, and withdrawal intentions repeatedly. They should account for the possibility that coaches respond differently to athletes who are already exhausted, which creates reciprocal association. Evidence from self-report surveys is valuable for identifying concerns and generating hypotheses, but the strength of a practical recommendation depends on whether changing the coaching environment reliably changes prospective welfare outcomes.
A scientifically sound program evaluates safeguarding and motivational support in parallel. The absence of recorded complaints may result from effective protection, underreporting, dependence on selectors, or weak reporting access. Similarly, a high mean team-satisfaction score may conceal concentrated harm to one subgroup. Confidentiality, attention to distributional effects, access to independent support, and recording of athletes who exit the program expand welfare assessment beyond the experiences of those who remain available for evaluation.
Conflict, Rupture, and Relational Repair
Disagreement is intrinsic to sustained coaching partnerships. Athlete and coach may differ over training goals, competition readiness, feedback style, injury risk, or the allocation of responsibility. The scientific question is how disagreements are interpreted, communicated, and resolved, and whether the relationship permits correction when either participant recognizes a mistake. A rupture can involve a single salient event or the accumulation of small interactions that undermine trust. Its consequences depend on previous experience, available recourse, mutual perceptions, and the institutional power held by each person.
The three-Cs framework provides a way to describe relational change without reducing disagreement to general dissatisfaction. An athlete may continue to respect a coach while losing commitment to the partnership; a coach may remain committed while misinterpreting the athlete's desire for greater independence. Direct and meta-perspective assessments can distinguish what each person feels from what each believes the other feels. Jowett, S (2009) demonstrated the scientific value of evaluating this relational network, although instrument validation does not establish the effectiveness of a specific repair strategy.
Communication practices relevant to repair include clarifying the contested issue, allowing both parties to describe their interpretation, providing reasons for decisions, identifying an agreed behavioral change, and assessing whether the change occurred. These practices are hypotheses about mechanisms that should be evaluated rather than represented as a universally proven conflict-resolution protocol. Sport-specific research has established that communication and relationship scores are associated with welfare outcomes, as in Choi, H., Jeong, Y., & Kim, S.-K (2020), but stronger tests of rupture-repair sequences require direct observation and prospective linked-dyad data.
Power asymmetry introduces an essential boundary. When a disagreement includes coercion, intimidation, discrimination, exploitation, or threats to an athlete's safety, a private reconciliation conversation is insufficient as the central safeguarding response. The Tuakli-Wosornu, Y. A., et al (2024) positions protection and accountability at institutional as well as interpersonal levels. Independent reporting, investigation, and protection from retaliation may be necessary. Researchers should avoid treating every reported harmful episode as a bilateral communication problem, because responsibility and the capacity to leave may be unequally distributed.
A more complete longitudinal design would identify important disagreements at their occurrence, collect independent reports from athlete and coach, record any third-party intervention, and assess subsequent trust, perceived safety, and behavioral cooperation. Such a design could test whether specific repair actions predict restoration of relationship quality and whether repeated failed attempts signal a need for change in coaching arrangements. The field has established useful dimensions for this inquiry; causal evidence for particular repair procedures remains an open research priority.
Digital Coaching and the Artificial Era
Digital tools now mediate some aspects of sport instruction. Video review, performance dashboards, wearable sensors, messaging platforms, remote coaching sessions, and algorithmic recommendations can provide feedback between practices or reshape what coaches observe. Their psychological significance follows from the same research questions that apply to face-to-face coaching: Who controls the performance criteria? Which forms of data are treated as legitimate evidence? Can the athlete question the interpretation? How is uncertainty expressed? Does monitoring increase competence and autonomy or create perceived surveillance and conditional approval?
A digitally mediated coaching environment can be analyzed at three connected levels. First, a technical system records and transforms information; reliability depends on sensor quality, calibration, missingness, and model validity. Second, a human coach or institution interprets the information and decides whether to modify training. Third, an athlete experiences the resulting evaluation, feedback, and opportunity to participate. Causal inference requires evidence about each link. A dashboard associated with higher training adherence does not by itself demonstrate that an algorithm increased autonomous motivation or relational trust. Coaching studies using older, entirely human communication settings cannot be transferred unchanged to technology-mediated conditions.
The distinction between measuring and governing performance is especially important. Continuous collection can expand the coach's informational access, while the athlete may have limited knowledge of how data are interpreted or shared. Data minimization, informed understanding, access controls, transparent criteria, and avenues to challenge inaccurate records therefore belong within a psychologically informed research design. These are relevant to fairness and trust whether decisions are made by a human, an automated scoring system, or their combination.
Research in the Artificial Era further raises questions about nonbiological systems that provide advice, feedback, and apparent interpersonal responsiveness. The relevant empirical constructs include perceived social presence, anthropomorphism, reliance, perceived agency, accountability, and the distribution of decision authority. A conversational assistant can issue encouraging messages or simulated coaching feedback; those observable behaviors do not establish that it possesses the affective experiences or interpersonal commitments measured in human coaches. The scientific task is to operationalize the interaction: whether athletes understand the system's limitations, how they allocate trust, whether advice is safe and effective, and whether human support remains accessible.
The framework From Homo to Artificial provides an intellectual horizon for investigating how organized psychological functions are increasingly distributed across humans, technical infrastructures, and artificial agents. Its research implications are formulated as empirical questions rather than presumed findings. Future comparisons could randomly assign athletes to human-only feedback, human coaching augmented by explainable analysis, or appropriately bounded artificial feedback, and measure technique learning, satisfaction, agency, reliance calibration, and welfare. Such studies should include baseline digital literacy, privacy preferences, error rates, and meaningful safeguards. The goal is a scientific account of how new coaching configurations transform motivational and relational experience.
Evidence Appraisal and Methodological Limitations
The scientific literature displays convergence across several distinct evidence categories. Historical leadership research establishes a differentiated vocabulary of sport-specific behavior and athlete preferences. Motivational climate and self-determination studies consistently relate perceived coaching practices to mastery orientation, need satisfaction, and motivational quality. The coach–athlete relationship literature provides validated multidimensional constructs and evidence of links with satisfaction, communication, and welfare. Intervention trials demonstrate that some coach behaviors are changeable, with varying persistence and downstream effects. These findings warrant an integrated research program, but the degree of certainty differs sharply by study design and outcome.
The central limitation of cross-sectional evidence is common-method dependence. The same athlete can report receiving supportive coaching, feeling close to the coach, and experiencing high motivation on a single questionnaire. Positive affect, social desirability, recent victories, and overlapping item content can inflate apparent links. Some published analyses label a sequence of correlated variables a pathway or mediation even though all were measured at one occasion. Statistical modeling clarifies relations among indicators under assumptions; it does not by itself reconstruct temporal causation. The Liu, H. J., De Jonge, K. M. M., Den Hartigh, R. J. R., & Van (2026) meta-analysis provides strong evidence of recurring associations and explicitly identifies the predominance of individual-level data as a limitation for claims about reciprocal dyadic loops.
The intervention evidence is more causally informative but narrower in scope. Smoll, F. L., Smith, R. E., Barnett, N. P., & Everett, J. J (1993) demonstrates outcomes after a youth baseball coach-training workshop, with small numbers of treated coaches and a specific athlete population. Tzioumakis, Y. S., Tessier, D., Smith, N., Fabra López, P (2026) provides a randomized, longitudinal, independently observed test of coach-created motivational climate in four European countries. Its persistent reduction in disempowering practice contrasts with less durable improvement in empowering practice. That is evidence for a change in observed coaching behavior, not proof that every anticipated athlete outcome followed. Replication across sports, coaching levels, genders, and seasons is necessary to estimate general effects.
The evidence base also faces selection, publication, and measurement bias. Successful athletes and high-status coaches are easier to recruit in some settings, while early dropout and reported harm may be missing from convenience samples. Research may prioritize positive associations and interventions with publishable effects. Measures vary in time frame, factor structure, and population validity, limiting the interpretation of pooled estimates. Team-level dependence is sometimes underrepresented in conventional statistical models, and culture-specific expectations may complicate comparisons between countries. Independent replication, preregistration, transparent analysis, appropriate effect-size reporting, and assessment of measurement invariance strengthen future inference.
A scientifically defensible evidence hierarchy is outcome-specific. For the claim that athletes recognize differences between task- and ego-involving climate, psychometric factor and cross-sample studies are pertinent. For the claim that coaching training changes observed conduct, coach-allocated experiments and blinded coding are more probative. For the claim that supportive leadership prevents burnout, longitudinal and intervention studies with independent welfare measures are needed. For the claim that safer institutions prevent violence, prevalence measurement, implementation research, and ethically designed evaluation of safeguarding procedures are essential. Each question has its own evidentiary demands.
An Integrated Research Agenda
A mature research program should begin with measurement that recognizes the organization of sport. Team climate, dyadic relationship quality, and individual motivational experience must be modeled as distinct but connected levels. Researchers should collect both coach and athlete reports, independently observe representative training and competition encounters, and link these measures across time. The design should allow athletes to report concerns privately and remain enrolled in follow-up even after changing or leaving the team. Differences across age, performance level, gender, disability, culture, and competitive institutions should be considered at sampling and measurement stages rather than added as post hoc discussion.
Mechanistic investigations should test whether specific coaching practices change autonomy, competence, relatedness, and interpretations of error before altering sustained participation or welfare. Intervention components may be experimentally varied to separate the effects of informational feedback, meaningful athlete input, reductions in coercive conduct, and institutional safeguards. Repeated sessions and long-term follow-up would indicate whether early improvements persist. Explicit multilevel causal models should account for coaches as exposure clusters and for the transfer of social norms among teammates.
Dyadic investigations should compare direct perceptions, meta-perceptions, and observed communication. One key unresolved question is whether increased relationship quality helps athletes interpret demanding feedback constructively or whether clear and respectful feedback builds subsequent relationship quality. Another is whether conflict can be repaired through specific communication behaviors when power differences remain, and when institutional intervention or separation is the more appropriate outcome. Such questions require prospective event-level data, not retrospective impressions restricted to successful partnerships.
The safeguarding agenda should connect psychological science with organizational evaluation. Research should assess whether reporting channels are accessible and trusted, whether protections against retaliation are functioning, whether coach training changes harmful behavior, and whether policies operate similarly for minoritized and vulnerable athletes. Success metrics must include safety, welfare, and athlete agency, with the distribution of outcomes examined rather than only average team ratings. Independent oversight strengthens both participant protection and the credibility of the evidence.
Digital coaching introduces additional experimental opportunities and responsibilities. Investigators should specify which decisions are human, which are automated, and who is accountable for errors. They should evaluate algorithmic accuracy independently of user satisfaction, test how explanations and uncertainty influence reliance, and measure whether athletes maintain access to informed human support. The Artificial Era expands the field's possible settings while retaining the established demand for operational definitions, causal designs, and reproducible evidence.
Conclusion
Coaching leadership, motivational climate, and athlete–coach relationships constitute three analytically distinguishable dimensions of the social organization of sport. Leadership concerns what coaches do and how they exercise authority; motivational climate concerns the recurring criteria through which success, competence, participation, and error are interpreted; relational quality concerns the evolving patterns of trust, commitment, cooperation, and disagreement between particular coaches and athletes. Their intersection influences the opportunities athletes have to learn, exercise agency, disclose problems, sustain effort, and experience sport as a meaningful human activity.
Research has developed from multidimensional leadership theory and measurement through achievement-goal and self-determination accounts to dyadic relational models and contemporary behavior-focused interventions. The historical literature establishes differentiated concepts; psychometric and observational work supports measurement of several components; and controlled coach-education studies show that specific practices can be changed. Recent cross-study syntheses reinforce associations between need-supportive behavior, need satisfaction, and motivational quality. The strongest claims remain specific to the methods and outcomes that produced them. Reciprocal mechanisms, universal intervention effects, long-term welfare consequences, and technology-mediated relationships require further direct testing.
The synthesis places coaching within systems of interpersonal power, group norms, institutional incentives, and increasing technical mediation. Good scientific models should represent the protective significance of autonomy and clear instruction alongside independent safeguarding and accountability. The principal research challenge is now to connect observed coach behavior, athlete and coach perceptions, relational trajectories, organizational conditions, and objectively recorded outcomes through adequately powered, ethically sound, longitudinal and experimental work. The historical movement from judging coaches by reputation or victory toward testing their psychological and developmental effects defines the continuing scientific significance of the field.
References
- Ames, C. (1992). Classrooms: Goals, structures, and student motivation. Journal of Educational Psychology, 84(3), 261–271. https://doi.org/10.1037/0022-0663.84.3.261
- Bartholomew, K. J., Ntoumanis, N., & Thøgersen-Ntoumani, C. (2010). The controlling interpersonal style in a coaching context: Development and initial validation of a psychometric scale. Journal of Sport and Exercise Psychology, 32(2), 193–216. https://doi.org/10.1123/jsep.32.2.193
- Chelladurai, P. (1980). Leadership in sports organizations. Canadian Journal of Applied Sport Sciences, 5(4), 226–231. https://pubmed.ncbi.nlm.nih.gov/7449038/
- Chelladurai, P., & Saleh, S. D. (1980). Dimensions of leader behavior in sports: Development of a leadership scale. Journal of Sport Psychology, 2(1), 34–45. https://doi.org/10.1123/jsp.2.1.34
- Choi, H., Jeong, Y., & Kim, S.-K. (2020). The relationship between coaching behavior and athlete burnout: Mediating effects of communication and the coach–athlete relationship. International Journal of Environmental Research and Public Health, 17(22), Article 8618. https://doi.org/10.3390/ijerph17228618
- Coatsworth, J. D., & Conroy, D. E. (2009). The effects of autonomy-supportive coaching, need satisfaction, and self-perceptions on initiative and identity in youth swimmers. Developmental Psychology, 45(2), 320–328. https://doi.org/10.1037/a0014027
- Conroy, D. E., & Coatsworth, J. D. (2007). Assessing autonomy-supportive coaching strategies in youth sport. Psychology of Sport and Exercise, 8(5), 671–684. https://doi.org/10.1016/j.psychsport.2006.12.001
- Duda, J. L. (2013). The conceptual and empirical foundations of Empowering Coaching™: Setting the stage for the PAPA project. International Journal of Sport and Exercise Psychology, 11(4), 311–318. https://doi.org/10.1080/1612197X.2013.839414
- Jowett, S. (2009). Validating coach-athlete relationship measures with the nomological network. Measurement in Physical Education and Exercise Science, 13(1), 34–51. https://doi.org/10.1080/10913670802609136
- Jowett, S., & Cockerill, I. M. (2003). Olympic medallists' perspective of the athlete–coach relationship. Psychology of Sport and Exercise, 4(4), 313–331. https://doi.org/10.1016/S1469-0292(02)00011-0
- Jowett, S., & Ntoumanis, N. (2004). The Coach–Athlete Relationship Questionnaire (CART-Q): Development and initial validation. Scandinavian Journal of Medicine & Science in Sports, 14(4), 245–257. https://doi.org/10.1111/j.1600-0838.2003.00338.x
- Liu, H. J., De Jonge, K. M. M., Den Hartigh, R. J. R., & Van Yperen, N. W. (2026). Basic psychological need support, need satisfaction, and autonomous motivation in coach-athlete relationships: A systematic review and meta-analysis. International Journal of Sports Science & Coaching, 21(2), 1178–1191. https://doi.org/10.1177/17479541251400621
- Mageau, G. A., & Vallerand, R. J. (2003). The coach–athlete relationship: A motivational model. Journal of Sports Sciences, 21(11), 883–904. https://doi.org/10.1080/0264041031000140374
- Newton, M., Duda, J. L., & Yin, Z. (2000). Examination of the psychometric properties of the Perceived Motivational Climate in Sport Questionnaire-2 in a sample of female athletes. Journal of Sports Sciences, 18(4), 275–290. https://doi.org/10.1080/026404100365018
- Ntoumanis, N. (2001). Empirical links between achievement goal theory and self-determination theory in sport. Journal of Sports Sciences, 19(6), 397–409. https://doi.org/10.1080/026404101300149357
- Olympiou, A., Jowett, S., & Duda, J. L. (2008). The psychological interface between the coach-created motivational climate and the coach–athlete relationship in team sports. The Sport Psychologist, 22(4), 423–438. https://research.birmingham.ac.uk/en/publications/the-psychological-interface-between-the-coach-created-motivationa/
- Reinboth, M., Duda, J. L., & Ntoumanis, N. (2004). Dimensions of coaching behavior, need satisfaction, and the psychological and physical welfare of young athletes. Motivation and Emotion, 28(3), 297–313. https://doi.org/10.1023/B:MOEM.0000040156.81924.B8
- Smith, R. E., Cumming, S. P., & Smoll, F. L. (2008). Development and validation of the Motivational Climate Scale for Youth Sports. Journal of Applied Sport Psychology, 20(1), 116–136. https://doi.org/10.1080/10413200701790558
- Smith, R. E., Smoll, F. L., & Cumming, S. P. (2007). Effects of a motivational climate intervention for coaches on young athletes' sport performance anxiety. Journal of Sport and Exercise Psychology, 29(1), 39–59. https://doi.org/10.1123/jsep.29.1.39
- Smoll, F. L., Smith, R. E., Barnett, N. P., & Everett, J. J. (1993). Enhancement of children's self-esteem through social support training for youth sport coaches. Journal of Applied Psychology, 78(4), 602–610. https://doi.org/10.1037/0021-9010.78.4.602
- Smoll, F. L., Smith, R. E., & Cumming, S. P. (2007). Effects of coach and parent training on performance anxiety in young athletes: A systemic approach. Journal of Youth Development, 2(1). https://doi.org/10.5195/jyd.2007.358
- Tuakli-Wosornu, Y. A., et al. (2024). IOC consensus statement: Interpersonal violence and safeguarding in sport. British Journal of Sports Medicine, 58(22), 1322–1344. https://doi.org/10.1136/bjsports-2024-108766
- Tzioumakis, Y. S., Tessier, D., Smith, N., Fabra López, P., Krommidas, C., Comoutos, N., Keramidas, P., Appleton, P. R., Sarrazin, P., Balaguer, I., Digelidis, N., Quested, E., Papaioannou, A. G., & Duda, J. L. (2026). Effect of a theory-based coach education program on the objectively assessed motivational climate in grassroots football across four European countries: Results from the Promoting Adolescent Physical Activity (PAPA) project. Frontiers in Psychology, 17, Article 1563931. https://doi.org/10.3389/fpsyg.2026.1563931
- Vella, S. A., Oades, L. G., & Crowe, T. P. (2013). The relationship between coach leadership, the coach–athlete relationship, team success, and the positive developmental experiences of adolescent soccer players. Physical Education and Sport Pedagogy, 18(5), 549–561. https://doi.org/10.1080/17408989.2012.726976