The Research Behind
Your Dressage Journey

David Kolb's experiential learning cycle identifies four stages that must all occur for experience to become lasting skill: concrete experience → reflective observation → abstract conceptualization → active experimentation.

Most riders complete stages 1 and 4 automatically — they ride, then they ride again. Stages 2 and 3 — reflection and conceptualization — are where learning stalls without deliberate support. The post-ride debrief creates stage 2. The AI coaching analysis performs stage 3, which riders cannot do alone because working memory cannot hold six months of training history simultaneously while also riding a horse.

Walker et al. (2002) demonstrated that a single night of sleep produces approximately 20% improvement in motor skill speed with no loss of accuracy — gains that wakefulness alone cannot produce. Motor skill consolidation correlates specifically with Stage 2 NREM sleep spindles in late-night sleep. The gains are largest for complex motor sequences — exactly the category dressage falls into.

This means the gap between rides is not dead time. It is consolidation time. What the rider reflects on, intends, and mentally rehearses in this gap directly influences what the nervous system consolidates overnight.

Keiser & Arthur (2020) meta-analyzed 61 studies of After-Action Reviews across 3,499 participants and found an average effect size of d = 0.79 on performance improvement — a large effect that exceeds most formal training interventions. The two factors that consistently drove results were alignment to the individual's actual performance and the use of objective performance review.

The core AAR questions — "What was supposed to happen? What actually happened? Why the difference? What will we do differently?" — are the structural logic of the post-ride debrief.

Roediger & Karpicke (2006) showed that students who used retrieval practice (active recall) scored 92% vs. 79% on delayed tests compared to students who reread material, with effects lasting 8 months. The mechanism: the act of retrieving a memory strengthens the neural trace far more than passively reviewing it.

Applied to riding: reconstructing what happened in a session from memory — rather than simply looking at notes — is itself practice. The ride arc picker requires the rider to form a holistic mental model of the session's shape. The confidence slider requires them to estimate their in-session capability. Both are retrieval acts.

Malcolm Knowles formalized six characteristics of adult learners that distinguish them from children: adults are self-directed, bring rich prior experience, learn when they perceive immediate relevance, orient toward problem-solving, are driven by intrinsic motivation, and need to understand "what's in it for me" before engaging.

For dressage instruction, this means adults need to understand why a technique works (not just what to do), want meaningful choice in what they work on, and connect new movements to problems they're actually experiencing — not to an abstract curriculum.

Deci & Ryan's Self-Determination Theory (1985; extensively validated since) identifies three basic psychological needs whose satisfaction predicts intrinsic motivation and persistence in adult learners: autonomy (volitional choice), competence (experienced mastery), and relatedness (belonging and connection). When all three are satisfied, intrinsic motivation flourishes. When thwarted, motivation erodes regardless of talent.

Adult amateur riders face particular challenges with all three: limited control over training conditions (autonomy thwarted), slow and non-linear progress (competence fragile), and often isolated training environments (relatedness limited).

Albert Bandura identified four sources of self-efficacy, in descending order of power: mastery experiences (your own successes), vicarious experiences (watching similar others succeed), verbal persuasion (encouragement from credible others), and physiological state interpretation (whether you read arousal as excitement or fear). Moritz et al.'s 2000 meta-analysis found a mean correlation of r = 0.38 between self-efficacy and sport performance across 45 studies.

For adult riders, the mastery experience source is most powerful but most fragile — because progress in dressage is genuinely non-linear and often invisible to the rider in the moment.

Kingston & Hardy's 54-week study of 37 club golfers is one of the most practically important findings in applied sport psychology. They distinguished three goal types — outcome ("win"), performance ("score 65%"), and process ("follow through smoothly") — and found that process goals uniquely improved self-efficacy (d = 0.87) and anxiety control (d = 0.68), while performance goals primarily reduced anxiety. Outcome goals were useful for direction but harmful for in-competition focus.

The mechanism: process goals focus attention on what the rider can control (their actions) rather than what they cannot (the horse's response, the judge's opinion). Working memory research (Cowan, 2001) establishes a cap of approximately 4 simultaneous chunks — three process goals is the evidence-based maximum for in-session attentional focus.

Gabriele Wulf's 25 years of research on attentional focus (UNLV) has produced one of the most robust and practically significant findings in motor learning: directing attention to movement effects (external focus) consistently and substantially outperforms directing attention to body movements (internal focus).

Chua et al.'s 2021 meta-analysis of 73 studies (1,824 participants) found:

The mechanism (constrained action hypothesis): internal focus constrains the motor system, which then requires more muscular effort to produce the same movement. External focus allows motor programs to run freely at the automatic level. This is why telling a rider "soften your elbow" (internal) produces a different — and typically worse — result than "follow the horse's mouth as if your hands float on water" (external).

Robert Bjork (UCLA) coined the term "desirable difficulties" to describe a family of conditions that slow acquisition and impede short-term performance while dramatically enhancing long-term retention and transfer: spacing (distributing practice across days), interleaving (mixing different movements within a session), and reduced feedback frequency.

The spacing effect alone produces 10–30% better retention than the same practice massed into fewer sessions (Cepeda et al., 2006, meta-analysis of 254 studies, d = 0.54). The critical implication: performance during practice is an unreliable indicator of learning (Soderstrom & Bjork, 2015). A lesson that feels difficult and inconsistent may be producing more durable learning than one that flows beautifully.

Driskell, Copper & Moran's 1994 meta-analysis found that mental practice produces an overall effect of d = 0.527 on motor performance. Holmes & Collins' PETTLEP model (2001) — Physical, Environment, Task, Timing, Learning, Emotion, Perspective — demonstrated that imagery is most effective when it closely simulates the actual performance environment. A practical ratio of approximately 75% physical to 25% mental practice produces outcomes comparable to 100% physical practice.

For adult amateurs riding 3–5 hours per week, mental rehearsal is the highest-leverage free practice available. Every day between rides is a practice opportunity that most riders leave unused.

What athletes feel they are doing often differs dramatically from what they are actually doing. Muscle spindles — the primary position sensors — adapt to habitual postures: a chronically collapsed left hip eventually feels "normal" and symmetrical. This is not a character flaw; it is how the proprioceptive system works. Sullivan et al. (2018) confirmed the Dunning-Kruger pattern in sport coaching: practitioners in the lowest skill quartile had significantly higher confidence than competence, while advanced practitioners often underestimated their own capability.

Adams (1971) showed that error estimation — making a prediction about your performance before checking external feedback — is the primary mechanism for building the "perceptual trace" that enables independent self-correction. The estimation process itself builds proprioceptive accuracy over time.

Wergin, Gröpel & Mesagno's 2021 meta-analysis of 112 effect sizes established that pre-performance routines produce a g = 0.70 effect under pressure conditions — a medium-to-large effect that is notably stronger in high-stakes situations than in low-stakes practice. The critical design principle: consistency of execution predicts performance consistency more strongly than the specific content of the routine. A routine the rider always follows outperforms a "better" routine followed inconsistently.

The mechanism: pre-performance routines function as attentional funnels, progressively narrowing focus from external environment to the task at hand. They reduce intrusive thoughts (cognitive anxiety), establish a consistent arousal state, and — critically for equestrian sport — regulate the physiological signals that transmit to the horse before the rider even makes contact.

The gap between sophisticated coaching analysis and what a rider actually does differently in the arena is the central failure mode of most coaching systems. Three research streams converge on how to close it.

Process goals (Kingston & Hardy, 1997): Working memory holds approximately four chunks simultaneously (Cowan, 2001). Three process focus points is the research-supported maximum for in-session attentional focus. More than three produces attentional splitting, not clarity. The Practice Card enforces this cap structurally.

External focus cues (Wulf et al.): Instructions focused on movement effects — what the horse feels like — outperform body-focused instructions neurologically. The "In the saddle — feel for" section of the Practice Card contains only external-focus cues, translated from the Technical Coach's analysis. "Let the swing travel through" rather than "engage your core."

Analogy/metaphor (Masters & Liao, 2001): Biomechanical metaphors produce implicit-like learning that resists stress-induced degradation. Riders who arrive at a show carrying vivid movement images — not technical checklists — are less vulnerable to reinvestment (the choking mechanism). The Practice Card's image section provides one memorable metaphor per week.

Jeannerod (1994) demonstrated functional equivalence between imagined and executed movement: vivid, first-person motor imagery activates the same neurophysiological processes as physical movement — motor cortex, cerebellum, and supplementary motor area all show measurable activation during mental rehearsal. The key variable is specificity: imagery that closely matches the actual performance environment in physical sensation, timing, and emotional context produces substantially stronger effects than generic visualization.

Holmes & Collins' PETTLEP model (2001) operationalized this: effective imagery must be Physical (match the actual physical posture), Environment-specific (the actual arena), Task-accurate (real-time, not accelerated), Learning-stage-appropriate, Emotional (include the feel of the movement), and Perspective-correct (internal/kinesthetic for most sport skills). Generic "picture yourself succeeding" visualization does not meet these criteria and produces correspondingly weaker effects.

Hatzigeorgiadis et al.'s 2011 meta-analysis of 32 studies found self-talk produces an overall d = 0.48 effect on sport performance — but with an important distinction between types. Instructional self-talk (technical cues like "outside rein" or "breathe through") shows d = 0.55 for fine motor and precision tasks — exactly the category dressage falls into. Motivational self-talk ("I've got this," "stay with him") shows d = 0.37 and is more effective for endurance and effort tasks.

The practical application: for technical execution in the arena, a 2–3 word instructional cue linked to an external focus is more effective than motivational encouragement. For managing show nerves and sustaining effort through a difficult test, motivational cues take precedence.

In the landmark study by Keeling, Jonare & Lanneborn (2009), 20 participants walked and rode horses between two points multiple times. Before one pass, participants were told an umbrella would open (it never did). The heart rates of both horses and humans increased in anticipation — the horse responded to the rider's physiological state, not to an actual threat.

This bidirectional feedback loop has direct practical consequences: rider anxiety → sympathetic activation → increased heart rate and muscle tension → horse detects physiological change → horse becomes alert → horse's tension amplifies rider's anxiety → escalating cycle. Expert riders (Von Lewinski et al., 2013) buffer this transmission — their horses do not show elevated heart rates even when the riders are physiologically stressed at competition. This buffering capacity is a skill that develops with experience and deliberate self-regulation practice.

Lagarde, Peham, Licka & Kelso (2005, Journal of Motor Behavior) compared expert and novice riders at trot using motion capture. Expert riders' trunk motions were continuously phase-matched with the horse's movements. Novice riders showed transient departures from synchrony, especially during the extension phase. Crucially: the expert's synchrony produced increased temporal regularity in the horse's own trunk oscillations — the skilled rider literally makes the horse move more consistently.

Viry et al. (2013, PLOS ONE) confirmed that saddle contact is the primary stabilizing channel for this coordination. Two-point position (rising up out of the seat) destabilizes coordination because the main information channel is severed. Peham et al. (2001) quantified that professional rider-horse pairs showed lower kinematic deviation (11.5% vs. 13%) correlating directly with higher dressage scores (7.3 vs. 4.1).

Beauchamp & Whinton (2005, Journal of Sport and Exercise Psychology) identified a phenomenon unique to equestrian sport: riders must maintain belief in both their own capability and their horse's capability simultaneously. These two dimensions of confidence are distinct, can diverge, and require different interventions. A rider with high self-efficacy but low horse-confidence needs different support than the reverse — yet most coaching (and most sport psychology) addresses only the rider's self-belief.

This dual-efficacy dynamic intensifies at shows, where the horse is in an unfamiliar environment and the rider must simultaneously manage their own anxiety and their assessment of the horse's emotional state. Much of what adult amateurs describe as "show nerves" is actually horse-confidence anxiety — and the intervention is not more self-belief, it's more horse-confidence data.

John Sweller's cognitive load theory (1988) established that working memory holds approximately four chunks simultaneously (Cowan, 2001). A rider cannot monitor their current ride, track six months of training history, plan next week's sessions, manage their horse's health patterns, and recall their trainer's last three corrections at the same time. This is not a personal limitation — it is a neurological one that applies universally.

For adult amateur riders who arrive at the barn after a full work day, the cognitive cost of reconstructing context before every ride is itself a performance tax. The guidance hypothesis (Winstein & Schmidt, 1990) adds a related finding: too much information delivered too frequently creates dependency rather than learning. Adult learners need curated information — the right focus, at the right moment, in a manageable quantity.

Embodied cognition research (Lakoff & Johnson, 1980; Barsalou, 1999) establishes that cognition emerges from brain-body-environment interaction — the rider's physical experience shapes their conceptual understanding of movement, not the other way around. This means physical limitations are not merely inconveniences; they are cognitive constraints. A rider who cannot feel when their hip is blocked cannot respond to coaching about throughness, regardless of how well they understand the concept.

Biomechanics research confirms the mechanism: Peham et al. (2001) demonstrated that expert riders show significantly lower kinematic variability than amateurs, with the pelvis identified as the primary — and most underresearched — coupling point between rider and horse. Clayton & MacKechnie-Guire (2023) showed that rider asymmetries, postural bracing, and restricted mobility each produce measurable changes in horse movement quality. Physical work that closes the proprioceptive gap between what the rider feels and what is actually happening directly improves the horse's ability to express correct movement.

Chase & Simon's chess research (1973) demonstrated that expert pattern recognition depends on having seen the same pattern many times — not on general intelligence. In riding, pattern fluency develops through thousands of rides. The problem: individual riders cannot hold six months of training data in working memory, so recurring patterns remain invisible even when they are consistent and significant. A horse who consistently loses relaxation after rest days, a rider whose confidence correlates with lesson frequency, a movement that improves in schooling but regresses at shows — these are detectable only across longitudinal data.

Bandura's mastery experiences research adds a visualization-specific mechanism: seeing progress documented across time is a distinct self-efficacy source, separate from the subjective feeling of improvement. When a rider can see their ride quality trend rising over twelve weeks, that visual representation produces confidence-building effects that memory alone cannot, because memory is unreliable and tends toward recency bias.

Donald Schön's distinction between reflection-in-action (real-time adjustment during the lesson) and reflection-on-action (deliberate review afterward) underpins the Lesson Notes function. During a lesson, a rider is operating near their cognitive capacity limit — they cannot simultaneously execute, receive instruction, and form lasting conceptual understanding of what the instruction means. The post-lesson note is where the instruction converts from instruction into understanding.

The deeper research rationale is triangulation. Augmented feedback research (Salmoni, Schmidt & Walter, 1984) establishes that what coaches observe (knowledge of performance) and what riders feel (proprioceptive experience) are systematically different data streams. When a coach says "you braced through the transition" and the rider felt "balanced and through," that discrepancy is not confusion — it is the most valuable data point in the session. The gap between external observation and internal experience is precisely where proprioceptive calibration needs to happen.

Jacques Theureau's self-confrontation methodology (1992) formally combines personal video with structured reflective questions: "What were you sensing at that moment? What do you see on the video?" The explicit juxtaposition of felt experience against external evidence is the mechanism — not just watching footage, but comparing what you felt with what the camera recorded. This directly targets the proprioceptive illusion: habitual asymmetries that feel normal are often starkly visible on video.

Ste-Marie et al. (2011) demonstrated that self-modeling video — watching your own best performances rather than errors — significantly improved competitive performance in gymnastics. Feedforward self-modeling (edited footage of successful attempts) proved especially beneficial for riders with lower visual imagery ability, because the video provides the reference point that mental imagery cannot yet construct independently. The mechanism: self-relevant video activates the same neural pathways as direct experience, but without the proprioceptive filtering that makes internal experience unreliable.