Frameworks for Medical Trainee Development: From Competence to Formation

Medical education has undergone a profound reconceptualization over the past two decades, shifting from time-based training models toward outcome-driven developmental frameworks that emphasize not just what physicians can do, but who they become. This research synthesis examines the theoretical foundations of medical student and resident development—spanning competency-based frameworks, professional identity formation, evaluation principles, self-directed learning, goal-setting, and rotation-based learning—to provide a comprehensive understanding of how learners progress from novice to physician.

The evidence reveals a fundamental insight: effective physician development requires integration of external competency frameworks with internal processes of self-regulation, reflection, and identity formation. Neither competencies alone nor self-assessment in isolation produces excellent physicians. Instead, the research points toward “informed self-assessment”—a process where learners take responsibility for seeking external feedback and integrating it with personal reflection to direct their own growth.

Competency-based medical education reoriented training around outcomes

The paradigm shift from structure-based to competency-based medical education (CBME) represents perhaps the most significant reform in medical training since the Flexner Report. Carol Carraccio and colleagues articulated this transformation as moving from knowledge acquisition to knowledge application, from teacher-directed to learner-centered education, and from fixed-time completion to variable-time progression based on demonstrated competence.

CBME rests on a core principle: competencies are derived from analysis of societal and patient needs, not from tradition or convenience. The International CBME Collaborators defined this approach as “fundamentally oriented to graduate outcome abilities and organized around competencies derived from an analysis of societal and patient needs.” Five essential components characterize well-designed CBME programs: clearly articulated outcome competencies, competencies sequenced along a developmental continuum, learning experiences that facilitate acquisition through workplace-based learning, teaching practices promoting individualized development, and programmatic assessment that documents progress.

The ACGME Milestones operationalize this framework through narrative descriptions of developmental progression across six core competencies—Patient Care, Medical Knowledge, Practice-Based Learning and Improvement, Interpersonal and Communication Skills, Professionalism, and Systems-Based Practice. Milestones function as criterion-referenced developmental markers, agnostic to training year, describing progression from novice (Level 1) through competent independent practice (Level 4) to expert performance (Level 5). The framework draws on the Dreyfus model of skill acquisition, which describes how professionals progress from rule-following novices who cannot prioritize to intuitive experts who grasp situations holistically without conscious analysis.

Entrustable Professional Activities (EPAs), developed by Olle ten Cate, bridge the gap between abstract competencies and clinical practice. While competencies describe qualities of individuals, EPAs describe units of work—activities that can be “fully entrusted to a trainee once he or she has demonstrated the necessary competence.” Each EPA integrates multiple competencies; a physician performing a patient discharge requires medical knowledge, communication skills, systems awareness, and professionalism simultaneously. The trust-based entrustment model recognizes five levels of supervision, from observation only through independent performance to supervising others. This framework acknowledges that clinical education fundamentally involves supervisors making continuous entrustment decisions about trainee responsibilities.

Professional identity formation goes beyond competency acquisition

While competency frameworks address what physicians can do, professional identity formation (PIF) addresses who physicians become. The Carnegie Foundation’s 2010 landmark report Educating Physicians identified PIF as one of four essential goals for medical education reform, calling it “the backbone of medical education.” Sandra Jarvis-Selinger’s influential 2012 paper made the case explicitly: “Competency Is Not Enough.”

PIF is defined as “the transformative journey through which one integrates the knowledge, skills, values, and behaviors of a competent, humanistic physician with one’s own unique identity and core values.” The goal is for students to come to “think, act, and feel like a physician.” This represents a shift from teaching professionalism as behaviors to supporting the developmental process of becoming. The concept draws from educational traditions in clergy formation, where “formation” emphasizes transformation of the whole person—character development alongside skill development, integrating personal values with professional expectations.

Richard and Sylvia Cruess at McGill University have been central contributors to PIF theory. Their foundational work describes professional identity formation through the lens of socialization—how newcomers are incorporated into professional communities. They draw on Lave and Wenger’s concept of “legitimate peripheral participation,” where novices begin at the periphery of a community, participating in simple tasks, and progressively move toward full membership. Medical students are engaged in this process throughout training, absorbing “modes of action and meaning as part of becoming a community member.”

The hidden curriculum plays a powerful role in this socialization. Frederic Hafferty defined it as anything outside formal learning dimensions—the unwritten socialization process transmitting norms and values through modeling, corridor conversations, and observed behaviors. Research documents concerning effects including loss of idealism, emotional neutralization, and acceptance of hierarchy, though the hidden curriculum can also transmit positive values when the clinical environment models excellence. More than half of medical students report disconnects between explicit teaching and faculty behaviors.

A 2024 reconceptualization by Cruess and colleagues—”Being, Becoming, and Belonging”—addresses limitations in earlier PIF frameworks, particularly regarding diversity and inclusion. This updated model redefines socialization as active “engagement with and critical examination” of professional norms rather than passive internalization. It acknowledges that learners may “subvert or suppress parts of themselves” when professional standards conflict with their identities, and emphasizes that meaningful belonging requires valuing diverse perspectives, not just achieving competence.

Phronesis—practical wisdom—emerges as central to understanding how values and clinical judgment develop. This Aristotelian concept describes “knowledge of what is good and what is bad for humans” and functions as an “executive virtue” adjudicating when other values conflict. Pellegrino and Thomasma identified phronesis as “medicine’s indispensable virtue.” A major UK study identified 15 virtue continua in practicing physicians’ decision narratives, suggesting practical wisdom represents collective wisdom of the medical community that can be learned early in careers through role models and reflective practice.

Effective feedback transforms evaluation into learning

Modern understanding of feedback has evolved from information delivery to dialogic conversation. Hattie and Timperley’s influential model identifies three fundamental feedback questions: “Where am I going?” (feed-up), “How am I going?” (feedback), and “Where to next?” (feedforward). Their research reveals that feedback operates at four levels—task, process, self-regulation, and self—with personal-level feedback (“you’re smart”) being least effective and often counterproductive because it deflects attention from the task.

Subha Ramani and colleagues have advanced a sociocultural approach emphasizing that feedback is not about commenting on observations but allowing learners to “tell their narrative, discover areas needing support, and articulate challenges and successes.” The shift is captured in the phrase “swinging the feedback pendulum from recipes to relationships.” Feedback credibility—whether learners accept and act on feedback—depends on source credibility (clinical expertise, direct observation), feedback characteristics (timeliness, specificity), relationship factors (educational alliance, mutual trust), and recipient characteristics (emotional readiness, self-assessment accuracy).

The distinction between formative and summative assessment is foundational. Summative assessment measures achievement and makes high-stakes decisions; formative assessment steers learning and occurs early and frequently with developmental intent. Yet the paradigm has shifted further—from “assessment of learning” to “assessment for learning” to “assessment as learning,” where evaluation becomes intrinsically educational.

Programmatic assessment, developed by Cees van der Vleuten and colleagues at Maastricht University, synthesizes these principles. The core insight is that “individual data points are maximized for learning and feedback value, whereas high-stakes decisions are based on the aggregation of many data points.” No single assessment determines promotion or graduation; instead, multiple low-stakes assessments feed into holistic judgments. This approach requires triangulation across methods, meaningful feedback culture with mentoring support, and human professional judgment to synthesize qualitative data. Implementation research reveals tensions—assessments designed as formative are increasingly perceived as summative when they contribute to final decisions—but the model offers a framework for assessment that serves learning rather than merely measuring it.

Self-directed learning requires structured support and metacognitive skill

Self-directed learning (SDL) is often invoked as an educational goal, but research reveals important nuances about what makes it effective. Malcolm Knowles defined SDL as “a process in which individuals take the initiative in diagnosing their learning needs, formulating learning goals, identifying resources, choosing strategies, and evaluating outcomes.” His theory of andragogy emphasizes that adult learners need to know why they are learning something, bring prior experience as a resource, are ready to learn when facing real problems, and are internally motivated.

Gerald Grow’s Staged Self-Directed Learning model provides crucial insight: learners advance through four stages of increasing self-direction, and teaching should match the learner’s current stage. Dependent learners need coaching and immediate feedback; interested learners need inspiring guidance and goal-setting support; involved learners benefit from facilitated discussion; only truly self-directed learners thrive with delegation and independent projects. Mismatching teaching approach to learner stage—particularly offering too much independence to learners not yet ready—undermines learning.

Self-regulated learning (SRL), articulated most influentially by Barry Zimmerman, describes a cyclical three-phase process. In the forethought phase, learners engage in task analysis (goal-setting, strategic planning) and activate self-motivation beliefs (self-efficacy, outcome expectations). In the performance phase, learners exercise self-control (attention focusing, task strategies) and self-observation (metacognitive monitoring). In the self-reflection phase, learners make self-judgments (evaluation, causal attribution) and experience self-reactions (satisfaction or adaptive responses). Research in medical education finds that self-regulated learning strategy use is associated with clinical knowledge on licensing exams and that high-achieving learners use multiple effective strategies while under-achievers have limited repertoires.

Metacognition—thinking about thinking—includes metacognitive knowledge (facts and strategies), metacognitive monitoring (knowing what you know and don’t know), and metacognitive control (regulating cognitive activity). Calibration, the accuracy of self-assessment, is particularly challenging: research consistently shows learners display overconfidence in self-chosen strategies, with low correlations between predicted and actual performance. This overconfidence is most pronounced in lower-performing learners, creating the troubling finding that those who most need improvement are least likely to recognize it.

Reflective practice, articulated by Donald Schön, distinguishes reflection-in-action (thinking on your feet during an activity) from reflection-on-action (retrospective examination). Schön argued that technical knowledge alone is insufficient for complex professional problems; excellence requires “artistry” developed through reflective practice. Kolb’s experiential learning cycle—concrete experience, reflective observation, abstract conceptualization, active experimentation—and Gibbs’ six-stage reflective cycle (description, feelings, evaluation, analysis, conclusion, action plan) provide structured frameworks for guiding reflection.

Self-assessment is limited; informed self-assessment works

Perhaps the most important finding for learner-driven development comes from Kevin Eva and Glenn Regehr’s research on self-assessment. Their work documents that self-assessment is remarkably inaccurate—”a ubiquitous finding in the research literature is that self-ratings are quite poor when compared with externally generated measures of ability.” Even more troubling, medical students’ accuracy in self-assessment did not improve after 2.5 years of experience and feedback. Cognitive biases—memory favoring successes over failures, self-serving attribution, lack of metacognitive insight in poor performers—systematically undermine pure self-assessment.

The Dunning-Kruger effect manifests clearly in medical education: students who don’t perform well typically overestimate their performance while high performers are more likely to assess themselves accurately or underestimate. A study of first-semester medical students found 35.5% overestimated performance while the correlation between actual score and self-assessment was strongly negative—those who did worst were most confident.

Eva and Regehr’s reformulation proposes “informed self-assessment”—a pedagogical process by which learners take personal responsibility for looking outward, seeking feedback and explicit information from external sources, then using these externally generated sources to direct improvement. The shift is from “How good am I?” to “How can I find out how good I am?” This reframes self-assessment as self-directed assessment seeking—the ability to actively seek formative external assessment and integrate it with self-perception.

Locke and Latham’s goal-setting theory provides guidance for effective goals: specific goals lead to better performance than vague “do your best” goals, and there is a positive linear relationship between goal difficulty and performance so long as the person is committed and has requisite ability. The SMART framework (Specific, Measurable, Achievable, Relevant, Time-bound) operationalizes these principles. Research on residents shows that coaching significantly improves goal quality, particularly on specificity and measurability dimensions, suggesting learners benefit from support in creating effective goals.

Self-determination theory, developed by Deci and Ryan, identifies three innate psychological needs: autonomy (ownership of actions), competence (feeling capable), and relatedness (connection to others). When these needs are satisfied, intrinsic motivation flourishes; when thwarted, motivation and well-being diminish. Medical education research shows that controlling strategies lead to worse performance while providing meaningful rationale, acknowledging feelings, and conveying choice promote internalization of learning goals.

Carol Dweck’s growth mindset research demonstrates that beliefs about ability affect learning behaviors. Students with fixed mindsets believe abilities are innate and dread failure as a statement about their limitations; those with growth mindsets understand talents can be developed through effort and view failure as opportunity. In medical education, growth mindset promotes feedback receptivity, resilience, and perseverance. However, Dweck cautions against “false growth mindset”—claiming the mindset without following through in actions or simply equating it with effort alone.

Rotation-based learning requires active engagement and goal-setting

Clinical rotations represent a fundamental shift from classroom to experiential learning, governed by workplace learning theory. Lave and Wenger’s concept of legitimate peripheral participation describes how newcomers learn by beginning at the periphery with simple tasks and progressively moving toward full participation. Stephen Billett’s framework emphasizes co-participation—learning is constructed through interdependence between what workplaces offer (affordances, opportunities, culture) and how individuals elect to engage.

Tim Dornan’s Experience-Based Learning model identifies “supported participation” as the core condition for clinical workplace learning. Medical students learn by participating in practice with appropriate support. Critically, 64% of learning experiences involve affective support, and 58% of key outcomes are affective—confidence, motivation, identity formation. The research reveals an important caution: “self-directed learning” without support was too often experienced by students as “lack of support.”

Beginning-of-rotation goal-setting is essential for maximizing learning in compressed rotation schedules. Research on rotation-specific goals shows that goal proximity matters—rotation-specific goals are more actionable than distant career goals. A dual approach combining standardized learning objectives (ensuring baseline experiences) with individualized goals (addressing personal development needs) optimizes learning. Using “implementation intentions”—specific if-then plans for obstacles—increases time spent on learning goals.

Effective rotations require attention to transitions. The pre-clinical to clinical transition is typically most challenging, requiring complete learning approach shifts. Research identifies cognitive strategies (setting explicit goals, seeking orientation information proactively), social strategies (building relationships quickly, seeking mentorship from senior trainees), and emotional strategies (maintaining perspective, practicing self-care, building resilience). Transfer of learning across contexts requires explicit discussion of how learning applies across settings and reflection on transferable versus context-specific skills.

The supervisor-learner relationship is central. Effective supervisors create opportunities for participation, provide guidance and support, and make entrustment decisions about learner autonomy. The SUPERB/SAFETY model provides guidance: supervisors should Set expectations, remain easily available, balance supervision with autonomy; trainees should Seek input early, communicate during Active clinical decisions, and when Feeling uncertain. Trust development depends on demonstrated competence, integrity, reliability, and humility—the willingness to ask for help.

End-of-rotation reflection consolidates learning. Portfolio approaches serve as repositories for assessments and reflections, platforms for tracking development, and tools for longitudinal growth documentation. Columbia University’s “signature reflection” model involves meta-reflection on portfolio entries—reviewing archives and reflecting on patterns—to construct professional identity through connecting past and future selves.

Conclusion: Integrating frameworks for learner-driven development

The research converges on several principles for effective medical trainee development. First, competency frameworks provide essential structure but are insufficient alone—professional identity formation, with its emphasis on becoming rather than merely doing, must complement competency-based approaches. Second, pure self-assessment is unreliable, but informed self-assessment—systematically seeking external feedback and integrating it with reflection—enables effective self-directed learning. Third, feedback must be credible and dialogic; the shift from information delivery to meaningful conversation, grounded in observed performance and trusting relationships, determines whether feedback influences behavior.

Fourth, self-directed learning is developmental, requiring scaffolding appropriate to the learner’s current stage; premature independence without support undermines learning. Fifth, specific, challenging goals improve performance, but learners often need coaching support to create high-quality goals. Sixth, intrinsic motivation depends on autonomy, competence, and relatedness—learning environments that support these needs foster deeper engagement than controlling approaches. Finally, rotation-based learning requires active engagement, beginning-of-rotation goal-setting, supported participation with graduated autonomy, and end-of-rotation reflection to consolidate experience into lasting development.

These frameworks support a vision of medical education where learners take ownership of their development—but ownership informed by external standards, scaffolded by mentoring relationships, and grounded in honest reflection about the gap between current performance and aspirational excellence. The physician who emerges from this process has not merely acquired competencies but has formed an identity, developed practical wisdom, and cultivated the capacity for lifelong self-directed learning that the profession demands.