A2PL Model: Framework for Self-Directed Learning

• A2PL is a conceptual framework that uses GAI and LA to empower learners' self-directed growth through aspiration, complex thinking, and self-assessment.
• It shifts the focus from task automation to developing learner agency, emphasizing long-term developmental progress over short-term task performance.
• The model operationalizes interventions at the Person, Task, and Person×Task levels, employing non-prescriptive analytics to support guided discovery.

The A2PL model most commonly denotes Aspire to Potentials for Learners, a conceptual framework that integrates Generative AI (GAI) and Learning Analytics (LA) to cultivate Self-Directed Growth: a learner’s ability to sustain an ongoing, iterative, and sustainable self-directed learning process independently through aspiration, complex thinking, and summative self-assessment. In this formulation, A2PL is not primarily a task-automation system or a conventional adaptive tutoring model; it is a developmental framework for helping learners continuously construct, assess, and refine their own learning pathways across contexts and time (Mao, 29 Apr 2025).

1. Definition and conceptual scope

A2PL, expanded as Aspire to Potentials for Learners, was introduced as a response to AI-rich, decentralized knowledge environments in which learners increasingly use generative systems for search, synthesis, feedback, and productivity. The framework argues that the central educational question is not whether AI can help learners complete tasks more efficiently, but whether it can help them become more self-directed over time. Its core outcome is Self-Directed Growth, defined as a competency through which learners become the initiators, executors, assessors, and primary agents of their own learning journey (Mao, 29 Apr 2025).

The model is explicitly broader than task-level self-direction. It is presented as an advanced extension of self-directed learning in which learners do not merely perform isolated self-directed tasks, but learn to continuously build and evaluate their own developmental pathway. This shifts the analytic unit from short-term task performance to person-level developmental progression.

The framework is also explicitly conceptual. The paper describes methodological implications for future intervention design and learning analytics applications, and states that empirical validation and technical infrastructure remain future work. A2PL therefore occupies a programmatic position: it specifies constructs, stages, and analytic functions for a class of interventions rather than reporting an implemented end-to-end platform.

2. Research problem and motivating critique

The model emerges from a critique of both current education systems and existing research on AI-mediated self-directed learning. The paper situates education within a decentralized, continuously changing digital knowledge ecosystem where knowledge is co-created, updated, and circulated rather than simply transmitted by teachers or institutions. Within that setting, traditional teacher-centered and institution-centered models are treated as insufficient for developing durable learner agency (Mao, 29 Apr 2025).

A second motivating concern is the paradox of generative AI in education. GAI can improve access and convenience, yet it can also produce over-reliance, shallow engagement, and reduced learner agency. A2PL addresses this by insisting that AI should function as a scaffold for guided discovery rather than as an adaptive answer provider. This is a decisive departure from uses of AI that optimize task completion while leaving developmental competence under-specified.

The paper also frames the issue as one of educational equity. Equity is not reduced to access to information or digital tools. Drawing on Appadurai’s notion of the capacity to aspire, the framework emphasizes that disadvantaged learners may lack not only resources but also the ability to imagine, articulate, and navigate pathways linking present action to future possibilities. In this sense, A2PL treats aspiration as a developmental and socially situated capacity rather than as diffuse motivation.

The critique of prior research is correspondingly specific. The paper identifies several recurrent problems: emphasis on task-level performance instead of long-term development, reliance on post-task questionnaires, conflation of self-regulated learning (SRL) with self-directed learning (SDL), use of AI as an adaptive answer provider, and insufficient integration of learning analytics for tracing development across cycles, contexts, and tasks. A2PL is proposed to fill this gap by making GAI and LA instruments for sustained developmental progression rather than short-horizon performance support.

3. Core constructs and their theoretical organization

A2PL is organized around three foundational constructs: Capacity to Aspire, Complex Thinking, and Self-Assessment. These are not presented as isolated variables but as an integrated developmental system centered on Self-Directed Growth (Mao, 29 Apr 2025).

Capacity to Aspire is the motivational and directional core. The paper describes it as the learner’s ability to envision future possibilities, connect present actions to future goals, understand the relation between specific wants and broader social or professional contexts, and navigate the “cultural map” of aspiration. It includes two layers: identifying aspirations and long-term visions for social achievement and professional growth, and reflecting on gaps in resources, skills, and strategies needed to align current reality with those aspirations. Aspiration is therefore treated as a structured orientation for pathway formation.

Complex Thinking is the constructivist core. It refers to engagement with authentic, multifaceted, real-world problems through analytical reasoning, adaptive reasoning, critical thinking, metacognition, and reflective problem solving. In the framework, complex thinking mediates between aspiration and evaluation. Learners work on meaningful tasks that stretch their reasoning and thereby acquire a basis for judging their own capabilities.

Self-Assessment, particularly summative self-assessment, is the pragmatic core. The paper distinguishes it from ordinary task self-efficacy or formative monitoring. Formative self-assessment concerns immediate judgments during a task; summative self-assessment is a post-task reflective judgment about ability, process, and outcome. A distinctive feature of A2PL is that learners are expected to build their own self-assessment framework, rather than merely complete a researcher-authored questionnaire. That framework should reflect perceived strengths, gaps, strategies, and progress.

The paper anchors these constructs in three philosophical traditions. Humanistic philosophy underlies Capacity to Aspire, constructivist epistemology underlies Complex Thinking, and pragmatic philosophy underlies Self-Assessment. A figure adapted from Morris (2019) is used to position these as foundational to SDL, while a separate framework diagram places Self-Directed Growth at the center of the three-construct system. The result is a developmental model in which aspiration provides direction, complex thinking develops reasoning, self-assessment supports evaluation, and repeated cycles of these processes produce growth.

4. Levels, stages, and cyclical operation

A2PL operationalizes its framework through a three-level intervention structure: Person, Task, and Person × Task (Mao, 29 Apr 2025).

At the Person level, the focus is the learner’s aspirations and long-term vision. Learners reflect on personal aspirations, professional goals, societal achievement, and alignment between current identity and desired future identity. This level establishes the directional and humanistic basis of the cycle.

At the Task level, learners and GAI co-create C2A-based complex thinking tasks rooted in real-world scenarios and aligned with the learner’s goals. These tasks are intended to cultivate analytical and adaptive reasoning rather than merely test knowledge. The paper’s formulation suggests that task design is developmental: authentic complexity is not ancillary, but the mechanism through which learners develop the reasoning required for later self-evaluation.

At the Person × Task level, learners build their self-assessment framework by reflecting jointly on task performance and personal aims. They compare what they can do with what they aspire to do, identify skill, strategy, and resource gaps, and construct a more refined summative self-assessment framework. This is the point at which the framework becomes explicitly dynamic: task engagement feeds back into the learner’s self-conception and developmental pathway.

The process is cyclical rather than linear. The sequence given in the paper is: reflect on aspirations and long-term goals; engage in a complex thinking task aligned with those aspirations; construct or adjust a self-assessment framework; receive analytics-based feedback; revise the framework; and begin a new cycle with a more refined pathway. The paper characterizes this as a spiral of guided discovery rather than direct instruction. A plausible implication is that the framework is designed to support cumulative internalization of self-directed competencies across repeated iterations rather than discrete mastery events.

5. Learning analytics, GAI support, and methodological implications

A2PL assigns a central role to Learning Analytics and rejects designs that rely on static questionnaires, short-term outcomes, or post-task self-report as primary evidence of development. According to the paper, such methods cannot adequately capture whether SDL competencies are developing across cycles and tend to confuse task engagement with person-level agency (Mao, 29 Apr 2025).

The proposed analytics layer has two functions. The first is diagnostic analytics. Here, GAI evaluates the learner’s summative self-assessment framework using the Aspire to Potential Scoring Rubric (APSR). The rubric is described qualitatively as incorporating component weights, relational logistics, and vectorized progression across SDL cycles. Diagnostic scoring is based on the relevance of components, the interrelationships among elements, and progression across phases of the SDL cycle. An important design choice is that this scoring is not disclosed to the learner, because the framework aims to preserve guided discovery rather than turn GAI into a visible judge.

The second function is interactive analytics. Using diagnostic results, GAI produces non-prescriptive feedback in the form of thought-provoking prompts, sequenced questions, personalized scaffolding, and reflection cues. The system should neither provide direct answers nor reveal the assessment logic. Instead, it should support learners in deepening and refining their own self-assessment frameworks.

These design commitments have methodological consequences. Future interventions, according to the paper, should collect learner data traces over time, capture change across multiple learning cycles, distinguish task performance from developmental growth, and use analytics to support reflection rather than replace it. The intended setting is self-paced, non-synchronous learning environments, where learners gradually internalize SDL competencies through structured interaction with GAI and analytics.

The paper also notes that there are no explicit mathematical formulas for the framework itself. Its formalization remains conceptual, expressed through labels such as A2PL, C2A-based, Person, Task, and Person × Task, and through the qualitative structure of the APSR.

6. Significance, limitations, and terminological ambiguity

The principal significance of A2PL lies in its reframing of AI in education around learner development rather than automation. It foregrounds agency by treating learners as decision-makers and evaluators of their own developmental path. It foregrounds equity by linking educational justice to the ability to aspire and to build navigable pathways toward future possibilities. It foregrounds adaptability because the model is iterative and personalized, and sustainability because the intended outcome is not short-term task success but an enduring capacity to continue learning, evaluating, and redirecting oneself (Mao, 29 Apr 2025).

The framework also clarifies several common misconceptions. It does not equate SDL with SRL; the paper explicitly criticizes prior work for blurring that distinction. It does not treat AI as an authoritative solver; it specifies AI as a scaffold for guided discovery. And it is not presented as an already validated technical system; the paper identifies it as a conceptual framework whose empirical validation remains to be done.

Its limitations are likewise clear from the source. Because the paper is conceptual, its claims are programmatic rather than experimentally verified. The APSR is described qualitatively rather than mathematically. The framework depends on future technical infrastructure capable of tracing learners across cycles and delivering non-prescriptive analytics at scale. The paper’s own emphasis on guided discovery also implies a demanding design problem: GAI must support reflection without collapsing into direct prescription.

A further source of confusion is terminological. The acronym A2PL is also used in a different domain to denote a two-level Plackett–Luce model for preference modeling in smart mobility platforms, where user and context features determine route-attribute weights and a Plackett–Luce softmax layer predicts route choice (Santos-Pascual et al., 7 May 2026). That model belongs to Bayesian discrete choice modeling rather than educational theory. In educational research, however, A2PL refers to Aspire to Potentials for Learners and to the associated framework for cultivating Self-Directed Growth.