Introduction
Following the reported learning losses resulting from the COVID-19 pandemic and its related restrictions, there has been a renewed focus on the efficacy of P–20 education among researchers, scholars, policymakers, and parents. This renewed focus has skewed toward evaluating the impacts of distance (online) learning during the pandemic lockdowns (Edirisingha, 2022; Rice et al., 2025). The attention on P–20 education resulting from COVID-19 coincides with several high-profile initiatives (both public and private) that serve as tests for broader implementation.
Two prime examples of the movement to reshape P–20 education are the Next Education Workforce and Integrity Through Design initiatives. Both led by Arizona State University, the Next Education Workforce initiative implements co-teaching models in partner schools across the nation (Basile & Maddin, 2022; Next Education Workforce: Research Headlines, n.d.), and the Integrity Through Design initiative outlines principles for teaching in the age of AI (EdPlus Digital Learning Team, 2025). Another key example from which we draw inspiration is the Khan Lab School, created by Khan Academy founder Salman Khan and designed to focus P–12 education on mastery of skills by the learner (Academic Programs, n.d.; Manning, 2018). Each of these examples, and the model presented in this paper, is derived from the principles of learning engineering, which is a cyclical, iterative process of creation, implementation, and investigation to solve challenges to learning and/or learning conditions (Goodell et al., 2023). Our research seeks to identify the current best practices in P–20 education through a narrative review methodology and propose a new theoretical model that combines these best practices in a cohesive manner.
This paper puts forth a new, multi-dimensional theoretical model that builds upon the work of the Next Education Workforce initiative, employing principles from the Integrity Through Design initiative to supplant traditional models that rely upon outdated perceptions of elementary, secondary, and postsecondary education. Bounded by three pillars of student development—ontology, epistemology, and pedagogy—the model we introduce fosters lifelong learning by leading students to become more intellectually curious through a series of five developmental phases. Like the Khan Lab School, our model refocuses progression through schools to be purely based on skill mastery and does away with conventional grading systems, which have been shown to be both ineffective in assessing skill mastery and, in some cases, harmful to the mental well-being of students (Tripp et al., 2025). The model builds on these features by grounding these principles in the three pillars of student development and by leveraging AI to allow pupils to engage in work through low-stakes exploration. Furthermore, our model applies beyond the P–12 levels, expanding to postsecondary and postbaccalaureate education that is designed to foster lifelong learning and intellectual curiosity.
This research study employs a narrative review methodology to review research on alternative education models conducted over the past fifteen years. As a research method, narrative reviews provide an overall summary of a given body of literature (Sukhera, 2022). They differ from systematic reviews in that they do not necessarily focus on a specific question within a specific context. A narrative review is appropriate for studying alternative learning models because, while there is a specific context–recent scrutiny of P–20 education–the research question is not narrowly focused. For our purposes, our research question is “what alternative education models exist, and which ones show the most promise in meaningfully reforming the P–20 education landscape?” For this review, the authors sought out and studied the currently available, peer-reviewed literature on alternative education models that have been published in the last fifteen years. This narrative review was functionally bounded by the seven foundations of learning engineering laid out by Goodell et al. (2023). The literature considered focused on the process of learning as a human-centered experience, the structure of which is adapted based on available data.
Traditional school models in the United States have long relied on a system intended to be time- and merit-based, but that focuses too heavily on age and rewards perceived adequacy (“box-checking”) over true skill mastery through exploration (Crow, 2020; Crow & Dabars, 2020; Gomes et al., 2024). This has resulted in phenomena such as “teaching to the test” and curriculum narrowing that are antithetical to fostering lifelong learning and intellectual curiosity. Some have argued that age-based progression is responsible for results showing that only 35% of high school seniors are proficient in reading, and 22% are proficient in math (DeMallie, 2025). In fact, DeMallie goes so far as to suggest that “we need to eliminate the K–12 framework that groups students and standards primarily by chronological age.” Importantly, DeMallie explains that certain assumptions embedded in this framework–namely that everyone in an age group is capable of learning and gaining proficiency in the same things within a given timeframe of one academic year—result in students being pushed to the next level even when they lack sufficient preparation to succeed.
Alternatively, progression through demonstrated skills mastery has been shown to increase student motivation to finish and avoid learning loss (Borchers et al., 2025). This concept extends beyond the K–12 sphere into higher education, as Kearney et al. (2024) found that commonalities existing between K–12 and higher education categories demonstrate the need for “movement toward a dimensional perspective that considers student development, support needs, learning strategies, and other domains along an educational spectrum.” Mastery-based learning has gained ground in K–12 education in particular and can take many forms, including the flipped classroom model (Bergmann, 2023); AI control, shared control, and full learner control (Borchers et al., 2025); and, per Sturgis et al. (2011), competency-based learning. Though one problematic aspect of the predominant K–12 education model, age-based progression is not the sole feature that could use reform. We turn now to the literature on grading systems versus the concept of “ungrading” to shed light on the ways in which a student’s skill mastery may be measured.
Traditional grading systems have also historically been shown to increase anxiety and decrease motivation for taking challenging courses, while evidence suggests that narrative evaluation supports motivation and the psychological needs of pupils through actionable feedback and trust-building among peers and instructors (Chamberlin et al., 2023). Other alternative grading practices, such as in-class quiz retakes, have been found to reduce the negative impacts of traditional letter-grading (Tripp et al., 2025). Even absent the negative impacts of traditional grading systems, the rationale behind these systems is unclear at best, and “not entirely thought out” at worst (Blum, 2020).
Another promising aspect of both ungrading and gradeless classrooms is that they align well with pedagogies that commit to critical, feminist, democratic, anti-racist, and abolitionist approaches. Per De Welde (2025), ungrading and gradeless classrooms are “often leveraged to increase equity, student agency, and inclusivity.” This is increasingly important as our schools become more diversified due to continued demographic shifts in the overall U.S. population.
The proposed model is grounded in three core components, or pillars, of student development and learning: ontology, epistemology, and pedagogy. Specifically, ontology asks us to contemplate what we believe is real and how we think about the world around us. Epistemology asks us to reflect on how we acquire or create knowledge. Ontology and epistemology are connected in the sense that, in the context of student development, one’s epistemology is focused on how one acquires and creates knowledge about the world around them. The third pillar, pedagogy, asks us to consider how we communicate knowledge to others. Pedagogy is related to ontology and epistemology in student development in that one’s pedagogy reinforces one’s knowledge and worldview. Our model articulates the themes of five distinct phases of development through consideration of how one’s ontology, epistemology, and pedagogy transform as one progresses through these phases.
The fundamental framework of the educational journey is established through three key areas of student development: ontology (how to think about the world), epistemology (how to learn about it), and pedagogy (how to share that knowledge with others). These three evolve over the course of the journey and are the lens that students will use when engaging with different subjects (like math, reading, etc). Table 1 provides an overview of the Five Phases of the Education Tree Model.
The structural redesign of the educational environment dismantles traditional grade levels and punitive assessment models in favor of a fluid, mastery-based system. Drawing on the Next Education Workforce model, students are grouped into "home rooms" based on social and cognitive development, while academic instruction occurs in proficiency-based clusters that transcend age. This approach eliminates rigid "gifted" or "remedial" tracking, allowing students to progress at their own pace in specific subjects. Concurrently, the evaluation system shifts from letter grades to a proficiency framework where failure is replaced by an iterative "getting there" status; students engage in continuous, low-stakes assessments and retakes until mastery is achieved, ensuring that credit is a true reflection of competence rather than compliance.
This personalized progression is operationalized through technological enablers—adaptive learning platforms and agentic LLMs—that shift the teacher's role from lecturer to facilitator, allowing for deep 1-on-1 support. This integration mirrors ASU President Michael Crow’s vision of "Education through Exploration," or what he terms "Realm 4" of teaching and learning, and even creates the possibility for “infinitely scalable learning” (“Realm 5”): a ubiquitous, democratic educational landscape akin to a "Final Encyclopedia” (Crow, 2020). In this model, technology acts as a scaffold for curiosity, transforming the learner’s journey from a passive receipt of information into an active, individualized exploration. By leveraging AI to organize knowledge into digestible, user-structured pathways, the system overcomes the "factory model" of education, empowering students to construct their own identity and maximize their human potential through continuous, exploration-driven inquiry.
The model redefines secondary and postsecondary education not as distinct endpoints but as continuous, interconnected phases of development. At the secondary level, any coursework completed beyond the mastery requirement for graduation is automatically treated as dual enrollment or stackable credentials, counting toward college degrees, trade programs, or apprenticeships. This fluidity extends into the postsecondary realm, where stackable credentials become ubiquitous, creating a flexible ecosystem of human capital development. Learners are encouraged to weave in and out of higher education, certificate programs, and the workforce, utilizing multiple on-ramps to upskill and reskill throughout their careers rather than following a rigid, linear path.
A critical paradox currently plagues higher education: many subject-matter experts, including PhD students and adjunct faculty, possess deep disciplinary knowledge but receive little to no formal training in how to teach (Robinson & Hope, 2013). This results in a system where those with the least pedagogical expertise are often tasked with teaching the most complex concepts. While the immediate solution requires universities to mandate rigorous pedagogical training for all instructors, the P–20 model offers a more profound, organic long-term fix. By integrating pedagogy, the act of teaching, as a core developmental pillar from early childhood through graduate school, the model ensures that future experts have spent years practicing the communication of knowledge. Thus, the system naturally produces scholars who are not only masters of their field but also effective, empathetic educators.
Building this system necessitates the use of learning engineering principles, moving beyond isolated instructional design to the architecture of an ecosystem of complex, adaptive, integrated modules (Goodell et al., 2023). The phases serve as the primary modules bounded by the three pillars of ontology, epistemology, and pedagogy. Branching off within and between the phases are subject-specific modules, further granularized for particular skills and capabilities. These modules are designed as closed-loop control systems, where constant data collection determine the precise mastery level of each student, allowing educators to individualize instruction (Barr et al., 2023). Educators serve as learning engineers in this model by customizing the interfaces and integrations of the modules, not just for individual students, but for their specific learning population as a whole. They then guide students through these modules, determining the thresholds of skill mastery, module transition, and phase promotion, ensuring the optimal balance of challenging coursework in areas of strength with alternative approaches and strategies where growth is especially needed.
This paper introduces the "Education Tree," a theoretical model that dismantles traditional age-based progression in favor of a mastery-based system grounded in the developmental pillars of ontology, epistemology, and pedagogy. By integrating adaptive artificial intelligence and restructuring classrooms into proficiency-driven clusters, the model supports a personalized learning journey where students evolve from passive observers into active creators and teachers. This comprehensive framework not only facilitates lifelong learning through stackable credentials but also systematically resolves the pedagogical gap in higher education by cultivating teaching skills from the earliest stages of development.
Table 1
The Five Phases of the Education Tree Model
Phase | Theme | Ontology (Worldview) | Epistemology (Learning Strategy) | Pedagogy (Act of Teaching) |
1. Early Formalized Development (early primary) | "The world is what I see" | Reality is immediate, sensory, and complete. Objects exist as they are perceived. | Knowing by experiencing, imitating, and being told. | Show-and-tell: Demonstrating mastery by pointing out and describing experienced objects. |
2. Advanced Assisted Development (late primary) | "The world is what we can discover" | Reality is a set of objective facts waiting to be found; there are "right answers." | Knowing by finding information (books, experts, organizing facts). | Presentation: Teaching by organizing facts and presenting findings (e.g., posters, reports). |
3. Early Self-Development (secondary) | "The world has multiple perspectives" | Reality is complex; facts are subject to interpretation, perspective, and bias. | Knowing by analyzing and comparing sources to build an argument. | Debate & persuasion: Defending a thesis or persuading peers; anticipating counter-arguments. |
4. Higher Education (post-secondary) | "The world is organized by disciplines" | Reality is organized into specialized systems (disciplines) to understand complexity. | Knowing by applying disciplinary methods (e.g., scientific method, historical analysis). | Tutoring & leading: Translating specialized knowledge into simpler terms for novices (TA work, peer tutoring). |
5. Graduate / Lifelong Learning | "The world Is something I can help create" | Reality is dynamic and constructible; it is something one can actively contribute to. | Knowing by creating new knowledge through original research, synthesis, and innovation. | Publishing & contributing: Sharing original knowledge via publication, dissertation, or mentorship. |
Figure 1
Visual Overview of the Education Tree Model
Note: Created using ChatGPT 5.2 (2025)
Disclosure of AI usage: Gemini 3.0 Pro (2025) was used to organize and structure the model and the writeup. ChatGPT 5.0 (2025) was used in the literature search to identify relevant research while ChatGPT 5.2 (2025) was used to generate Figure 1.
EdTech Archives