PETSP-ML: a four-stage classroom process model for explanatory thinking in machine learning, adapted from CRISP-DM¶

Design · All designs
Evidence · no claims cited

Description¶

PETSP-ML (Promotion of Explanatory Thinking Standard Process for Machine Learning) is a process model the authors developed "to promote explanatory thinking in a structured manner", adapted from the Cross-Industry Standard Process for Data Mining (CRISP-DM). It reorganizes CRISP-DM's six phases into four classroom-aligned stages: Task Understanding, Data Phase (merging data understanding and preparation), Model Understanding, and Model Training and Evaluation, with an iterative structure encouraging students to revisit earlier stages. Its deliberate departure from professional practice is making model understanding a distinct, prerequisite stage before model training and shifting focus away from performance metrics.

Design Implications¶

Context¶

Requirements¶

Model Understanding is treated as a distinct, prerequisite stage before model training, reflecting the pedagogical priority of building conceptual foundations before procedural skills

Constraints¶

CRISP-DM's original formulation is not directly suited to educational settings, where objectives, constraints, and prior knowledge differ substantially from industrial projects

Target Learners¶

K-12 students learning machine learning

Learning Goals¶

explanatory thinking about how and why ML systems behave as they do
understanding of training data, algorithmic mechanisms, evaluation metrics, and sources of error or bias

Claims¶

ICE-T: a didactic framework integrating intermodal transfer, computational thinking, and explanatory thinking for ML education

Examples¶

-

Key Sources¶

Pierre Haritz, Hendrik Krone, and Thomas Liebig. (2026). Addressing Trust in AI Systems through Education: A Didactic Perspective. https://arxiv.org/abs/2609.02453

Page metadata

Field	Value
`type`	`design`
`id`	`petsp-ml-process-model`
`title`	`PETSP-ML: a four-stage classroom process model for explanatory thinking in machine learning, adapted from CRISP-DM`
`description`	`PETSP-ML (Promotion of Explanatory Thinking Standard Process for Machine Learning) is a process model the authors developed "to promote explanatory thinking in a structured manner", adapted from the Cross-Industry Sta...`
`status`	`review`
`generated`	by `process:settle-candidates` on `2026-10-11`
`trust`	unverified — no one has confirmed this page's content
`sources`	1 — `pierre-haritz-2026`

Frontmatter is the page's machine-readable record — see the Schema & Guide.