How can AI help North Dakota evaluators write CP²R and Danielson evaluations?
How can AI help North Dakota evaluators write CP²R and Danielson evaluations?
North Dakota does something most states don't: it lets each district pick its own teacher evaluation model from a short list the state approves, and one of those models — CP²R — was built in North Dakota. That local-choice design is good for districts and awkward for generic AI, which has never heard of CP²R and reaches for outdated Danielson labels. This post walks through North Dakota's model landscape, what CP²R actually is, how the current Danielson framework fits, the evaluation cadence state law sets, and where AI genuinely helps with the write-up.
North Dakota is a local-choice state
The North Dakota Department of Public Instruction does not mandate a single statewide evaluation instrument. Instead, it approves a set of models and lets each district choose one. For the 2025-2026 school year, the approved models are CP²R, Danielson, Marshall, and Marzano, with an innovation-waiver path for a district that wants to use something else. A district picks one model and applies it consistently.
This matters for documentation because the language, domains, and rating scale differ from one model to the next. A CP²R write-up and a Danielson write-up don't look alike, and blending them is exactly the mistake a general-purpose tool makes.
What CP²R is
CP²R is North Dakota's own model, and it stands for Capacity, Passion, Presence, and Relevance. Those four words are the four domains, and beneath them sit eleven categories in all:
Capacity (four categories: Intellectual, Physical, Social, Emotional) — the teacher's readiness to meet students' needs, and the students' capacity to meet instructional expectations.
Passion (two categories: Like, Energy) — visible care, enthusiasm, and energy toward teaching and learning.
Presence (three categories: Attention, Awareness, Acceptance) — focused, aware, and accepting presence during instruction.
Relevance (two categories: Self, Others) — connecting learning to students' lives and to the world beyond the classroom.
Two features set CP²R apart. First, it is deliberately formative — designed for evidence-based reflection and professional growth rather than compliance, and schools may focus on the domains that fit their local priorities. Second, it scores both the teacher's practice and the students' response in each domain. The evidence you capture has to speak to both.
The CP²R rating scale
CP²R uses a four-level scale: Not Evident, Somewhat Evident, Evident, and Strongly Evident. The levels describe how consistently and strongly the characteristics of a category show up in the lesson, for teacher and students alike. Because the model is growth-oriented, the point of a rating is the conversation and the next step, not a compliance stamp.
The Danielson option
Districts that prefer a national standard can choose the Danielson Framework for Teaching. North Dakota references the current 2022 framework, which organizes twenty-two components across four domains: Planning and Preparation, Learning Environments, Learning Experiences, and Principled Teaching. Danielson is rated on the familiar four-level scale — Unsatisfactory, Basic, Proficient, Distinguished.
The 2022 edition renamed several domains from the version many evaluators learned years ago, which is one more place a generic tool goes wrong: it produces the old labels. The domain names above are the current ones.
The cadence state law sets
North Dakota law (NDCC 15.1-15-01) sets how often teachers are evaluated. Teachers in their first three years are evaluated at least twice per school year; more experienced teachers at least once. Evaluations are completed by April 15, and districts then report teacher-effectiveness data to the state through the STARS system. North Dakota reports on ineffective teaching rather than labeling individual teachers, and that data is used only at the state level.
For an evaluator, that cadence means real volume — especially in buildings with newer teachers — and every one of those observations needs a written record in the model's own language.
After the observation: where the work piles up
The observation is familiar work. The write-up that follows is where the hours go: turning fragmentary notes into evidence organized against the right CP²R category or Danielson component, settling on a defensible level, and writing feedback a teacher can act on — across every observation the cadence requires, and before the April 15 deadline.
Where AI helps with the North Dakota write-up
The strongest fit between current AI and North Dakota's system is that model-based documentation.
AI handles three parts of it reliably. First, it turns fragmentary notes into coherent, evidence-anchored prose in the model's language — and for CP²R, it can hold onto both the teacher evidence and the student evidence the model asks for. Second, it maps evidence to the right category or component, including evidence that supports more than one. Third, it holds consistency across a caseload, so similar evidence lands in similar territory from teacher to teacher.
For North Dakota specifically, the value is drafting in the right model's language — CP²R's or Danielson's — rather than a generic blend, and doing it fast enough to keep pace with the cadence.
Where AI doesn't help — and what stays with the evaluator
The honest scope follows from how the models work.
AI cannot make the professional judgment at the center of an evaluation — what the evidence really shows, how a lesson fit a longer arc, what the right next step is for this teacher. That stays with the evaluator, who remains the last set of eyes on every rating and comment. CP²R in particular is built around a coaching conversation, and the conversation is the point.
AI also cannot choose your model or file your STARS report. The model is a district decision, and the state reporting happens in STARS, outside any drafting tool.
And general-purpose AI has no built-in knowledge of CP²R and defaults to stale Danielson labels. For a model as specific as CP²R, that isn't a small gap — it's the difference between a usable draft and a rewrite.
What to look for in an AI tool for North Dakota
A few questions worth asking before committing a tool to this work.
Does it actually know CP²R — its four domains, its eleven categories, and its Not Evident to Strongly Evident scale — or does it fall back on a generic rubric?
Does it capture both teacher and student evidence the way CP²R intends?
For Danielson, does it use the current 2022 domains, or the old names?
Can it keep the two models straight, drafting to whichever one your district adopted without blending them?
Where does your observation data live — on the vendor's servers, or used to train models?
How EvalScribe handles North Dakota evaluations
EvalScribe is built around both of North Dakota's main approved models. Whichever your district has adopted, an evaluator captures notes by typing, dictating, or photographing handwriting (Smart Scan OCR converts it to text), EvalScribe maps that evidence to the right CP²R category or Danielson component, and drafts a rating and evidence-anchored feedback in that model's own language — CP²R across its four levels, or Danielson across its four. Every rating and comment is fully editable before you finalize it, and the evidence stays traceable to the note it came from.
Two scope notes. First, North Dakota also approves the Marshall and Marzano models; EvalScribe focuses on CP²R and Danielson today, and if your district uses one of the others, telling us at [email protected] helps us prioritize. Second, EvalScribe drafts the observation write-up; your district's model choice, its summative decisions, and the state's STARS reporting are handled outside the app.
Beta testers report saving 30 to 60 minutes per evaluation versus writing the documentation by hand. Across the cadence North Dakota's law requires, that adds up to dozens of hours back — hours that can go to the coaching conversation CP²R is built to support. More detail on the North Dakota workflow is available at evalscribe.com/northdakota.
Frequently asked questions about North Dakota evaluations and AI
Which models does EvalScribe support? North Dakota approves four — CP²R, Danielson, Marshall, and Marzano. EvalScribe is built around the two most substantial: CP²R and Danielson.
What does CP²R stand for? Capacity, Passion, Presence, and Relevance — four domains and eleven categories, scored Not Evident to Strongly Evident.
Which Danielson edition? The current 2022 framework, with its four domains and the Unsatisfactory to Distinguished scale.
How often are teachers evaluated? Under NDCC 15.1-15-01, first-three-year teachers at least twice a year, experienced teachers at least once, completed by April 15.
Does EvalScribe file the STARS report? No. It drafts the observation write-up; the STARS teacher-effectiveness report is handled separately by the district.
If you're evaluating teachers in North Dakota, see how EvalScribe drafts for CP²R and Danielson at evalscribe.com/northdakota. Questions, or a school or district license? Reach the team at [email protected].
References
North Dakota Department of Public Instruction, Teacher Effectiveness Reporting (approved models CP²R, Danielson, Marshall, Marzano; CP²R domains, categories, and rating scale; STARS reporting)
North Dakota Century Code, NDCC 15.1-15-01 (teacher evaluation frequency and the April 15 completion requirement)
The Danielson Group, The Framework for Teaching (the current 2022 framework domains and components North Dakota references)
Related articles
Page maintained by Anthony D. Neely, Ph.D. — practicing K-12 educator with nearly 20 years in the classroom, 2025–2026 Walker County Distinguished Teacher of the Year, and co-founder of EvalScribe. Framework details verified against the North Dakota Department of Public Instruction's Teacher Effectiveness Reporting page and NDCC 15.1-15-01.
Last updated on August 6, 2026.
