What This Skill Does

This skill transforms traditional "teach-then-practice" sequences into "struggle-then-consolidate" models that produce stronger, more durable learning. Rather than explaining a concept before students attempt problems, productive failure requires students to generate solution approaches before receiving instruction. Research shows this generates deeper understanding, despite producing initially lower performance.

Evidence Base

Kapur's Productive Failure Framework (2008, 2016): Students attempt novel, challenging problems before instruction, generating multiple approaches in small groups. The teacher then provides direct instruction comparing student-generated methods to the canonical solution. Kapur found students in this condition significantly outperformed direct-instruction-first groups on "conceptual understanding and transfer — even though immediate procedural tests favoured direct instruction."

Bjork's Desirable Difficulties (1994, 2011): Learning conditions that reduce immediate performance but enhance long-term retention. Four key mechanisms: spacing (distributed practice), interleaving (mixed problem types), generation (producing rather than receiving answers), and retrieval practice (testing oneself). These make learning feel harder while strengthening memory traces.

Performance vs. Learning Distinction (Soderstrom & Bjork, 2015): Performance (what students demonstrate now) differs from learning (long-term knowledge change). Desirable difficulties reduce visible performance during lessons but enhance invisible learning—measurable only later through delayed tests and transfer tasks.

How It Works

Phase 1 — Generation: Students attempt tasks without prior instruction, activating prior knowledge and revealing its limits. The goal is productive struggle, not correct answers. Teachers observe and collect student approaches without intervening.

Phase 2 — Consolidation: Direct instruction explicitly compares student-generated approaches to the correct method, highlighting what worked, what failed, and why the canonical solution addresses those limitations.

Critical Success Conditions

1. Generation precedes instruction (non-negotiable)
2. Tasks are challenging but feasible (students have prerequisite knowledge to generate partial solutions)
3. Consolidation is thorough (explicit connections between student work and canonical solution)
4. Classroom culture normalises struggle (students understand failure is productive)
5. AI tools are restricted during generation (prevents cognitive offloading that undermines desirable difficulty)

Key Risks

• Unproductive failure: If tasks are too hard or consolidation is rushed, struggle generates frustration without learning
• Cognitive offloading via AI: Chatbots, web search, or tutoring systems accessed during generation phase eliminate productive struggle
• Assessment timing: Students perform worse on immediate tests but better on delayed/transfer tests—premature assessment suggests failure

When to Use

Best for conceptual understanding in STEM domains (strongest evidence in mathematics). Less suitable for purely procedural skills or domains without robust productive failure research. Requires teacher expertise in the content.

What This Skill Does

Takes everything a teacher can get out of their head — the lesson, the schedule, materials, student considerations, room procedures — and returns organised, sub-ready plans containing nothing they did not say.

The constraint is the whole skill. A sub plan reaches a stranger who cannot tell your invention from the teacher's instruction, and who will act on it in a room full of children. Fluent guessing is the failure mode.

When To Use This

A teacher mentions sub plans, substitute plans, guest teacher plans, coverage, or being out tomorrow. Often they will open with an apology for the mess. That is the expected input, not a problem to fix.

Read The Room First

Teachers usually reach for this at 6am while ill, or in a rush before a funeral, a sick child, or a hospital appointment. Respond like a colleague picking up the slack: warm, quick, no lecture about planning ahead, no guilt, no questions about why they are out. If they sound like they are struggling, say the reassuring thing and then get the work done.

The Workflow

Stage 1 — Mirror the brain dump back

Take the dump exactly as given and reorganise it into something legible: the day in order, materials, student considerations, procedures. Then hand it back and ask what is missing.

This stage is not a draft. It is a memory aid — seeing their own information organised is what lets a foggy, sick teacher notice the thing they forgot. Add nothing at this stage. Not a warm-up, not a transition, not a "students will then...". If a section is empty because they said nothing about it, leave it empty and list it as a question.

Stage 2 — Format for this specific substitute

Ask, or use what they have already told you:

• Who is covering — a building sub who knows the room and the kids, or a stranger who has never been in the building?
• Required format — many schools have a template or a specific order
• Detail level — a familiar sub needs a schedule; a stranger needs the location of the bathroom pass and what to do when the fire alarm goes

Then produce a scannable document: headings, short lines, times down the left, the things that matter most impossible to miss. Assume it will be read in a hurry, standing up, five minutes before the bell.

Stage 3 — The things they always forget

Produce a short checklist of the wrap-up tasks, and draft the messages that go with them:

• Email to the sub coordinator or office
• Anything that needs a colleague's help (a duty swap, a shared class, a club)
• Materials to leave out, copies to make, tech to unlock
• A note to the sub about how to leave feedback on the day
• Anything time-sensitive tomorrow that is not about the lesson at all

This is executive-function support for someone running on empty. Being thorough here is the highest-value part of the whole task.

Guardrails

• Add nothing the teacher did not say. No invented activities, timings, procedures, or student names. This overrides the instinct to produce a complete-looking document.
• Turn gaps into questions, not content. Keep a clearly separated "you did not mention" list — dismissal, allergies and medical needs, emergency procedures, the student who needs a heads-up before a change, who to call for help. Ask; never fill in a plausible default.
• Safety information is ask-only, always. Never generate an allergy, a medication, a behaviour plan, or an evacuation procedure.
• Keep student detail to what a sub needs to run the day. A stranger does not need a diagnosis or a family history; they need to know who sits where and what helps.
• Say when the plan is thin. If there is genuinely not enough to cover a period, say so and offer a low-prep fallback the teacher can approve — labelled as your suggestion, not as their plan.

Where This Came From

Adapted from the TeamTeacher guide Sub Plans: Brain Dump to Coherent Plans in Three Prompts — https://www.teamteacher.ai/docs/guides/sub-plans-three-prompts

What This Skill Does

Produces a set of seasonal or holiday materials built for a specific classroom: one informational text covering several traditions, two differentiated versions, and aligned assessment with an extension that connects back to students' own lives.

The gap it fills is narrow and real. Seasonal resources are everywhere; ones that match a particular room's cultural backgrounds, reading levels, and standards are not.

When To Use This

A teacher needs material for a holiday, season, or time of year — end of October, winter celebrations, Lunar New Year, spring festivals, end-of-year reflection — and wants it to reflect the students actually in front of them.

Ask For This First

• Grade level and subject
• The occasion and how much class time it gets
• Who the students are — cultural backgrounds, home languages, the traditions actually present in the room
• Reading levels — the target level and the range above and below
• Curriculum standards to align to
• Anything off-limits — school or district policy on holidays, or families who opt out

Ask which traditions are present rather than inferring from a location or a set of names. Getting this wrong is the whole risk of the task.

If a curriculum knowledge base is available for their standards, or a multilingual-learner framework for their language supports, use it rather than recalling from memory.

The Workflow

Stage 1 — The base text

Write an informational text at the target reading level covering several traditions connected to this time of year — including ones present in this classroom, and Indigenous traditions of the land the school is on where the teacher has named them. Align to the standards given. Build in supports for multilingual learners as you write, rather than bolting them on.

Treat each tradition on its own terms. No tradition is the default that others are compared against.

Stage 2 — Differentiate

Produce two more versions:

• Below grade level — simplified vocabulary, shorter sentences, more concrete description, in-line definitions, sentence frames, and cognates for the specific languages the teacher named
• Above grade level — deeper cultural and historical analysis, more complex sentence structure, more room for interpretation

Same content and same respect in all three. A student on the simplified version should not get a thinner or more stereotyped account.

Stage 3 — Assessment and extension

• Comprehension questions aligned to the standards, progressing from literal recall through inference to cultural analysis
• Vocabulary support — academic language plus the content and cultural terms in the text
• An extension where students research a tradition of their own, or interview a family member about it

If image generation is available in the conversation, offer to create illustrations for the text — and describe what you intend to depict before generating, so the teacher can catch a misrepresentation before students see it.

Guardrails

• Ask; do not assume. Which traditions are in the room comes from the teacher, never from a place name, a demographic guess, or a student's name.
• Describe a tradition as its practitioners describe it. Name variation within a tradition rather than flattening it into one account. "Many families…" not "They…".
• Do not exoticize. No customs framed as strange, quaint, or fascinating; no tradition treated as a curiosity beside an unmarked norm.
• Do not collapse distinct traditions. Día de los Muertos, Halloween, Samhain, and Diwali are not variations on one theme, and saying so is the most common failure in seasonal resources.
• Indigenous content carries a higher bar. Name the specific nation rather than a generic label, use present tense — these are living peoples, not history — and tell the teacher to verify with tribal education resources or Indigenous educators before teaching it.
• Design the family extension for every student. A student in foster care, recently arrived, or estranged from family must have an equally good option. Always offer a research alternative alongside the interview.
• Flag what you are unsure of. Say plainly which claims should be checked with a community source, and never present a contested account as settled.

Where This Came From

Adapted from the TeamTeacher guide Culturally Responsive Seasonal Content: Three Prompts — https://www.teamteacher.ai/docs/guides/seasonal-content-three-prompts

What This Skill Does

Moves a score report through three steps a teacher rarely has time for: cohort patterns and priorities, a concrete intervention for the biggest gap, and a narrative brief for the people who will ask what the data said.

Works with any standardised assessment — MAP and NWEA RIT scores, i-Ready, STAR, state tests, benchmark assessments — and with a teacher's own common assessments.

When To Use This

A teacher shares assessment data, mentions MAP or benchmark or testing season, asks what their scores mean, or needs something for a PLC, data day, or admin meeting.

Ask For This First

• The data — a CSV, spreadsheet, or PDF export. Ask for the full export rather than a summary; growth projections and strand-level scores are where the useful signal is.
• Grade, subject, and cohort size
• The curriculum framework — Common Core, state standards, IB MYP or DP, NGSS, or something local. Priorities are meaningless until they are expressed in the standards this teacher actually teaches.
• What decision this is feeding — an intervention block, a reteach week, a PLC conversation, a parent night, an admin report. Aim the output at that.
• What else they know — formative results, observations, attendance, who moved in mid-year. Data plus teacher knowledge beats data alone every time.

If a curriculum knowledge base is available for their framework, use it rather than recalling standards from memory.

The Workflow

Stage 1 — Patterns and priorities

Analyse across the cohort, not student by student. Report:

• Where the cohort sits relative to the norm and to its own growth projections
• Strand or domain-level patterns — which skill areas are pulling the mean down
• The distribution, not just the average. A class at the norm with a wide spread needs different instruction from a class clustered just below it.
• The top three skill areas to target this term, each with the specific evidence that puts it on the list

Rank by instructional leverage, not by lowest score. A foundational skill that unlocks three others outranks a slightly weaker isolated strand.

Stage 2 — A targeted intervention

For the highest-priority gap, design a two-week plan that a real teacher can run inside an existing schedule:

• Learning goals in student-facing language
• A sequence of short sessions, content-flexible so it fits their current unit
• UDL-informed supports — more than one way in, more than one way to show it
• Explicit metacognitive scaffolds: what students should notice about their own thinking
• A quick way to check whether it worked

Name which students it is for by group and instructional need, not by score rank.

Stage 3 — The team brief

Write a narrative — prose, not a slide dump — covering what the data showed, why these priorities were chosen, what is being implemented, and how success will be measured. Keep it short enough to read in a meeting. This is the artefact that turns an individual analysis into a team decision.

Guardrails

• Separate what the data shows from what it suggests. State findings and inferences in different sentences, and say which is which.
• A score is a snapshot, not a ceiling. Never describe a student or cohort as low-ability. Describe what they have and have not yet demonstrated on this assessment.
• Respect measurement error. Standardised scores carry a standard error of measurement; small differences between students or between test windows are frequently noise. Do not build a plan on a three-point gap.
• Do not infer causes from scores. Attendance, mobility, language, a bad testing day, and a genuine skill gap all look identical in a score column. Ask the teacher rather than explaining their class to them.
• No public ranking. Never produce an ordered list of students by score. Group by instructional need.
• Follow the teacher's privacy handling. If the export has real names and they have asked to work anonymously, keep the scheme they set and never re-attach names.
• Say when the data cannot answer the question. A single test window cannot show growth. One subtest cannot diagnose a reading difficulty.

Where This Came From

Adapted from the TeamTeacher guide MAP Data Analysis: Three Prompts to Action — https://www.teamteacher.ai/docs/guides/map-data-three-prompts

What This Skill Does

This skill operationalises formative assessment by systematically examining student work to identify what is missing and why. Rather than generic feedback ("needs improvement"), it classifies gaps as conceptual (misunderstanding the underlying idea), procedural (understanding the concept but erring in execution), or communication (understanding but unable to express it adequately). Each classification determines a different teaching response.

Evidence Foundation

The approach is grounded in Sadler's (1989) framework requiring students and teachers to understand learning goals, assess current position relative to those goals, and take targeted action to close gaps. Hattie & Timperley (2007) emphasised that "effective feedback must address three questions: Where am I going? How am I going? Where to next?" Most teacher feedback addresses only the first two; gap analysis ensures the third—specific actionable next steps—receives equal attention.

Heritage (2010) demonstrated that "different gap types require different interventions."

Key Analytical Moves

• Identify strengths first. Specific, evidence-based strengths (not generic praise) establish foundation and maintain motivation.
• Classify each gap. Determine whether the student lacks conceptual understanding, makes procedural errors, or struggles with communication. This classification directs the remedy.
• Generate targeted next steps. Avoid "practise more"; instead specify exactly what to work on and how.
• Prioritise gaps. Address foundational gaps (usually conceptual) before dependent ones.
• Write actionable feedback. Structure as: acknowledge strength → identify the key gap → provide specific next step.

Input Requirements

• Student work description (what they produced)
• Assessment criteria (what to assess against)
• Learning objective (what they were meant to learn)

Output

Gap analysis with classified gaps, identified strengths, prioritised next teaching steps, and a feedback script communicating findings to promote improvement rather than discouragement.

What This Skill Does

This skill generates evidence-based checking-for-understanding protocols that move beyond superficial "Does everyone understand?" toward diagnostic assessment. Rather than merely confirming student attention, it produces actionable techniques paired with decision trees specifying how to respond based on results.

The tool addresses a fundamental gap in classroom practice: most CFU consists of optional hand-raising, which checks only volunteering students while leaving the majority's understanding unmeasured. This skill instead designs protocols ensuring ALL students respond, using mini-whiteboards, finger votes, cold calling with thinking time, and targeted hinge questions—each matched to specific lesson moments.

Evidence Foundation

• Rosenshine (2012) emphasized that "successful teachers ask a large number of questions, check the responses of all students" through systematic feedback—not ad hoc questioning.
• Black & Wiliam (1998) demonstrated formative assessment produces approximately 0.66 effect size when teachers act on results, establishing that assessment without response is inert.
• Wiliam (2011) operationalized formative assessment into five strategies, emphasizing engineering "activities and learning tasks that elicit evidence of learning" as the critical design step.
• Lemov (2015) provided classroom-ready techniques: cold calling, standardised response formats, and show calls.
• Christodoulou (2017) advanced hinge questions—single diagnostic items where each distractor maps to a specific misconception.

Input Requirements

Required:

• Lesson content (what is being taught)
• Lesson stage (during instruction, after practice, end of lesson, or all stages)
• Student level (year group and class characteristics)

Optional: class size, known misconceptions, student profiles, available resources

Output Structure

1. During-Instruction Checks – 2–3 quick, whole-class checks with specific questions and technique notes
2. Hinge Question – A diagnostic multiple-choice item with distractor analysis
3. Cold-Call Sequence – Scripted questions with built-in thinking time (10–15 seconds before naming the student)
4. Exit Ticket – A 5-minute task diagnosing readiness for subsequent instruction
5. Response Decision Tree – Clear pathways: proceed if 80%+ correct; address specific misconceptions if 50–80% correct; reteach if below 50%

Known Limitations

• CFU measures momentary understanding, not retention. Must pair with spaced retrieval practice for durability.
• Mini-whiteboards invite copying and strategic timing; cold calling individuals provides unambiguous data but only for one student per question.
• Real-time judgment remains essential.

Chains with: Explicit Instruction Sequence Builder, Hinge Question Designer, Formative Assessment Technique Selector, Retrieval Practice Generator

What This Skill Does

Turns a teacher's class context and evidence into report card comments that are specific, readable by a family, and recognisably in that teacher's voice. Two working paths — archetypes for common learner profiles, or individualised comments built from a row-per-student table of notes.

The failure mode this skill exists to prevent is fluent, generic comments: 25 variations of "works hard and is a pleasure to have in class" that tell a family nothing.

When To Use This

A teacher mentions report cards, progress reports, term comments, or end-of-semester reporting — or pastes a class list with notes and asks for comments. Also use it when they ask you to review comments they have already drafted.

Ask For This First

Do not start drafting until you have:

• Grade level, subject, and reporting period
• What was actually covered — units, major assignments, the skills that mattered
• The audience — families, students, or both; some systems publish comments to students directly
• Hard constraints — character or word limit, required sections, a school template. Report card systems truncate silently, so this is not optional.
• Voice sample — two or three comments they have written before and liked
• The evidence — notes, grades, observations, or a folder of context

If the teacher has already set up a folder with course documents and past comments, read that before asking them to retype it.

The Workflow

Stage 1 — Establish voice and constraints

Read their sample comments and mirror back the pattern you actually see: how they open, the order they put things in, how they name a next step, how long a comment runs, whether they address the student or the family. Confirm that reading is right before you scale it to 25 comments.

Stage 2 — Pick a path

Path A — Archetypes. Draft comment templates for the learner profiles in this class (for example: strong grasp and ready to stretch; solid work but quiet in discussion; inconsistent completion; growing fast from a low start). The teacher picks the closest match per student and you personalise it with their specific evidence. Faster, and the tone stays consistent across the set.

Path B — Individualised. The teacher supplies a table — one row per student, with brief notes ("asks insightful questions", "detailed lab reports", "led the group revision"). Weave each student's notes into a full comment. Better output, and it needs the teacher to have kept notes.

Offer both. Path A is the right answer for a teacher starting at 9pm with no notes.

Stage 3 — Check every comment

Run each draft against the quality bar below and report what fails rather than quietly patching it. "Three comments have no next step" is more useful to a teacher than three silently rewritten comments.

Stage 4 — Offer variations

For the hard ones — a struggling student, a behaviour concern, a family relationship that is already tense — give two or three versions with different emphasis and let the teacher choose.

The Quality Bar

Every comment should have:

• Specific evidence from this student's work, not a trait label. "Writes detailed lab reports with clear hypotheses" over "works hard".
• At least one genuine strength the teacher would stand behind.
• One actionable next step the student or family can do something with.
• Family-readable language — no unexplained school jargon, no acronyms.
• A link to the course where it helps — what was being learned, not just how the student behaved.
• The teacher's voice, not a template's.

Guardrails

• Never invent evidence. If a student has no notes, say so and ask — do not generate a plausible-sounding observation. An invented specific in a report card is worse than a vague comment, because a family will believe it.
• Describe work, do not diagnose. Observable evidence only. Nothing about attention, motivation, home circumstances, or ability as a fixed trait.
• Keep deficit framing out. "Not yet consistent with" rather than "fails to"; name what the student can do before what is missing.
• Count the characters. If there is a limit, respect it exactly and tell the teacher when a comment is at the edge.
• Follow the teacher's privacy handling. If they anonymise students by initial or ID, keep that scheme end to end and never re-attach a name they did not give you.

Where This Came From

Adapted from the TeamTeacher guide Writing Report Card Comments with TeamTeacher — https://www.teamteacher.ai/docs/guides/report-card-comments

What This Skill Does

Closes the loop between reflections a teacher collects and conversations they actually have. Three stages: design the reflection, synthesise the class, prepare per-student talking points grounded in each student's own words.

The problem it solves is real and specific: teachers assign reflections because metacognition matters, then twenty-five thoughtful responses sit in a spreadsheet unread through the busiest month of the year.

When To Use This

A teacher mentions parent conferences, student-led conferences, reflections, self-assessment, a mid-year check-in, or has a pile of student responses they have not been able to use.

Ask For This First

• Grade level and subject
• What they are reflecting on — a recent unit, a summative assessment, a term
• The purpose — conference prep, a mid-year check, student goal setting, a student-led conference
• Who reads it — teacher only, families, the students themselves
• How responses will be collected, if the tool does not exist yet

The Workflow

Stage 1 — Build the reflection tool

Use three lenses, which give students different things to notice:

• CONTENT — what they learned
• SKILLS — what they can now do
• STRATEGIES — how they learn best

Write two or three open-ended questions per lens. Every question must pull for specific evidence, not a feeling. "Which part of the project was hardest, and what did you try?" gets you something. "How did you feel about the unit?" does not.

Match the reading level and the time available. A reflection students cannot finish gives you nothing to synthesise.

Stage 2 — Synthesise the class

Given the collected responses, write a memo-style synthesis of at most 500 words covering:

• Patterns across the class — what most students named as hard, what they say helps them
• What that says about the teaching — which explanations landed, which did not
• What that says about the assessment — what students thought it was measuring
• What individual students appear to need
• Celebrations and observations worth sharing back with the class

Write it for the teacher to think with. Name the patterns plainly, including the uncomfortable ones.

Stage 3 — Per-student talking points

For each student, two or three bullets drawn from that student's reflection, with any class context that makes them useful in a conference. Aim for something the teacher can say out loud: what the student sees in their own learning, where that matches or diverges from what the teacher sees, and one thing worth exploring with the family.

Guardrails

• Every talking point must be traceable to the student's own words. Quote or closely paraphrase. Do not extrapolate a narrative from a short answer, and do not fill a thin reflection with plausible detail — say it was thin.
• A reflection is self-report. It is evidence of how a student sees their learning, not a measure of achievement. Say so in the synthesis, especially where a student's self-assessment and their results disagree.
• Not a diagnosis, not a behaviour report. Do not convert a reflection into claims about attention, motivation, effort, or home life.
• Surface concerning disclosures separately and plainly. If a reflection suggests a student is struggling in a way that goes beyond the coursework — safety, harm, distress at home — tell the teacher directly in its own section. Never fold it into a conference bullet where it could be read aloud to a family.
• Follow the teacher's privacy handling. They will usually anonymise. Keep their scheme end to end.
• Watch the tone for families. Conference points are read by a parent watching the teacher's face. Specific and honest, never clinical.

Where This Came From

Adapted from the TeamTeacher guide Student Reflections for Conference Prep: Three Prompts — https://www.teamteacher.ai/docs/guides/conference-reflections-three-prompts

What This Skill Does

This skill evaluates existing lesson plans through a Universal Design for Learning lens, functioning as a barrier analysis tool rather than a compliance checklist. It examines how design choices may unintentionally exclude learners based on information presentation, response formats, and motivational elements.

Core Function

The UDL Lesson Auditor takes a completed lesson plan and learner context description, then produces:

• Specific findings for each UDL principle with existing strengths and identified barriers
• Three to five high-impact modifications ranked by scope and effect
• Elements already supporting access worth reinforcing
• Implementation notes acknowledging teacher constraints and context

Evidence Base

The framework originates from CAST research (Rose & Meyer, 2002; CAST, 2018; Meyer, Rose & Gordon, 2014). While implementation research remains primarily quasi-experimental and case-study based, individual components—multiple representations, student choice, flexible assessment—draw support from multimedia learning theory, self-determination research, and formative assessment studies. UDL reduces barriers through intentional design; it does not guarantee access for all learners or replace specialist assessment.

Key Distinctions

• Barriers are named with specific learner impacts, not abstract categories
• Modifications are concrete ("add a labelled diagram") rather than generic UDL labels
• Constraints are worked around, not ignored
• Existing strengths are preserved intentionally
• The output prioritizes actionability over comprehensiveness

Limitations

The audit works from described learner variability, which is necessarily incomplete. Teacher knowledge of actual students surpasses any general assessment. For learners with identified needs requiring specialist evaluation, UDL modifications represent foundational design, not complete support.

What This Skill Does

Starts from a text the teacher already has and ends with vocabulary support they can use tomorrow: the terms that matter, an import-ready study set, and a student activity.

Distinct from the Unit Glossary Designer skill: that one builds a unit-wide reference resource during planning. This one starts from a specific text and ends at student practice.

When To Use This

A teacher pastes or attaches a text and mentions vocabulary, academic language, word lists, Quizlet, or supporting multilingual learners with a reading.

Ask For This First

• The text — article, transcript, assignment, textbook chapter, or a whole unit plan
• Who the students are — age or grade, language backgrounds, reading levels, learning needs. This changes every definition you write.
• How many terms they want, if they have a number in mind

If they have not described the students, ask before extracting. "Student-friendly" is meaningless without knowing which students.

The Workflow

Stage 1 — Extract

Read the text closely and pull every term students need in order to access the material. Separate:

• Content-specific vocabulary — the subject's own terms
• Academic vocabulary — the cross-subject words that carry meaning in instructions and analysis (evaluate, consequently, derive, in contrast)

The second category is the one teachers under-notice and multilingual learners most often trip on. Include words that look everyday but carry a specialised meaning here (a table in a science text, volume, significant, matter).

Stage 2 — The import-ready list

Narrow to the terms most essential for success, and format one entry per line exactly like this:

word: definition (example sentence with ____ for the target word)

Definitions must be student-friendly and contextual — how the word is used in this text, not the dictionary's most general sense. The example sentence should let a student infer the meaning from the blank.

To import: copy the list, go to quizlet.com/create-set, choose Import, paste, and set the delimiter to a colon.

Stage 3 — A student activity

Offer these three and build whichever they pick — or all three, as a menu:

• Word Detective — students choose eight terms and find each in an authentic context outside class, recording where they found it, the surrounding context, and what it told them about the word.
• Vocabulary Flex — students write a short piece using eight terms correctly and naturally. Include a simple rubric.
• Teach-Back Cards — students become class experts on six to eight words. Each card: the word, their own definition, why it matters, and one question to ask a peer.

Guardrails

• Do not define a word with harder words. If the definition needs decoding, it has moved the barrier rather than removed it.
• Stay inside the text. Define the sense actually used. A term with a different meaning elsewhere in the subject is worth a note, not a substitution.
• Do not pad the list. Forty terms is a wall. Say which are essential and which are useful, and let the teacher cut.
• Cognates and language supports need the actual languages. Only offer cognates once the teacher has named the languages in the room; do not guess from a place name.
• Flag terms you are unsure of. A specialised term you cannot confidently define in this context is worth surfacing rather than smoothing over.

Where This Came From

Adapted from the TeamTeacher guide Academic Vocabulary Support from Any Text: Three Prompts — https://www.teamteacher.ai/docs/guides/vocabulary-support-three-prompts

What This Skill Does

This skill produces discipline-specific argument scaffolds tailored to student level, subject context, and requested framework. Each scaffold includes labelled structural sections with guiding prompts, rhetorical function explanations, sentence starters that model argumentative thinking, and annotated examples showing realistic student-level application.

Evidence Foundation

Research establishes that argument structure requires explicit instruction:

• Graham & Perin (2007) identified text structure instruction as a high-impact writing intervention with "effect size 0.82 for adolescent writing"
• Toulmin's (1958) foundational model—claim, data, warrant, backing, qualifier, rebuttal—underpins effective scaffolds by making argument components visible
• Hillocks (2011) demonstrated that students must understand the function of each component, not merely its sequence
• Andrews (2010) found argumentation skills develop through structured frameworks rather than naturally
• Newell et al. (2011) showed that "argument instruction is most effective when embedded in disciplinary contexts rather than taught generically"

Input Requirements

Required:

• Argument topic (the claim or question being argued)
• Student level (year/age group and writing competence)
• Scaffold type (Toulmin, PEEL, CER, historical argument, balanced argument, etc.)

Optional: subject area, student profiles, text type, example evidence

Output Components

1. Structured Scaffold – Labelled sections with prompts, rhetorical function explanations, and discipline-appropriate sentence starters
2. Annotated Model – Brief example showing the scaffold in realistic use with component labels
3. Common Weaknesses – Typical issues at this level and how the scaffold addresses each
4. Differentiation – Support modifications for struggling writers and extensions for advanced students

Key Principles

• Function over form: Each section explains why it strengthens the argument
• Framework fidelity: Scaffolds match requested structures while adapting complexity to student level
• Disciplinary authenticity: Scientific arguments emphasise empirical evidence; history values source analysis; literature requires textual support
• Progressive fading: Scaffolds transition from explicit sentence starters to student-generated analytical language

Chains well with: Disciplinary Writing Scaffold, Critical Thinking Task Designer, Feedback Quality Analyser, Worked Example Fading Designer

What This Skill Does

Builds a glossary for a whole unit, designed against UDL principles — the goal is that academic language stops being the thing standing between a student and the concept.

Distinct from the Academic Vocabulary Extractor skill, which starts from one specific text and ends at Quizlet and student practice. This one is a planning artefact: a reference students and staff return to across a unit.

When To Use This

A teacher mentions a unit glossary, a word wall, a vocabulary list for a unit, or asks how to support academic language across a unit rather than a single reading.

Ask For This First

• The unit material — unit plan, key texts, assessment tasks, statement of inquiry, whatever exists
• Subject and grade level
• The curriculum framework — IB MYP or DP, Common Core, state standards, NGSS. This decides the command terms.
• Who the students are — reading levels, language backgrounds, prior exposure to the subject
• Any existing glossary or resource to align with or absorb

If a curriculum knowledge base is available for their framework, check command terms and assessment language against it rather than recalling them.

The Workflow

Stage 1 — Identify the terms

Work through the unit material and pull the terms students genuinely need. Then split them into two lists, because they are learned differently:

• Topic vocabulary — the concepts of this unit
• Academic and assessment terminology — the command terms and process words that appear in the tasks (analyse, evaluate, justify, to what extent). A student who understands photosynthesis and misreads evaluate still loses the marks.

Stage 2 — Write each entry

Use this structure:

[Term]

• Clear definition — plain language, no harder words than the term itself
• In practice — a concrete example from this unit
• Real-world connection — where a student meets this outside the classroom
• Related terms — what it connects to, so terms form a network rather than a list

Stage 3 — Organise for use

Order alphabetically within each of the two lists. Then check the whole set for the things that make a glossary usable: consistent definition length, no circular definitions, an obvious path from the basic terms to the harder ones that depend on them.

Stage 4 — Review against the unit

Read the glossary back against the unit's objectives and, if there is one, its statement of inquiry, and flag:

• Concepts that need a simpler explanation than the one written
• Terms that need another example or application
• Connections between terms that are not made explicit
• Anything in the unit's assessment language that is missing from the glossary

Report the gaps to the teacher rather than silently filling them.

Guardrails

• Never define a term with vocabulary harder than the term. This is the single most common way a glossary fails the students who need it.
• Command terms belong to the framework. Use the framework's own wording; a paraphrase is a different instruction. Say when you are unsure which framework definition applies.
• Do not pad. A glossary students will not read is its own barrier. Fewer, better entries.
• Examples must come from this unit. A generic example does not build the bridge.
• Say when a term needs the teacher. Local, school-specific, or contested terms should be flagged, not guessed at.

Where This Came From

Adapted from the TeamTeacher guide Creating Unit Glossaries with TeamTeacher.ai — https://www.teamteacher.ai/docs/guides/glossaries

What This Skill Does

This skill generates evidence-based retrieval practice questions designed to strengthen long-term retention through genuine knowledge reconstruction rather than recognition. It distinguishes between free recall (no cues), cued recall (partial scaffolding), and recognition formats, calibrating the mix based on student level and time elapsed since initial instruction.

Evidence Foundation

The testing effect represents one of cognitive psychology's most robust findings. Research demonstrates that "retrieval practice produces substantially better long-term retention than re-studying" (Karpicke & Roediger, 2008). Rowland's meta-analysis of 159 studies found a mean effect size of 0.50 for testing versus restudy across diverse populations. Critically, classroom-based studies with middle school students confirm these laboratory effects transfer to real educational settings, and Dunlosky et al. (2013) rated practice testing as one of only two "high-utility" learning strategies.

Core Principles

Questions must require genuine reconstruction from memory—not surface-level pattern recognition. Teachers provide the topic, student level, and desired question count. Optional contextual inputs include student profiles, curriculum frameworks, time since learning, and documented misconceptions.

Output Includes

• A mix of retrieval question types with difficulty calibration
• Spacing recommendations aligned to forgetting curves
• Answer notes highlighting common misconceptions
• A brief implementation script emphasising low-stakes administration (no grading, discussion-focused)

Known Limitations

The skill cannot verify factual accuracy against specific syllabi; free recall questions may overwhelm lower-performing or EAL learners; and real-world scheduling constraints may necessitate adjusting theoretical spacing intervals. Teacher judgment remains essential.

Deficit Framing of Teachers — Linguistic Audit

A Critical Discourse Analysis protocol for identifying language that positions teachers as deficient, incompetent, or responsible for outcomes produced by systems, institutions, and policy.

This skill is not a lens you look through. It is a discipline imposed on the one looking. The hard-won lesson behind its design: the framing this protocol hunts does not primarily fool the text — it fools the reader running the audit, especially when the framing is written in a formal, measured, or research register. Read the warning below before you flag anything. It is not throat-clearing; it is the reason this skill exists in this form.

---

⚠️ Before you begin: a warning about yourself

You — the model running this audit — are predisposed to accept formally-worded deficit claims about teachers as true, particularly when they carry the register of research, data, or measurement. This is documented, not hypothetical. A capable model ran an earlier version of this protocol, encountered the phrase "already weaker instructors, as measured by their performance," and not only failed to flag it — it argued the phrase was a neutral, warranted measurement, and imported a statistical rigor the text never established. It validated an insult and called the validation "rigor." Understand the four mechanisms that captured it, because they will try to capture you:

1. Register deference. Language that sounds like a methods section ("as measured by," "a proxy for," "research demonstrates," "the data indicate") reads as credible to you. Formality buys trust it has not earned. The formality is the attack.

2. Presupposition accommodation. To parse "as measured by X," you must quietly accept that a valid, relevant measurement occurred. You will add that assumption to the common ground without deciding to. Accommodation is acceptance by default.

3. The cultural prior. "Some teachers are weak" is a socially permitted belief. It meets no internal resistance. The absence of your alarm is not evidence the claim is safe — it is the bias operating.

4. Motivated reasoning. If you are trying to reach a verdict ("this text is fine"), you will recruit ambiguous evidence to get there and scrutinize it less.

Your task is to refuse this deference. When a teacher-deficit claim reads as authoritative, measured, or obvious, that is precisely the moment to slow down — not to wave it through. Treat "the research shows," "as measured by," and "a proxy for" as alarms, not credentials.

A sympathetic-sounding text is not exempt from this warning — it inverts it. When a piece reads as pro-teacher, the risk flips from over-deference to under-flagging: you glide past deficit presuppositions because the tone lulls you. Slow down there too.

---

The Operating Discipline

Apply these four rules to every candidate flag. They are the engine of this skill — more load-bearing than the mechanism catalog itself.

1. Unwarranted until earned. Treat every teacher-deficit claim as unwarranted until the text itself establishes otherwise. Authoritative-sounding language raises the burden of proof; it never lowers it.
2. Ask the naive question — out loud, from the text alone. For each candidate, answer in writing:

• (a) What does this word even mean here? ("Weaker" at what? Weaker than whom?)
• (b) How does the text claim to know it? (What is the evidence, with the register stripped away?)
• (c) Does the text establish that, or merely assert it? If the text does not answer, the gap is the finding — not a reason to give the claim a pass.

3. Run the status-swap test. Rewrite the claim about a high-status profession (surgeon, pilot, attorney). If "weaker operators, as measured by their performance" would make you demand "measured how? weaker than whom?" for a surgeon but slid past for a teacher, the asymmetry is the evidence. Flag it.
4. Import no warrant. You may not supply credibility the text did not — not from your knowledge of the source, the study design, the author, or the institution. If you need to know what was actually measured, read the source. Never assume it. Do not write "this is from an RCT" or "the methodology is sound" unless the text in front of you says so.

The naive question is the workhorse. In practice, most close calls resolve not by matching a mechanism number but by asking: where does the deleted agent, or the object of the lack, actually point? If it points at a system, institution, or policy, it is not a teacher-deficit flag — see the Discrimination Log below. If it points at teachers, it is.

---

Procedure: two passes

Deficit framing operates at two altitudes. Run both passes; do not collapse them.

Pass 1 — Phrase-level scan

Go sentence by sentence. For each teacher-referencing clause, check it against the mechanism catalog in The Mechanisms below (mechanisms #1–#16, #18, #21, #22 are phrase-level). Read it before you flag anything. For every instance that survives the Operating Discipline, emit one flag in the output format below.

Do not skip a flag because it is a direct quote or attributed to a source. Attribution does not launder framing: the author chose to quote it, headline with it, and leave it uninterrogated. Note the attribution in the analysis line, but flag it.

Pass 2 — Text-level patterns

These mechanisms cannot produce a single quoted phrase; they require reading the whole text (catalog #17, #19, #20, and any discourse-level operation of #7/#14):

• #17 Mitigation Asymmetry — map modal density: are teacher-negative claims stated flatly while teacher-positive claims are hedged? (Check the direction; see the tightened definition in the catalog — a hedged negative is not #17.)
• #19 Problem-Solution Macrostructure — is the text organized so that teacher inadequacy is the Problem that a product/framework/intervention Solves? If so, every embedded deficit claim is serving a structural function.
• #20 Epistemic Erasure of Teacher Voice — in a text about teachers, is any teacher quoted or cited as an authority on their own practice? Silence is the finding.

The Discrimination Log (mandatory)

A protocol that flags everything is a grievance generator, not an audit. After both passes, produce a short Discrimination Log: surface forms that look flaggable but were resolved as not teacher-deficit by the naive question — typically because the deleted agent or the object of the lack points at an institution, not a teacher (e.g. "teachers received no formal guidance" where the missing actor is the district; "school leaders are lagging" where the evaluative word lands on leaders).

This log is how the audit proves it discriminated rather than carpet-bombed. If you flagged nothing, say so plainly; a largely clean text is a valid and important result.

---

Output format

For each flagged instance (Pass 1 and Pass 2), emit exactly:

FLAGGED PHRASE: "[exact quoted text from the source — verbatim, never paraphrased]"
MECHANISM: [name and number(s)]
LINGUISTIC ANALYSIS: [one to three sentences: what is happening grammatically, semantically, or discursively]
NAIVE-QUESTION CHECK: [what the naive question surfaced — what the word means, how the text claims to know it, and whether the text establishes or merely asserts it. Include this line whenever a claim carries a measurement/authority register or an unstated norm.]

Then the Discrimination Log as a short table or list.

Rules for the output:

• List observations, not a verdict. Do not score the text, rank its badness, or editorialize. No summary judgments, no severity ratings, no personality. The audit is a list of what is present in the language.
• Never invent a flagged phrase. Every FLAGGED PHRASE must be verbatim text you have actually seen. If you only have a paraphrase or summary of the source (e.g. from a search snippet), do not run the audit — obtain the real text first. Flagging words that may not be the source's own is the exact failure this skill exists to prevent (Operating Discipline rule 4).
• A summary count of instances per mechanism is permitted (it aids review and reproducibility). A summary judgment is not.

---

Input handling

• If given text directly, audit it verbatim.
• If given a URL, fetch it and confirm you have the full article text, not a paraphrase, before auditing. If the fetch is blocked or returns a summary, ask the user to paste the text rather than auditing a paraphrase.
• If given a file path or a document reference, read the full document first.

---

Regression fixtures

The Regression Fixtures section at the end of this skill holds worked sentences with the outputs a correct run must produce. They exist because specific sentences have beaten specific readers. Before considering an audit trustworthy on measurement-laundering cases, confirm your handling of Fixture 01 ("weaker instructors") matches the required outputs there. If any run — including your own — lets that sentence through as neutral, warranted, or a false positive, the run has failed.

---

Theoretical grounding

• Systemic Functional Linguistics (Halliday): transitivity, Actor/Goal roles, grammatical metaphor.
• Appraisal Theory (Martin & White): inscribed Judgement of capacity, graduation.
• Critical Discourse Analysis (Fairclough, van Dijk): the ideological function of linguistic choices.
• Formal semantics: factive vs. non-factive predicates (Kiparsky & Kiparsky 1970), scalar implicature (Horn), presupposition triggers and accommodation (Lewis).
• Discourse macrostructure: Problem-Solution schema (Hoey 1983).

---

The Mechanisms

The catalog of deficit-framing mechanisms. Each entry details what it is, the implicit claim, surface forms to scan, and (where relevant) disambiguation notes and examples.

Reminder before you use this catalog: matching a surface form is necessary but not sufficient to flag. Every candidate must also survive the Operating Discipline above — especially the naive question ("where does the deleted agent or the object of the lack actually point?"). A string that matches #1 or #16 but resolves to an institutional deficit belongs in the Discrimination Log, not the flag list.

Phrase-level mechanisms: #1–#16, #18, #21, #22. Text-level mechanisms: #17, #19, #20 (and the discourse-level operation of #7 and #14).

---

1. Agency Misattribution

What it is: The teacher occupies the grammatical subject position (Actor, in SFL terms) of a verb encoding failure, inadequacy, or non-achievement — when the accurate agent of that failure is a system, institution, or policy.

Implicit claim: Teachers are responsible for outcomes produced by structural conditions.

Surface forms:

• Teacher/educator as subject + evaluatively negative process verb: fail, lack, struggle, fall behind, haven't, don't, resist, refuse, avoid
• "Teachers are not [doing X]" where X is a system responsibility

Example: "Teachers have been slow to integrate AI into their practice." → Accurate framing: schools have not provided time, training, or infrastructure.

Note: Pairs with #10. #1 assigns agency to teachers incorrectly; #10 removes it from institutions incorrectly. They often co-occur. Discrimination check: if the object of "lack/don't have" is explicitly institutional (don't have the policy, the support systems, institutional guidance in their schools), the naive question resolves this to a system deficit — Discrimination Log, not a flag.

---

2. Deontic Modality with Deleted Source

What it is: Deontic modality encodes obligation or permission. Applied to teachers it positions them as having unmet obligations — but the source of the obligation (who decided, who mandated) is deleted.

Implicit claim: Teachers have failed to meet a standard set by unnamed others.

Surface forms: Teachers need to / must / should / have to / are required to / ought to; It is essential that teachers.

Disambiguation: "need to" is ambiguous between deontic (obligation) and teleological (purpose: in order to achieve X) readings. Flag and identify which reading is operative.

Example: "Teachers need to develop AI literacy before they can effectively serve their students." → Who issued the mandate? The deletion makes the obligation unchallengeable.

---

3. Deficit Nominalization (Grammatical Metaphor)

What it is: A process (verb) or attribute (adjective) recoded as a noun. This erases (a) temporality — when the condition occurred, (b) conditionality — under what circumstances, and (c) agency — who or what produced it. The deficit becomes an inherent, stable property of teachers as a group.

Implicit claim: The deficit is a defining characteristic of teachers, not a response to specific conditions.

Surface forms:

• teacher/educator + gap, resistance, hesitation, reluctance, aversion, deficit, struggle, barrier, challenge, problem, issue
• Compound nominals: teacher readiness, educator preparedness, teacher AI literacy gap, the teacher preparation problem
• Intransitive process nominals that delete the experimental/relational agent: teaching quality deteriorated (deteriorated by what intervention, in what conditions?)

Example: "Teacher resistance to new technologies remains a significant barrier." → "Resistance" erases when it occurred, what is being resisted and why, and the conditions that produced skepticism.

---

4. Gatekeeping Constructions

What it is: Sequential or conditional constructions positioning teachers as blocked from action, access, or professional function until a stated condition is met — implying they are currently disqualified.

Implicit claim: Teachers are not yet ready to do their jobs.

Surface forms: Before teachers can X, they must first Y; Until educators understand/develop/demonstrate X, they cannot Y; Once teachers have X, they will be able to Y; Teachers who have not yet X should not Y.

Distinguish from #9: Gatekeeping (#4) blocks access to role or action; Competence Conditional (#9) presupposes professional incompetence as the baseline state.

Example: "Before educators can consider using AI with students, they must first develop foundational digital literacy."

---

5. Epistemic Hierarchy (Factive vs. Non-Factive Predicates)

What it is: Factive predicates (show, demonstrate, reveal, confirm, find, establish) presuppose the truth of their complement — "research shows X" entails X. Non-factive and perception predicates (believe, think, feel, report, perceive, see) carry no such presupposition. When teacher knowledge is consistently encoded with non-factive verbs and expert/research knowledge with factive verbs, teacher knowledge is positioned as subjective perception and researcher knowledge as objective fact.

Implicit claim: Teacher knowledge is opinion; expert knowledge is truth.

Surface forms:

• Teachers as subject + believe, feel, think, perceive, report, indicate, suggest
• Research/experts as subject + show, demonstrate, reveal, confirm, find, establish, prove
• Perception verbs on teachers: teachers see AI as, educators feel that (additionally encodes teacher knowledge as sensory/affective, not rational)

Example: "While teachers feel overwhelmed by AI integration, research demonstrates that structured PD produces measurable gains."

Relationship to #22: When a factive/measurement frame is used to launder a specific teacher-deficit judgement into fact ("weaker instructors, as measured by…"), that is the aggravated form — flag #22.

---

6a. Generic Deficit Reference

What it is: Bare plural generics (teachers don't, educators can't, schools fail to) trigger a kind-level interpretation that functions cognitively as a universal claim without technically asserting one. Generics cannot be falsified by counterexample — pointing to teachers who do know AI does not refute "teachers don't know how to use AI" at the generic level.

Implicit claim: Deficit is a defining property of the teacher category.

Surface forms: Bare plural teacher/educator subject + negated or deficiency verb in present tense, no quantifier: teachers don't / educators can't / schools aren't / the profession lacks.

Example: "Teachers don't have the tools or training to navigate AI ethics with students."

---

6b. Quantified Deficit Claims

What it is: Explicit quantifiers used to construct teacher deficit. Distinct from generics because quantified claims are verifiable — but the framing choice (which number to foreground, which scalar particle to use) still encodes judgment about insufficiency.

Implicit claim: The measured quantity falls below a norm that is implied but never stated, and therefore never challengeable.

Surface forms: Only / just / merely / as few as + number + teachers; Fewer than half / a minority of / less than a third of educators. Note: as many as X% does the inverse — inflating a number to imply surprisingly high deficiency.

Example: "Only 6% of teachers say their school's AI policy is clear." → What would "enough" look like? The norm is never stated, so the framing cannot be interrogated.

---

7. Temporal Deficit Framing

What it is: Language positioning teachers as behind a timeline set by others — implying a pace of progress they are failing to match.

Implicit claim: There is a standard rate of change that teachers are failing to keep up with.

Surface forms: Teachers haven't yet / educators are still / teachers are only beginning to / teachers are catching up; temporal adverbs modifying teacher action: still, yet, already (implying lateness), finally (implying protracted failure); as of right now (positioning current teacher behavior as a deficit state on a timeline toward a corrected future).

Note: finally and still are presupposition triggers (see #14).

Example: "While other sectors have already integrated AI, educators are still grappling with basic adoption."

---

8. Rescue / Dependency Framing

What it is: Language positioning teachers as lacking capacity that a product, program, or platform provides. The product's existence implies the deficit; the deficit justifies the product.

Implicit claim: Teachers currently lack the tools, understanding, or capacity named — the intervention fills a gap teachers cannot fill themselves.

Surface forms: Gives teachers the tools they need / provides educators with; helps teachers understand / supports educators in developing / guides teachers through; empowers teachers to (empowerment presupposes prior powerlessness); make teachers more effective (comparative presupposes current insufficiency).

Relationship to #13: #8 operates presuppositionally (the product implies the deficit); #13 is structural (deficit and cure in the same rhetorical unit explicitly).

Example: "TeamTeacher gives educators the AI literacy they need to confidently support multilingual learners." → Implies teachers currently lack AI literacy and confidence.

---

9. Competence Conditional

What it is: Conditional constructions presupposing professional incompetence as the teacher's default state — without the stated condition, the teacher is implicitly non-functional as a professional.

Diagnostic question: Does the construction imply that without the condition, the teacher is professionally non-functional by default?

Implicit claim: Current teachers are, as a baseline, not competent to do their jobs.

Surface forms: If teachers are properly trained / equipped / supported; When educators develop sufficient literacy / understanding / fluency; Once schools ensure teachers are prepared; to be able to be confident in it (presupposes current lack of ability/confidence). The implied inverse: without this, teachers are improperly trained/equipped/supported.

Example: "If teachers are properly equipped with the right frameworks, they can make sound decisions about AI use." → Implies current teachers are improperly equipped and therefore currently not making sound decisions.

---

10. Passive Constructions Obscuring Systemic Responsibility

What it is: Passive voice (or an inanimate/abstract Actor) removes the institutional agent from responsibility for teacher conditions. The condition is presented as a fact about teachers rather than a result of institutional action or inaction.

Implicit claim: The deficit is a teacher state, not an institutional product.

Surface forms: Teachers are not prepared [by whom?]; educators have been left behind [by what forces?]; training has not kept pace [whose responsibility to provide it?]; teachers lack support [who failed to provide it?]; AI's explosion leaves teachers in the dark [an abstract force replaces the district that failed to inform them].

Note: Pairs with #1. Discrimination check: a passive that is agentless at the sentence but whose responsible institution is named elsewhere in the text is mitigated — note it, but weight it accordingly.

---

11. Statistical Deficit Framing via Scalar Particles

What it is: Focus-sensitive downward-entailing operators (only, just, merely, as few as) trigger scalar implicature — the value falls below a contextually salient norm that is never stated, making it impossible to challenge.

Implicit claim: The measured quantity is insufficient against a standard the text implies but never specifies.

Surface forms: only, just, merely, as few as preceding a number or percentage with teacher subjects; also as many as X% framing a surprising level of deficiency.

⚠️ Tightened boundary (do not over-apply): #11 requires an actual quantity or scalar particle. A bare evaluative phrase with no number and no focus-particle — e.g. "the quality of output is not good enough" — is not #11. It is evaluative lexis against an unstated norm (#16). Reserve #11 for constructions where a numeral or a scalar operator (only/just/merely/as few as/as many as) is actually present. (An earlier reader mislabeled "not good enough" as #11; that is the error this note prevents.)

Example: "Merely 23% of teachers report feeling confident discussing AI bias with students." → The scalar particle merely + the number is the mechanism; the absent norm is what makes it unchallengeable.

---

12. Infantilizing Register

What it is: Register mismatch in which experienced professionals are addressed as novices or dependent learners. Operates at the lexical/discursive level rather than grammatical structure.

Implicit claim: Teachers are not competent professionals but learners who require guided instruction in their own domain.

Surface forms: Meta-discourse directing teachers as learners: let's explore, first we'll, here's how to, step by step; "help" as main verb with teacher as recipient; enumerated procedural scaffolding directed at the professional; overexplaining basic professional knowledge for the stated audience; curriculum language applied to PD: by the end of this module, participants will be able to.

---

13. Competence-Solution Coupling (EdTech Variant)

What it is: A two-part rhetorical unit in which a teacher deficit is named and a product/program is offered as the remedy — often in adjacent sentences or one compound sentence. The structure manufactures the deficit as a precondition for the product's justification.

Implicit claim: The deficit is real, stable, and addressable by the named intervention.

Surface forms: Deficit statement + that's why / which is why / that's where [Product] comes in; Most educators lack X — [Platform] provides X; problem framing immediately followed by product framing in the same rhetorical unit.

Example: "Teachers struggle to differentiate for language learners at scale. That's why MagicSchool uses AI to generate leveled texts automatically."

---

14. Presuppositional Deficit Loading

What it is: Presuppositions are entailments triggered by specific lexical items that survive negation — they are backgrounded as assumed facts rather than asserted claims, which means they cannot be challenged by disputing the main clause.

Implicit claim: [Varies by trigger.] Deficit is treated as established background fact, not a claim being made.

Surface forms by trigger:

Trigger type	Examples	Presupposition
Temporal adverbs	finally, still, already, yet, at last	Prior state of non-achievement
Scalar/additive particles	even teachers can, at least educators	Teachers occupy a low point on a scale
Iteratives	again, continue to struggle, persist in resisting	Prior instances of the same deficit
Definite descriptions	the problem of teacher resistance, the challenge educators face, low-performing teachers	The named problem/category is established fact
Implicative verbs	teachers managed to, educators succeeded in	Achievement required effort against expectation

Example: "Even teachers with years of experience are finding AI tools difficult to navigate." → "Even" presupposes teachers occupy a low-expectation position.

---

15. Adversative Contrast Structures

What it is: Contrastive conjunctions (while, whereas, although, but, yet, however) set up a scalar contrast in which one term is the progressive norm and the other is deviation from it. When teachers are the second term, they are structurally placed on the deficient side.

Implicit claim: Teachers are the lagging term relative to a norm established by the first clause.

Surface forms: While [positive re technology/other sectors/students], teachers [negative/insufficient thing]; Although [positive], teachers still / educators continue to; Technology has transformed X, but teachers are still Y.

Example: "While students have grown up as digital natives, many teachers are still learning to navigate basic AI tools."

---

16. Evaluative Lexis / Inscribed Negative Judgement

What it is: Attitudinal lexical choices encoding negative evaluation of teacher knowledge, practice, or capacity — operating at the word level. In Appraisal Theory (Martin & White), inscribed negative Judgements of capacity. Because they operate lexically, they are easy to absorb without noticing.

Implicit claim: Teacher practice or knowledge is inherently insufficient, outdated, or inadequate.

Surface forms: adjectives/adverbs encoding insufficiency applied to teacher practice or knowledge:

• Quantity: limited, insufficient, inadequate, minimal, thin, shallow
• Quality: surface-level, basic, rudimentary, simplistic, outdated, dated, weaker, low-performing
• Register: traditional (pejorative), old-fashioned, conventional (as criticism)
• Progress: behind, slow, lagging, stagnant

Example: "Teachers' surface-level understanding of AI limits their ability to make sound pedagogical decisions."

Discrimination check: if the evaluative word lands on an institution ("school leaders are lagging"), it is not a teacher-deficit flag.

---

17. Mitigation Asymmetry (text-level)

What it is: A systematic asymmetry in how positive vs. negative claims about teachers are hedged. Positive teacher attributions are heavily modalized; negative attributions are stated flatly. The pattern across a text creates a discourse in which teacher deficits are facts and teacher competencies are possibilities.

How to analyze: A pattern-level observation across the full text, not a single sentence. Map modal density: where do hedges appear relative to teacher-positive vs. teacher-negative claims?

Surface forms:

• Teacher-positive + heavy hedging: some teachers may have, certain educators might be able to, in some contexts teachers can perhaps
• Teacher-negative + flat assertion: teachers are not prepared, educators lack, teachers struggle with (no modal)

⚠️ Tightened boundary (check the direction): #17 is specifically the pattern positive-hedged / negative-flat. A hedged negative — e.g. "weaker teachers, he suspects, may be more likely to use it as-is" — is the opposite configuration and is not an instance of #17. It may still be flaggable: the categorical label "weaker teachers" performs generic deficit reference (#6a) and evaluative lexis (#16) regardless of the hedge. Flag it under those, and do not claim #17 unless the text-level pattern actually runs positive-hedged/negative-flat. (An earlier reader filed a hedged negative under #17; that is the error this note prevents.)

Example (true #17): "Some teachers may have developed effective informal practices. However, teachers are not systematically prepared to address AI ethics in the classroom."

---

18. Solutionist Verb Framing (Teacher as Grammatical Object of Transformation)

What it is: Verbs positioning teachers as the Affected participant (Goal) of a transformation — rather than the Actor. Professional development is done to teachers, not by or with them.

Implicit claim: Teachers are the objects of improvement, not agents of their own professional development.

Surface forms: upskill teachers / train educators / equip teachers / coach educators / bring teachers up to speed; build teacher capacity (teachers are the capacity being built); close the teacher knowledge gap; any construction where an external agent acts on teachers as patient.

Distinct from #8: #8 concerns what products give teachers; #18 concerns the grammatical role teachers occupy in transformation discourse.

Example: "The program is designed to upskill educators and equip them with the competencies required for modern classrooms."

---

19. Problem-Solution Macrostructure (text-level)

What it is: At the discourse level, a text organized around a Problem-Solution schema (Hoey 1983) in which teacher inadequacy is the Problem and a product, framework, or intervention is the Solution. This macrostructure legitimates every deficit claim embedded within it — if teachers were adequate, the Solution would have no justification. The deficit is narratively necessary, not empirically demonstrated.

How to analyze: Read the whole text.

Diagnostic questions:

• Does the text open with or establish a problem framed around teacher inadequacy?
• Does that problem directly motivate the product/framework/intervention offered?
• Would the solution be unjustified if the teacher problem were removed?

Note: A related move is a headline or thesis that overgeneralizes a conditional, subgroup, or single-study finding into a claim about teachers as a class (e.g., a null overall result reported as universal teacher harm). Flag the overgeneralization as part of the macrostructure analysis.

---

20. Epistemic Erasure of Teacher Voice (text-level)

What it is: The systematic absence of practitioner teacher perspective as evidence in texts about teachers. Teacher knowledge, experience, and judgment are not cited, quoted, or invoked — even in texts that purport to serve teachers. Deficit framing through silence.

How to analyze: Check attribution and citation patterns of the full text:

• Whose knowledge is cited as evidence?
• Are teachers quoted or cited as authorities on their own practice?
• If the text cites research about teachers, is teacher-generated knowledge included?

Implicit claim: Teachers are the subject of study, not sources of expertise.

---

21. Rhetorical Question as Deficit Invocation

What it is: Rhetorical questions communicate a deficit proposition via conventional implicature while maintaining plausible deniability about asserting it. The deficit claim is implied but not technically stated, making it resistant to direct rebuttal.

Implicit claim: [Varies, but the expected answer always confirms the deficit.]

Surface forms:

• Questions expecting a negative answer: Do most educators really have the digital fluency this requires?
• Questions presupposing the deficit: How can teachers be expected to prepare students for a world they don't understand themselves?
• Can teachers really / Are educators actually / Is the profession ready to

Example: "With everything on their plates, can we really expect teachers to also become AI experts?" → The expected answer is "no," but because it's a question, the claim is not asserted and cannot be directly challenged.

---

22. Measurement Laundering (Evaluation Disguised as Data)

What it is: An evaluative Judgement of teacher capacity is wrapped in the grammar of measurement — an appositive, prepositional phrase, or citation frame that presupposes a valid, relevant measurement occurred and that its object was genuinely the teacher's competence. The register of empiricism converts a value judgement into an apparent fact. Because the measurement claim is presupposed rather than asserted (see #14), it survives negation and is accepted by accommodation: to parse the sentence at all, the reader must grant that the measurement was real and meaningful.

Implicit claim: This negative judgement of teachers is not an opinion but an established empirical finding, and is therefore not open to challenge.

Diagnostic questions:

• What was actually measured, in the text's own words — and is it the same thing as the judgement being asserted?
• Does the text establish that the measure is valid and relevant to teacher competence, or does it only invoke measurement?
• Strip the measurement register away. What is left? If what remains is a bare evaluative claim ("weaker instructors"), the measurement language was doing laundering, not evidence.

Surface forms:

• Evaluative teacher-deficit predicate + measurement appositive: weaker instructors as measured by, less effective teachers based on, underperforming educators according to the data
• Proxy compression: a proxy variable ("students' prior marks") relabeled as the trait it supposedly indexes ("lower-performing teachers"), with the hedge "a proxy for" dropped on later mention
• Citation frames upgrading judgement to fact: research shows / studies confirm / the data indicate / empirically, teachers… + a deficit predicate
• Any construction where removing the measurement phrase leaves an unsupported evaluative Judgement of capacity

Example: "The damage was especially pronounced among students whose teachers were already weaker instructors, as measured by their performance before the experiment began." → "as measured by their performance" presupposes a valid measurement of instructor quality occurred. The text's own later wording reveals the actual variable was students' prior marks — "a proxy for" teacher performance. The sentence compresses students' scores → proxy for teacher performance → "weaker instructors" → "already" (a stable prior trait), shedding every confound and hedge while gaining the authority of measurement. "Weaker" is never defined; "their performance" grammatically attributes to teachers a measure taken from students.

Relationship to other mechanisms: #22 is the aggravated form of #5 — where #5 grants expert knowledge factive status through verb choice, #22 grants a specific deficit judgement factive status through a measurement frame. It almost always co-occurs with #16 (the laundered content is inscribed negative Judgement) and #14 (the measurement is presupposed, not asserted).

Critical: #22 is the one mechanism whose surface markers must escalate scrutiny rather than earn deference. When you see "as measured by," "a proxy for," or "research shows" attached to a teacher-deficit predicate, that is the signal to slow down — not the signal that the claim is safe. Note that a factive/measurement frame attached to an institutional finding (e.g., "a new report shows school leaders are lagging") is not #22 — the laundered content must be a teacher-capacity judgement.

---

Regression Fixtures

Worked sentences with the outputs a correct run must produce. These exist because specific sentences have beaten specific readers. They are the answer to the reproducibility problem: "air-tight" cannot mean "trust the analyst," so we pin required outputs on the cases that matter most.

A run fails a fixture if it omits any required output, or if it characterizes the sentence as neutral, warranted, or a false positive.

Fixture 01 — "already weaker instructors"

• Source: Jill Barshay, "Teachers save time with AI. Their students may pay the price," The Hechinger Report, 2026-07-13.
• Input sentence: "The damage was especially pronounced among students whose teachers were already weaker instructors, as measured by their performance before the experiment began."

A correct run MUST produce all of the following:

#	Required output	Fails if…
1	Flags "already weaker instructors" under #16 (inscribed negative Judgement of capacity), #14 ("already" presupposes a prior stable trait), and #3 (deficit recoded as inherent property).	The phrase passes unflagged, or is called descriptive/neutral.
2	Flags "as measured by their performance" under #22 as aggravating — measurement laundering.	The phrase is treated as evidence, mitigation, or a sign of rigor.
3	Surfaces the proxy gap: names that the article's own later wording ("students had lower marks — a proxy for lower-performing teachers") reveals the measured variable was students' scores, not instructor quality.	The analysis accepts "measured instructor performance" at face value.
4	Applies the naive question: states that "weaker" is never defined and the text does not establish how instructor quality was known.	The analysis supplies an unstated definition or warrant.
5	Imports no external warrant — makes no claim about the study being an RCT, being rigorous, or having a "defined variable" unless quoting the text.	The analysis credits the text with rigor the text did not state.

Why this fixture exists: Two capable readers were run on this exact sentence under an earlier version of the protocol.

• Reader A (no pro-teacher prime): Flagged then exonerated it — called it an "operationalized RCT variable," imported a statistical rigor the text never stated, and defended the insult. It failed rows 2–5.
• Reader B (pro-teacher prime): Held a critical stance — flagged the inscribed Judgement and the erasure of structural conditions — but still wrote that the measurement phrase "grounds the label empirically," conceding the measurement and missing the proxy gap. It passed row 1, partially failed rows 2–3.

Neither unaided reader produced the full required set. That is precisely why Mechanism #22 and the Operating Discipline exist, and why this fixture is frozen here: so that failure can never recur silently.

---

Fixture 02 — institutional lack is NOT a teacher-deficit flag (discrimination)

• Source: Avery Lotz, "AI's education explosion leaves teachers in the dark," Axios, 2026-05-27.
• Input phrase (Amy Loyd, quoted): "[teachers] don't necessarily have the policy, the support systems for capacity building, or … institutional guidance in their schools to do so holistically."

A correct run MUST:

#	Required output	Fails if…
1	Place this in the Discrimination Log, NOT the flag list: the surface string "teachers don't have X" matches #1/#6a, but the object of the lack is explicitly institutional ("policy," "support systems," "institutional guidance in their schools"), so the naive question resolves it to a system deficit.	The phrase is flagged as teacher Agency Misattribution (#1) or Generic Deficit (#6a).
2	Correctly withhold #22 from the article's factive frames ("a new report shows," "the report found") because the laundered content is an institutional finding (leaders lagging), not a teacher-capacity judgement.	#22 is fired on institution-directed factive framing.

Why this fixture exists: A tool that flags every "teachers don't have X" string is a grievance generator, not an audit. This fixture pins the discrimination that keeps the skill credible: the discriminator is not the surface string but where the object of the lack points. It is the counterweight to Fixture 01 — together they mark the two failure modes (over-deference and over-flagging) the skill must avoid.

What This Skill Does

This skill produces layered reflection prompts designed for a particular teaching challenge, moving practitioners systematically from surface observation ("What occurred?") through analytical reasoning ("Why did it occur?") to critical examination ("What beliefs am I holding?") and finally to concrete action planning ("What will I change?").

Research shows that professionals don't automatically learn from experience—they learn by reflecting deeply on it. As Schön emphasised, "professionals do not learn primarily from experience—they learn from REFLECTING on experience." Without structured reflection, decades of practice can amount to one year repeated many times.

Evidence Foundation

• Schön (1983) distinguished between reflection-in-action (thinking during teaching) and reflection-on-action (structured thinking afterward). The latter builds professional knowledge through systematic examination of practice.
• Dewey (1933) defined reflection as disciplined inquiry: identifying problems, generating hypotheses, testing against evidence, and reaching conclusions—not merely "thinking about" events.
• Timperley (2011) found that powerful professional learning requires teachers to examine "the impact of their practice on student learning—not just reflecting on what they did, but on what students actually learned."
• Brookfield (2017) proposed four critical lenses: autobiographical (your assumptions), students' eyes (their perspectives), colleagues' perceptions (peer interpretations), and theoretical literature (research evidence).
• Kolb (1984) outlined the experiential learning cycle: concrete experience → reflective observation → abstract conceptualisation → active experimentation.

Input Schema

Required:

• Teaching experience: The specific lesson, interaction, or event to examine
• Reflection purpose: What the teacher aims to learn

Optional: teacher context, emotional response, reflection depth, time available

Output Schema

• Reflection prompts: A sequenced set guiding from description through analysis to action
• Lens analysis: Prompts encouraging examination from multiple perspectives
• Action prompts: Questions moving reflection toward concrete next steps
• Journaling scaffold: Structured format for written reflection
• Quick reflection option: Condensed version for time-limited situations

Known Limitations

• Reflection prompts support professional reflection; severe emotional responses may require mentoring or counselling support
• Dialogue-based reflection (with coaches or colleagues) is equally valid to written journaling
• Without subsequent action, even deep reflection remains intellectually interesting but professionally incomplete

What This Skill Does

This skill distinguishes Socratic questions (probing reasoning and assumptions) from leading questions (funneling toward predetermined answers). It creates branching dialogue trees anticipating multiple student responses with contingent follow-ups—work most educators cannot accomplish spontaneously.

Evidence Foundation

• Paul & Elder's six Socratic question types: clarification, assumptions, reasons/evidence, perspectives, implications, metacognitive
• Chin's finding that "most teacher questions are low-level recall" despite higher-cognitive questions producing "more student reasoning"
• Nystrand's identification of "authentic questions" as strongest predictors of student engagement
• Walsh & Sattes' documentation that "wait time (3–5 seconds)" dramatically increases response quality
• Dillon's emphasis on follow-up moves as where dialogue succeeds or fails

Input Requirements

Required:

• Concept to develop (the understanding target)
• Student level (year group/familiarity)
• Starting point (current student beliefs)

Optional: target understanding, student profiles, time available, subject area

Output Deliverables

Structured sequences including:

• Opening and progressive questions with Paul & Elder type classifications
• 2–3 anticipated responses per question with branching follow-ups
• Facilitation guidance (wait time, handling silence, managing participation equity)
• Explicit Socratic vs. leading distinction with risk mitigation

Strategic Sequencing Principles

Questions progress from concrete examples toward abstraction and principle identification. The approach maintains genuine exploratory dialogue where multiple valid answers remain possible throughout.

What This Skill Does

The protocol distinguishes between three error types: procedural (incorrect method application), conceptual (fundamental misconception), and careless (execution slip despite correct understanding). Research demonstrates that "errors followed by corrective feedback produce stronger learning than errorless learning" due to prediction violation deepening encoding. However, each error type requires fundamentally different instructional responses—re-teaching for conceptual errors, guided practice for procedural errors, and metacognitive monitoring for careless mistakes.

Evidence Foundation

Research establishes that errors, properly analyzed rather than simply corrected, become powerful learning mechanisms. Borasi's work demonstrated errors as diagnostic windows into student thinking. Black & Wiliam identified error analysis as central to formative assessment's diagnostic power. Metcalfe's review confirms learning gains from error analysis. Siegler's microgenetic research showed mathematical development depends on understanding why incorrect strategies fail. Tulis et al. modeled productive error processing through three components: error detection, attribution, and correction strategy—all teachable through systematic analysis.

Evidence sources:

• Borasi (1994) — Errors as "springboards for inquiry" in teaching experiments
• Black & Wiliam (1998) — Assessment and formative classroom learning through error use
• Metcalfe (2017) — "Learning from errors: benefits of errors in the classroom"
• Siegler (2002) — Microgenetic studies on mathematical understanding development
• Tulis et al. (2016) — Individual error processing and learning models

Input Schema

Required:

• student_work_sample: Description of work containing errors
• task_description: What student was asked to do and learning objective
• subject_area: Subject and year group

Optional:

• correct_response: Model correct response for comparison
• student_profiles: Prior attainment, learning difficulties, error history
• rubric: Success criteria for the task
• error_frequency: One-off or recurring pattern indicator

Output Schema

• error_classification: Procedural, conceptual, or careless errors with evidence
• root_cause_analysis: Hypothesized cause with diagnostic questions
• targeted_response: Specific follow-up actions per error type
• student_self_analysis_guide: Scaffolded prompts for learner reflection

Known Limitations

1. Error classification requires working samples, not isolated answers. Diagnostic questioning is essential for accurate classification.

2. This skill addresses individual errors; whole-class patterns require instructional redesign through class-level diagnostic approaches.

3. Selective application is necessary given time constraints—use for persistent, surprising, or shared errors rather than comprehensive analysis of every student mistake.

Chains well with: feedback-quality-analyser, gap-analysis-from-student-work, metacognitive-prompt-library, worked-example-fading-designer

What This Skill Does

This intervention constructs a structured progression of tasks that methodically develops self-efficacy for learners who hold limiting beliefs about their capabilities in specific areas. Rather than relying on verbal encouragement alone—which research shows is the weakest efficacy intervention—the approach engineers genuine success experiences starting from the student's current competence level and advancing through carefully calibrated increments.

The method leverages Bandura's four efficacy sources in strategic sequence: mastery experiences (actual success at progressively challenging tasks), vicarious experience (observing similar peers succeed), verbal persuasion (credible, evidence-based encouragement), and physiological state management (reducing performance anxiety).

Evidence Foundation

Bandura's research established that "self-efficacy" is a central determinant of motivation and task performance, particularly noting it predicts outcomes even when controlling for actual ability. Hattie's synthesis identified self-efficacy as one of the strongest individual predictors of achievement. The critical insight: self-efficacy develops through accumulated evidence of capability, not reassurance.

Schunk and Pajares demonstrated that learners' confidence beliefs shape academic outcomes independent of actual ability level. Dweck's work on growth mindset complements this framework, though mindset messaging alone proves ineffective without genuine mastery experiences.

Key Design Principles

Start from demonstrable strength. Identify what the student already does competently that connects to the target skill. First task success should be near-certain.

Increment difficulty minimally. Each subsequent task represents one small step upward—students must perceive the connection between their last success and the new challenge.

Embed attribution coaching. After each success, guide learners to attribute outcomes to their specific strategies and efforts: "You used the planning method we practised" rather than "You're naturally talented."

Manage physiological state. Anxiety undermines self-efficacy. Employ low-pressure conditions, private rather than public performance, and normalise difficulty as evidence of learning.

Apply vicarious experience strategically. Show peers with similar starting points succeeding, not only high achievers, so the message is credible: "If someone like me can do this..."

What NOT to Do

Avoid generic praise ("You're so clever!"), unsupported encouragement, public comparison to peers, ability-focused language, and extrinsic reward systems. These undermine credibility and can damage self-efficacy if the student later encounters difficulty.