Sketch, Spec, Swap

At A Glance

The scene in 30 seconds: Groups of three (pre-paired A↔B) study a different real website for 5 minutes on one shared device, then put it away. From memory, each student sketches the page and writes a four-section spec - Project, Structure, Style, Behavior. Specs travel to the paired group; receiving students reconstruct the page from words alone, having never seen the source. Reconstructions come back, the live site reopens, and groups lay out the seven artifacts (3 originals, 3 outsider-reconstructions, the live page) to see where vague language broke down.

The scene teaches that a specification is empirically testable: when the reader has no prior knowledge of the artifact, only the words on the page can produce it - precise language reproduces, vague language does not. The cross-group exchange is what makes the test empirically clean; any divergence between the original and the reconstruction is purely a function of the words on the spec, with no memory of the live page filling in what the spec failed to capture.

One teacher rule: Don't write spec content for students. Don't pick design choices for them. Circulate to support, not to prescribe. If a student is stuck, ask "What did you see?" rather than telling them what to write.

What students leave with: A four-artifact packet (original sketch, completed spec, reconstruction of a peer's spec, written reflection) plus a take-home assignment to spec a different website at home for next class.

Paper conventions: Sketches and reconstructions are drawn on lightly-ruled graph paper (~5mm gray grid); the spec template is standard lined paper. Block print is the drafting convention - recommended for any handwriting on the in-class artifacts, required on the take-home spec so the artifact can be scanned and read by downstream digital tools later in the semester. The grid and the block print are the same convention engineers and architects used for paper specifications before CAD; same paper, same writing, same purpose.

---

Scene Schedule (Quick Reference)

Beat	Minutes	What students do	What teacher does
1. Settle	0-1	Find group station, glance at assignment card	Deliver 60-second framing (script in Beat 1)
2. Observe	1-6	Group observation of assigned website on shared device - no notes, no photos	Time-keep; signal at minute 5; devices go away
3. Sketch	6-14	Individually sketch the page from short-term memory	Circulate; support without prescribing
4. Spec	14-29	Complete the four-section spec template	Circulate; help with rough-notes use
5. Swap & Reconstruct	29-43	Hand spec stack to paired group; reconstruct a stranger's spec, knowing nothing about their site	Coordinate the exchange; enforce "no contact with the spec's author"
6. Compare & Close	43-50	Reconstructions return home; reopen own website; compare sketch / outsider's reconstruction / live site; write reflection; post one sentence on wall	Cluster posted sentences into 4 gap-types; deliver bridging statement

---

Materials Checklist (Day-Of)

Run through this list before students arrive:

• Assignment cards at each station - 1 URL per group of 3, with paired-group ID
• Shared device per station, pre-loaded to the group's URL, screen awake
• Backup-URL list on the teacher's desk
• Four-section spec template - 1 per student + accommodated formats (see Section V)
• "Rough notes - not assessed" scratchpad - 1 sheet per student
• Engineering graph paper (~5mm grid) - 2 sheets per student
• Pencils with erasers (no pens - students must be able to revise)
• Classroom timer visible to all students
• Wall / whiteboard cleared for Beat 6 sentence-posting
• Noise-canceling headphones available from the supply tray
• Low-traffic seating positions identified

---

I. Basic Information & Goals

• Scene Title: Sketch, Spec, Swap - A Paper Primer on Website Decomposition

• Concept/Topic to Teach: Decomposing a pre-assigned website, observed briefly with a small group at the start of the scene, into a four-section structural specification (Project, Structure, Style, Behavior), and empirically testing the precision of that specification through peer reconstruction.

• Standards Addressed:

  • H.AP.3A - Decompose problems into smaller components using constructs such as procedures, modules, and/or objects. (Students decompose a familiar website into the four structural categories of the course's spec format.)
  • H.AP.4B - Develop and utilize test cases to verify that a program performs according to the program's design specifications. (The peer's reconstruction operates as a test case: it verifies whether the spec, as written, is precise enough to reproduce its intended artifact.)
  • Core Practice 2 - Collaborating around computing. (Group work of three students; peer-to-peer specification exchange; collective gap analysis in Beat 6.)
  • Core Practice 3 - Recognizing and defining computational problems. (Translating a remembered visual artifact into a precisely-defined, decomposable description.)
  • Core Practice 4 - Developing and using abstraction. (Students abstract a particular visual page into descriptive language at four levels of granularity, discarding incidental visual detail in favor of structural essence.)
  • Core Practice 5 - Creating computational artifacts. (The four-section specification is itself a computational artifact, produced on paper.)

• General Goal(s): Establish, before any computer is touched, that the durable artifact at the center of this course is the specification rather than any particular execution of it. Introduce the four-section spec format the course will use throughout the year. Demonstrate, through paper-only peer exchange, that the precision of a specification is empirically testable: ambiguous language produces visibly divergent output, and that divergence is data rather than failure.

• Specific Objectives: By the end of the scene, each student will be able to:

  • (a) Analyze a pre-assigned website, observed briefly with their group at the start of the scene, by decomposing it from short-term memory into the four structural categories of the course's spec format - Project, Structure, Style, Behavior.
  • (b) Create a written four-section specification of that website using only language and structural notation, without visual representation.
  • (c) Apply the same decomposition process in reverse by reconstructing a peer's described website from specification alone, without prior knowledge of the original.
  • (d) Evaluate the precision of the original specification by comparing it against the peer's reconstruction and identifying, in writing, the specific words, phrases, or omissions that produced each observed divergence.

  All four objectives target the upper three tiers of Bloom's revised taxonomy (Analyzing, Evaluating, Creating), satisfying the MC35 Exemplary expectation that learning objectives center higher-order thinking.

II. Instructional Materials & Setup

• Required Materials:
  • Teacher: projector or wall-mounted display for the scene schedule and closing prompt; whiteboard or designated wall area for posting Beat 6 reflections; clock or timer visible to the class; a stack of lightly-ruled engineering graph paper (~5mm gray grid, two sheets per student) for the sketch and reconstruction beats; a pre-curated list of group website assignments (one URL per group of three, selected by the teacher in advance - see notes below).
  • Per group: one shared computer or tablet at each group's station, positioned in the center of the desk so all three students can see one screen; a printed assignment card at each station naming the group's URL and the scene's six-beat schedule.
  • Students: pencil and eraser (pens excluded - students must be able to revise); one printed four-section spec template per student on standard lined paper (see Section V for accommodated versions); two sheets of lightly-ruled graph paper for the original sketch and the reconstruction.
  • Prepared in advance: the four-section spec template handout, produced in the default printed format for the general roster and in three accommodated formats prepared in coordination with the campus accessibility coordinator (see Section V). All printed materials produced at least one week prior to the scene to allow time for braille embossing turnaround. The group website list selected at least three days prior, with criteria favoring single-page sites whose four structural categories (Project, Structure, Style, Behavior) are visible from the homepage without scrolling; sites with auto-playing media, aggressive interstitials, or chaotic feeds avoided; mix of structural complexity distributed across groups so that Beat 6's wall posting produces a variety of gap-types rather than a single repeated observation. The class is also pre-organized into paired groups (Group A ↔ Group B, Group C ↔ Group D, etc.) for the Beat 5 cross-group spec exchange; assignment cards include not only the group's URL but the identity of its paired group, so students know where their specs travel and where their reconstruction work comes back from.
• Anticipatory Set: The hook is the assignment itself. As students enter, each group's station already has a printed assignment card naming the URL it will examine, and the shared device at the station is open to that URL. Curiosity does the work that an oral prompt would otherwise have to do - students sit down, see the card, glance at the screen, and immediately want to know what they are about to study. The teacher delivers a single sixty-second framing as students settle: "Each of your groups has been assigned a different website. In five minutes you will look at it together, no notes, no photographs. Look at it the way a person who will have to describe it from memory should look at it. After the five minutes I will say stop, and the devices go away for the rest of class." The constraint - observing without recording - is the scene's first cognitive demand and is named explicitly at the outset so students understand they are practicing structured attention, not browsing.

III. Procedures

Timed for a 50-minute scene. All times below are minutes elapsed from the start of class. If you are on a longer block, keep the beat order and stretch each beat proportionally.

Beat 1 - Settle & Frame (0-1 min)

What to do:

1. As students enter, direct each to their group's station. The assignment card on each desk names their URL; the shared device is already on that page.
2. Once everyone is seated, deliver the 60-second framing aloud (see Script).
3. Start the Beat 2 timer (5 minutes).

Script (teacher, 60 seconds):

"Each of your groups has been assigned a different website. In five minutes, you're going to look at it together - no notes, no photographs. Look at it the way a person who is going to have to describe it from memory should look at it. After five minutes I'll say stop, and the devices go away for the rest of class."

Teacher Note (rationale). The constraint of observation without recording is the scene's first cognitive demand - students are practicing structured attention, not browsing. Naming the constraint explicitly at the outset is itself an accommodation for any student who benefits from having implicit expectations lifted (see Section V).

Beat 2 - Group Observation (1-6 min)

What to do:

1. Each group of three sits around its shared device on the URL printed on its assignment card.
2. Students may scroll the page and hover over elements, but they may not navigate beyond the homepage, take photographs, or write anything down during this beat.
3. Conversation within the group is permitted and actively encouraged.
4. At minute 5, signal time aloud. Chromebooks are closed (lids shut) and pushed to the side of the station, out of view for the rest of the scene - until Beat 6.

Teacher Note (rationale). Group conversation during observation is part of the cognitive work; surfacing what to look for with peers helps students decompose the artifact more deliberately than they would alone. This beat carries Louisiana standards H.AP.3A (decomposition) and Core Practice 4 (abstraction).

Beat 3 - Sketch from Short-Term Memory (6-14 min)

What to do:

1. Pass out one sheet of lightly-ruled graph paper (~5mm grid) per student.
2. Each student sketches the homepage of their group's assigned website from memory, top to bottom.
3. Deliver the sketching framing aloud (see Script).
4. Pencil only. Students must be able to revise.

Script (teacher, before sketching begins):

"This is not an art class. Show me where things sit and roughly how big they are - let the grid help you keep your proportions honest. Label anything the sketch can't show on its own - search field, button, video embed, sign-in form. Labels in block print, please."

Teacher Note (rationale). The graph paper is itself an engineering and architectural drafting convention - the same paper engineers used for system schematics before CAD. The grid does two things in this scene: it scaffolds spatial fidelity ("the hero spans the top three rows; the nav fits in two cells"), and it sets up consistent letter sizing for block-print labels, which makes the artifact machine-readable for downstream digital tools later in the semester. All three students in a group sketched the same source artifact, so small divergences between their sketches are normal - and informative. Don't smooth them over.

Beat 4 - Translate Sketch into a Written Specification (14-29 min)

What to do:

1. Pass out the printed four-section spec template (on standard lined paper), one per student. Also pass out one "rough notes - not assessed" scratchpad sheet per student.
2. Deliver the block-print framing aloud (see Script).
3. Each student translates their sketch into language, completing each section in full sentences:
   • Project - What is this website for, and who visits it? Two to three sentences.
   • Structure - What's on the page, top to bottom? A numbered list.
   • Style - Colors, type, density, mood. Three to five sentences.
   • Behavior - What's clickable, what moves, what changes when a user interacts? Three to five sentences.
4. Students may refer to their own sketch but not their peers' sketches. The specification has to stand on its own.
5. The rough-notes sheet is for drafting only - final, polished language goes on the formal template. Tell students they can scribble anything on the rough sheet without it counting against them.
6. Scribe option (available to any student). Within each group, students may decide among themselves to pair as author + scribe - one student dictates their spec aloud, the other writes it down on the dictator's template. The cognitive authorship of the spec (deciding what it says) stays with the dictator; the scribe transcribes only what they hear, without paraphrase. Pairings are chosen by the students, not assigned by the teacher. A student who wants their work to be in their own hand simply writes it themselves; a student who would benefit from delegating the writing act asks a groupmate to scribe.
7. Circulate. Support, do not prescribe. If a student is stuck, ask "What did you see?" rather than telling them what to write.

Script (teacher, after templates are out):

"Write your spec in block print - the same writing convention engineers used on paper specifications before there were computers to do it for them. Recommended for the in-class spec, required for the take-home version."

Teacher Note (rationale). This beat carries the scene's central cognitive load. Standards H.AP.3A (decomposition) and Core Practice 5 (creating computational artifacts). The rough-notes scratchpad and the scribe option are in the default flow for every student - not as special accommodations - because the cognitive separation between generating an idea, organizing it into the section structure, and physically writing it benefits most writers, dyslexic or otherwise. The scribe option also makes the act of dictating a spec into a low-stakes precursor to Beat 5: you have to be precise enough that another human writes what you actually meant. Letting students self-organize who scribes is itself the metacognitive read of "where do my strengths fit in this group?" - a Core Practice 2 (collaborating around computing) skill in practice.

Beat 5 - Cross-Group Swap & Reconstruct (29-43 min)

What to do:

1. Each group of three stacks its three specs together and hands the stack to its paired group (Group A → Group B, Group B → Group A). The exchange happens simultaneously and physically - students walk specs to the paired group's station, no digital transfer.
2. Within the receiving group, distribute the three incoming specs among the three students (one spec per student). Every student now holds a stranger's spec describing a website they have never seen.
3. The receiving group's shared Chromebook stays closed. The originating group's shared Chromebook also stays closed. No student in the room can see the source artifact during this beat.
4. Each student reconstructs the page on a fresh sheet of lightly-ruled graph paper, drawing only from what the spec says. Labels in block print (same drafting convention as Beat 3).
5. No contact with the spec's author. Deliver the developer-framing aloud (see Script).
6. At the end of the beat, each group stacks its three reconstructions together and walks the stack back to the originating group. Group A now has Group B's reconstructions of Group A's specs (and vice versa). Both groups carry both their original sketches and their outsider-produced reconstructions into Beat 6.

Script (teacher, as the swap begins):

"You're playing the role of the developer who receives the spec. You don't know the source. You can't ask the author what they meant. Draw only what's written. If something feels ambiguous, that's the data we're collecting - interpret it and move on."

Teacher Note (rationale). This is the scene's empirical test of specification precision and the activity that anchors Louisiana standard H.AP.4B (test cases verify program specifications). The cross-group exchange is what makes the test empirically clean: when reconstructors have no prior knowledge of the source website, any divergence between original and reconstruction is purely a function of the words on the spec - there is no memory of the live page filling in what the spec failed to capture. This is the design correction that the earlier within-group version of this scene lacked. The peer reconstruction is now a true test case in the H.AP.4B sense; divergence from the original is signal, not noise.

Teacher Note (logistics). Pair-group structure means uneven group counts cause friction. Plan for it: if the class has an odd number of groups, one trio of groups can be set up as a three-way ring (A → B → C → A) so every group has both an outgoing and an incoming exchange partner. If a group of three is reduced to two students by absence, their spec stack still goes to the paired group; the paired group simply distributes two specs across two reconstructors. The math works.

Beat 6 - Three-Way Comparison, Reflection & Close (43-50 min)

What to do:

1. Reopen the originating group's shared Chromebook. Return to the URL on the assignment card. The live website is now visible again, for the first time since minute 6.
2. Each group lays out seven items on the desk: 3 original sketches (their own), 3 reconstructions returned from the paired group, and the live website on the screen.
3. Each student walks through their own triplet - their original sketch ↔ the outsider's reconstruction of their spec ↔ the live site - and asks two questions aloud to their groupmates:
   • What in my spec produced the gap between my sketch and a stranger's reconstruction of it?
   • What did I miss about the live site even when I was looking at it?
4. Each student writes a one-paragraph reflection on the back of their original spec: "The words I would add to my spec to close those gaps." The same scribe option from Beat 4 remains available - a student who chose dictation earlier may continue with the same pairing for the reflection.
5. With 3 minutes remaining, one spokesperson per group writes a single sentence on the designated wall area: "The gap that surprised us most was - [completion]."
6. Once all groups have posted, read the sentences aloud and cluster them into four named gap-types: ambiguity, omission, implicit assumption, observational error.
7. Close with the bridging statement aloud (see Script).

Script (teacher, closing):

"When we work with the AI tool next week, you'll see the AI produce these same gap-types - faster and more visibly. The work we did on paper today is the durable work. The tool is the test apparatus that surfaces, in real time, what we already learned to look for here."

Teacher Note (rationale). The three-way comparison (sketch ↔ reconstruction ↔ live site) is the scene's richest moment. Students see what they thought they saw, what a stranger drew from their words, and what was actually there. Because the reconstructor was in another group and has never seen the source artifact, the divergence between sketch and reconstruction is uncontaminated by memory - it is a clean measurement of specification precision. The four named gap-types become the rubric vocabulary for the next scene (Website Observation and Description Activity). Reflections and sketches remain posted on the wall or filed in the student's folder. The reconstructions made by this group of the paired group's specs go back to the paired group's portfolio; each student keeps the seven artifacts that belong to their own work.

Independent Practice (Take-Home)

Assign each student: pick a different familiar website than the one your group used in class, and produce a fresh four-section spec on a clean template - this time with the page held open in front of you on a device rather than from memory. The take-home spec is written in block print throughout (not recommended this time - required). Bring it to the next class.

Teacher Note (rationale). This take-home spec is the entry artifact for the upcoming Website Observation and Description Activity, where students move from working from short-term memory after brief group observation (this scene) to sustained individual observation of a live artifact under a defined protocol (next scene). The block-print requirement here, with no in-class time pressure, prepares the artifact for a downstream tool feature already on the v2 roadmap: students will eventually photograph their paper specs and upload them, with a vision-LLM transcribing the four sections directly into the build_tool's spec panels. Block print on lined paper is what makes that OCR pathway reliable. By the time the feature ships, students will already have a portfolio of artifacts in the right format.

Contingencies - What to do if...

Situation	What to do
A group's website won't load.	Swap to a backup URL from the teacher's desk list. Don't spend more than 30 seconds troubleshooting; the scene is paper-based and any clean homepage works as the source.
A group finishes early.	See Section VI (Extensions). Direct early finishers to write a second spec for a different familiar website with a deliberately planted ambiguity - they predict on the back where the ambiguity will produce a reconstructor error.
A student is absent and a group is down to two.	The two-student group still produces two specs; both specs go to the paired group. The paired group has three reconstructors but only two incoming specs, so one student in the paired group doubles up with one of the other two (reads the same spec and produces an independent reconstruction). This yields a small bonus - two reconstructions of the same spec - which is itself informative data about how different readers diverge on the same words.
A student does not want to share their sketch.	The sketch never leaves the originating group; only the spec travels to the paired group. So the student's spec gets reconstructed by a stranger but their sketch stays with them. This actually lowers the social cost of the scene - the artifact being evaluated by outsiders is the words, not the drawing. Encourage participation.
A paired group falls behind and reconstructions don't return in time for Beat 6.	The originating group can still run a partial Beat 6 with the materials they have - their original sketches and the live website - and the missing reconstructions can be slid in during the next class's warm-up. Don't let the missing piece block the rest of the closure.
Tech failure across the room (no devices working).	Pre-printed screenshot packets of each assigned website serve as the fallback source. Students get 5 minutes with the packet during Beat 2 instead of the live device. Keep one packet per assigned URL in your file drawer.
Pacing is running behind.	Protect Beat 5 (swap & reconstruct) at all costs - it is the scene's lever. Trim Beat 6 first; if needed, push the reflection writing to the take-home assignment.
A student becomes overwhelmed by Beat 5 (rare).	Walk them through the stranger's spec aloud, helping them parse what it says without telling them what to draw. The cognitive act of interpreting the spec remains theirs.

IV. Assessment & Evaluation

• Assessment Based On Objectives: A four-row rubric, applied to each student's complete four-artifact packet - original sketch, completed specification, reconstruction-of-a-peer's-spec, and closing reflection. Each row is scored on a three-point scale (1 - does not meet, 2 - meets, 3 - exceeds), weighted equally. Maximum possible: 12 points.

Row	Criterion	Cognitive Level	Standards
(a)	Decomposition. Specification organizes the website into all four required categories, each populated with content appropriate to that category (Project content does not appear in Structure, Style choices are stated rather than implied, etc.).	Analyzing / Creating	H.AP.3A; CP3; CP4
(b)	Specification precision. Specification contains sufficient unambiguous detail that a peer with no prior knowledge of the site can reconstruct it recognizably. Precision is judged against the peer's actual reconstruction, not against a hypothetical perfect reader.	Creating	H.AP.4B; CP5
(c)	Reconstruction fidelity. Student-as-reconstructor renders their peer's specification accurately, drawing only what is written and resisting the impulse to import personal assumptions about how the page "must" look.	Analyzing	H.AP.4B; CP2
(d)	Reflective analysis. Closing paragraph identifies specific words, phrases, or structural omissions in the original specification that produced each observed divergence in the reconstruction. Generic reflections ("I should have written more") do not meet; targeted reflections ("I wrote 'a search bar at the top' but did not specify whether it spanned the full width, which is why my partner drew a narrow centered search field") meet or exceed.	Evaluating	H.AP.4B; CP3

The rubric is shared with students at the start of Beat 2 so the assessment criteria are visible throughout the scene, not revealed after the fact.

V. Differentiation & Modifications

The unifying differentiation strategy in this scene is the printed handout. All students receive the four-section spec template on paper. The format of that printout - its typography, its embossing, its supplementary scaffolding - is the lever adjusted per learner profile. This design avoids singling individual students out (every student in the room receives a printed handout, by design) and instead makes accommodation a property of the artifact, not of the recipient. The strategy also models, structurally, a principle the course teaches: a good specification survives translation into multiple formats without losing meaning.

1. Dyslexia

Research basis: arXiv:2207.02308 (generative-AI writing accommodations for dyslexic, autistic, and ADHD learners). The research identifies the cognitive load of simultaneously generating, formatting, and surface-correcting written content as a primary barrier for dyslexic writers. Accommodations in this scene target the delivery channel and the load distribution rather than the cognitive demand of the task itself.

• Accommodation 1: Dyslexia-Supportive Print Format of the Spec Template

  • The Strategy: The student receives the same four-section spec template as the rest of the class, printed in a dyslexia-supportive format - OpenDyslexic or Lexend font face, 14-point body type, increased letter- and line-spacing, left-aligned text rather than justified, and off-white (cream) paper to reduce visual contrast strain.
  • MC35 Exemplary Rationale: This adjustment changes the visual presentation of the scene's input artifact without altering its cognitive demand. The student decomposes the same website, completes the same four spec sections, and produces the same assessed artifact as peers, with the visual barrier to reading the template's prompts removed.

• Accommodation 2: Flexible Drafting Workspace ("Rough Notes" Scratchpad)

  • The Strategy: A separate sheet of plain paper, marked rough notes - not assessed, sits beside the formal spec template during Beat 4. The student may draft sentence fragments, exploratory phrasing, and false starts on this sheet before transcribing finalized language into the four-section template. The separation lets the student focus first on the cognitive act of expressing what they want to say, and only second on placing it within the assessed section structure. This is the paper analogue of a digital scratchpad: it isolates gist composition from form composition, which research suggests reduces overall cognitive load for dyslexic writers.
  • MC35 Exemplary Rationale: Simultaneous demands of generating ideas and formatting them produce a compounded cognitive load that disproportionately affects this student. Separating the drafting act from the formatting act reduces the executive function burden without altering what the student is being assessed on. The assessed artifact (the completed template) remains the student's own original composition; the scratchpad is a temporary scaffold for getting there.

• Accommodation 3: Peer Scribe Option for Beats 4 and 6

  • The Strategy: If handwriting latency is a barrier during Beat 4 (composing the specification) or Beat 6 (writing the closing reflection), the student may dictate their content aloud to a designated peer scribe at the group station. The cognitive act of authoring - choosing the words, organizing the sections, selecting the precision - remains entirely the student's own. The peer scribe transcribes verbatim, without paraphrase or editorial intervention.
  • MC35 Exemplary Rationale: This separates the motor and orthographic acts of writing from the cognitive act of composition. The student demonstrates the same decomposition, the same precision, and the same reflective analysis as their peers - they are simply dictating rather than handwriting their output.

• Accommodation 4: Opt-Out from the Block-Print Convention

  • The Strategy: The class-wide convention of writing handwritten artifacts in block print - a drafting-tradition convention that scaffolds both precision and downstream OCR - is recommended but not required for this student on either the in-class artifact or the take-home spec. The student may write in whatever handwriting feels natural to them. The downstream OCR pathway that block print was meant to enable can still be reached for this student through the peer-scribe pathway (Accommodation 3) on the in-class artifact, and on the take-home through one of: a family member or assistive technology acting as scribe, a typed submission entered directly into the digital spec template, or speech-to-text dictation captured on the student's device.
  • MC35 Exemplary Rationale: Block print is a handwriting-style convention, not a cognitive-content convention. Requiring it of the dyslexic student would add a motor-and-letterform load that competes with - rather than supports - the cognitive work the scene is actually assessing (decomposition, precision, reflective analysis). The opt-out removes the competing demand without changing what the student is assessed on. The downstream machine-readability that the convention was designed to scaffold remains reachable for this student through alternative production formats, so the student is not excluded from the eventual paper-to-tool pipeline; they reach it by a different path.

• Modification: Reduced Spec Scope (Two-Section Specification)

  • The Strategy: If the cumulative writing load across all four spec sections exceeds the student's capacity within the scene's window, the assigned scope is reduced from four sections to two. The student selects, in consultation with the teacher, the two sections they will complete in full - typically Structure plus one of the remaining three. The reduction is in breadth rather than depth: each completed section must meet the same precision standard the full-scope rubric demands of that section. The scope reduction is documented explicitly in the assignment record so the student and any reviewer know what was produced and what was scoped out.
  • MC35 Exemplary Rationale: This is a modification because the assessed scope changes; the student produces less material than peers, not the same material through a different access path. Defensible against the International Dyslexia Association's reduced-workload modification example, with the standard explicitly named (number of sections, not depth within sections). The scene's central cognitive objectives - decomposition into structured language, peer-testable precision, evaluative reflection on observed gaps - are each demonstrable within a two-section specification. A student who produces a focused, precise two-section spec has met the modified standard; a student who produces a vague or technically incoherent two-section spec has not.

2. Asperger's Syndrome

Research basis: arXiv:2403.03297 (writing-accommodation research for autistic learners). The research identifies navigating ambiguity as the most commonly cited classroom and workplace barrier for autistic adults and adolescents, with judgmental reactions to clarification requests producing exhaustion, anxiety, and overthinking. Accommodations in this scene remove implicit instruction and sensory load without lowering the standard the student is held to; modifications adjust assessed criteria away from cultural-rather-than-technical conventions of writing and interface design.

• Accommodation 1: Explicit Written Assignment Specification

  • The Strategy: The assignment card placed at each group's station is structured into named, unambiguous sections - Goal (what we are producing today), Required Components (the four-artifact packet: sketch, spec, reconstruction, reflection), Acceptance Criteria (the rubric, attached), Examples of Strong Work (one annotated sample packet from a prior cohort, if available), and Common Pitfalls (the four named gap-types and what they look like). The student who reads the assignment card has the same information as the student who relies on verbal aside, classroom Q&A, or hallway clarification.
  • MC35 Exemplary Rationale: This accommodation changes the access path to the scene's expectations, not the expectations themselves. The student is held to the same rubric and produces the same artifacts; what changes is that ambiguity that would otherwise be resolved through social channels is resolved structurally, in writing, before the scene begins. The student who acts on the written assignment correctly has met the standard regardless of whether they caught every verbal aside.

• Accommodation 2: Section-Header Prompt Scaffolds on the Spec Template

  • The Strategy: The student's printed spec template includes a small bordered prompt card under each of the four section headers identifying the kind of content that belongs there - "Project: who visits and what they came for"; "Structure: top-to-bottom list of every named element on the page"; "Style: colors, type, density, mood - name them"; "Behavior: what is clickable, what moves, what changes when a user interacts."
  • MC35 Exemplary Rationale: The four-section structure of the scene is itself a strong cognitive scaffold for a student who benefits from explicit organizational frames. The additional in-template prompts extend that scaffold by removing the implicit step of inferring what content fits each header, without altering the cognitive demand of generating the content itself. The student performs the same Analyzing, Creating, and Evaluating work as the rest of the class.

• Accommodation 3: Sensory and Environmental Control

  • The Strategy: The student may select their group's workstation from designated low-traffic positions along the back wall of the lab. Noise-canceling headphones are available from a small classroom supply for student checkout at the start of the scene. The student's individual monitor brightness may be dimmed without justification. Eye contact is not required during one-on-one teacher feedback. These options are framed as available to any student in the class so the autistic student is not visibly singled out.
  • MC35 Exemplary Rationale: Research identifies bright overhead lights, unpredictable group interaction, and forced social contact as environmental factors that compound communication challenges and consume working memory the student would otherwise direct toward the task. Reducing sensory load does not change the assessed task; it removes competing cognitive demand from the student's working-memory budget. The student produces the same artifact under the same standard, in conditions that do not compete with the cognitive work for attention.

• Accommodation 4: Pre-Scripted Clarifying Questions for Beat 5

  • The Strategy: A printed clarifying-questions card sits at the student's station with three pre-written question stems they may use within their group during the Beat 5 swap-and-reconstruct activity: "Can you point to where in your spec it says ___?", "I drew ___ here - was that what you meant?", "What was the next-most-important detail you wanted to convey?" Group composition is determined in advance with attention to social compatibility.
  • MC35 Exemplary Rationale: The script removes conversational ambiguity from a beat of the scene that would otherwise demand on-the-fly social interpretation. The student exchanges the same artifacts and performs the same comparative reasoning as their peers; the script removes the procedural overhead of figuring out how to enter the conversation.

• Modification 1: Acceptance of Direct, Unhedged Technical Prose

  • The Strategy: The rubric does not penalize the student for spec language that reads as blunt, technical, or unconventionally direct in tone, provided the language remains precise and unambiguous. A student who writes "the search field at the top of the page is approximately 480 pixels wide and aligned to the left edge of the header" meets the precision standard at least as well as a student who writes the same observation in softer or more conversational language. What the rubric drops, for this student, is the cultural expectation of conversational warmth in technical writing - a neurotypical default rather than a property of effective specifications.
  • MC35 Exemplary Rationale: This is a modification because the assessed criteria change; specific writing conventions that exist in the standard rubric are removed from the modified standard. The cognitive work the student demonstrates - decomposition, precision, evaluative reflection - remains identical and is, if anything, more directly assessable when conversational filler is not penalized. Many downstream uses of structured specifications (in industry technical writing, in code-generation pipelines, in design documentation) actively prefer the directness the modified rubric permits.

• Modification 2: Alternative Production Formats for the Specification

  • The Strategy: Where the prose-composition act itself is the barrier rather than the underlying competency, the student may demonstrate the same decomposition through one of three alternative formats: (a) an annotated sketch in which each labeled element corresponds to a brief spec-style description written beside it on the same page; (b) a five-minute recorded verbal walkthrough on the student's phone in which they describe each of the four spec sections aloud; or (c) a structured Q&A defense in which the teacher reads four pre-defined questions (one per spec section) and the student responds in whatever format works for them. The technical content being assessed is identical to what would be assessed in a written prose spec - decomposition into the four categories, precision sufficient for peer reconstruction, internal consistency.
  • MC35 Exemplary Rationale: This is a modification because the production format changes; the same cognitive competency is evaluated through a different medium. A student whose strength is in structural thinking but whose prose composition burns disproportionate cognitive effort gets to demonstrate what they actually know. The modified standard is not easier; it is rendered in a different format. The student must still produce a specification sufficient for peer reconstruction in Beat 5 - what the peer receives is a transcript of the verbal walkthrough or a written summary of the Q&A defense rather than a handwritten prose spec.

3. Visual Impairment (Computer Lab Focus)

• Accommodation 1: Specification Handout in Braille
  • The Strategy: The same four-section spec template, produced in advance of the scene on the campus braille embosser in coordination with the school accessibility coordinator. A refreshable braille display may be substituted if the student's IEP designates one as their working medium. Production lead time is one school week to accommodate embosser scheduling.
  • MC35 Exemplary Rationale: This provides the student with the scene's central artifact - the structured spec template - in their primary reading medium, ensuring that the cognitive work of Beats 3 and 5 (composing and reflecting on a specification) occurs without a translation barrier. The student is challenged by the scene's core concepts rather than by its delivery format.
• Accommodation 2: Screen-Reader Audio Observation in Beat 2 (group-shared)
  • The Strategy: The student's group's shared device is configured in advance with a screen reader (VoiceOver, NVDA, or JAWS depending on platform and IEP designation), and the group's assigned website is verified by the teacher to be accessible - heading structure intact, alt text present, no inaccessible interactive elements. During Beat 2, the group's five-minute observation is conducted primarily through audio: the screen reader narrates the page through a small speaker or shared headphone splitter, and the sighted students in the group experience the website's structure the way a screen-reader user does. The visually impaired student leads the group's navigation by directing which heading or landmark to read next. Sighted group members may also glance at the visual layout, but the primary cognitive channel for the group is the audio narration.
  • MC35 Exemplary Rationale: This accommodation does not merely include the visually impaired student in Beat 2; it places them in the group's navigational lead. The sighted students experience accessibility from the user's side and learn, embodiedly, that a well-structured website is not a visual achievement but a semantic one. The visually impaired student exercises the same Analyzing skills as the rest of the class, and arguably operates at the highest level of fluency within their own group during this beat. The activity reinforces, rather than dilutes, the scene's emphasis on higher-order thinking - and produces a cross-cohort learning outcome (sighted students gain a primary experience of screen-reader navigation) the scene would otherwise not offer.
• Accommodation 3: Spec Authoring in Braille (Beats 3 and 4)
  • The Strategy: The student composes their sketch-equivalent and four-section specification in braille - either via the campus braille notetaker (e.g., BrailleNote, Mantis) if the student's IEP designates one, or on a Perkins brailler producing physical embossed output. The student's spec is exchanged in Beat 5 with a sighted peer, who either reads the braille directly if proficient or works from a transcribed printed copy produced by the teacher or accessibility coordinator. In Beat 6, the peer's reconstruction is described aloud to the student so the comparison remains accessible. The student's closing reflection is composed in braille and remains in their portfolio as the durable artifact of the scene.
  • MC35 Exemplary Rationale: The student authors the scene's central artifact - the four-section specification - in their primary literacy medium and exchanges it with a peer through the same artifact-swap mechanism as the rest of the class. The cognitive work of Creating, Analyzing, and Evaluating is performed in full, and the durable artifact the student leaves with is in their own medium of fluency rather than a translated approximation.
• Modification: Structural Inventory in Lieu of Visual Sketch (Beat 3)
  • The Strategy: If a student cannot produce the visual sketch (Beat 3), they make a numbered list instead of a drawing. The list describes the page from top to bottom: what sections are there, what headings appear, what buttons or search fields or sign-in forms exist, and in what order a user would encounter them. Same rules as the sketch - done from memory after the device is closed at minute 5, pencil so they can revise, labels in block print, no looking at the live site. They can dictate the list to a groupmate using the scribe option from Beat 4 if they prefer. The list stays with the student through Beat 4 the same way a sketch would - they use it to help write their spec, but it does not get handed to the paired group in Beat 5. In Beat 6, lay out the seven items as usual: their list counts as the "original sketch," the reconstruction comes back from the paired group as a drawing, and the live site reopens for comparison. The student answers the same two reflection questions as everyone else. Note: review the student's IEP (or consult with the team) for the production format the student has demonstrated the most success with for "written work" - handwritten, brailled, typed, or dictated.
  • MC35 Exemplary Rationale: This is a modification because the production format of the Beat 3 artifact changes - structural language replaces a visual drawing - and the change is documented in the assignment record. The cognitive work the scene actually assesses (decomposition of a remembered web page into ordered structural elements) is performed in full; if anything, the format shift makes the decomposition step more explicit, since the student names each element rather than relying on spatial arrangement to imply it. The modification preserves the spec-traveling-to-the-paired-group dynamic of Beat 5 (the paired group still receives a written four-section spec, not the inventory list) and the seven-artifact layout of Beat 6.

VI. Extensions & Connections

• Adaptations: See Section V.

• Extensions (Gifted Students / Early Finishers): A student who completes the four-artifact packet ahead of their group may compose a second specification - this time for a different familiar website, and deliberately written with a single planted ambiguity in one specific section. The student writes, on the back of the second template, a sealed prediction identifying where the ambiguity lives and what specific divergence they expect a reconstructor to produce. This extension invites the student to use the specification format generatively - to probe and exploit the limits of language rather than merely to describe - and produces a self-designed test case in the H.AP.4B sense. The sealed predictions can be opened and verified as a warm-up at the start of the next class.

• Possible Connections:

  • Immediate next scene (Course-internal): The take-home specification produced in Independent Practice becomes the entry artifact for the Website Observation and Description Activity. The transition is from working from short-term memory after a brief group observation (this scene) to sustained, individual observation of a live artifact with a defined protocol (next scene). The four named gap-types surfaced in this scene's closure (ambiguity, omission, implicit assumption, observational error) become the rubric vocabulary the next scene uses.
  • Course arc: This is the founding paper-scene of the course. Every digital scene that follows - the introduction to the build tool, the spec-revise loop, the commenting-practice unit - refers back to the four-section format students learned to write here, on paper, in their own handwriting.
  • English / Technical Writing: The four-section specification is a genre of technical writing. The scene parallels English composition's instruction on precision and audience awareness, and can be paired with a cross-curricular reflection if the school offers integrated humanities programming.
  • Industry connection: Software engineers and product designers compose specifications professionally - variously called Product Requirements Documents (PRDs), design briefs, or technical specifications. The cognitive work of this scene is the cognitive work of those professional documents, scaled to a single page of paper.
  • The transferable claim: This scene is the first opportunity to make explicit, in language students can hold in their hands, the curriculum's central claim - expressing intent precisely enough that another reader (a person, a model, or future-you) can produce the right artifact from it is the most transferable AI-coding skill, and it operates the same way whether the reader is a peer, a small local language model, or a frontier system. Pencil and paper are the appropriate medium for the first encounter with that claim.