Chernobyl: Causes and the Changing Verdict

This is a ready-to-teach unit built around a String Theory board, the third in a series that began with The Great Gatsby and Romeo and Juliet. Like those, it is written in Understanding by Design (backward design) at the unit level. The daily sequence uses the 5E cycle (Engage, Explore, Explain, Elaborate, Evaluate) rather than gradual release, because this is inquiry social studies and 5E is the register those classrooms already use.

The unit has two organizing tasks, and the board is built so that the second grows out of the first. First, students string every cause of the explosion of Reactor 4 into the event and label how deep each cause runs: Underlying, Long-term, Short-term, or Trigger. Second, they put the official explanations on the board as sources and watch the verdict move: the 1986 account blamed the operators and six men went to prison on it; the 1992 account, from the same international body, found the reactor’s design principally at fault. The closing argument asks which account the evidence supports, and the board is where the evidence lives.

Before anything else: this is a real event

The literature units in this series put fictional people on a board. This one puts real people on it, most of whom died badly, some of whose children and spouses are alive, and six of whom were convicted on an account of events the record later revised. Three things follow for the classroom.

• The operators are not suspects. String Theory began as a tool for true-crime hobbyists, and the vocabulary of that world (suspects, motives, the pinboard with red string) will reach for the operators of Unit 4 if nobody stops it. The History template exists partly to stop it: its strings are causes, sources and affiliations, not accusations, and its Basis field asks what the record supports rather than who did it. Hold the class to that frame. A student who writes “Dyatlov’s fault” on a card has written a 1987 verdict, not a historical claim, and the unit’s whole second half is about what happened to that verdict.
• The dead are named respectfully and described clinically. Acute radiation syndrome is described in one clinical sentence; doses are numbers on a card, not prose; no hospital or autopsy imagery goes on any board, licensed or not. Students may have seen such images. The rule is still no.
• The site is in a war. Russian forces occupied the Chernobyl site from 24 February to 31 March 2022, and a drone struck the New Safe Confinement on 14 February 2025. Some students will have family in Ukraine, Belarus or Russia. Keep the 2022 and 2025 events on the board, Documented and IAEA-sourced, one sentence each, and keep the unit’s argument where the evidence is: 1986 to 1992.

Stage 1: Desired results

Enduring understandings

Students will understand that:

• Causes have depth. A design decision made in the 1960s and a button pressed at 01:23:40 are both causes of the same event, and a historian’s first job is to say which kind each one is. Calling everything “a cause” explains nothing.
• The official account of an event is a source, with a date and an author, and it can be contradicted by a later official account from the same body. “Documented” means a document says so, not that the matter is settled.
• An explanation can be adopted, acted on, and later revised, and the actions taken on it (a trial, a prison sentence, a safety reform) do not get revised with it.
• What people were told, and when, is a historical event in its own right, with its own causes and its own consequences, separate from the thing they were told about.

Essential questions

• How deep does a cause have to run before the people at the controls stop being the explanation?
• The same agency published two accounts six years apart. What changed: the evidence, or who was allowed to say it?
• Forsmark detected the release before Moscow admitted it. What does it mean when the first reliable source about an event is a thousand kilometers away?
• The 1987 court convicted six men on the 1986 account. What, if anything, does a historian owe people who were judged on a verdict that moved?

Standards

C3 Framework for Social Studies (grades 9 to 12):

• D2.His.14.9-12: analyze multiple and complex causes and effects of events in the past.
• D2.His.10.9-12: detect possible limitations in various kinds of historical evidence and differing secondary interpretations.
• D2.His.16.9-12: integrate evidence from multiple relevant historical sources and interpretations into a reasoned argument about the past.
• D3.1.9-12: gather relevant information from multiple sources representing a wide range of views while using the origin, authority, structure, context, and corroborative value of the sources.
• D4.1.9-12: construct arguments using precise and knowledgeable claims, with evidence from multiple sources, while acknowledging counterclaims and evidentiary weaknesses.

Common Core Literacy in History/Social Studies:

• RH.11-12.1: cite specific textual evidence, connecting insights from specific details to an understanding of the text as a whole.
• RH.11-12.2: determine the central ideas of a primary or secondary source and summarize how they develop.
• RH.11-12.3: evaluate various explanations for actions or events and determine which best accords with textual evidence, acknowledging where the text leaves matters uncertain.

Social studies standards are a different family from ELA, so this header is not interchangeable with the literature units’. Outside C3, the unit’s skills sit in the historical-inquiry strand every state framework carries: TEKS World History Studies (§113.42) social studies skills, Florida SS.912.W.1, Virginia WHII.1. Swap the codes in the header and the unit is unchanged.

Stage 2: Assessment evidence

Summative, part one: the board

Each student (or pair) submits one String Theory project, which must contain at least the following:

Every causal string carries a Cause type. The History template puts a Cause type field (Long-term / Short-term / Trigger / Underlying) on contributes_to, triggers, results_in and results_from. A causal string with the field blank is graded as absent, exactly as a string with no Significance is in the literature units.

Every card carries a Basis and a Source. The History template’s Basis is a strict four-way split: Documented, Inferred, Disputed, Legendary. A board where everything is Documented has not read the sources closely enough to find where they disagree. A board with no Legendary cards has not noticed what students arrived already believing.

Every Disputed card has at least two Source cards strung to it, with a contradicts string between the sources. Disagreement is drawn, not described.

Every image or media item carries a License and an Attribution. This is a gate rather than a rubric row; see the ground rules below.

Summative, part two: the argument

A two-page document-based argument, due after Day 5: To what extent were the operators of Unit 4 responsible for the explosion? The essay must cite at least two Source cards from the student’s own board that contradict each other, must use the cause-type labels as analytical vocabulary (a claim that a cause is Short-term rather than Underlying is a claim, and has to be defended), and must say where the record leaves the matter uncertain rather than resolving the uncertainty by assertion. The AP causation rubric grades it cleanly; so does any state argument rubric.

Formative

• Day 1 exit ticket: one Person card from the night shift with a correctly assigned Basis on every field and the Source filled.
• Day 2 column check: a screenshot of the student’s causes board with every string into the explosion labeled, and the card sitting in the column its label says.
• Day 3 peer audit: students swap projects and flag any Disputed card with fewer than two sources, any Cause type they would relabel, and any Basis they would dispute.
• Day 4 Timeline check: a screenshot of the Timeline view with the night placed to the second and the first week placed to the hour.

Rubric (4 points per row, 20 total)

Before you start: media, citation and AI ground rules

Set these rules before the first image goes onto a board, not after. Twenty minutes on Day 0 is the whole cost, and with this subject the rules do more than protect the school: they keep the board honest about real people. The Gatsby unit walks the License dropdown value by value; what follows is the same material in the form this subject needs.

Use the License dropdown as the mini-lesson

Every media item in String Theory carries License, Attribution and Source fields, and the License dropdown offers exactly seven values. Walk the list on the projector, because students will be choosing from it all unit.

The contrast worth teaching: the pictures everyone knows are the ones nobody may use

Chernobyl is the sharpest version of this lesson in the series, because the subject’s visual memory is almost entirely copyrighted.

• The photographs everyone recognizes (the reactor from the helicopter, the liquidators on the roof, the abandoned classrooms) are overwhelmingly the work of Igor Kostin, who photographed the site from the first day until his death in 2015. They are All Rights Reserved, held by his estate and agency. A student who “finds one on Google” has found a copyrighted work, and “it’s for school” is an argument, not a permission.
• The places and the machine are well covered under free licenses. The IAEA Imagebank publishes photographs of the plant, the 1986 Vienna meeting, the sarcophagus and the New Safe Confinement under CC-BY-SA; Wikimedia Commons has thousands of Exclusion Zone photographs under CC-BY and CC-BY-SA, and the RIA Novosti archive’s Soviet-era photographs there carry CC-BY-SA as well. This is why a well-built board puts its visual weight on the reactor, Pripyat and the confinement rather than on faces.
• The people are poorly covered, and the honest result is a card with no photograph. Memorial portraits of the firefighters and operators circulate widely and almost never carry a license anyone can establish. A Person card without a photo is a deliberate card, not a gap, and the Description field does the work a portrait would. Say this to the class on Day 0 so nobody treats an empty photo well as a failure.
• The HBO series is fully in copyright. A still of Jared Harris is All Rights Reserved. It is also a still of an actor playing a real man, which is a second problem the license does not touch.

Send students somewhere good: the IAEA Imagebank and Wikimedia Commons will cover nearly everything this unit needs, and each states a license on the item page. A short whitelist makes most of the Unknown problem disappear before it starts.

Fair use, said honestly

Fair use is a four-factor balancing test: the purpose of the use, the nature of the original, how much is used, and the effect on the market for it. Educational context helps under the first factor. It does not decide the question, and it is not a checkbox. Students should reach for Fair Use rarely and be able to say which factors they are relying on. Kostin’s photographs are licensed commercially to this day, which is the fourth factor arguing against a fair-use claim on them.

One thing that is specific to this tool and genuinely matters: a String Theory project can be published to a public link. A board seen by one class of thirty is a very different fair-use posture from the same board on the open web. Decide, as a class and in advance, whether boards get published at all. If any will be, hold those to a stricter standard: public domain and Creative Commons only, no fair-use claims. A project can publish the whole project or only the boards you pick, and what a published board hides never leaves the server.

AI use: four separate questions, decided in writing

The literature units ask three. This subject adds one, and it comes first.

1. AI-generated likenesses of real people: no. The literature units offer an AI-portrait activity, because Daisy Buchanan and Mercutio have no faces until a reader supplies one. Leonid Toptunov had a face, a mother, and a grave in Mitino cemetery. A generated “portrait” of him is a fabrication placed in the historical record next to his name, and the “AI-generated” label does not fix that. The same goes for generated scenes of the control room, the roof of Reactor 3, or Hospital No. 6. This unit has no AI-image path, by design.
2. AI-generated facts. A chatbot supplies a date, a dose, a quotation, or a sequence of events, and it goes on a card. The rule is simple: nothing goes on a card without a Source card it traces to. The Source field on every card is where the trace lives, and “ChatGPT” is not a source. Language models are confidently wrong about this event in specific, recurring ways (they reproduce the series’ timeline, invent quotations for Legasov, and move the helicopter crash to the first week), and students should be told that in advance so they treat a fluent answer as a lead to check rather than a finding.
3. AI-generated analysis. A chatbot assigns the Cause types, writes the Significance notes, or decides which sources contradict which. This guts the assignment: the cause-type judgment, the Basis call and the contradicts string are the thinking being assessed.
4. AI as a study aid. A student asks what a void coefficient is, what glasnost meant in 1986, or how to pronounce Shcherbina. This is fine, and drawing the line explicitly is what makes the other three rules credible.

About the app itself: the only AI feature a student can reach in String Theory is the Appearance dialog’s “Generate with AI” tab, which hands you a prompt to run in your own AI tool and paste back, and what it produces is a skin: colors, fonts, density. It generates no images and writes no analysis. Any AI content on a student’s board arrived from outside the app.

Names, and which spelling goes on the card

The place names and most of the personal names in this story have Ukrainian and Russian forms, and the Soviet-era documents use the Russian ones. Agree on a convention on Day 0 and apply it everywhere: the usual one is conventional English for the title and the familiar names (Chernobyl, Pripyat, Kyiv), the Ukrainian form for Ukrainian people (Pravyk rather than Pravik), and the other forms recorded in the card’s Also known as field, so the search box finds the card under any spelling. When quoting a 1986 source that says “Kiev,” quote it as written. A student who notices that the spelling itself has a politics has noticed something true, and it is worth five minutes of class.

Citation, and one limitation to disclose up front

Source cards (an Object / Artifact with Object type set to source) plus the per-media license and attribution fields feed the Sources view, which assembles a bibliography as a by-product of building the board. Show students this on Day 1.

Be straight with them about the limit: the Sources view is not Chicago or MLA. It renders a readable one-line citation per source, not a style-guide entry. If your department requires Chicago notes or MLA 9, students will reformat the list by hand, and they should hear that at the start of the unit.

Make attribution a gate, not a rubric row

Any board submitted with an attached media item marked Unknown, or with a blank Attribution, is returned unmarked rather than marked down. Attribution is a condition of submission in the same way a name at the top of a paper is.

The gate is not a trap, because there is always a way through it. By the deadline every attached image is in one of three states: its license was determined, or the student is making a Fair Use claim they can defend on the factors, or the image is not on the board. With this subject the third option is the common and correct one for people, and a board whose Person cards are mostly photo-less is a board that did the research.

Stage 3: Learning plan

Day 1: Engage. “Not great, not terrible.”

Open with what the room already believes. Ask for everything students know about Chernobyl and write it on the board without comment: the number 3.6, the divers, the bridge, the helicopter falling, the scientist who told the truth. Most of it comes from the 2019 series, and some of it is false. Leave the list up all week; it is the Legendary column waiting to happen.

Create the project on the History template and build the one card the whole project turns on, Reactor 4 explodes, together on the projector. Type = Disaster. When = 26 April 1986, 01:23:43, and show the class that the field takes a time: the night will be ordered to the second. Level = Primary. Then do the Basis move explicitly, because this template has four values and the fourth is the one students have not met:

• “AZ-5 was pressed at 01:23:40” is Documented. The SKALA computer logged it; INSAG-7 reproduces the record.
• “Toptunov pressed AZ-5 expecting it to shut the reactor down” is Inferred. It is what every account concludes and what no record states, because the man who could have confirmed it died on 14 May.
• “The second explosion was a hydrogen explosion” is Disputed. The sources genuinely divide (steam, hydrogen, a small prompt-criticality excursion), and a Disputed card gets two Source cards and a contradicts string between them before the week is out.
• “The dosimeter read 3.6 roentgen, ‘not great, not terrible’” is Legendary. The plant’s dosimeters did max out at 3.6 roentgen per hour, which is Documented and is the real story; the line is a screenwriter’s. A Legendary card is not a mistake to delete. It is a record of what people believe and where it came from, and it carries its source like any other: in this case, the series.

Students: build two Person cards from the night shift independently (Akimov, Toptunov, Dyatlov, Tregub, Khodemchuk, Shashenok are the candidates), with Role, Occupation, Significance, a Source, and a Basis on every claim. The exit ticket is one of them. Watch for the first “Dyatlov’s fault” and name it for what it is: a verdict, dated July 1987, that the unit will get to.

Homework: the first reading, a two-page excerpt from INSAG-1’s conclusions (the “remarkable range of human errors” paragraph, and the paragraph that coined “safety culture”).

Day 2: Explore. How deep does each cause run?

The causes board. Lay it out as four columns under four headings that are the History template’s own vocabulary: Underlying, Long-term, Short-term, Trigger, with the explosion card alone on the right and every string on the board running into it.

Model one string per column, out loud, with the Cause type set on the string and the card dragged into the matching column:

• Underlying: Soviet safety culture (a Concept, Domain = Idea, which INSAG-1 named and INSAG-7 made the first conclusion) contributes_to the explosion.
• Long-term: Positive void coefficient (a Concept, Domain = Technology) contributes_to the explosion. Built into the RBMK design in the 1960s.
• Short-term: Emergency core cooling system disconnected (an Event, 25 April) contributes_to the explosion.
• Trigger: AZ-5 pressed (an Event, 01:23:40) triggers the explosion.

Then release. Students build the rest of the causes from the record and must label every string. The material is rich enough that two careful students will build different boards: the graphite displacers on the control rods and the positive scram effect; the 1983 Ignalina incident that revealed it and was never turned into operator training; the reactor’s lack of a containment building; the test rescheduled into the night shift by the grid dispatcher; the power collapse to about 30 MW and the xenon poisoning that followed; the operating reactivity margin at six to eight rods against a minimum of fifteen; the extra circulation pumps; the blocked protection signals at 01:19; the production targets and the plan; the closed-city secrecy that kept the 1975 Leningrad incident from the people who needed it.

Push on the hard cases, because they are the lesson. Is disconnecting the emergency cooling system Short-term or Trigger? (Short-term: it made the explosion possible without making it happen.) Is the void coefficient Long-term or Underlying? (Long-term, by the usual reading; Underlying is reserved for the conditions that made such a design acceptable.) Is “safety culture” a cause at all, or a label for the other causes? (That is exactly the argument INSAG had with itself between 1986 and 1992, and a student who raises it has arrived at Day 3 a day early.)

Column check: a screenshot in which every string into the explosion is labeled and every card sits in the column its label names. Geometry and data have to agree.

Homework: the second reading, the corresponding two pages from INSAG-7, including the sentence that the accident “flowed from deficient safety culture, not only at the Chernobyl plant, but throughout the Soviet design, operating and regulatory organizations.”

Day 3: Explain. The verdict changes.

The sources board, laid out left to right by year. Students create Source cards (Object / Artifact, Object type = source) for, at minimum: INSAG-1 (IAEA, 1986), the trial verdict (Chernobyl, 30 July 1987), the Legasov tapes (recorded 1987 to 1988), the Shteynberg commission and the Abagyan commission (both 1991), and INSAG-7 (IAEA, 1992). Each carries its publisher, its date, and a link in Source. The two readings from homework are two of the cards.

Build the first historiographic string together: INSAG-7 contradicts INSAG-1, with the Significance saying on what point (where the principal cause lay), and the Specific relationship reading revised. Then Shteynberg corroborates Abagyan and both corroborate INSAG-7; INSAG-7 reproduced them as annexes, which is as literal as corroboration gets. Each report documents the explosion.

Now the card that makes the board an argument: a Concept, Operator error as the sole cause, Domain = Idea, Basis = Disputed. It gets a documents string from INSAG-1, a contradicts string from INSAG-7, and the board’s ruled-out mark. Then the trial verdict results_from INSAG-1, Cause type = Short-term, and the six defendants (Bryukhanov, Fomin, Dyatlov, Rogozhin, Kovalenko, Laushkin) get the person-of-interest mark, not because they are suspects but because the board is recording whom the 1986 account was applied to.

Discussion: the men were sentenced in July 1987. The reassessment came in 1991 and the IAEA adopted it in 1992. Dyatlov was released in 1990 for ill health and spent the rest of his life writing that the reactor was at fault. What does Documented mean when the documents disagree? Which of the two accounts was better evidenced at the time it was written, and does that change how a historian should treat the trial?

Second half of the period is the peer audit. Swap projects and flag three things: a Disputed card with fewer than two sources, a Cause type you would relabel, and a Basis you would dispute. Give the last ten minutes to defending flagged calls to the person who flagged them.

Day 4: Elaborate. Three clocks, and a map.

The structural day, and the reason the unit is built in this tool rather than on paper.

Students have been filling in When on every Event. Switch to the Timeline view and read it at three scales. The first is the forty seconds: the test starting at 01:23:04, AZ-5 at 01:23:40, the explosions at 01:23:43 and after, each a separate card ordered to the second. The second is the first week: Pravyk’s watch on scene at 01:28; the “under control” call to Moscow at about 03:00 while the dosimeters read their maximum; the fires out except the core at 06:35; the evacuation announcement in Pripyat at 14:00 on the 27th, about thirty-six hours after the explosion, with its promise of three days; the Forsmark alarm at about 09:00 Swedish time on the 28th; the twenty-second statement on Soviet television at about 21:00 that evening; the May Day parade in Kyiv; Shcherbina’s first press conference on 6 May admitting the thirty-six hours; Gorbachev on television on 14 May. The third is the decades: the trial in 1987, Legasov’s death on 27 April 1988, the 1991 commissions and the end of the Soviet Union the same year, INSAG-7 in 1992, Unit 3 shut down in December 2000, the Chernobyl Forum in 2006, the New Safe Confinement slid into place in November 2016, the occupation in 2022, the drone strike in 2025.

Ask: on which of the three clocks do the causes live, and on which do the consequences? Students will find that the Trigger lives on the first clock, the Short-term causes on the hours before it, and the Long-term and Underlying causes before the Timeline’s main axis even starts, in the backstory band. The depth they labeled on Day 2 is visible as distance on Day 4.

Then switch to the Navigator. Place cards carry coordinates, so the map draws itself: the plant, Pripyat, Chernobyl town, Kyiv, Moscow, Vienna, and Forsmark, eleven hundred kilometers away in Sweden. Ask: why is a Swedish power plant on a board about a Ukrainian one? Because the first reliable public source about the release was a radiation alarm on a worker’s shoes in Uppland, and the second was Swedish wind-pattern analysis pointing southeast, and Moscow’s statement came only after Stockholm asked. Have students string Forsmark linked_to the TASS statement with the Specific relationship field reading forced, and the Significance saying how.

Assign: every student adds the two information strings that Day 4 turns on, using linked_to with the Specific relationship filled in: who the Soviet state informed, and when, and whom it withheld from, and for how long. Pripyat’s 49,000 residents and the May Day crowds in Kyiv are the obvious cards on the receiving end. Then each student drafts the thesis of the Day 5 essay as the Significance note on one string of their choosing.

Timeline check: a screenshot with the night to the second and the first week to the hour.

Day 5: Evaluate. To what extent?

Students give four-minute walkthroughs in small groups: the causes board first, then the sources board, ending on the one string they would rest their essay on. The group’s job is to challenge one Cause type and one Basis per board, and the presenter’s job is to defend both from the sources.

Close with a whole-class seminar on the essay question. To what extent were the operators of Unit 4 responsible? Require every contribution to name a Source card. The seminar usually finds its way to the uncomfortable middle: the operators did violate the operating regulations (Documented, in both reports), and the reactor they were operating would have survived those violations if it had been designed to the standards every Western reactor of the day was built to (Documented, in the second). Both are true. The question is which kind of cause each one is, and that is the vocabulary students have spent the week learning to use.

Written follow-up: the two-page argument described in Stage 2, due at the next class. The Significance notes are the thesis drafts; the sources board is the document packet; the cause-type labels are the analytical vocabulary.

Supplement: the series your students have already seen

Most of the room arrives having watched Chernobyl (HBO, 2019), and a teacher has two choices: fight it, or put it on the board. Put it on the board.

The series goes on the sources board as a Source card: Object / Artifact, Object type = source, with its producer and year, and a Significance that says what it is: the single most influential account of the event in the English-speaking world, and a dramatization. It gets an influences string to a Concept called Public memory of Chernobyl, and then it gets contradicts strings to the record on specific, named points, each with the record’s Source card on the other end. This is historiography done on the thing students actually consume, and it is the best Legendary lesson in the unit.

The points worth stringing

Have students find these themselves from the record; the list is the answer key, not the handout.

What the series got right, and why that belongs on the board too

A corroborates string is as instructive as a contradicts one, and the series earns several: the forty seconds, in sequence; the reactor’s design flaws, correctly described; the thirty-six hours before the evacuation; the twenty-second statement; Legasov’s Vienna presentation; the tapes. A student who strings only the errors has built a gotcha, not a historiography. A series can be wrong about a helicopter and right about a safety culture, and knowing which is which is the skill.

Ground rules specific to this supplement

• No stills. The series is All Rights Reserved and every frame is an actor playing a real person. The Source card carries a title, a year and a link, and no image.
• The invented character is not a lie to be exposed. Composites are a standard device of dramatization and the series discloses it. The lesson is that a dramatization is a source with its own conventions, and a reader needs to know them.
• Credit the series for the room’s interest. Half the class is in the unit because they watched it. A teacher who treats that as contamination loses them on Day 1.

Differentiation and accommodations

• Scaffolded entry: provide a project with the explosion card, the four column headings, and the six required Source cards already created but empty. Students fill fields and string causes rather than deciding what counts as a source.
• Extension for advanced students: require a third board. The 1986 to 1992 revision is one historiographic chain; the death toll is another, with the Chernobyl Forum’s figure of about four thousand eventual cancer deaths among the most exposed populations, UNSCEAR’s thyroid-cancer findings, and the far higher estimates from advocacy groups as three Source cards and two contradicts strings. Alternatively, have students build the same causes board for a second industrial disaster they research themselves (Bhopal in 1984, the Challenger in 1986, Fukushima in 2011) and compare the column counts.
• IEP / 504: the board is a visual-spatial alternative to the document-based essay. Reduce the card minimums rather than the field requirements; six well-sourced causes beat fifteen unlabeled ones. The essay can be delivered as the Day 5 walkthrough with notes, if writing is the accommodation.
• ELL: the vocabulary load here is technical (void coefficient, reactivity margin, scram) rather than literary, and a glossary on the wall from Day 0 helps more than it would in a novel unit. Permit first-language notes in Description, accept quoted evidence without paraphrase, and let the Pin caption, one short phrase, stand in for a paragraph. Students who read Ukrainian or Russian can be the class’s resource on the names, and should be asked.
• Students with a personal connection: given the region and the current war, assume at least one. Offer the option of working on the machine, the places and the sources boards rather than the people, and say so privately before Day 1 rather than in front of the room.
• Absent students: the project persists and syncs, so a returning student picks up their own boards rather than reconstructing a poster.

Materials

To run the unit you will need:

• The two readings: a two-page excerpt from INSAG-1’s conclusions and the corresponding two pages from INSAG-7. Both reports are free from the IAEA’s publications site, and the excerpts fit on one double-sided handout.
• One String Theory project per student or pair, on the History template pack.
• A projector for Days 1 through 5.
• A printed rubric.
• A worked exemplar. The Chernobyl project shown on the History template page is a six-board build of this unit (the night, the causes in four columns, the cost, what the world was told, the verdict changing, the long tail), with every causal string labeled and every Disputed card carrying its two sources. Its screenshots on that page show a class what “finished” looks like on Day 1 and Day 2. It is a model, not an answer key: its cause types are arguable and students should be invited to argue with them.
• A one-page handout of the media rules, the naming convention, and the AI policy. Hand it out on Day 0 and put it on the wall.

Notes for the teacher

Setup. Create the project and choose the History pack. The cards read Person, Group / Institution, Place, Object / Artifact, Event and Concept, and the causal strings carry the Cause type field without any configuration. A project can add its own attributes if you want them; the exemplar adds Fate and Dose received (Sv) to Person cards, which keeps the doses as numbers in a field rather than sentences in prose.

Dates. Every date in this unit is Gregorian and after 1900, and the record gives the night to the second, so none of the “circa” or BCE questions a classical history unit raises come up here. The Event card’s When field takes a time; AZ-5 at 01:23:40 and the explosion at 01:23:43 sort correctly. Agree on one clock: Moscow summer time, which is what the plant’s logs and the Soviet record use, with the Forsmark alarm converted from Swedish time (09:00 there was 11:00 in Moscow). Put the convention on a sticky note on the board.

Two boards, not one. The causes and the sources want different layouts: four columns for the first, a left-to-right decade for the second. Build them as two boards in one project. The Timeline and Navigator read the whole project regardless of board, so nothing is lost.

Pin caption is not a name field. When set, it replaces the entity’s name on the card face, so it should say why the card matters: “the report that moved the blame,” not “INSAG-7”; “the alarm that forced the announcement,” not “Forsmark.” Anything past roughly fifty characters stops fitting a card cleanly.

Let the Cause type argument happen. The four values will produce disagreement in the column check and the peer audit, and that disagreement is the AP causation rubric happening out loud. Resist arbitrating it. The same goes for Basis: a class that argues about whether a claim is Inferred or Disputed is doing the work.

On “everyone was responsible.” Students reach for it by Day 3, and this subject makes it more tempting than most because it is close to what INSAG-7 said. It is still not an argument. The essay question says to what extent, and the cause-type vocabulary exists so that “the operators and the designers and the system” can be said with the proportions filled in.

On classroom accounts: each student needs their own account, and String Theory accounts are for people 13 and older, which a grade 10 to 12 class clears. Each account is its own workspace, and the free plan allows two projects and 250 entities per workspace, which is comfortably more than this unit needs. Publishing is per project and can be narrowed to the boards you pick; see the ground rules above for why that matters to the publishing decision as well as to collection.

Collecting the boards. On free accounts, each student publishes their project and sends you the link, so the publishing rules above apply to every board you collect. Our Classroom plan replaces that with a class link. You make a class for the unit and hand out its link. Each student opens it, signs in to their own free account, and picks the project they built for the unit, or starts it there. The project stays theirs (no one but you can see it unless they publish it), stops counting against their two free projects, and shows up on your roster, where you can open it read-only until the class ends. You can’t edit it, and you can’t see anything else in their account. Classroom is for students 13 and older, like the rest of String Theory.

The media rules are not an add-on. With this subject the license question and the honesty question are the same question. The photographs students want are copyrighted, the portraits they want do not exist under any license, and the generated images they could make would be fabrications placed next to real names. A board that respects all three constraints is a board that has already learned what the unit is for.