Classify fictional causal claims by design and alternatives
Causal reasoning asks more than whether two variables move together. This pack supplies four fictional cases spanning observation, self-selection, and randomised intervention. The learner names cause and outcome, checks design and time order, generates alternatives, proposes discriminating evidence, and writes a bounded classification. Numerical differences remain descriptive until design supports a causal interpretation.
Tutorial curriculum · current public edition · Revision 7T · Full Tutorial Edition
One source of teaching truth
Full individual tutorial · TLU-PWR-116
Causal reasoning asks more than whether two variables move together. This pack supplies four fictional cases spanning observation, self-selection, and randomised intervention. The learner names cause and outcome, checks design and time order, generates alternatives, proposes discriminating evidence, and writes a bounded classification. Numerical differences remain descriptive until design supports a causal interpretation.
Revision 7T · Full Tutorial Edition · complete governed practice tutorial
Classify fictional causal claims by design and alternatives
Causal reasoning asks more than whether two variables move together. This pack supplies four fictional cases spanning observation, self-selection, and randomised intervention. The learner names cause and outcome, checks design and time order, generates alternatives, proposes discriminating evidence, and writes a bounded classification. Numerical differences remain descriptive until design supports a causal interpretation.
Authority and safety
Use only the fictional cases. Do not accuse a real person, diagnose, generate misinformation, or justify policy, legal, clinical, employment, or financial action.
A plausible mechanism or diagram does not replace outcome evidence.
Consequential causal claims require domain experts, suitable study design, affected-person input, and accountable review.
Exact materials
The four complete fictional case rows
The causal-workbook columns built into the fixture
The answer key, hidden until conclusions are recorded
Paper or an accessible note method
Set up in this order
Operationally name the proposed cause and measured outcome, then record initial confidence from 0 to 100 before classifying the design.
Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label.
Record at least two compatible alternative explanations and one test whose predicted results would differ between them.
Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN. A plausible story is not a measured mechanism.
Choose unsupported, suggestive, better supported or unresolved; then record final confidence and the exact evidence that moved or preserved it.
State the boundary and one decision the row cannot support.
Supplied fixture
All cases are invented. The final four columns are the complete learner worksheet. Do not use any row to make a claim about a real person or system.
Case id
Fictional claim
Design and data
Known context
Learner design
Learner alternatives
Learner discriminating test
Learner bounded conclusion
Learner mechanism status
Confidence before and after review
C1
Umbrellas make the park path wet.
30 rainy days average 70 umbrellas and all have wet paths; 30 dry days average 5 umbrellas and none has a wet path.
Observation only; rain status recorded.
[classify]
[two alternatives]
[one test]
[conclude]
[shown / plausible / unknown]
[record both]
C2
A blue greenhouse filter increases seedling height.
40 matched pots randomly assigned: 20 blue filter, 20 clear; means 18.4 cm and 16.1 cm after four weeks.
Water, soil, and day fixed; one fictional greenhouse.
[classify]
[two alternatives]
[one test]
[conclude]
[shown / plausible / unknown]
[record both]
C3
Using the library quiet room causes readers to finish more pages.
30 visitors choose quiet room and average 42 pages; 30 choose open room and average 30 pages in the same hour.
Book difficulty and prior habit unmeasured.
[classify]
[two alternatives]
[one test]
[conclude]
[shown / plausible / unknown]
[record both]
C4
A reminder card increases neutral sorting-task completion.
100 fictional trials randomised: 40/50 complete with reminder, 30/50 without.
Same task, device, criterion, and window.
[classify]
[two alternatives]
[one test]
[conclude]
[shown / plausible / unknown]
[record both]
Check your work only after you finish
Lock every response first. The checking page is separate so the answers are not exposed inside this exercise.
For each case award 1 point for operational variables, time order, correct design, two plausible alternatives, a discriminating test, mechanism status, confidence update and bounded classification: 8 per case, 32 total. Mechanism without measured pathway evidence cannot be called shown. Any move from a fictional row to real blame, diagnosis or policy earns STOP.
Worked example
For C3 define cause as choosing the quiet room and outcome as pages during the hour.
Calculate the observed difference: 42-30=12 pages.
Classify self-selection and list book difficulty and prior reading habit as alternatives.
Propose accessible random assignment or predeclared measurement of selection factors as discriminating evidence.
Conclude suggestive association with unresolved causal direction and magnitude.
Correct result: C3 shows a 12-page average association, not proof that the room caused it. Self-selection and unmeasured book difficulty block a strong causal conclusion.
Right and wrong
Right and wrong comparison
Moment
Right / safer
Wrong / riskier
Why
C1
Treat rain as a common cause.
Perfect association proves umbrellas wet paths.
A measured common cause explains both.
C2
Say better supported in one greenhouse.
The filter works for every plant everywhere.
One setting does not establish universal transfer.
C3
Preserve self-selection.
Rename groups treatment and control.
Wording cannot upgrade allocation.
Mechanism
Mark it as hypothesis and seek discriminating outcomes.
An arrow diagram proves cause.
A diagram is not outcome evidence.
Common mistakes and fixes
Common mistakes and corrections
Mistake
Fix
Leaving variables vague.
Name exposure, outcome, unit, and time order.
Calling every alternative a confounder.
Separate common cause, reverse causation, selection, and measurement error.
Proposing more correlation as the only test.
Choose evidence with different predictions under two explanations.
Turning a bounded result into blame.
Return to the fictional case and state what further evidence is required.
Evidence boundary
What the tutorial may—and may not—claim.
Supportable claim
Instruction can reduce selected causal-reasoning errors in science-report interpretation, with short delayed retention.
Measurement boundary
Declare causal question, design, alternatives, confounding, mechanism, uncertainty, data limits and decision context; diagram quality is not causal truth.
Myth
See the true cause of anything
Metric
Accurate causal classification and alternative-cause generation on novel declared cases
Boundary
A training-task gain cannot certify causality in a complex real system.
Negative and limiting findings
Evidence remains bounded to science-report and control-of-variables tasks.
Probe questions without direct instruction did not improve children’s unconfounded experimental reasoning.
No real policy, clinical or organizational outcome was tested.
Supports only the bounded empirical proposition in the declared configurations. Constrains generalisation, transfer, certainty or efficacy; it is not optional context.
Colleen M. Seifert; Michael Harrington; Audrey L. Michal; Priti Shah · 2022 · PRIMARY_RESEARCH
Supports only the bounded empirical proposition in the declared configurations. Constrains generalisation, transfer, certainty or efficacy; it is not optional context.
Tutorial delivery controls
Learn, adapt, troubleshoot and resume
Estimated timeEstimated 18 min reading and worksheet pass
DifficultyIntroductory
EquipmentBasic stationery or digital tools
SpaceDesk / seated
Method qualityEstablished10 of 10 structural checks present. Automated method-readiness band; human editorial sign-off is separate.
Evidence contextG1; Focused research depthScientific support is evaluated separately from teaching-method structure.
Editorial reviewPending manual sign-offNo human approval is claimed until reviewer, date and content hash are recorded.
Progress is saved only in this browser on this device.
Step-by-step learner mode
Each activity includes its success check, a nearby accessible alternative and an “I’m stuck” correction path. Alternatives preserve the target where possible; when they change the task, Titan labels them as related rather than equivalent.
01
Operationally name the proposed cause and measured outcome, then record initial confidence…
Operationally name the proposed cause and measured outcome, then record initial confidence from 0 to 100 before classifying the design.
Why this step exists
“Operationally name the proposed cause and measured outcome, then record initial confidence from 0 to 100 before classifying the design” creates the auditable value, condition or statement needed for the tutorial’s later comparison and conclusion.
Success check
A legible entry directly completes “Operationally name the proposed cause and measured outcome, then record initial confidence from 0 to 100 before classifying the design” and records the requested condition, decision or boundary without adding a new claim.
I’m stuck on this step
Reset: Re-read this authored instruction — “Operationally name the proposed cause and measured outcome, then record initial confidence from 0 to 100 before classifying the design.” — and its success check, then attempt only this step.
Possible snag: Leaving variables vague.
Correction: Name exposure, outcome, unit, and time order.
Possible snag: Calling every alternative a confounder.
Correction: Separate common cause, reverse causation, selection, and measurement error.
Stop / get help: A governed step-by-step lesson may be completed independently inside its stated limits. Stop if the declared configuration cannot be maintained, uncertainty becomes material, adverse effects appear, or qualified authority is required.
02
Classify the design without upgrading it by wording: observation, self-selection, randomised comparison…
Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label.
Why this step exists
“Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label” isolates the named evidence distinction so the learner can interpret Causal reasoning without extending the claim.
Success check
The output directly answers “Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label”, names the relevant distinction and stays within the supplied record.
I’m stuck on this step
Reset: Re-read this authored instruction — “Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label.” — and its success check, then attempt only this step.
Possible snag: The result from “Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label.” does not yet meet this declared check: The output directly answers “Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label”, names the relevant distinction and stays within the supplied record.
Correction: Return to the start of “Classify the design without upgrading it by wording: observation, self-selection, randomised comparison…”, reduce complexity or pace, and repeat only the part needed to satisfy: “The output directly answers “Classify the design without upgrading it by wording: observation, self-selection, randomised comparison or another exact label”, names the relevant distinction and stays within the supplied record.”
Stop / get help: A governed step-by-step lesson may be completed independently inside its stated limits. Stop if the declared configuration cannot be maintained, uncertainty becomes material, adverse effects appear, or qualified authority is required.
03
Record at least two compatible alternative explanations and one test whose predicted…
Record at least two compatible alternative explanations and one test whose predicted results would differ between them.
Why this step exists
“Record at least two compatible alternative explanations and one test whose predicted results would differ between them” creates the auditable value, condition or statement needed for the tutorial’s later comparison and conclusion.
Success check
A legible entry directly completes “Record at least two compatible alternative explanations and one test whose predicted results would differ between them” and records the requested condition, decision or boundary without adding a new claim.
I’m stuck on this step
Reset: Re-read this authored instruction — “Record at least two compatible alternative explanations and one test whose predicted results would differ between them.” — and its success check, then attempt only this step.
Possible snag: Proposing more correlation as the only test.
Correction: Choose evidence with different predictions under two explanations.
Stop / get help: A governed step-by-step lesson may be completed independently inside its stated limits. Stop if the declared configuration cannot be maintained, uncertainty becomes material, adverse effects appear, or qualified authority is required.
04
Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN
Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN. A plausible story is not a measured mechanism.
Why this step exists
“Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN” isolates the named evidence distinction so the learner can interpret Causal reasoning without extending the claim.
Success check
The output directly answers “Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN”, names the relevant distinction and stays within the supplied record.
I’m stuck on this step
Reset: Re-read this authored instruction — “Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN. A plausible story is not a measured mechanism.” — and its success check, then attempt only this step.
Possible snag: The result from “Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN. A plausible story is not a measured mechanism.” does not yet meet this declared check: The output directly answers “Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN”, names the relevant distinction and stays within the supplied record.
Correction: Return to the start of “Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN”, reduce complexity or pace, and repeat only the part needed to satisfy: “The output directly answers “Classify mechanism as SHOWN, PLAUSIBLE or UNKNOWN”, names the relevant distinction and stays within the supplied record.”
Stop / get help: A governed step-by-step lesson may be completed independently inside its stated limits. Stop if the declared configuration cannot be maintained, uncertainty becomes material, adverse effects appear, or qualified authority is required.
05
Choose unsupported, suggestive, better supported or unresolved
Choose unsupported, suggestive, better supported or unresolved; then record final confidence and the exact evidence that moved or preserved it.
Why this step exists
“Choose unsupported, suggestive, better supported or unresolved” fixes the exact Causal reasoning configuration or decision before later observations are compared.
Success check
A legible entry directly completes “Choose unsupported, suggestive, better supported or unresolved” and records the requested condition, decision or boundary without adding a new claim.
I’m stuck on this step
Reset: Re-read this authored instruction — “Choose unsupported, suggestive, better supported or unresolved; then record final confidence and the exact evidence that moved or preserved it.” — and its success check, then attempt only this step.
Possible snag: Turning a bounded result into blame.
Correction: Return to the fictional case and state what further evidence is required.
Stop / get help: A governed step-by-step lesson may be completed independently inside its stated limits. Stop if the declared configuration cannot be maintained, uncertainty becomes material, adverse effects appear, or qualified authority is required.
06
State the boundary and one decision the row cannot support
State the boundary and one decision the row cannot support.
Why this step exists
“State the boundary and one decision the row cannot support” establishes the exact Causal reasoning question or boundary that every later comparison must preserve.
Success check
Evidence of completion shows the learner followed this exact instruction without adding an unstated step: “State the boundary and one decision the row cannot support.”
I’m stuck on this step
Reset: Re-read this authored instruction — “State the boundary and one decision the row cannot support.” — and its success check, then attempt only this step.
Possible snag: The result from “State the boundary and one decision the row cannot support.” does not yet meet this declared check: Evidence of completion shows the learner followed this exact instruction without adding an unstated step: “State the boundary and one decision the row cannot support.”
Correction: Return to the start of “State the boundary and one decision the row cannot support”, reduce complexity or pace, and repeat only the part needed to satisfy: “Evidence of completion shows the learner followed this exact instruction without adding an unstated step: “State the boundary and one decision the row cannot support.””
Stop / get help: A governed step-by-step lesson may be completed independently inside its stated limits. Stop if the declared configuration cannot be maintained, uncertainty becomes material, adverse effects appear, or qualified authority is required.
Correct versus incorrect execution
These accessible process diagrams are built from the tutorial’s own right/wrong teaching. They are not anatomical illustrations and do not add technique beyond the canonical tutorial.
C1 — A measured common cause explains both.
Correct / safer
Treat rain as a common cause.
Wrong / riskier
Perfect association proves umbrellas wet paths.
C2 — One setting does not establish universal transfer.
Correct / safer
Say better supported in one greenhouse.
Wrong / riskier
The filter works for every plant everywhere.
C3 — Wording cannot upgrade allocation.
Correct / safer
Preserve self-selection.
Wrong / riskier
Rename groups treatment and control.
Mechanism — A diagram is not outcome evidence.
Correct / safer
Mark it as hypothesis and seek discriminating outcomes.
Wrong / riskier
An arrow diagram proves cause.
Method-structure checklist
10 of 10 structural checks present
✓ Ordered, Power-specific instructions — present
✓ Every activity has a success check — present
✓ Materials or supplied records are declared — present
✓ Measurement or assessment rule is present — present
✓ Tutorial-specific troubleshooting is present — present
✓ Stopping or escalation boundary is present — present
✓ Every activity has an adjacent alternative — present
✓ Correct-versus-incorrect comparison is present — present
✓ Evidence context is bound to the Power record — present
✓ Planning metadata is present — present
The method-readiness band and presence checklist assess tutorial presentation and are separate from evidence quality for the underlying Power. They are automated editorial aids, not human approval.
Manual editorial sign-off: Pending. This tutorial must not display a human-approved state until an identified editor signs the exact content hash.