Revision 7T · Full Tutorial Edition · Updated 1 September 2026
PWR-134 · SELF GUIDED full tutorial
Generate competing explanations and a benign test that could rule one out
A scientific hypothesis is a possible explanation that leads to observations we can check. This lesson uses similar-sized ice cubes melting on metal, wood and folded cloth. You will describe what happened without explaining it first, write four genuinely different explanations, state what each predicts, run one safe comparison and keep results that do not fit your favourite explanation.
What you will produceThe learner produces at least four plausible hypotheses, distinct predictions and one benign controlled test capable of weakening at least one explanation.
Full step-by-step individual tutorial · TLU-PWR-134
The learner produces at least four plausible hypotheses, distinct predictions and one benign controlled test capable of weakening at least one explanation.
Use fictional datasets or ordinary low-risk physical observations with household-safe materials.
Run the ice-cube test on a stable tray with spill control.
Qualified help is required for
Human or animal research, clinical testing, hazardous materials, environmental release, security probing or regulated experiments.
Domain interpretation when findings could affect other people.
Never do this from the page alone
Conduct covert experiments, diagnose, ingest test substances or use fire, pressure, electricity, pathogens or dangerous chemicals.
Call the first compatible result proof or discard a discrepant result.
2 · Get ready
Gather what you need and check the starting conditions
What you need
Three ice cubes from the same tray that look similar in size, plus a metal plate or spoon, a wooden board and a folded dry cloth.
A waterproof tray, towel, timer and either a ruler or a piece of cardboard on which the starting outline of each cube can be drawn.
A paper table with columns for explanation, predicted result, result that would weaken the explanation, test result and remaining uncertainty.
Before you start
Work away from electricity and walking paths; contain meltwater and wipe spills immediately.
Use ordinary ice only, do not ingest test material and stop if cold contact is uncomfortable.
Define “melted” by one visible criterion before starting.
3 · The method
Follow these steps in order
Record the observation neutrally
Write what appears to differ in melt time, where, when and under which starting conditions without explaining it.
Why: A neutral observation prevents the preferred story from entering the data.
Check: The statement contains measured conditions and no causal word.
Generate competing mechanisms
List at least four explanations: heat transfer, contact area, cube-size difference and timing or placement error.
Why: A hypothesis space reduces premature attachment to one cause.
Check: Four mechanisms differ in what produces the observation.
Derive discriminating predictions
For each hypothesis, state what extra pattern should appear if it is true and what result would weaken it.
Why: Predictions make explanations testable against one another.
Check: Every row has support and weakening outcomes.
Specify the measurement
Choose the outcome, unit, observation interval, number of repeats and rule for missing or ambiguous observations before testing. Record the practical precision of the timer or size measure.
Why: A vague outcome can be reinterpreted to favour any hypothesis.
Check: The plan contains a fixed measurement rule that another learner could follow without guessing.
Choose one safe contrast
Design a test that changes or measures one relevant factor while holding timing, cube size and room location as constant as practical.
Why: A controlled contrast helps separate alternatives.
Check: The plan names changed variable, controls and residual confounds.
Precommit and run
Write predicted order and stop criteria, start all cubes together and record the declared outcome without changing the rule.
Why: Precommitment limits hindsight and selective measurement.
Check: Raw times or observations remain unchanged beside the prediction.
Interpret without proof language
Compare each hypothesis with the result, retain alternatives, note anomalies and design the next benign discriminating check.
Why: One result can favour an explanation without proving it universally.
Check: The conclusion uses favours, weakens or unresolved and preserves discrepancies.
Repeat before extending
Run the same safe test again on a different day before changing the surface or outcome. Compare direction and size of the pattern and investigate any reversal.
Why: A repeat can expose timing, cube-size or room-condition effects before a more elaborate story is built.
Check: Both raw runs remain visible, and extension occurs only when the predeclared pattern is sufficiently consistent for the learning question.
4 · Worked example
See the whole method used once
Scenario
Three similar ice cubes are placed simultaneously on metal, wood and folded cloth in one room.
Walkthrough
Ivo records that the metal cube appears to lose height fastest during the first three minutes.
He lists heat transfer, contact area, size mismatch and a late start as alternatives.
Heat-transfer predicts the order should remain when cube size and contact area are matched; timing error predicts no stable repeat.
He traces equal cube outlines, starts from one video timestamp and precommits the metal-fastest prediction.
The repeat favours heat transfer but one cube was still slightly smaller, so he keeps size mismatch unresolved and proposes a measured-size repeat.
Result
Ivo has used competing hypotheses and a discriminating benign test without calling the first pattern proof.
5 · Right and wrong
Compare correct or safer execution with the common wrong version
Right and wrong comparison
Moment
Right / safer
Wrong / riskier
Why it matters
Observation
Describe the measured pattern before explanation.
Write “metal conducts heat better” as the observation.
Mechanism and data must remain distinguishable.
Hypotheses
Generate several mechanisms that could differ in prediction.
Write four phrasings of the preferred cause.
Quantity without mechanistic diversity does not challenge the theory.
Prediction
State what would weaken each hypothesis.
Predict only confirming outcomes.
An unfalsifiable story cannot be discriminated.
Conclusion
Say the result favours one explanation under these conditions.
Say one kitchen trial proves a law or discovery.
Measurement limits and alternatives remain.
6 · Common mistakes
Spot the error and apply the correction
Common mistakes and corrections
Mistake
Fix
Cube size is assumed identical.
Trace or measure the starting dimension and retain mismatch as a confound.
Start times differ.
Use one launch cue and document timing error.
Outcome definition changes mid-test.
Predeclare visible-height, mass or melt-time criterion and keep it.
Discrepant data are deleted.
Preserve the raw row and classify hidden variable, measurement error, boundary or alternative explanation.
7 · Practice
Turn the steps into a usable skill
First session
Write one neutral observation.
Generate four mechanisms.
Complete prediction and falsifier rows.
Run one contained ice-cube contrast.
Interpret and design one next test.
Repeat plan
Use one different benign observation every two weeks for three cycles. Delay the next-test design by a day and ask another learner whether the predictions genuinely differ before running it.
Progress when
Hypotheses are mechanistically distinct.
Each has a weakening outcome.
Tests change one main factor and preserve discrepant data.
Do not progress when
The question requires people, animals, dangerous materials or covert action.
The learner changes outcomes after seeing results.
Spills, cold exposure or equipment make the test unsafe.
8 · Check the result
Measure what changed
Distinct falsifiable hypotheses and discriminating predictions for a benign observation.
Good result: At least four distinct hypotheses, four weakening predictions and one test that genuinely discriminates at least two, confirmed by a second reader.
This does not prove: It does not establish discovery, causal truth, cross-domain scientific skill or ethical permission for other experiments.
Self-check
Is the observation free of causal language?
Do hypotheses predict different patterns?
What result would weaken the favourite?
Which confound remains after the test?
9 · Stop, adapt or get help
Keep the safety boundary practical
Stop and get help
Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
Stop for unsafe cold contact, water near electricity or uncontrolled spill.
Seek domain, ethics and safety review before moving beyond the supplied benign fixture.
Accessibility and adaptations
Use prerecorded fictional data instead of handling ice.
Allow speech-to-text, pictorial matrices or a scribe.
Use large timers, stable containers and an assistant for placement without changing hypotheses.
10 · Evidence and limits
Why these instructions are here
primary research
A small fMRI study reported changes associated with training hypothesis-generation skills but did not establish active-comparator, delayed, cross-domain or discovery outcomes.
The Understanding Science resource explains that scientific testing connects an idea to what we would expect to observe if that idea were correct, then compares the expectation with the evidence.
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
Record the observation neutrally
Write what appears to differ in melt time, where, when and under which starting conditions without explaining it.
Why this step exists
A neutral observation prevents the preferred story from entering the data.
Success check
The statement contains measured conditions and no causal word.
I’m stuck on this step
Reset: Re-read this authored instruction — “Write what appears to differ in melt time, where, when and under which starting conditions without explaining it.” — and its success check, then attempt only this step.
Possible snag: Outcome definition changes mid-test.
Correction: Predeclare visible-height, mass or melt-time criterion and keep it.
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
02
Generate competing mechanisms
List at least four explanations: heat transfer, contact area, cube-size difference and timing or placement error.
Why this step exists
A hypothesis space reduces premature attachment to one cause.
Success check
Four mechanisms differ in what produces the observation.
I’m stuck on this step
Reset: Re-read this authored instruction — “List at least four explanations: heat transfer, contact area, cube-size difference and timing or placement error.” — and its success check, then attempt only this step.
Possible snag: Cube size is assumed identical.
Correction: Trace or measure the starting dimension and retain mismatch as a confound.
Possible snag: Start times differ.
Correction: Use one launch cue and document timing error.
Possible snag: Discrepant data are deleted.
Correction: Preserve the raw row and classify hidden variable, measurement error, boundary or alternative explanation.
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
03
Derive discriminating predictions
For each hypothesis, state what extra pattern should appear if it is true and what result would weaken it.
Why this step exists
Predictions make explanations testable against one another.
Success check
Every row has support and weakening outcomes.
I’m stuck on this step
Reset: Re-read this authored instruction — “For each hypothesis, state what extra pattern should appear if it is true and what result would weaken it.” — and its success check, then attempt only this step.
Possible snag: The result from “For each hypothesis, state what extra pattern should appear if it is true and what result would weaken it.” does not yet meet this declared check: Every row has support and weakening outcomes.
Correction: Return to the start of “Derive discriminating predictions”, reduce complexity or pace, and repeat only the part needed to satisfy: “Every row has support and weakening outcomes.”
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
04
Specify the measurement
Choose the outcome, unit, observation interval, number of repeats and rule for missing or ambiguous observations before testing. Record the practical precision of the timer or size measure.
Why this step exists
A vague outcome can be reinterpreted to favour any hypothesis.
Success check
The plan contains a fixed measurement rule that another learner could follow without guessing.
I’m stuck on this step
Reset: Re-read this authored instruction — “Choose the outcome, unit, observation interval, number of repeats and rule for missing or ambiguous observations before testing. Record the practical precision of the timer or size measure.” — and its success check, then attempt only this step.
Possible snag: The result from “Choose the outcome, unit, observation interval, number of repeats and rule for missing or ambiguous observations before testing. Record the practical precision of the timer or size measure.” does not yet meet this declared check: The plan contains a fixed measurement rule that another learner could follow without guessing.
Correction: Return to the start of “Specify the measurement”, reduce complexity or pace, and repeat only the part needed to satisfy: “The plan contains a fixed measurement rule that another learner could follow without guessing.”
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
05
Choose one safe contrast
Design a test that changes or measures one relevant factor while holding timing, cube size and room location as constant as practical.
Why this step exists
A controlled contrast helps separate alternatives.
Success check
The plan names changed variable, controls and residual confounds.
I’m stuck on this step
Reset: Re-read this authored instruction — “Design a test that changes or measures one relevant factor while holding timing, cube size and room location as constant as practical.” — and its success check, then attempt only this step.
Possible snag: The result from “Design a test that changes or measures one relevant factor while holding timing, cube size and room location as constant as practical.” does not yet meet this declared check: The plan names changed variable, controls and residual confounds.
Correction: Return to the start of “Choose one safe contrast”, reduce complexity or pace, and repeat only the part needed to satisfy: “The plan names changed variable, controls and residual confounds.”
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
06
Precommit and run
Write predicted order and stop criteria, start all cubes together and record the declared outcome without changing the rule.
Why this step exists
Precommitment limits hindsight and selective measurement.
Success check
Raw times or observations remain unchanged beside the prediction.
I’m stuck on this step
Reset: Re-read this authored instruction — “Write predicted order and stop criteria, start all cubes together and record the declared outcome without changing the rule.” — and its success check, then attempt only this step.
Possible snag: The result from “Write predicted order and stop criteria, start all cubes together and record the declared outcome without changing the rule.” does not yet meet this declared check: Raw times or observations remain unchanged beside the prediction.
Correction: Return to the start of “Precommit and run”, reduce complexity or pace, and repeat only the part needed to satisfy: “Raw times or observations remain unchanged beside the prediction.”
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
07
Interpret without proof language
Compare each hypothesis with the result, retain alternatives, note anomalies and design the next benign discriminating check.
Why this step exists
One result can favour an explanation without proving it universally.
Success check
The conclusion uses favours, weakens or unresolved and preserves discrepancies.
I’m stuck on this step
Reset: Re-read this authored instruction — “Compare each hypothesis with the result, retain alternatives, note anomalies and design the next benign discriminating check.” — and its success check, then attempt only this step.
Possible snag: The result from “Compare each hypothesis with the result, retain alternatives, note anomalies and design the next benign discriminating check.” does not yet meet this declared check: The conclusion uses favours, weakens or unresolved and preserves discrepancies.
Correction: Return to the start of “Interpret without proof language”, reduce complexity or pace, and repeat only the part needed to satisfy: “The conclusion uses favours, weakens or unresolved and preserves discrepancies.”
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
08
Repeat before extending
Run the same safe test again on a different day before changing the surface or outcome. Compare direction and size of the pattern and investigate any reversal.
Why this step exists
A repeat can expose timing, cube-size or room-condition effects before a more elaborate story is built.
Success check
Both raw runs remain visible, and extension occurs only when the predeclared pattern is sufficiently consistent for the learning question.
I’m stuck on this step
Reset: Re-read this authored instruction — “Run the same safe test again on a different day before changing the surface or outcome. Compare direction and size of the pattern and investigate any reversal.” — and its success check, then attempt only this step.
Possible snag: The result from “Run the same safe test again on a different day before changing the surface or outcome. Compare direction and size of the pattern and investigate any reversal.” does not yet meet this declared check: Both raw runs remain visible, and extension occurs only when the predeclared pattern is sufficiently consistent for the learning question.
Correction: Return to the start of “Repeat before extending”, reduce complexity or pace, and repeat only the part needed to satisfy: “Both raw runs remain visible, and extension occurs only when the predeclared pattern is sufficiently consistent for the learning question.”
Stop / get help: Stop any test involving people, animals, clinical claims, covert observation, hazardous materials or environmental release.
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.
Observation — Mechanism and data must remain distinguishable.
Correct / safer
Describe the measured pattern before explanation.
Wrong / riskier
Write “metal conducts heat better” as the observation.
Hypotheses — Quantity without mechanistic diversity does not challenge the theory.
Correct / safer
Generate several mechanisms that could differ in prediction.
Wrong / riskier
Write four phrasings of the preferred cause.
Prediction — An unfalsifiable story cannot be discriminated.
Correct / safer
State what would weaken each hypothesis.
Wrong / riskier
Predict only confirming outcomes.
Conclusion — Measurement limits and alternatives remain.
Correct / safer
Say the result favours one explanation under these conditions.
Wrong / riskier
Say one kitchen trial proves a law or discovery.
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.