Stages of a Civilisation | Evidence Atlas and Confidence

Did you know a genuine archaeological artefact can be used in an inaccurate story, while a beautifully detailed AI illustration can show something that never happened at all? Evidence quality is the gatekeeper of civilisation-stage classification. This guide helps parents, tutors and older children distinguish an observation from a reconstruction, a measured quantity from a promise, and a narrow experiment from a society-wide operating capability.

The eduKate Civilisation Atlas defines Foundation 0 and eight numbered capability stages. These are teaching coordinates, not a universally compulsory historical chronology or a score of human worth. A responsible diagnosis names the place, period, source and function and permits “not enough evidence” where material is incomplete. The evidence reviewer and the assumption checker serve as fictional evidence challengers in the classroom activity, while the lesson facilitator records the children’s considered conclusions. None is a real archaeologist, space engineer or autonomous publisher.

Previously published AI-generated concept art reused for evidence-literacy teaching. Not a photograph of an actual excavation, genuine deciphered tablets or an authenticated historical institution.

Nine evidence questions across Foundation 0–Stage 8

Position	What evidence must establish
Foundation 0	Adaptive practices and transmission of knowledge
Stage 1	Sustained settlement and usable provisioning
Stage 2	Recurring records and administration
Stage 3	Accepted industrial output and mass services
Stage 4	Sustained planet-spanning interdependence
Stage 5	Verified correction of harm across time
Stage 6 (hypothetical)	Continuing human functions in a second celestial environment
Stage 7 (hypothetical)	Substantial local autonomy of inhabited worlds
Stage 8 (hypothetical)	Widely distributed and resilient interplanetary human continuity

From source to claim: how to avoid confident but unsupported stories

Begin with four different labels: observation, interpretation, uncertainty and next test. A museum’s description may identify a pottery vessel and its archaeological period. The existence of the object is different from an unproven claim about who owned it or what laws governed the settlement. A dated mission release may demonstrate an engineering experiment but not a completed inhabited planet. A graph can show a change but not establish its cause without relevant methods and comparison.

No source type is universally perfect. Clay survives when fibre may not. Official statistics have defined denominators and collection procedures. Historical accounts can preserve only certain social voices. A generated image can illustrate mechanisms but cannot document real people or buildings. A robust stage diagnosis combines relevant observations and reports the limits rather than producing a fake confidence score from one source.

Part II. Twenty-four evidence checks for civilisation classification

1. Source identity and provenance

Did you know? Where exactly did the record originate, and who describes it today? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A reposted photograph can lose the excavator, collection and date that gave it meaning. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children compare a museum catalogue card with an unattributed social-media image. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover museum accession, context, dated cataloguing and the original source. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

For a younger learner, use three cards labelled fact, possible meaning and question. Upper-primary students can produce a small table and explain a unit or date. Secondary learners should identify source provenance, sampling limits, alternatives and the distinction between correlation and cause. The tutor should avoid demanding a numerical confidence percentage unless a defensible quantitative method has actually been supplied.

The case can also be used to teach transfer. Show a new object, graph or illustration in a different period or environment, then ask the child to apply the same source checks. A claim about an ancient settlement cannot use a later-period tablet without justification; a claim about a lunar city cannot use a synthetic rendering as documentary proof. The skill is portable precisely because the evidence standards remain explicit while the subject matter changes.

Final diagnostic: describe the smallest conclusion the current source supports, the wider capability claim it does not yet justify, and one observation that would change that result. Name a potential social or environmental harm that a narrow statistic might conceal. The evidence reviewer’s evidence rule, the assumption checker’s challenge and the lesson facilitator’s revision help pupils practise intellectual honesty rather than reward confident guesses.

2. Geography and system scope

Did you know? Does evidence from one site justify a claim about an entire civilisation? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: One village or spectacular ruin may not represent a whole region. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: The classroom activity distinguishes one house, a village and a regional river system. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover site boundary, sample representativeness and comparative sites. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

3. Chronology and period boundaries

Did you know? Do items thought to belong together actually come from the same period? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A later inscription can be projected into a much older occupation layer. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children assemble three imaginary excavation layers and keep dates separate. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover dated contexts and evidence of reuse or movement. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

4. Stratigraphy and deposit formation

Did you know? How did an object or feature reach the layer where it was recovered? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Objects from disturbed deposits can be mistaken for intact contemporary practices. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: A fictional river flood moves older objects into a new deposit. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover archaeological context, disturbance and specialist interpretation. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

5. Preservation bias

Did you know? Which objects survive more easily, and what disappeared? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Stone or fired clay surviving more often does not prove perishable knowledge was absent. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: The tutor removes paper cards from a mock archive before a second group examines it. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover material survival, excavation coverage and alternative traces. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

6. Absence of evidence versus evidence of absence

Did you know? Does an unexcavated or poorly preserved site prove a practice never existed? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A missing building or document may reflect sampling, decay or relocation. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: An imaginary site has one excavated quarter while the rest remains unknown. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover survey coverage, material detectability and competing explanations. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

7. Recurring occupation and settlement

Did you know? Do repeated floors and repairs demonstrate the claimed kind of habitation? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Several huts are not always permanent year-round homes. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children compare seasonal return visits with uninterrupted fictional occupancy. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover building phases, repairs, subsistence remains and dating. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

8. Storage and usable surplus

Did you know? Does a large storage object reveal dependable reserves and access rules? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: One granary can be mistaken for evidence of universal abundance or a king. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: An invented classroom activity starts with 200 tokens and records spoilage and committed reserves. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover replenishment, usability, ownership uncertainty and duration. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

9. Symbols, tallies and writing

Did you know? What does a marked object encode, and what cannot be translated? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Scratch marks or undeciphered signs may be treated as an exact legal text. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Learners use a fictional numerical key and test whether a new reader understands it. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover specialist decipherment, context and repeated conventions. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

10. Oral knowledge and under-recorded labour

Did you know? What valuable practices might not be captured by surviving written sources? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A large archive can privilege rulers while excluding everyday care or practical expertise. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children record a safe paper skill passed by demonstration rather than text. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover multiple source types, craft traces and cultural context. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

11. Industrial objects versus operating output

Did you know? Does surviving machinery prove years of reliable production? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: One factory photograph may show a building but not delivered goods. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: A paper line records attempted, rejected and accepted pieces. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover output series, quality, energy, maintenance and transport. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

12. Population estimates and averages

Did you know? Are houses or artefacts contemporary, and what denominator is used? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Every excavated house may not have been occupied at the same moment. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Imaginary villages have 20, 40 and 100 accepted goods, producing a skewed mean. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover sampling period, simultaneous occupation and indicator definition. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

13. Global network reach and inclusion

Did you know? Does a physical global connection demonstrate meaningful access for all? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A cable or large port can coexist with excluded users and local bottlenecks. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Students trace a toy shipment from port to inland school. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover last-mile service, affordability and verified delivery. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

14. Graphs, units and changing methods

Did you know? Are two apparent trends truly comparable? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A graph can change units or data collection and create artificial improvement. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children compare readings before and after a fictional instrument change. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover baseline, method, uncertainty, denominator and data provenance. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

15. Causation and confounding

Did you know? Is the new policy genuinely responsible for the observed change? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A better river score after a wet season may be attributed to an unrelated project. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: The classroom activity supplies an intervention and a comparison location. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover counterfactual alternatives, timing and causal mechanism. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

16. Restoration outputs versus ecological outcomes

Did you know? What did the activity change in the world after implementation? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Saplings planted may be counted as a restored forest. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: A fictional habitat has planting counts and later survival observations. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover species suitability, ecosystem functions and follow-up. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

17. Institutional records and accountability

Did you know? Can the recorded decision be retrieved, corrected and acted upon? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A signed repair form can replace verification that the repair worked. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children compare a maintenance claim with a later inspection card. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover version history, actual service and accessible review. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

18. Official statistics and denominators

Did you know? What exact population or trade volume does the percentage describe? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Eighty percent of maritime merchandise volume can be misquoted as eighty percent of everything. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Pupils distinguish TEUs, goods and households using fictitious labels. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover source year, denominator and the source’s methodological notes. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

19. Current date and freshness

Did you know? Is a mission target or public statistic current at the specified date? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: An older Artemis plan can be treated as an event that already happened. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Students compare NASA’s October 23, 2026 plan with an outdated timeline. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover publication date, planned status and later official updates. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

20. Component experiment versus complete system

Did you know? Which narrow technical function did a scientific demonstration actually prove? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: MOXIE producing oxygen is not a staffed and autonomous Mars city. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children compare toy oxygen cards with a broader imagined life-support system. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover mission scope, operating duration and integration. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

21. Photography, illustration and synthetic media

Did you know? Is a compelling image a document or an artistic representation? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: AI-generated human settlements can be mistaken for authenticated archaeology or space photographs. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: the evidence reviewer asks children to caption a previously disclosed AI concept image correctly. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover original media provenance, disclosure and verifiable historical claims. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

22. Triangulation and independent sources

Did you know? Do several different kinds of evidence point to the same conclusion? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: Many sites repeating one press release are not necessarily independent confirmation. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Children compare an artefact, a source catalogue and a fictional secondary summary. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover independent observations, dates and method diversity. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

23. Confidence without invented precision

Did you know? Can uncertainty be expressed accurately without fake numerical probabilities? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: A worksheet assigns 97.3 percent confidence from one attractive picture. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Learners write supported, possible, insufficient and contradicted with reasons. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover quality and quantity of evidence, known alternatives and missing data. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

24. Revision and transparent correction

Did you know? Does a new fact improve the final report instead of being silently ignored? The first task of the researcher is to specify the claimed fact, its source and the period when the claim applies. A dramatic headline may be memorable, but the strongest evidence often consists of unglamorous dates, measurements, provenance and repeated records. In the eduKate Atlas a classification is about sustained operating capacity, which usually demands more than one artefact or announcement.

The trap in this case is: An outdated explanation can remain online after contrary evidence emerges. A pupil who spots it should not automatically assume the opposite interpretation is true. Ask what can be safely observed, what is inferred, and which alternatives remain plausible. Sources vary in completeness, and absence may reflect preservation, access or sampling as well as actual history. A careful explanation protects the dignity of the people involved and resists converting technology into a moral hierarchy.

classroom activity experiment: Pupils preserve a version-one claim and write a corrected version-two. The evidence reviewer checks whether the description is traceable to a trustworthy source. The assumption checker introduces a possible bias or missing condition. The lesson facilitator records a revised claim with a date and a note about confidence. The children must identify which details are invented teaching conditions and which are verifiable through a museum, scientific agency or report.

The next evidence request should cover change history, dated evidence and fair explanation. A strong pupil explains how each observation could strengthen or weaken a stage inference. They should compare at least two different classes of information where appropriate—for example a material feature with a dated written account, or an output statistic with a verified delivery. Repetition of a press release on several websites is not necessarily multiple independent confirmations.

Part III. Eighteen case dossiers and careful source interpretations

Case 1: Çatalhöyük’s occupation layers

UNESCO identifies multiple Neolithic occupation levels and houses with roof access. Begin by locating the best original source and checking the date and exact terminology. The site offers rich settlement evidence but cannot automatically supply every imagined political role. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

The assumption checker challenges the broader inference. Is the report describing an artefact, a repeated practice, an institution, a population or a planned future operation? What happens when we change the unit from one site to a whole society? The same object might support one narrow claim while leaving an ambitious stage inference unproven. Competing explanations should be judged against available evidence rather than treated as interchangeable guesses.

The lesson facilitator prepares a bounded conclusion with observation, inference, uncertainty and next test. Younger children can make a three-column card; older learners should record methods, relevant denominators and source provenance. The ethical safeguard is to avoid reducing the worth of a people to its technological tools or labelling a hypothetical off-world town as a current reality. A successful case study produces a verifiable paragraph, not the grandest possible title.

Case 2: Jomon sedentary lifeways

UNESCO documents long-lived pre-agricultural settled hunter-fisher-gatherer societies in northern Japan. Begin by locating the best original source and checking the date and exact terminology. Settled life is not universally identical to cereal agriculture. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 3: A fictional storage pit

A single pit contains grain-like material without a complete use history. Begin by locating the best original source and checking the date and exact terminology. The object may suggest storage without proving a reliably replenished communal surplus. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 4: Inka khipu accounting

The Smithsonian describes knotted-cord records used in Inka administration. Begin by locating the best original source and checking the date and exact terminology. Record keeping need not resemble alphabetic English writing. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 5: An undeciphered Indus sign

An inscription exists, but a specific translation is not securely available. Begin by locating the best original source and checking the date and exact terminology. Invented legal or tax translations must not be presented as verified history. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 6: A Meiji industrial site

UNESCO documents industry and locally adapted technology transfer in Japan. Begin by locating the best original source and checking the date and exact terminology. Industrialisation took more than one historical and institutional route. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 7: A famous steam machine

A photograph documents particular equipment at a particular time. Begin by locating the best original source and checking the date and exact terminology. It cannot alone prove continuous factory throughput or fair labour. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 8: A global shipping figure

UNCTAD describes a major role for sea transport in international merchandise by volume. Begin by locating the best original source and checking the date and exact terminology. The denominator should not become all economic activity or every household’s goods. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 9: Singapore container throughput

An official TEU figure describes handling activity at a major port. Begin by locating the best original source and checking the date and exact terminology. TEUs are not individual products or verified final household deliveries. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 10: Internet availability

ITU estimates broad internet use alongside substantial offline populations. Begin by locating the best original source and checking the date and exact terminology. Backbone links do not guarantee meaningful access or learning outcomes. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 11: A green restoration photograph

A vegetation image may show a scene or intervention on one day. Begin by locating the best original source and checking the date and exact terminology. It does not demonstrate causal ecological recovery across years. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 12: A Montreal Protocol assessment

WMO and UNEP report progress in ozone protection and expected future recovery. Begin by locating the best original source and checking the date and exact terminology. Projections are not already completed global restoration. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 13: NASA’s Artemis II flyby

NASA’s April 2026 release describes a completed crewed lunar flyby. Begin by locating the best original source and checking the date and exact terminology. A completed flyby is not a permanent human settlement on the Moon. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 14: The October Artemis plan

NASA’s October 23, 2026 update describes upcoming Artemis tests and intended landing targets. Begin by locating the best original source and checking the date and exact terminology. Planned missions must not be described as completed achievements. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 15: MOXIE’s Mars oxygen

NASA reports 122 grams generated during sixteen experimental runs. Begin by locating the best original source and checking the date and exact terminology. A robotic oxygen experiment cannot prove a resident Mars community. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 16: ISS life-support recovery

NASA reported a scoped water-recovery demonstration aboard the Earth-orbital station. Begin by locating the best original source and checking the date and exact terminology. One high subsystem recovery percentage is not total off-world autonomy. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 17: A fictional AI Moonbase picture

The rendering depicts domes and residents without authenticated operations. Begin by locating the best original source and checking the date and exact terminology. Concept art is not documentary evidence of inhabitants or safety. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Case 18: An imagined three-world network

Several paper settlements communicate through delayed routes in a classroom. Begin by locating the best original source and checking the date and exact terminology. A toy demonstration of the concept is not an existing Stage 8 civilisation. The evidence reviewer asks the student to identify which part can be observed or supported by an authoritative description. The learner should not invent a missing household, technology or date to complete the story.

Part IV. Twelve evidence-literacy lessons

Lesson 1: Source identity and provenance and Geography and system scope

Learning objective: combine source identity and provenance with geography and system scope in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Children compare a museum catalogue card with an unattributed social-media image. Round two: The classroom activity distinguishes one house, a village and a regional river system. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

In round three, the tutor provides a second source with a different kind of evidence. Learners compare dates, geography, denominators, methods and how independent the sources are. If they disagree, pupils should document the conflict and seek a way to resolve it. A later-period find or planned NASA mission should not be silently moved into an earlier historical period.

For young primary pupils, use picture cards with fact and guess labels. For upper primary, add a table and one arithmetic step. For secondary learners, add sampling, causal interpretation, preservation bias and source independence. Finish with a teach-back: can another student explain what the evidence shows and what it leaves unknown without seeing the original lesson title?

The ten-mark check awards two marks each for observation accuracy, provenance, justified inference, a meaningful alternative or limitation, and clear respectful explanation. The child should write which stage coordinate the finding might inform, but never force a positive stage label simply because the exercise is part of this series.

Lesson 2: Chronology and period boundaries and Stratigraphy and deposit formation

Learning objective: combine chronology and period boundaries with stratigraphy and deposit formation in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Children assemble three imaginary excavation layers and keep dates separate. Round two: A fictional river flood moves older objects into a new deposit. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 3: Preservation bias and Absence of evidence versus evidence of absence

Learning objective: combine preservation bias with absence of evidence versus evidence of absence in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: The tutor removes paper cards from a mock archive before a second group examines it. Round two: An imaginary site has one excavated quarter while the rest remains unknown. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 4: Recurring occupation and settlement and Storage and usable surplus

Learning objective: combine recurring occupation and settlement with storage and usable surplus in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Children compare seasonal return visits with uninterrupted fictional occupancy. Round two: An invented classroom activity starts with 200 tokens and records spoilage and committed reserves. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 5: Symbols, tallies and writing and Oral knowledge and under-recorded labour

Learning objective: combine symbols, tallies and writing with oral knowledge and under-recorded labour in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Learners use a fictional numerical key and test whether a new reader understands it. Round two: Children record a safe paper skill passed by demonstration rather than text. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 6: Industrial objects versus operating output and Population estimates and averages

Learning objective: combine industrial objects versus operating output with population estimates and averages in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: A paper line records attempted, rejected and accepted pieces. Round two: Imaginary villages have 20, 40 and 100 accepted goods, producing a skewed mean. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 7: Global network reach and inclusion and Graphs, units and changing methods

Learning objective: combine global network reach and inclusion with graphs, units and changing methods in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Students trace a toy shipment from port to inland school. Round two: Children compare readings before and after a fictional instrument change. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 8: Causation and confounding and Restoration outputs versus ecological outcomes

Learning objective: combine causation and confounding with restoration outputs versus ecological outcomes in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: The classroom activity supplies an intervention and a comparison location. Round two: A fictional habitat has planting counts and later survival observations. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 9: Institutional records and accountability and Official statistics and denominators

Learning objective: combine institutional records and accountability with official statistics and denominators in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Children compare a maintenance claim with a later inspection card. Round two: Pupils distinguish TEUs, goods and households using fictitious labels. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 10: Current date and freshness and Component experiment versus complete system

Learning objective: combine current date and freshness with component experiment versus complete system in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Students compare NASA’s October 23, 2026 plan with an outdated timeline. Round two: Children compare toy oxygen cards with a broader imagined life-support system. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 11: Photography, illustration and synthetic media and Triangulation and independent sources

Learning objective: combine photography, illustration and synthetic media with triangulation and independent sources in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: the evidence reviewer asks children to caption a previously disclosed AI concept image correctly. Round two: Children compare an artefact, a source catalogue and a fictional secondary summary. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Lesson 12: Confidence without invented precision and Revision and transparent correction

Learning objective: combine confidence without invented precision with revision and transparent correction in a defensible claim. The tutor begins with a source card, its date and a short observation. Every child must explain the observation in their own words before interpreting it. The historical or scientific example may be real, but any teaching numbers introduced later must be explicitly fictional.

Round one: Learners write supported, possible, insufficient and contradicted with reasons. Round two: Pupils preserve a version-one claim and write a corrected version-two. The evidence reviewer asks which statements come from the source; the assumption checker proposes an alternative interpretation or missing data; the lesson facilitator records a revised answer. A child is allowed to mark a conclusion as insufficiently supported. That is better than inventing an estimate or claiming that one impressive source proves an entire civilisational system.

Part V. Twenty-four worked source and calculation exercises

Worked evidence challenge 1: Source identity and provenance

A fictional report related to source identity and provenance predicts 100 units, but an appropriate count verifies only 90. The difference is 10, and 90.0% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A reposted photograph can lose the excavator, collection and date that gave it meaning. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. museum accession, context, dated cataloguing and the original source Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 2: Geography and system scope

A fictional report related to geography and system scope predicts 104 units, but an appropriate count verifies only 91. The difference is 13, and 87.5% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. One village or spectacular ruin may not represent a whole region. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. site boundary, sample representativeness and comparative sites Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 3: Chronology and period boundaries

A fictional report related to chronology and period boundaries predicts 108 units, but an appropriate count verifies only 92. The difference is 16, and 85.2% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A later inscription can be projected into a much older occupation layer. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. dated contexts and evidence of reuse or movement Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 4: Stratigraphy and deposit formation

A fictional report related to stratigraphy and deposit formation predicts 112 units, but an appropriate count verifies only 93. The difference is 19, and 83.0% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Objects from disturbed deposits can be mistaken for intact contemporary practices. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. archaeological context, disturbance and specialist interpretation Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 5: Preservation bias

A fictional report related to preservation bias predicts 116 units, but an appropriate count verifies only 94. The difference is 22, and 81.0% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Stone or fired clay surviving more often does not prove perishable knowledge was absent. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. material survival, excavation coverage and alternative traces Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 6: Absence of evidence versus evidence of absence

A fictional report related to absence of evidence versus evidence of absence predicts 120 units, but an appropriate count verifies only 95. The difference is 25, and 79.2% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A missing building or document may reflect sampling, decay or relocation. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. survey coverage, material detectability and competing explanations Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 7: Recurring occupation and settlement

A fictional report related to recurring occupation and settlement predicts 124 units, but an appropriate count verifies only 96. The difference is 28, and 77.4% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Several huts are not always permanent year-round homes. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. building phases, repairs, subsistence remains and dating Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 8: Storage and usable surplus

A fictional report related to storage and usable surplus predicts 128 units, but an appropriate count verifies only 118. The difference is 10, and 92.2% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. One granary can be mistaken for evidence of universal abundance or a king. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. replenishment, usability, ownership uncertainty and duration Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 9: Symbols, tallies and writing

A fictional report related to symbols, tallies and writing predicts 132 units, but an appropriate count verifies only 119. The difference is 13, and 90.2% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Scratch marks or undeciphered signs may be treated as an exact legal text. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. specialist decipherment, context and repeated conventions Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 10: Oral knowledge and under-recorded labour

A fictional report related to oral knowledge and under-recorded labour predicts 136 units, but an appropriate count verifies only 120. The difference is 16, and 88.2% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A large archive can privilege rulers while excluding everyday care or practical expertise. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. multiple source types, craft traces and cultural context Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 11: Industrial objects versus operating output

A fictional report related to industrial objects versus operating output predicts 140 units, but an appropriate count verifies only 121. The difference is 19, and 86.4% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. One factory photograph may show a building but not delivered goods. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. output series, quality, energy, maintenance and transport Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 12: Population estimates and averages

A fictional report related to population estimates and averages predicts 144 units, but an appropriate count verifies only 122. The difference is 22, and 84.7% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Every excavated house may not have been occupied at the same moment. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. sampling period, simultaneous occupation and indicator definition Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 13: Global network reach and inclusion

A fictional report related to global network reach and inclusion predicts 148 units, but an appropriate count verifies only 123. The difference is 25, and 83.1% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A cable or large port can coexist with excluded users and local bottlenecks. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. last-mile service, affordability and verified delivery Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 14: Graphs, units and changing methods

A fictional report related to graphs, units and changing methods predicts 152 units, but an appropriate count verifies only 124. The difference is 28, and 81.6% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A graph can change units or data collection and create artificial improvement. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. baseline, method, uncertainty, denominator and data provenance Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 15: Causation and confounding

A fictional report related to causation and confounding predicts 156 units, but an appropriate count verifies only 146. The difference is 10, and 93.6% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A better river score after a wet season may be attributed to an unrelated project. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. counterfactual alternatives, timing and causal mechanism Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 16: Restoration outputs versus ecological outcomes

A fictional report related to restoration outputs versus ecological outcomes predicts 160 units, but an appropriate count verifies only 147. The difference is 13, and 91.9% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Saplings planted may be counted as a restored forest. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. species suitability, ecosystem functions and follow-up Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 17: Institutional records and accountability

A fictional report related to institutional records and accountability predicts 164 units, but an appropriate count verifies only 148. The difference is 16, and 90.2% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A signed repair form can replace verification that the repair worked. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. version history, actual service and accessible review Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 18: Official statistics and denominators

A fictional report related to official statistics and denominators predicts 168 units, but an appropriate count verifies only 149. The difference is 19, and 88.7% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Eighty percent of maritime merchandise volume can be misquoted as eighty percent of everything. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. source year, denominator and the source’s methodological notes Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 19: Current date and freshness

A fictional report related to current date and freshness predicts 172 units, but an appropriate count verifies only 150. The difference is 22, and 87.2% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. An older Artemis plan can be treated as an event that already happened. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. publication date, planned status and later official updates Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 20: Component experiment versus complete system

A fictional report related to component experiment versus complete system predicts 176 units, but an appropriate count verifies only 151. The difference is 25, and 85.8% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. MOXIE producing oxygen is not a staffed and autonomous Mars city. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. mission scope, operating duration and integration Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 21: Photography, illustration and synthetic media

A fictional report related to photography, illustration and synthetic media predicts 180 units, but an appropriate count verifies only 152. The difference is 28, and 84.4% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. AI-generated human settlements can be mistaken for authenticated archaeology or space photographs. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. original media provenance, disclosure and verifiable historical claims Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 22: Triangulation and independent sources

A fictional report related to triangulation and independent sources predicts 184 units, but an appropriate count verifies only 174. The difference is 10, and 94.6% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. Many sites repeating one press release are not necessarily independent confirmation. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. independent observations, dates and method diversity Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 23: Confidence without invented precision

A fictional report related to confidence without invented precision predicts 188 units, but an appropriate count verifies only 175. The difference is 13, and 93.1% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. A worksheet assigns 97.3 percent confidence from one attractive picture. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. quality and quantity of evidence, known alternatives and missing data Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Worked evidence challenge 24: Revision and transparent correction

A fictional report related to revision and transparent correction predicts 192 units, but an appropriate count verifies only 176. The difference is 16, and 91.7% of the expected units were verified. Those numbers are not a historical, medical, financial or spaceflight measurement. The arithmetic is only a transparent classroom example of why expected, recorded and observed quantities must be distinguished.

The evidence reviewer asks whether the discrepancy has a cause already established. An outdated explanation can remain online after contrary evidence emerges. The assumption checker provides an alternative involving missing records, a change in measurement or some other relevant input. The lesson facilitator records “difference observed; cause unresolved” unless evidence supports a narrower explanation. The learner should not accuse an individual or invent a physical process simply because the difference is inconvenient.

The task then becomes an evidence plan. change history, dated evidence and fair explanation Would more measurements, a different source, a dated follow-up or specialist interpretation help? The best answer identifies a sensible next test and explains how its result might strengthen or weaken a specific stage claim. A correct number is necessary for the toy calculation but insufficient for a conclusion about an entire civilisation.

Part VI. Frequently asked questions and vocabulary

What is the difference between observation and inference?

Where exactly did the record originate, and who describes it today? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A reposted photograph can lose the excavator, collection and date that gave it meaning. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Does one artefact establish a civilisation stage?

Does evidence from one site justify a claim about an entire civilisation? The useful response starts with provenance, a defined unit and an appropriately narrow claim. One village or spectacular ruin may not represent a whole region. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Why does provenance matter?

Do items thought to belong together actually come from the same period? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A later inscription can be projected into a much older occupation layer. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

What is preservation bias?

How did an object or feature reach the layer where it was recovered? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Objects from disturbed deposits can be mistaken for intact contemporary practices. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

What does a missing record show?

Which objects survive more easily, and what disappeared? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Stone or fired clay surviving more often does not prove perishable knowledge was absent. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Can we use an AI illustration as a historical source?

Does an unexcavated or poorly preserved site prove a practice never existed? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A missing building or document may reflect sampling, decay or relocation. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

How do we identify undeciphered writing?

Do repeated floors and repairs demonstrate the claimed kind of habitation? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Several huts are not always permanent year-round homes. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Does an archive represent everyone?

Does a large storage object reveal dependable reserves and access rules? The useful response starts with provenance, a defined unit and an appropriately narrow claim. One granary can be mistaken for evidence of universal abundance or a king. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Why must the date be visible?

What does a marked object encode, and what cannot be translated? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Scratch marks or undeciphered signs may be treated as an exact legal text. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

How can a percentage be misleading?

What valuable practices might not be captured by surviving written sources? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A large archive can privilege rulers while excluding everyday care or practical expertise. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Do two news sites mean two independent sources?

Does surviving machinery prove years of reliable production? The useful response starts with provenance, a defined unit and an appropriately narrow claim. One factory photograph may show a building but not delivered goods. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Can a projection be counted as achieved?

Are houses or artefacts contemporary, and what denominator is used? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Every excavated house may not have been occupied at the same moment. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

What if official sources change their plans?

Does a physical global connection demonstrate meaningful access for all? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A cable or large port can coexist with excluded users and local bottlenecks. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

What is a proper alternative explanation?

Are two apparent trends truly comparable? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A graph can change units or data collection and create artificial improvement. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Does a greenhouse picture establish a lunar society?

Is the new policy genuinely responsible for the observed change? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A better river score after a wet season may be attributed to an unrelated project. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Is a negative experiment useful?

What did the activity change in the world after implementation? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Saplings planted may be counted as a restored forest. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

When should we say insufficient evidence?

Can the recorded decision be retrieved, corrected and acted upon? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A signed repair form can replace verification that the repair worked. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Can we give an exact confidence score?

What exact population or trade volume does the percentage describe? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Eighty percent of maritime merchandise volume can be misquoted as eighty percent of everything. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

How do we assess claims about Stage 8?

Is a mission target or public statistic current at the specified date? The useful response starts with provenance, a defined unit and an appropriately narrow claim. An older Artemis plan can be treated as an event that already happened. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Why include rights and fairness?

Which narrow technical function did a scientific demonstration actually prove? The useful response starts with provenance, a defined unit and an appropriately narrow claim. MOXIE producing oxygen is not a staffed and autonomous Mars city. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Can students safely study real environmental topics?

Is a compelling image a document or an artistic representation? The useful response starts with provenance, a defined unit and an appropriately narrow claim. AI-generated human settlements can be mistaken for authenticated archaeology or space photographs. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

What is triangulation?

Do several different kinds of evidence point to the same conclusion? The useful response starts with provenance, a defined unit and an appropriately narrow claim. Many sites repeating one press release are not necessarily independent confirmation. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

What do the evidence reviewer and the assumption checker do?

Can uncertainty be expressed accurately without fake numerical probabilities? The useful response starts with provenance, a defined unit and an appropriately narrow claim. A worksheet assigns 97.3 percent confidence from one attractive picture. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

What is the last numbered stage?

Does a new fact improve the final report instead of being silently ignored? The useful response starts with provenance, a defined unit and an appropriately narrow claim. An outdated explanation can remain online after contrary evidence emerges. A trustworthy explanation states one observation, one interpretation and a missing piece of evidence that could change the result. Source quality does not depend on making the story sound dramatic.

Evidence vocabulary

Observation: information directly documented by a source or suitable instrument. Inference: an interpretation supported by observations. Provenance: origin, context and date of a source. Stratigraphy: relationships among archaeological layers. Preservation bias: the uneven survival of materials. Denominator: the total against which a ratio is measured. Confounder: another factor affecting an apparent relationship. Counterfactual: a relevant alternative condition for causal reasoning. Confidence: a reasoned assessment of evidence strength rather than an invented precision score. Triangulation: comparing different appropriate evidence types. Data freshness: whether observations are current for the stated question. Synthetic media: generated or altered media that must not be mistaken for documentary proof. Correction history: records of how and why a claim changes.

Sources and nine companion stage guides

Confidence is earned by checkable explanations

A well-supported civilisation-stage description shows a connected operating function within a specified place and period, without hiding contradictory information. A good investigator preserves the difference between observation, interpretation, uncertainty and a next test. A photograph may capture a real object while leaving its function unknown. An AI illustration may be useful for explaining a concept but cannot document a real settlement. NASA experiments can be remarkable without creating an inhabited second world.

The evidence reviewer, the assumption checker and the lesson facilitator are fictional teaching roles that help children practise this method. They do not substitute for original museums, scientific agencies or research methods. The Atlas contains Foundation 0 and eight numbered stages, with Stage 8 a hypothetical human interplanetary threshold. Responsible curiosity continues beyond the last numbered article, but honest evidence must always guide the claim.