Listening Comprehension News Reports and Documentary Audio | Separate Events, Sources, Claims, Evidence and Commentary

Advanced News and Documentary Listening Casebook | Provenance, Evidence and Uncertainty Laboratory

The core article establishes the source-rich listening map. This advanced casebook tests it under conditions where polished narration can make every proposition sound equally authoritative. The listener must repeatedly ask: who owns this claim, what kind of evidence supports it, how certain is it, what has changed, and what does the report itself currently conclude? The cases below turn those questions into a practical examination system.

Case 1 | The narrator reports a claim without endorsing it

The report says, “Local campaigners say the decline was caused by the new fee.” The next sentence notes that officials have not confirmed the cause. A learner who hears “caused by the fee” and ignores “campaigners say” converts an attributed claim into report fact. Provenance words are part of the proposition.

Case 2 | An official statement is treated as independent confirmation

A company spokesperson repeats the organisation’s own explanation for an incident. The report then says investigators are still examining the cause. Official status does not automatically turn a self-reported explanation into independently established fact. Listen to how the report frames the source.

Case 3 | Witness reaction becomes causal evidence

A witness says she was frightened by a loud noise. The learner concludes the noise caused the technical failure. The testimony establishes experience of a sound, not the mechanism of failure. Evidence type controls what can be inferred.

Case 4 | Expert explanation becomes narrator conclusion

A researcher says the pattern is “consistent with” one explanation. The narrator later notes that other explanations remain possible. The expert’s interpretation is one source layer; the documentary’s final synthesis is broader and more cautious.

Case 5 | The narrator’s synthesis is stronger than any single source

Several independent observations point in the same direction. No source alone states the full conclusion, but the narrator says, “Taken together, the evidence suggests…” The report-level proposition is a synthesis. It still inherits the narrator’s chosen degree of certainty.

Case 6 | The narrator’s synthesis is weaker than a confident source

One interviewee says she is certain about the cause. The documentary ends by saying the evidence remains inconclusive. The final report state is uncertainty despite the source’s confidence.

Case 7 | Two witnesses agree on event, disagree on sequence

Both saw the same incident, but one says the alarm sounded before the lights failed and the other remembers the reverse. The report confirms the incident and leaves sequence disputed. Agreement on event does not imply agreement on chronology.

Case 8 | Two experts agree on data, disagree on interpretation

Both accept the measured increase. One attributes it to policy; the other to a broader trend. The data are shared; causal interpretation differs. Questions may target either layer.

Case 9 | Two sources disagree because they use different time periods

An official cites the last month; a researcher cites five years. Their claims sound contradictory until the time frames are restored. Source comparison requires shared measurement scope before disagreement can be judged.

Case 10 | Historical statistic mistaken for current state

A documentary describes a 40 per cent rate from ten years ago and later gives a current figure of 22 per cent. A learner who remembers the dramatic first number can answer the wrong period. Time belongs to data.

Case 11 | Preliminary number becomes final number

An early bulletin says “about 200” people were affected. A later update gives a confirmed count of 176. Both values belong in the report history, but only one is current. Exact-detail listening and status tracking must cooperate.

Case 12 | Final number remains approximate

The report ends with “nearly 200”. The learner writes exactly 200. Polished broadcast delivery does not make approximate language exact.

Case 13 | The same number refers to different populations

Thirty per cent of young respondents and thirty per cent of all respondents appear in different studies. Identical values do not mean identical evidence. Group and source labels remain necessary.

Case 14 | Sample size mistaken for result

The study involved 500 people and found 62 per cent supported the change. A question asks what proportion supported it. Both numbers sound important; their roles differ.

Case 15 | Percentage change mistaken for percentage-point change

The report uses precise wording to distinguish them. The listener should preserve the source’s phrasing rather than perform unasked calculations. When the task is about the reported statement, fidelity outranks improvisation.

Case 16 | Correlation becomes causation

The documentary notes that two trends rose together and explicitly says the relationship does not prove one caused the other. A learner who selects a causal option has strengthened the evidence beyond the source.

Case 17 | A causal study remains limited

An experiment supports a causal effect under specific conditions. The learner generalises it to all contexts. Evidence can be strong and still scope-limited.

Case 18 | One case study is treated as universal proof

A city successfully changed one system. The documentary uses it as an example, not proof that every city will achieve the same result. Example and general evidence differ.

Case 19 | One counterexample defeats an absolute claim

A speaker says a method “always” works; the report presents a clear failure. The counterexample is powerful against universality even if most cases still succeed.

Case 20 | One counterexample does not defeat a qualified claim

The report says the method “usually” works. One failure is compatible with the claim. Scope words determine evidential force.

Case 21 | A source’s emotional certainty exceeds evidential certainty

A witness is completely convinced about what caused an event. The report notes she could see only one part of the scene. Confidence and access are separate dimensions.

Case 22 | Expert hedging is mistaken for weakness

A researcher says “the evidence strongly suggests” rather than “proves”. The learner interprets this as indecision. In evidence-rich discourse, calibrated caution can reflect precision rather than ignorance.

Case 23 | “According to” disappears from memory

The phrase is short and low-salience, yet it determines ownership. Notes should make source visible when several claims compete.

Case 24 | “Reportedly” disappears from memory

The learner stores the proposition but loses its reduced evidential status. Certainty words are not decorative journalism; they control what the report commits to.

Case 25 | “Officials confirmed” updates an earlier claim

The report begins with uncertainty and later confirms the same proposition. The model should strengthen status rather than store two separate claims.

Case 26 | “Officials denied” creates competing source claims

A campaign group alleges one cause; officials deny it. The report may not resolve the dispute. Both source positions remain, with attribution intact.

Case 27 | “No evidence has been found” becomes “proved false”

Absence of evidence is not always evidence of absence. Unless the report draws that conclusion, the answer should preserve the narrower status.

Case 28 | “No longer under investigation” becomes “proved true”

An investigation can close for several reasons. Use the specific conclusion stated, not assumptions about procedural outcomes.

Case 29 | A correction is buried late

The report repeats an early figure several times and corrects it near the end. Repetition creates memory strength, but final status still controls.

Case 30 | A correction changes only one field

The location remains correct while the time changes. Updating the whole event can introduce errors. Source-rich reports need selective state revision just like practical announcements.

Case 31 | Reporter paraphrase is broader than direct quote

A source says “some sites were affected”; the reporter says “the network experienced disruptions”. Depending on context, the broader summary may still be accurate. Questions should distinguish exact quote from report-level synthesis.

Case 32 | Direct quote is narrower than narrator claim

The narrator presents a general trend and uses one participant quote to illustrate it. The quote does not carry the whole generalisation; the report structure does.

Case 33 | Quote selected for human impact

A witness says, “We had nowhere to go.” The line illustrates lived consequence, not the technical cause. Function matters more than emotional salience.

Case 34 | Quote selected as counterpoint

Most evidence supports one interpretation, but the report includes a dissenting expert. The source’s function is contrast. Do not mistake inclusion for equal evidential weight unless the report presents it that way.

Case 35 | Quote selected as historical voice

An archival recording explains how people understood the issue decades ago. That historical viewpoint is not necessarily the documentary’s current conclusion.

Case 36 | Archive audio is treated as present commentary

Voice quality or transition cues may signal historical material. Time-source binding is essential.

Case 37 | Documentary shifts from past to present without date repetition

“Today, however…” is the reset. A listener still anchored in historical context can misassign later evidence.

Case 38 | Current narrator evaluates past policy

The policy existed historically; the judgement is contemporary. Event time and commentary time differ.

Case 39 | Historical analogy mistaken for causal connection

The documentary compares two events to show similarity. It does not say the earlier event caused the later one. Comparison is not causation.

Case 40 | Historical context is the cause

In another report, a long-running institutional rule directly shapes the current event. Context can be causal when the audio establishes the link. No one universal rule replaces evidence.

Case 41 | Documentary uses reconstruction

A narrator describes what likely happened based on records. The learner treats the scene as direct observation. Reconstructed events should retain their evidential framing.

Case 42 | Documentary uses simulation or model

A model shows how a process could unfold. The learner hears it as a report of what did happen. Hypothetical demonstration and historical event need separate status.

Case 43 | Expert prediction becomes event fact

A scientist says a system “could fail under continued stress”. A learner answers that failure occurred. Prediction and event are different information roles.

Case 44 | Recommendation becomes prediction

An expert says authorities should change policy. The learner records that authorities will change policy. Normative and predictive claims differ.

Case 45 | Prediction becomes recommendation

A report says a trend is likely to continue; the learner answers that the expert recommends continuing it. Same topic, wrong speech act.

Case 46 | A source has expertise but poor access to this event

An expert explains general mechanisms but was not present. A witness knows the event but not the technical mechanism. Each source contributes a different evidential role.

Case 47 | Witness has first-hand access but limited view

The witness saw one entrance, not the entire building. Do not generalise beyond access if the report highlights the limitation.

Case 48 | Official has broad data but institutional perspective

The report may present the official source alongside independent analysis. The listener’s job is attribution, not political judgement.

Case 49 | Researcher comments outside own evidence

The documentary can distinguish research findings from the researcher’s personal recommendation. Source identity is the same; claim type changes.

Case 50 | Participant story is used as evidence of impact

A personal narrative can show consequence while broader data establish prevalence. One source provides depth, another breadth.

Case 51 | Several sources converge

A witness, data set and independent expert all support one conclusion. The learner can note convergence without losing which source contributes what.

Case 52 | Several sources repeat one original source

Three speakers cite the same report. Apparent source count is larger than independent evidence count. Unless the audio highlights this distinction, do not infer independence merely from repetition.

Case 53 | Source disagreement is factual

Two accounts give different event times. The conflict concerns what happened.

Case 54 | Source disagreement is causal

Both agree on events and disagree about why.

Case 55 | Source disagreement is evaluative

Both agree on facts and disagree whether the outcome is good or bad.

Case 56 | Source disagreement is predictive

Everyone accepts current data but forecasts differ. The future remains uncertain.

Case 57 | Report resolves one disagreement

Later evidence confirms the time but leaves cause disputed. Update only the resolved field.

Case 58 | Report opens a new uncertainty

New data solve the original question and create another. Current synthesis includes both progress and remaining unknown.

Case 59 | Final sentence is not final synthesis

The report ends with a human-interest quote after the narrator has already stated the current conclusion. Last-line bias can distort main point.

Case 60 | Ideal provenance map

The learner can state the central event, major sources, each source’s claim, the evidence type, certainty, areas of agreement and disagreement, current report synthesis and unresolved question. Names can fade if roles remain sufficient; provenance cannot.

Advanced Provenance Laboratories

Lab 1 | Source-arrow drill

Listen to short reports and write source → claim. Add one certainty marker. This is the smallest provenance unit.

Lab 2 | Source-role contrast

Use witness, official, researcher and participant discussing the same event. Ask what each source can directly know and what each is interpreting.

Lab 3 | Quotation-function sorting

Classify quotes as evidence, human impact, counterpoint, historical voice, explanation or recommendation. Function becomes more useful than memorising wording.

Lab 4 | Claim-status ladder

Order confirmed, strongly supported, likely, possible, disputed and unknown. Then place report statements on the ladder using exact source language.

Lab 5 | Attribution-removal test

Give a statement with “according to” removed and ask how meaning changes. This makes provenance words visible.

Lab 6 | Reported-speech nesting

A reporter quotes an official who cites a researcher. The learner tracks three levels of source. Keep notes minimal but hierarchical.

Lab 7 | Independent-versus-repeated-source test

Several speakers repeat one study, while another independent dataset appears. Ask how many distinct evidence sources actually exist.

Lab 8 | Current-status update

Use reports that begin tentative and later confirm, correct or retract one field. The learner updates status without rewriting everything.

Lab 9 | Historical/current split

Alternate archival material and present commentary. The listener labels time and source for each statement.

Lab 10 | Witness-versus-expert split

One source describes experience; another explains mechanism. Questions alternate which type of evidence is required.

Lab 11 | Fact-versus-interpretation sorting

After a clip, sort statements into observation, claim, interpretation, prediction and recommendation. This prevents evidential flattening.

Lab 12 | Competing-explanation matrix

Rows are explanations; columns are supporting evidence, source and limitation. Leave the final state unresolved when the report does.

Lab 13 | Evidence-strength calibration

Provide one case study, one survey and one controlled experiment, each with explicit limitations. The learner matches conclusion strength to what the report says, not to generic assumptions about research hierarchy.

Lab 14 | Number-provenance drill

Every number must have source, group, time and unit where relevant. Naked statistics are banned during training.

Lab 15 | One-listen source-switch stress

Use rapid transitions among narrator, witness and expert. Measure whether one missed name causes source cascade. Role labels should support recovery.

Lab 16 | Replay provenance repair

Where audio repeats, target one uncertain source or certainty field. The second hearing should not become another full transcription pass.

Lab 17 | Unknown-name tolerance

Use unfamiliar source names while roles remain clear. The learner practises role-first notation when spelling is irrelevant.

Lab 18 | Similar-name stress

Two experts have similar surnames. Role, institution or claim fingerprint must prevent blending.

Lab 19 | Source-convergence laboratory

Several independent sources support a proposition with different evidence. The learner writes a synthesis that preserves source diversity and calibrated certainty.

Lab 20 | Source-divergence laboratory

Sources disagree. The learner states the disagreement precisely—fact, cause, value or prediction—without deciding beyond the report.

Advanced Documentary Laboratories

Lab 21 | Framing-question map

Identify the opening documentary question, then label which sections provide context, evidence, testimony, counterpoint and synthesis.

Lab 22 | Narrator-versus-source map

Use two columns for narrator statements and attributed source claims. Then identify where the narrator synthesises them.

Lab 23 | Archive-context test

Mark every historical clip or quotation with date and function. Ask whether it explains cause, provides comparison or illustrates past belief.

Lab 24 | Testimony-depth test

Use one participant story and broader statistics. Ask what the testimony adds that numbers do not, without treating it as prevalence evidence.

Lab 25 | Expert-caution test

Use specialists who clearly separate known, likely and unknown. The learner mirrors their certainty language in answers.

Lab 26 | Narrator-synthesis test

After several sources, ask for the documentary’s current conclusion rather than any single quote. This trains report-level modelling.

Lab 27 | Unresolved-question test

The documentary ends without closure. The learner states what is known and what remains open in one calibrated sentence.

Lab 28 | Correction-history test

Track how the report’s account changes as new evidence arrives. Earlier states remain historical, final state controls current questions.

Lab 29 | Commentary-versus-description test

Separate what happened from what the narrator says it means. This is especially important in polished documentary prose where the two can sit close together.

Lab 30 | Full unseen source-rich simulation

Use a new report or documentary-style audio under target playback rules. Before checking, require event map, source map, certainty state and unresolved question. Review every error by provenance, evidential role, certainty, time, number binding or recovery.

Advanced Diagnostic Profiles

Profile A | Strong event comprehension, weak provenance

The learner knows what happened but not who claims why. Train source arrows.

Profile B | Strong source tracking, weak evidence roles

Names are accurate, but observation, interpretation and recommendation blur. Train role sorting.

Profile C | Strong facts, certainty inflation

The learner removes may, likely and disputed. Train claim-status ladders.

Profile D | Strong single-source audio, weak multi-source reports

Interference is active. Add provenance matrices gradually.

Profile E | Strong current events, weak historical layering

Past context is answered as current state. Train time-source binding.

Profile F | Strong qualitative claims, weak numbers

Data lose group, source or time. Use number-provenance drills.

Profile G | Strong early sections, weak final synthesis

Source details remain but report-level conclusion disappears. Train narrator synthesis.

Profile H | Strong familiar topics, weak unfamiliar documentaries

World knowledge is carrying comprehension. Rotate domains while preserving source architecture.

Advanced measurement framework

Metric 1 | Provenance accuracy

Can the learner attach each important proposition to the correct source?

Metric 2 | Evidential-role accuracy

Are observation, interpretation, prediction and recommendation distinguished?

Metric 3 | Certainty preservation

Does the learner retain may, likely, confirmed and disputed?

Metric 4 | Source-convergence accuracy

Can the learner recognise genuine independent support without merging sources?

Metric 5 | Source-divergence accuracy

Can disagreements be classified precisely?

Metric 6 | Number-provenance accuracy

Do values retain group, period, unit and source?

Metric 7 | Historical/current accuracy

Are archival and current states kept separate?

Metric 8 | Narrator-synthesis accuracy

Can the learner state the report-level conclusion beyond individual quotes?

Metric 9 | Recovery latency

After a difficult source switch or unfamiliar name, how quickly is provenance restored?

Metric 10 | Late-report decay

Do source, certainty and synthesis errors rise near the end?

Advanced teacher protocol

Before asking whether a statement is correct, ask who says it. Then ask what kind of statement it is and how strongly it is supported. This sequence turns a polished report back into its evidence architecture. Use one visible source matrix early, fade it over time, and keep outside fact-checking separate from examination comprehension. The exam question asks what this source says and supports; real-world verification is another valuable but different task.

Advanced self-study protocol

After one report, write a source map before checking answers: narrator, witness, expert, official, document. Add one claim and one certainty marker per important source. Then state the current report conclusion and one unresolved question. On review, classify errors by provenance, evidence role, certainty, chronology, number binding or answer transfer. Use the transcript only on the corridor that broke.

Advanced parent protocol

Parents can use age-appropriate news or documentary audio to ask “Who says that?”, “What evidence did they give?”, and “What is still uncertain?” Avoid turning the exercise into a debate about whether the source is politically or personally agreeable. The learning goal is provenance and evidential understanding. Broader media literacy can grow from that foundation later.

Advanced final checklist

• I can hear a quoted claim without turning it into narrator fact.
• I separate event, explanation, interpretation and recommendation.
• I preserve the certainty level of every important claim.
• I know what each source contributes and what it cannot directly know.
• I distinguish witness experience from expert explanation.
• I keep historical context separate from current status.
• I attach every important number to source, group, time and unit.
• I can hold competing explanations without forcing a winner.
• I know when a later update confirms, corrects or leaves an earlier claim unresolved.
• I can recover after unfamiliar names by using roles and claims.
• I can state the report’s current synthesis separately from the most vivid quotation.
• I can transfer the provenance system to unfamiliar topics and documentary structures.

Advanced closing synthesis | Source-rich listening is the art of keeping voices nested correctly

A polished report can make many sources sound like one voice. The mature listener mentally unfolds that voice again. This proposition belongs to the witness. That interpretation belongs to the researcher. This figure comes from an official record. This historical quotation explains context. This sentence is the narrator’s synthesis. This point remains disputed. This later update changes the certainty of what came before.

Once those layers stay separate, news and documentary audio becomes much easier to reason about. The listener no longer asks only “What did I hear?” She asks “Who said it, on what basis, how strongly, and what does the report now allow me to conclude?”

Practice Bank | 64 Provenance and Evidence Stress Tests

This bank develops source-rich listening through repeated provenance decisions. Each stress test asks the learner to preserve who says what, on what basis, how strongly and at what point in the report. The surface can be news, documentary, historical reporting, expert commentary or participant testimony. The deep system remains source → claim → evidence → certainty → current synthesis.

Bank A | Source attribution

A1 | Reporter or witness?

The reporter paraphrases a witness account and then gives a separate confirmed update. Ask which statement belongs to the witness. Score whether reported speech remains attributed.

A2 | Official or reporter?

An official claims the disruption was unavoidable. The reporter later says the cause remains under investigation. Ask what is confirmed. The official explanation must not become narrator fact.

A3 | Researcher or institution?

A university report is summarised, then one researcher gives a personal interpretation. Ask which claim belongs to the study and which to the individual.

A4 | Participant or narrator?

A participant says the change improved her daily routine; the narrator says broader results are mixed. Ask for each perspective separately.

A5 | Historical source or current narrator?

An archival voice describes past expectations. Current narration explains what later happened. Score time-source binding.

A6 | Quoted source inside another source

An official says “our engineers tell us…” and then gives an explanation. Ask whose technical assessment is being reported. Track nested provenance.

A7 | Anonymous source

The report says “people familiar with the process say…” while an official source says something else. Preserve the level of attribution given without inventing identity or credibility.

A8 | Source switch without repeated name

The report moves from witness to expert using a role phrase rather than a name. Ask who owns the next claim. Score transition sensitivity.

Bank B | Evidential roles

B1 | Observation versus interpretation

The report states that attendance fell and one expert says the decline reflects changing habits. Ask which is directly observed and which is interpretation.

B2 | Data versus recommendation

A survey result is followed by a recommendation to change policy. Ask which statement is evidence and which is normative action.

B3 | Witness experience versus general claim

A participant describes one difficult journey. The narrator then says the problem affects many users based on broader data. Ask what each source contributes.

B4 | Example versus proof

A documentary presents one successful project to illustrate a possible solution. A distractor claims the example proves the method always works. Score scope control.

B5 | Historical context versus cause

The narrator explains earlier policy and then discusses a current failure. Ask whether the historical material directly causes the event or simply provides background. Use the audio’s stated relationship.

B6 | Prediction versus current fact

An expert says demand may rise next year. Ask what is true now. Score future-status control.

B7 | Recommendation versus forecast

A source says authorities should expand the service. Another says they probably will. Ask which is advice and which is prediction.

B8 | Explanation versus evaluation

An expert explains why a system changed; a participant calls the change unfair. Ask which statement is causal and which is evaluative.

Bank C | Certainty and status

C1 | Possible versus confirmed

Early audio says a technical fault may have caused the issue. Later evidence confirms it. Ask what the report knows at each stage.

C2 | Likely versus certain

The narrator says one explanation is most likely but not proven. Options differ only in certainty. Score modal preservation.

C3 | Disputed claim

One source alleges a problem; another denies it. The report offers no resolution. Ask the current status. The answer is disputed, not true or false.

C4 | Preliminary result

A first study suggests an effect. The narrator stresses that findings are preliminary. Ask how strongly the conclusion should be stated.

C5 | Retraction

An earlier figure is corrected publicly. Ask which number is current. Historical truth and current accuracy differ.

C6 | Unconfirmed witness account

A witness reports hearing an explosion, while investigators have not identified the source of the sound. Ask what is established. The experience is reported; cause is not.

C7 | Strong evidence with limited scope

A controlled study supports an effect in one context. Ask whether the result applies universally. Certainty and scope must both survive.

C8 | No evidence yet

The report says no evidence has been found for one theory. Ask whether the theory has been disproved. The answer may remain no, depending on source wording.

Bank D | Competing sources and explanations

D1 | Same evidence, two causes

Two experts accept the same data and offer different causal explanations. Ask what they agree on and what they dispute.

D2 | Different evidence, same conclusion

A witness account and a dataset independently support one conclusion. Ask how the sources converge without merging their evidential roles.

D3 | Source A stronger on event, Source B stronger on mechanism

A witness saw what happened; an engineer explains how it could happen. Ask which source supports which question.

D4 | Official versus independent analyst

Both discuss the same outcome but interpret responsibility differently. The learner must attribute rather than decide from personal trust.

D5 | Two studies, different populations

Results differ because the groups differ. Ask whether the findings truly conflict. Scope can explain apparent disagreement.

D6 | Two time periods

An older report and a new report produce different numbers. Ask what changed over time rather than treating one as wrong.

D7 | Majority sources versus dissenting source

Several experts support one explanation; one disagrees. Ask what the documentary concludes. Source count alone should not replace narrator synthesis.

D8 | Unresolved explanation

The report presents two plausible causes and ends without choosing. Ask what remains uncertain. Score tolerance for unresolved status.

Bank E | Historical and temporal layering

E1 | Past event, current consequence

A decades-old decision still shapes today’s system. Ask which information is historical and which describes current effect.

E2 | Historical analogy

The documentary compares the current event with an earlier one. Ask why history is included. Comparison is not necessarily causation.

E3 | Archive quote

An old recording expresses a belief later shown to be wrong. Ask what people believed then, not what is known now.

E4 | Historical statistic

A past figure is dramatically higher than the present one. Ask for current value. Time label protects against salience.

E5 | Investigation timeline

Initial suspicion, new evidence and final confirmed cause appear over several updates. Ask how the report changed.

E6 | Policy timeline

Proposal, implementation and later review are all discussed. Ask what is current policy versus earlier plan.

E7 | Research timeline

An early study suggests an effect; a later study qualifies it. Ask the present research state. Final synthesis should include the qualification.

E8 | Return-to-present marker

Use a long historical section followed by “Today, however…”. Score whether the learner resets the time frame promptly.

Bank F | Numbers, data and provenance

F1 | Value plus source

Two sources give different estimates. Ask which number belongs to which source and whether the report resolves the discrepancy.

F2 | Value plus group

The same percentage appears for two populations. Ask about one subgroup. Score group binding.

F3 | Value plus year

Several annual figures are mentioned. Ask for current year. Recency should not be inferred without labels.

F4 | Sample size versus result

A study of 800 people finds 54 per cent support. Ask which number describes sample size and which describes finding.

F5 | Approximate versus exact

The narrator says “around 1,000” while an official gives 982. Ask which figure the official reports and which the narrator uses for summary.

F6 | Rate versus count

The report gives both total cases and rate per thousand. Ask one specifically. Unit and metric must remain attached.

F7 | Change versus level

A figure rises by 12 per cent to a new total. Ask whether 12 describes change or final level. Score role.

F8 | Statistic as evidence, not answer

A number supports a conclusion, but the question asks what the narrator concludes. The learner must move from data to proposition without replacing one with the other.

Bank G | Transfer across report types

G1 | Breaking news bulletin

Information is tentative and updated quickly. Score certainty and correction tracking.

G2 | Investigative report

Several sources disagree and documents are introduced. Score provenance and evidential role.

G3 | Science documentary

Expert explanation, data and limitations dominate. Score claim strength.

G4 | Historical documentary

Archive voices, current narration and reinterpretation interact. Score time-source layering.

G5 | Human-interest report

Participant testimony is vivid while broader data provide context. Score testimony function.

G6 | Public-service report

Official instructions and background explanation coexist. Score event, recommendation and authority source separately.

G7 | Environmental report

Competing causal explanations and model projections appear. Score uncertainty and prediction status.

G8 | Full novelty documentary

Use unfamiliar topic, several source types and mixed evidence only after component skills are stable. Review the first source or certainty breakdown.

Bank H | Endurance and final integration

H1 | Long source-rich report

Use several sources over sustained duration. Compare early and late provenance accuracy. Endurance failure often appears as source flattening.

H2 | Late correction

Place a decisive update near the end. Score whether earlier repeated information remains incorrectly dominant.

H3 | Late source disagreement

Introduce a dissenting expert after a long apparent consensus. Ask current synthesis. The listener must update without overreacting.

H4 | Late historical switch

Move into archive material near the end. Score whether current/past labels remain stable under fatigue.

H5 | One missed source name

Use a difficult name followed by a clear role and claim. Score whether the learner recovers provenance through the role.

H6 | Mixed evidence questions

Alternate event, source, evidence, certainty, number and narrator-synthesis questions. The learner must answer at the correct evidential layer.

H7 | One-listen final documentary

Use target playback rules and sparse source notes. Score containment after one unresolved quotation.

H8 | Full unseen final simulation

Use new topic, source mix and realistic timing. Before checking, require one event statement, source map, current synthesis and unresolved question. This is the final transfer test for the 200 architecture.

Scoring rubric for this bank

Score Source, Role, Certainty, Time and Synthesis. Source asks who owns the proposition. Role asks fact, evidence, interpretation, prediction or recommendation. Certainty asks how strongly it is supported. Time asks historical versus current status. Synthesis asks what the report as a whole currently allows the listener to conclude.

Learner review template

For each significant error, record: proposition; source; evidential role; certainty; time/status; wrong interpretation; corrected map; one repair rule; one new report type for transfer. Keep review source-centred rather than rewriting the entire article or documentary.

Practice-bank closing synthesis

Source-rich listening becomes robust when polished narration no longer erases provenance. The learner can let names fade where roles are enough, but she cannot let claims lose owners, evidence lose status or tentative interpretations become facts. These stress tests succeed when unfamiliar reports still produce a stable map of who says what, why, how strongly and what remains unresolved.