What Is Evidence-Based Revision in G3 History? | Change an Interpretation When Evidence Changes

CHAPTER 1 OF 19 · Locate the change

1. Revising a view is not the same as abandoning standards

Back to contents

A student may feel that a strong answer must defend its first interpretation consistently. Consistency matters, but not when it protects an error. Historical reasoning is accountable to evidence. Changing an account for a stated reason can make it stronger, while refusing to change despite contrary evidence can make confidence less justified. The standard stays steady even when the conclusion changes.

Fictional teaching example: an initial source suggests a market relocation reduced travel costs. Later records show that the improvement applied mainly to central traders, while outer-district traders faced longer journeys. A revised interpretation can recognise the documented gain and its uneven distribution. It need not move from complete success to complete failure simply because the first view was too broad.

The important question is what changed. Did the fact change, did the available evidence expand or did your interpretation of existing evidence improve? These are different processes. Naming the process helps you explain the revision and avoid presenting every correction as the discovery of an entirely different history.

Standards include accurate quotation, relevant context, consistent scope and fair testing of alternatives. Apply them to the new material as carefully as to the old. A source that challenges your preferred answer deserves attention, but challenge alone does not make it trustworthy. Revision should follow evaluation, not novelty or surprise.

A useful reflection states, “I changed X because Y, while Z remains supported.” This keeps revision specific. It preserves findings that still hold and identifies the exact part needing repair. The student learns that an interpretation can improve through controlled adjustment rather than dramatic reversal.

CHAPTER 2 OF 19 · Locate the change

2. Record the original claim clearly enough to test it

Back to contents

You cannot revise a claim accurately if its meaning keeps shifting. Write the original proposition with its actor, period, place and degree of certainty. A vague statement such as “the policy worked” is difficult to test because success could mean many things. Specify whether the claim concerns access, cost, implementation, support or another criterion.

Fictional teaching example: “The transport trial improved access for all residents during its first month” contains several testable parts. It concerns a trial, a population, access and a period. A later letter from one excluded district challenges universal reach. It does not necessarily challenge every claim about service use or the experience of groups already served.

Identify the evidence supporting the original view. Was it a launch speech, a register or a later summary? A revision may reveal that the original support never established the claim in the first place. That is different from a well-supported account being overturned by genuinely new information. The distinction helps diagnose the learning issue.

Keep the initial version briefly in notes rather than overwriting it without explanation. You do not need an elaborate history of every sentence. A simple before-and-after record can show the missing operation: qualifier control, date checking, source comparison or causal evaluation. This makes feedback transferable to later tasks.

Before accepting a proposed revision, ask whether it addresses the original claim or a neighbouring one. Evidence of high attendance may challenge a claim of non-use but not a claim about affordability. Evidence about later outcomes may not settle immediate effects. Accurate revision begins with accurate identification of what is being tested.

CHAPTER 3 OF 19 · Locate the change

3. Evaluate the new source before letting it change the answer

Back to contents

Newly encountered evidence is not automatically stronger than familiar evidence. It may be more direct, better documented or broader in coverage, but it may also be mistaken, selective or dependent on an earlier source. Ask what the material adds and why its particular claim deserves weight. The same analytical standards apply on both sides.

Fictional teaching example: a late memoir claims that a council secretly planned to close a clinic. The memoir challenges an earlier account of financial pressure, but the claim needs evaluation. Was the writer involved? Does the chronology fit? Is there corroboration? A dramatic allegation should not replace a documented explanation merely because it feels revelatory.

Distinguish authenticity from accuracy. A genuine letter can contain a rumour. A correctly attributed report can use a flawed measure. Conversely, a later account can accurately synthesise several records. The source’s existence and origin matter, but its evidential value concerns the claim you intend to use it for.

Check independence as well. Three new articles may repeat one old allegation. Their agreement does not necessarily provide three separate confirmations. Trace the origin where possible and state uncertainty where the pack does not establish it. Revision should not be driven by the apparent volume of repeated information.

A practical note records the new detail, its relevance, a reason for confidence and an important limit. Only then decide how it affects the interpretation. This sequence slows the jump from surprise to certainty while keeping the enquiry open to genuine correction. Good revision is responsive without being credulous.

CHAPTER 4 OF 19 · Test its consequences

4. Separate a factual correction from an interpretive change

Back to contents

A factual correction repairs something such as a date, identity, sequence or quoted wording. An interpretive change alters the explanation or judgement drawn from facts. Sometimes one leads to the other. A corrected date may make a proposed cause impossible; a corrected figure may narrow a claim about coverage. State both steps when they matter.

Fictional teaching example: an essay places a protest before a fee increase, but records establish that the increase came first. Correcting the sequence changes which causal explanations are plausible. It does not automatically prove that fees caused the protest. That further inference still needs evidence connecting the increase to participants’ reasons or actions.

Interpretive changes can occur without new facts. You may notice that a speech expresses an aim rather than an outcome, or that a source covers only one district. The material was already present; the reasoning improves. This is genuine progress and should not be described as simply changing wording if the claim’s meaning has changed.

Avoid treating every disagreement as a factual error. Two historians may accept the same chronology but weigh causes differently. Their interpretations should be tested through mechanisms, scope and explanatory fit. Calling one account wrong requires identifying what fails, not merely noting that another conclusion exists.

For practice, label a teacher’s correction as fact, inference, scope, weighting or communication. These are learning categories, not official mark labels. The label helps choose a repair. A factual check will not by itself solve an unsupported causal link, while a clearer sentence will not rescue an incorrect date.

CHAPTER 5 OF 19 · Test its consequences

5. New evidence may narrow a claim without reversing it

Back to contents

Revision often changes the reach of an interpretation rather than its direction. Evidence may support improvement for one group, during one phase or on one criterion while challenging a universal conclusion. A precise narrowing preserves the valid finding and removes the overclaim. This is usually more informative than switching from entirely positive to entirely negative.

Fictional teaching example: a housing report records lower rents for new tenants, while later accounts show existing tenants received no comparable reduction. The revision should distinguish the groups. “The policy reduced rents for the documented new tenants” may remain secure. “The policy made housing cheaper for everyone” does not. The new account is more differentiated, not simply less favourable.

Time narrowing works similarly. A scheme may operate well initially but deteriorate later. A source from the later phase does not necessarily invalidate evidence of early success. It challenges a claim that success persisted throughout. The revised explanation should show the sequence and, where possible, investigate what changed.

Do not use a qualifier as a cosmetic patch. Replacing all with some may be necessary, but explain which group the evidence covers and why the broader statement fails. The reader should understand the boundary rather than receive a weaker-sounding version of the same vague claim.

A useful revised paragraph names the retained finding first, then the new limit and its significance. This structure helps students see that correction does not erase all previous work. Historical understanding often develops through a sharper distribution of effects, not through a single verdict that labels the whole event successful or unsuccessful.

CHAPTER 6 OF 19 · Test its consequences

6. Contradictory sources may concern different things

Back to contents

Before treating a new source as a direct challenge, compare the claim, group, date and measure. A diary about construction disruption may coexist with a later report about improved access. A register of visits may differ from a survey of satisfaction because they measure different things. Apparent contradiction should be investigated before it drives revision.

Fictional teaching example: one source says a new ferry was heavily used, while another says many residents disliked its timetable. These claims can both be true. People may use an inconvenient service when alternatives are limited. The revised interpretation may recognise substantial use alongside dissatisfaction rather than choose one source as wholly false.

Differences in perspective can also explain variation. An official may describe administrative achievement while a worker describes household cost. Neither automatically answers the other’s question. An account that integrates both may become more complete without requiring one voice to be discredited. The revision concerns dimensions of the event.

If the sources do conflict on the same claim, test access, method, purpose and corroboration. Do not resolve the conflict through automatic rules such as contemporary beats later or private beats public. Explain why a particular source is better placed for the specific issue. Sometimes the conflict remains unresolved, and the conclusion should reflect that.

A practical comparison grid has four short fields: what, who, when and how measured. Fill them before writing “contradicts.” This simple step prevents unnecessary reversals and helps students distinguish new complexity from new disproof. Revision becomes more accurate when the relationship between sources is understood first.

CHAPTER 7 OF 19 · Test its consequences

7. Reweight causes through mechanisms rather than source counts

Back to contents

A new source may reveal a previously overlooked cause or change the importance of one already considered. The task is to explain its causal job. Did it create pressure, enable action, influence timing or shape implementation? Counting how many documents mention a factor is not the same as judging how much it contributed to the outcome.

Fictional teaching example: several public speeches stress popular demand for a clinic, while one funding record shows construction depended on a donation. The donation may be crucial to explaining when the project became possible. Popular demand may still explain why the project was pursued. The revised account should distinguish these roles rather than award importance by document count.

A single document can be highly significant if it provides strong evidence about a decisive mechanism. It can also be narrow or misleading. Evaluate its content and context. Do not assume that rarity proves importance or that repetition proves causal weight. The explanation must connect conditions to decisions and outcomes.

Consider alternatives fairly. Would the project plausibly have proceeded without the donation, given documented borrowing options or other funds? Such a test needs known constraints rather than invented possibilities. You may conclude that funding was an enabling condition while pressure shaped priorities. That is a more useful revision than simply replacing one main cause with another label.

In writing, explain what the new evidence changes in the causal account. “The funding record strengthens the explanation of timing but does not eliminate evidence of public pressure” is clear and proportionate. It tells the reader how the account developed and why several factors may remain necessary to understand the event.

CHAPTER 8 OF 19 · Evaluate alternatives

8. Revise confidence without inventing a numerical probability

Back to contents

Evidence can strengthen or weaken confidence without completely changing the conclusion. A second independent record may support a date. A provenance problem may make a motive less secure. Students should express the resulting confidence through reasons and appropriate wording, not invented percentages. Saying an interpretation is 90% certain adds false precision without a suitable basis.

Fictional teaching example: an initial account of a strike date relies on one later recollection. A contemporary notice and an independent diary then agree on the date. Confidence in the chronology can increase because the evidence is more direct and independently convergent. That does not automatically increase confidence in every claim about the strike’s motives or significance.

Keep confidence claim-specific. A document may securely establish that an order was issued while leaving its implementation uncertain. A source pack can establish that a protest occurred while leaving participation numbers disputed. Do not assign one overall reliability score to an entire event and apply it to every question.

Useful wording includes well supported, plausible, uncertain and not established, accompanied by the reason. Avoid using cautious phrases mechanically. If the evidence firmly establishes a limited fact, state it clearly. The aim is neither maximal doubt nor maximal confidence, but a conclusion whose strength matches its support.

A short revision note can say that the date is now more secure because independent contemporary records agree, while the motive remains uncertain. This is a meaningful update. It shows that learning can improve the resolution of an account unevenly, with some parts clarified and others still open.

CHAPTER 9 OF 19 · Evaluate alternatives

9. Missing information should create a boundary, not an invented story

Back to contents

A gap can limit an interpretation, but absence does not automatically prove a competing explanation. A missing register does not establish that no service operated. An absent complaint does not establish universal satisfaction. When new reading reveals a gap, revise the claim to reflect what remains unestablished rather than fill the space with a convenient narrative.

Fictional teaching example: an essay argues that every district received relief because three distribution lists survive. The discovery that lists for two other districts are missing limits the claim’s coverage. It does not prove that those districts received nothing. The revised answer should distinguish documented distribution from unresolved coverage.

Ask whether the missing information would normally be recorded and whether the relevant record survives completely. Silence in a complete payment ledger carries a different implication from silence in a celebratory brochure. The source’s function shapes the expectation. Explain that relationship before treating an omission as evidence against a claim.

Do not let one gap dissolve everything established. If three districts have sound records of distribution, those findings remain useful. The revision concerns the extension to every district. A well-bounded conclusion can be more informative than a general statement that the archive is incomplete and therefore nothing can be known.

A useful sentence is “The evidence establishes X in these cases, but the missing material prevents a complete judgement about Y.” This names the consequence of the gap. It also suggests the next evidence needed, making uncertainty a focused research issue rather than an invitation to speculation.

CHAPTER 10 OF 19 · Evaluate alternatives

10. Test a rival interpretation with the same standards

Back to contents

A new interpretation may explain evidence that the original account left unresolved. Consider it fairly, but do not judge it by a lower standard simply because it is new or critical. Compare factual accuracy, source use, mechanism, scope and treatment of contrary evidence. The better explanation is the one better supported for the enquiry, not the one with the most striking wording.

Fictional teaching example: an initial account explains low class attendance through fees. A rival points to inconvenient hours. Compare payment records, timetables and participants’ statements where available. Both factors may contribute, or one may fit the pattern better. The task is to explain why, not merely announce that multiple interpretations exist.

Ask what each account explains and what it leaves difficult to explain. If the fee account fits one group but not another, its scope may need narrowing. If the timetable account relies on a date error, correct that before weighing it. A fair comparison can retain useful parts of both without forcing an equal compromise.

Avoid treating an alternative as proven because the first explanation has a weakness. Rejecting one account does not automatically establish another. The rival still needs positive support. This distinction matters when students move from “not fully explained by fees” to “therefore entirely caused by political hostility” without evidence for the new claim.

A practical revision paragraph states the original explanation, the evidence challenging it and why the revised account fits better. Include an unresolved point if consequential. The reader should be able to follow the change as reasoning rather than a change of allegiance between two authoritative voices.

CHAPTER 11 OF 19 · Evaluate alternatives

11. Correct a quotation before building a larger reinterpretation

Back to contents

Sometimes the most important revision begins with a small reading error. A missing qualifier, incorrect pronoun or truncated quotation can change the source’s meaning. Before constructing a new historical argument, check whether the extract has been understood accurately. Interpretive sophistication cannot compensate for a claim the source did not make.

Fictional teaching example: a report says “some workers initially welcomed the arrangement.” An essay quotes “workers welcomed the arrangement” and concludes that support was universal and lasting. Restoring some and initially corrects both scope and time. The first repair is accurate comprehension, not a new theory about why the report changed its mind.

Check attribution as well. A source may report what opponents alleged rather than endorse the allegation. A later historian may quote an official before challenging that official’s claim. If those voices collapse, the answer can assign a statement to the wrong person. Mark quotation boundaries and explanatory context before revising the larger interpretation.

Paraphrases need the same care. May become will, proposed become implemented or complaints become proof of failure. Compare the paraphrase against the original and identify what changed. If the revision restores an important distinction, explain it in your notes so that the correction transfers to later sources.

A useful final check is to read the sentence before and after the quotation. Does your explanation preserve the source’s actual claim? If not, fix the foundation first. Evidence-based revision often becomes much easier once the words being evaluated are accurate and their limits remain visible.

CHAPTER 12 OF 19 · Rewrite with evidence

12. Recognise when a category error changes the answer

Back to contents

A category error occurs when evidence about one thing is treated as evidence about another without a justified bridge. Enrolment becomes learning, spending becomes success, attendance becomes approval or a public promise becomes an outcome. New evidence may expose the error, but the repair requires clarifying the categories rather than merely adding another source.

Fictional teaching example: an essay treats increased clinic visits as proof that the town became healthier. A new report shows repeat visits by the same patients. The correction should identify that visits measure service activity, not unique access or health outcomes. The original figure can remain accurate while the conclusion drawn from it changes.

Ask what each source actually measures or records. Then compare it with the claim’s criterion. If the two differ, identify the additional evidence needed. A bridge may be possible through several sources, but it must be argued. Do not let a familiar word such as improvement hide different dimensions of the enquiry.

This kind of revision can produce a more useful structure. One paragraph may assess activity, another access and another outcomes, if the task and evidence justify those dimensions. The change is not merely stylistic organisation. It prevents one measure from standing in for every aspect of historical change.

For practice, write the source measure beside the paragraph’s conclusion. If they are different, explain the reasoning connection aloud. If you cannot, narrow the claim or find appropriate evidence. The aim is not to distrust all indicators, but to use each for the conclusion it can support.

CHAPTER 13 OF 19 · Rewrite with evidence

13. Keep a brief change log for substantial revisions

Back to contents

A change log records the original claim, the new evidence or correction and the revised conclusion. It makes learning visible and prevents repeated errors. It need not become an administrative burden. A few lines are enough when a revision changes scope, causal weight or an important factual premise.

Fictional teaching example: the original claim says all districts benefited. The new evidence is a dated petition from an excluded district. The revised claim recognises benefits in documented central districts while rejecting universality. The reason is coverage, not a general loss of trust in every source. The log preserves that distinction.

Include what did not change. Perhaps the opening date remains secure, the stated aim remains documented and use still increased. This prevents a local correction from becoming a total reversal without justification. It also helps students retain valid work instead of feeling that an entire answer must be discarded.

A useful log can support feedback discussions. Show the teacher the change and ask whether the new claim matches the evidence more closely. That is more focused than asking whether the whole essay is now correct. It invites a conversation about the operation that improved and any remaining limit.

Review a few entries before the next task. Look for patterns such as overgeneralisation, chronology errors or unsupported motives. Choose a short exercise targeting the recurring issue. The log’s purpose is transfer: a correction becomes valuable when it changes how you read and reason next time, not merely when it improves one submitted paragraph.

CHAPTER 14 OF 19 · Rewrite with evidence

14. A complete fictional revision workshop

Back to contents

Use this invented pack, not historical evidence. A town opens evening buses. Source A is a launch speech promising access for everyone. Source B records 900 passenger journeys during the first month. Based on those sources, a student writes that the trial successfully served all residents. Source C is then introduced: a letter from an outer district whose last service remained too early for shift workers.

First, inspect the original support. A establishes a stated ambition, not achieved universal access. B records journeys, not necessarily different passengers or coverage of every district. The original conclusion was already broader than its evidence. C adds a documented barrier, but the revision should acknowledge the earlier inference problem as well as the new information.

Second, evaluate C. Check its date, the group represented and the specific claim. If it concerns the same trial month and timetable, it directly qualifies universality. It may not establish how many residents were excluded or whether central users benefited. Preserve its scope rather than making one letter speak for the whole town.

Third, write a revised conclusion: the trial recorded use and may have improved access for some groups, but the available evidence does not establish town-wide provision, and C documents a specific scheduling limit. The judgement recognises activity without equating it with universal success. It does not reverse into a claim that nobody benefited.

Finally, add a fourth fictional source: a later timetable showing expanded outer-district services. This may support improvement in a later phase, but it does not erase the first-month barrier. The revised account now needs a chronology of change. The exercise demonstrates how evidence can refine an interpretation in several steps without making each new source the final unquestionable truth.

CHAPTER 15 OF 19 · Rewrite with evidence

15. Rewrite the paragraph, not only the concluding adjective

Back to contents

A revision that changes the interpretation should appear in the paragraph’s reasoning, not just its last word. If the conclusion shifts from complete success to uneven improvement, the evidence and explanation should now distinguish groups, periods or criteria. Otherwise, the body may continue defending a claim the conclusion has nominally abandoned.

Fictional teaching repair: an original paragraph quotes a launch promise, cites high attendance and announces universal benefit. Adding “partly” at the end does not fix the link between attendance and access. Rewrite the claim, explain the measure and incorporate the documented barrier. The revision must reach the inference, not merely soften the verdict.

Preserve accurate material where useful. A quotation may still establish the stated aim; the attendance figure may still establish recorded activity. Move them into the roles they can support. This is often more efficient than deleting everything and starting again. Historical revision is selective repair guided by the enquiry.

Check transitions after the change. A paragraph that once said therefore may now need however or a more specific causal connection. Pronouns and time phrases may also need adjustment. These language changes matter because they carry the revised reasoning. Editing should make the new relationship clear rather than leave traces of the old argument.

Read the final paragraph without consulting the change log. Can a reader understand the claim, evidence, explanation and limit on their own? The submitted answer should stand alone. The revision history is useful for learning, but the finished prose needs a coherent account rather than a record of every stage of the student’s thinking.

CHAPTER 16 OF 19 · Transfer the correction

16. Use feedback as evidence about your reasoning

Back to contents

Teacher feedback can identify a missing operation: inaccurate comprehension, weak explanation, irrelevant evidence or an unsupported judgement. Treat the comment as a prompt to inspect the relationship between source and claim. Do not respond only by adding length or more historical facts. More material will not necessarily repair the identified problem.

Fictional teaching example: feedback says a paragraph is descriptive. The student adds another quotation, making the description longer. A better response explains what the existing detail suggests and why it answers the question. If feedback identifies a scope problem, the repair may be a narrower claim rather than a larger evidence list.

Ask a focused follow-up when needed. “Does this source establish use rather than approval?” or “Which comparison would justify calling this the main cause?” makes the uncertainty clear. A specific question is easier to answer productively than asking for a model paragraph to memorise. The aim is understanding the missing reasoning step.

Apply the correction to a different example after repairing the original. This checks transfer. A student who learns visits are not unique users should recognise the same issue in a library register or transport count. The factual setting changes, but the analytical distinction remains useful. That is the lasting value of revision.

Parents can support this process by asking what changed and why the new version is better supported. Avoid treating correction as a personal failure or demanding that every comment produce an entirely rewritten essay. One well-understood repair can be a meaningful advance when it becomes a habit the student can use independently.

CHAPTER 17 OF 19 · Transfer the correction

17. An honest stopping point for a revisable conclusion

Back to contents

An answer can be well supported without being permanently final. Stop when the task has been answered proportionately using the available evidence, with consequential uncertainty stated clearly. Endless revision is not the goal. The goal is a coherent account that remains open to better evidence without pretending that every imaginable possibility deserves equal space.

Fictional teaching example: several independent records establish a decision date, while motives remain partly uncertain. You can state the date confidently and give a qualified explanation of motives. There is no need to weaken the secure finding simply because another part of the enquiry remains open. Confidence should stay attached to particular claims.

Before submitting, compare the conclusion with the body. Has a tentative inference become a certainty without additional support? Has an important qualification disappeared? Have you become more doubtful than the evidence warrants because a contradiction felt unsettling? Check both overconfidence and unnecessary retreat. The strongest answer matches its level of assurance to its reasons.

In an assessment, use time sensibly and follow instructions. A quick reread can catch a missing qualifier or an overstated conclusion. You cannot conduct unlimited research during a timed task, and you should not rewrite the question to avoid its demands. Use the evidence and knowledge available as carefully as the task allows.

The lasting principle is simple: preserve what remains sound, change what the evidence changes and explain the reason. Historical revision is not a weakness in knowledge. It is one of the ways knowledge becomes more accurate. A student who can revise a claim precisely is learning to reason with evidence rather than defend an answer simply because it was written first.

CHAPTER 18 OF 19 · Transfer the correction

18. Distinguish discovering a new source from asking a new question

Back to contents

An interpretation can change because previously unknown material appears. It can also change because a different enquiry makes existing material relevant in a new way. These are related but distinct developments. A source that seemed marginal to an administrative history may become central to a question about household experience. The past has not changed; the enquiry and use of evidence have.

Fictional teaching example: a relocation file has always contained letters about journeys to work. An early study focuses on construction costs, while a later enquiry examines everyday access. The letters now carry more explanatory weight. This is not proof that the earlier researcher concealed them or that the later account is automatically complete. Evaluate each account against its stated question and coverage.

A useful revision note identifies the source of change: new material, corrected reading, new comparison or a different research question. This prevents an exaggerated story in which every new interpretation supposedly overturns everything before it. Some accounts answer different questions and can coexist; others genuinely disagree on the same claim and need adjudication.

When comparing interpretations, establish whether their enquiries align before deciding that one refutes the other. An account of lower construction costs and an account of residents’ longer journeys may both be supported. Revision can integrate dimensions without erasing either. The student’s task is to make the relationship visible and judge what the actual question requires.

CHAPTER 19 OF 19 · Transfer the correction

19. Check the timeline after a substantial revision

Back to contents

A change in interpretation can affect several parts of an answer’s chronology. If a decision occurred earlier than first thought, some proposed influences may no longer precede it. If a report was written later, its author may have access to outcomes unavailable at the time. Recheck these connections rather than correcting one date in isolation.

Fictional teaching example: an essay says public complaints caused a council decision. A corrected chronology shows the decision was approved before those complaints were received. The complaints may still explain later implementation changes or public pressure, but they cannot play the same role in the initial decision without further evidence. The revision changes the causal pathway.

Update the plan as well as the paragraph. A heading such as reasons for adoption may need different evidence once the sequence is corrected. Check the introduction and conclusion for claims inherited from the old chronology. Small factual repairs can have large interpretive consequences, and those consequences should be followed through consistently.

Finally, state what the correction does not establish. Removing one proposed cause does not prove another automatically. Return to the remaining evidence and rebuild the explanation at the strength it supports. This is a disciplined revision: the timeline constrains the argument, and the argument changes rather than forcing the dates to fit the original story.