How to Answer Source-Based Questions | Interpret Evidence, Evaluate Provenance and Build Supported Judgements

How to answer source-based questions is not to label every source ‘biased’ and move on. Strong responses interpret what the source says, identify what it is useful for, consider who produced it and why, compare it with contextual knowledge or other sources, and build a judgement tied to the exact question.

Source evaluation becomes weak when students repeat generic phrases about reliability without examining content, provenance, audience, purpose and historical context.

This guide explains how to move from source comprehension to inference, comparison, usefulness, reliability and supported judgement without relying on formulaic evaluation.

This guide is designed for students, parents and educators who want a practical answer to how to answer source based questions effectively. The objective is not to add another slogan to revision. It is to turn the assessment demand into a visible sequence that can be practised, checked, corrected and repeated until the learner can execute it independently.

The 60-Second Answer

Read the question before the source. Identify what claim or issue is being tested. Read the content closely. Note creator, date, audience, purpose and context. Use provenance to explain what the source can or cannot show—not to dismiss it automatically. Cross-check against contextual knowledge or other sources. Answer the exact source task with evidence from the source and a clear judgement.

Working definition: A source-based question asks the learner to interpret, compare or evaluate evidence using the source’s content, provenance, context and relationship to other knowledge.

Why Source-Based Questions Matters

Sources are evidence with origins and purposes, not transparent windows into reality.

A source can be partial, persuasive or limited and still be useful for understanding attitudes, priorities or context.

The strongest answers therefore evaluate what a source can support rather than applying automatic labels such as biased or reliable.

A useful exam strategy changes what the learner can actually do under assessment conditions. It should improve interpretation, decision quality, pacing, checking and recovery rather than simply make revision feel more organised.

The Hidden Problem: Calling a Source Biased Does Not Explain Its Evidential Value

Students often treat bias as a reason to reject a source.

But a partisan speech may be extremely useful evidence for what the speaker wanted the audience to believe.

The better question is: useful for what claim, limited in what way, and how does provenance affect interpretation?

The Operating Model

Decode Task → Read Content → Infer → Examine Provenance → Contextualize → Cross-Check → Judge → Support.

Each stage has a separate job. Preparation reduces uncertainty. Interpretation clarifies what is being asked. Execution converts knowledge into marks. Monitoring protects time. Checking catches high-risk mistakes. Delayed retesting verifies that the improvement survives beyond one practice session.

How to answer source-based questions: Step by Step

1. Read the question first

Identify whether the task asks inference, comparison, usefulness, reliability, purpose or evaluation.

The same source supports different answers depending on the task.

2. Read source content closely

Identify explicit claims, tone, emphasis, omissions and details.

Use quotations or paraphrase selectively.

3. Make an inference

Explain what the content suggests beyond simple copying.

Tie inference to evidence in the source.

4. Examine provenance

Consider creator, date, audience, purpose, medium and situation.

Use provenance to interpret value and limits.

5. Add contextual knowledge

Use accurate outside knowledge to confirm, challenge or deepen the source where the question permits.

Do not dump unrelated facts.

6. Cross-check

Compare the source with another source or established knowledge.

Look for agreement, contradiction or complementary evidence.

7. Build a judgement

State how far the source supports the claim or how useful/reliable it is for the specific purpose.

Avoid all-or-nothing conclusions.

8. Return to the wording

Make sure the final answer addresses exactly what the question asks about this source.

Do not drift into generic history writing.

What Strong Performance Looks Like

The source task is identified.
Content evidence is used.
Inference goes beyond copying.
Provenance has a purpose.
Contextual knowledge is relevant.
Cross-checking is specific.
Judgements are qualified.
Bias is not used as an automatic dismissal.

These indicators are behavioural rather than personal. That matters because they can be demonstrated, practised and retested. The student does not need to become a different person; the student needs a more reliable sequence of decisions.

Common Failure Modes

1. Bias label

The source is called biased without explaining evidential consequences.

Ask what it is useful for despite or because of its perspective.

2. Content paraphrase only

The response restates the source and stops.

Add inference or evaluation.

3. Provenance checklist

Author, date and audience are listed mechanically.

Explain how each matters to the claim.

4. Context dump

Large amounts of outside knowledge overwhelm source evidence.

Use only knowledge that tests or clarifies the source.

5. Reliability equals truth

One accurate fact makes the whole source reliable.

Evaluate claim by claim and purpose by purpose.

6. Unreliable equals useless

A propagandistic source is dismissed completely.

Consider usefulness for attitudes, intentions or messaging.

7. Comparison without criteria

Two sources are called similar or different without specifying what aspect.

Compare content, purpose, tone or context explicitly.

8. No judgement

The answer contains analysis but never concludes.

State the degree of support or usefulness.

How the Strategy Changes Across Subjects

History

Use provenance, context, cross-checking and purpose to evaluate evidence rather than relying on generic bias language.

The disciplinary standard still controls quality. General exam technique can protect performance, but it cannot replace knowing what counts as relevant evidence, valid reasoning, precise terminology or complete working in this subject.

Social Studies

Connect source claims to issues, stakeholders, purpose and contextual evidence while maintaining relevance to the question.

English and media literacy

Analyse author, audience, purpose, tone, evidence and reliability of claims, especially in persuasive or informational sources.

General research skills

Treat every source as evidence produced in a context and ask what claims it can responsibly support.

Primary School, Secondary School and Examination Years

Primary school

Use simple questions: Who made this? What does it say? Why might they have made it? What can we learn from it? Avoid teaching children that biased automatically means useless.

Keep the routine short, concrete and easy to rehearse. Adults can model it first, then reduce prompts as the learner becomes able to perform the sequence independently.

Secondary school

Students should practise different source question types and learn to use provenance analytically rather than as a memorized checklist.

Secondary learners should increasingly own the preparation, timing and checking decisions themselves because the assessment environment becomes more varied and time pressure more consequential.

Before major examinations

Use authentic source-based papers under time and mark whether content, provenance, contextual knowledge and judgement are all doing useful work.

Near major exams, use the closest available simulation of the real format. Students should not be discovering basic pacing, question-reading or response-format decisions on the actual paper.

A Focused 60-Minute Practice Session

10 minutes — classify five source questions by task type.
10 minutes — annotate one source for content, tone and inference.
15 minutes — analyse provenance and purpose.
10 minutes — cross-check with contextual knowledge or a second source.
10 minutes — write a qualified judgement.
5 minutes — audit for generic bias language.

The exact timings are adjustable. The stable sequence is more important: independent attempt, feedback, targeted repair and later retest. Without those elements, practice can feel productive while leaving performance unchanged.

Diagnostic Checklist

Can the learner identify the source task?
Can they infer from content?
Can provenance be linked to value or limitation?
Can contextual knowledge be used selectively?
Can sources be cross-checked?
Can bias be discussed without automatic dismissal?
Is the final judgement qualified?
Does the answer remain tied to the question?

Use the checklist to choose the next practice target. If several items are weak, repair the earliest breakdown in the chain or the one that creates the largest downstream cost.

Practice Laboratory

Practice 1: Inference pair

Write one explicit statement from a source and one supported inference.

Explain the evidence link.

Practice 2: Provenance-to-value

Choose creator, audience or purpose and explain how it affects what the source can show.

Avoid listing facts without analysis.

Practice 3: Bias repair

Take a weak sentence saying ‘this source is biased’ and rewrite it into a claim about usefulness or limitation.

Tie it to the question.

Practice 4: Cross-check grid

Compare source A, source B and contextual knowledge on one claim.

Record agreement and contradiction.

Practice 5: Purpose question

Identify what the creator wanted the audience to think, feel or do.

Use source evidence.

Practice 6: Usefulness boundary

State one thing the source is useful for and one thing it cannot establish confidently.

Explain why.

Practice 7: Same-source new task

Answer an inference question and a reliability question on the same source.

Notice how analysis changes.

Practice 8: Judgement ladder

Write weak, medium and strong qualified conclusions.

Choose the one best supported by evidence.

Three Student Cases

Case 1: Maren says every source is biased

Maren has learned a checklist that ends with ‘therefore unreliable’.

Her tutor asks ‘unreliable for what claim?’ and ‘useful for what purpose?’

Her evaluations become specific instead of automatic.

Case 2: Iona ignores provenance

Iona quotes source content well but treats every document as context-free.

She adds one provenance point only when it changes interpretation.

Her answers become deeper without becoming formulaic.

Case 3: Leonie dumps contextual knowledge

Leonie knows the topic well and writes long background paragraphs that barely use the source.

Her tutor requires every outside fact to test or clarify one source claim.

Relevance improves and answers shorten.

How Parents Can Help Without Taking Over

Parents can support source-based questions by asking the learner to explain the next decision instead of supplying the answer. Useful prompts include: “What is the task?”, “What will you look for first?”, “What would make this answer incomplete?”, “How will you check?”, “What will you do if you miss something?”, and “When should you move on?”

When the learner can perform a step reliably, remove the prompt. The long-term goal is transferred control before the real assessment.

How Tutors Can Use a Three-Student Small Group

A three-student tutorial can make source-based questions visible because one learner can attempt, another can question the decision, and the third can check against a rubric, source or model. Roles can rotate quickly without anyone becoming a permanent observer.

After discussion, every learner should complete a fresh individual task. Shared insight becomes educationally useful only when it transfers into solo performance.

How to Measure Progress

Generic bias language declines.
Inferences become better supported.
Provenance analysis becomes purposeful.
Contextual knowledge becomes more selective.
Cross-checking improves.
Judgements become qualified.
Source task types are recognized faster.
Timed answers become more focused.

Marks remain important, but process improvements often appear first: faster starts, fewer missed instructions, more precise evidence use, fewer repeated errors, better pacing and more effective checking. Track those behaviours and confirm that performance eventually follows.

A Four-Week Implementation Plan

Week 1 — Separate task types

Practise inference, purpose, comparison and usefulness separately.

Learn what each question actually asks.

Week 2 — Improve provenance reasoning

Use creator, audience, purpose and context only when they affect interpretation.

Avoid checklist writing.

Week 3 — Add cross-checking

Use contextual knowledge and multiple sources to test claims.

Practise qualified judgements.

Week 4 — Simulate

Complete timed source-based sections and mark recurring weaknesses.

Retest the same question type on fresh sources.

Evidence and Responsible Use

The Historical Association has long argued against simplistic source evaluation that reduces reliability to a generic bias label, encouraging students to consider what a source is reliable or useful for, its context and how it compares with other evidence. Its teaching materials emphasize provenance, source content and contextual knowledge together rather than treating one feature as decisive.

No exam strategy can manufacture knowledge that was never learned. These routines help students express and protect what they know, but they still depend on subject teaching, retrieval, feedback and authentic practice.

Frequently Asked Questions

Does bias make a source useless?

No. Perspective can limit some claims while making the source valuable evidence for attitudes, intentions or messaging.

What is provenance?

Information about who produced the source, when, for whom, why and in what context.

Should I always use contextual knowledge?

Use it when the question and assessment expect it, and only when it tests or clarifies the source.

How do I compare sources?

Use a clear criterion such as claim, tone, purpose, evidence or context.

What is a good reliability judgement?

One that is specific to the claim and explains both supporting and limiting evidence.

Should I quote the source?

Use brief, relevant evidence rather than copying long passages.

Can an inaccurate source still be useful?

Yes, for example as evidence of beliefs, propaganda, misunderstanding or intended persuasion.

How do I avoid formulaic answers?

Let the question determine which source features matter rather than forcing every source through the same checklist.

Helpful Reading on eduKateSingapore

A Source Is Evidence for a Claim, Not a Label to Approve or Reject

Strong source work asks what the evidence shows, why it was produced, what context changes its meaning and how far it supports the specific judgement.

Read the task. Use the source. Infer. Examine provenance. Cross-check. Qualify. Judge. Then return to the exact question.

Properly taught kids shine a bright light into the future.

Extended Diagnostic Workshops

Workshop 1: Read the question first × Content paraphrase only

Create one fresh exam-style task in which the learner practises read the question first while watching specifically for content paraphrase only. Identify whether the task asks inference, comparison, usefulness, reliability, purpose or evaluation. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

The response restates the source and stops. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Take a weak sentence saying ‘this source is biased’ and rewrite it into a claim about usefulness or limitation. Tie it to the question. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 2: Read source content closely × Context dump

Create one fresh exam-style task in which the learner practises read source content closely while watching specifically for context dump. Identify explicit claims, tone, emphasis, omissions and details. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

Large amounts of outside knowledge overwhelm source evidence. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: State one thing the source is useful for and one thing it cannot establish confidently. Explain why. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 3: Make an inference × Unreliable equals useless

Create one fresh exam-style task in which the learner practises make an inference while watching specifically for unreliable equals useless. Explain what the content suggests beyond simple copying. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

A propagandistic source is dismissed completely. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Write one explicit statement from a source and one supported inference. Explain the evidence link. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 4: Examine provenance × No judgement

Create one fresh exam-style task in which the learner practises examine provenance while watching specifically for no judgement. Consider creator, date, audience, purpose, medium and situation. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

The answer contains analysis but never concludes. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Compare source A, source B and contextual knowledge on one claim. Record agreement and contradiction. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 5: Add contextual knowledge × Content paraphrase only

Create one fresh exam-style task in which the learner practises add contextual knowledge while watching specifically for content paraphrase only. Use accurate outside knowledge to confirm, challenge or deepen the source where the question permits. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

The response restates the source and stops. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Answer an inference question and a reliability question on the same source. Notice how analysis changes. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 6: Cross-check × Context dump

Create one fresh exam-style task in which the learner practises cross-check while watching specifically for context dump. Compare the source with another source or established knowledge. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

Large amounts of outside knowledge overwhelm source evidence. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Choose creator, audience or purpose and explain how it affects what the source can show. Avoid listing facts without analysis. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 7: Build a judgement × Unreliable equals useless

Create one fresh exam-style task in which the learner practises build a judgement while watching specifically for unreliable equals useless. State how far the source supports the claim or how useful/reliable it is for the specific purpose. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

A propagandistic source is dismissed completely. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Identify what the creator wanted the audience to think, feel or do. Use source evidence. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 8: Return to the wording × No judgement

Create one fresh exam-style task in which the learner practises return to the wording while watching specifically for no judgement. Make sure the final answer addresses exactly what the question asks about this source. Begin under a defined condition or time limit and require the learner to commit to a response before receiving feedback.

The answer contains analysis but never concludes. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Write weak, medium and strong qualified conclusions. Choose the one best supported by evidence. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 9: Read the question first × Content paraphrase only

Workshop 10: Read source content closely × Context dump

Workshop 11: Make an inference × Unreliable equals useless

Workshop 12: Examine provenance × No judgement

Workshop 13: Add contextual knowledge × Content paraphrase only

The response restates the source and stops. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Answer an inference question and a reliability question on the same source. Notice how analysis changes. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 14: Cross-check × Context dump

Large amounts of outside knowledge overwhelm source evidence. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Choose creator, audience or purpose and explain how it affects what the source can show. Avoid listing facts without analysis. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 15: Build a judgement × Unreliable equals useless

A propagandistic source is dismissed completely. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Identify what the creator wanted the audience to think, feel or do. Use source evidence. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 16: Return to the wording × No judgement

The answer contains analysis but never concludes. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Write weak, medium and strong qualified conclusions. Choose the one best supported by evidence. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 17: Read the question first × Content paraphrase only

Workshop 18: Read source content closely × Context dump

Workshop 19: Make an inference × Unreliable equals useless

Workshop 20: Examine provenance × No judgement

Workshop 21: Add contextual knowledge × Content paraphrase only

The response restates the source and stops. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Answer an inference question and a reliability question on the same source. Notice how analysis changes. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.

Workshop 22: Cross-check × Context dump

Large amounts of outside knowledge overwhelm source evidence. After correction, change one meaningful feature—wording, data, source, sequence, timing, evidence or representation. Then use this related exercise: Choose creator, audience or purpose and explain how it affects what the source can show. Avoid listing facts without analysis. Finish with a delayed retest so the next success cannot be explained only by immediate familiarity.