Research and Source Evaluation
Year 9 HEAP • Persuasion • Research
Research is not collecting links. It is deciding which claims deserve trust, what a source can actually show and where the gaps are. A campaign becomes more persuasive when it can show its working.
Source types
| Source | Strength | Limit | Best used for |
|---|
| Government or council report | Official data and stated method. | May be dated, aggregated or written for a different purpose. | Policy, funding and population data. |
| Peer-reviewed research | Checked by specialists and method is visible. | Can be narrow, technical or hard to generalise. | Establishing whether an approach works. |
| School records | Directly relevant to the local context. | Privacy restrictions and small samples. | Local counts, surveys and observations. |
| News report | Accessible and current. | May simplify, sensationalise or rely on one source. | Finding an issue and locating primary sources. |
| Advocacy group | Deep knowledge and lived experience. | Has a position to argue and may select evidence. | Understanding the case for a change. |
| Personal testimony | Shows experience and consequence. | Not proof of a general pattern. | Making an issue concrete, with consent. |
| Social media post | Fast and immediate. | Often unverified, decontextualised or fabricated. | Finding leads, not settling facts. |
| AI-generated summary | Can point to useful language and questions. | May invent sources or facts; not evidence in itself. | Generating search terms, then checking the originals. |
Lateral reading
Do not judge a source only by how convincing its own page looks. Open another tab and find out who is behind it.
| Move | Question | What you are looking for |
|---|
| Leave the page | What do other sources say about this organisation? | Independent descriptions, not just its own "About" page. |
| Check the author | Who wrote this, and what is their expertise? | Relevant qualifications, experience and affiliations. |
| Check the date | When was this published or updated? | Whether the evidence is current for the issue. |
| Trace the claim | Where did this number or quotation come from? | The original study, report or dataset. |
| Look for disagreement | Who disputes this, and why? | The strongest counter-evidence, not just the loudest. |
Reading a statistic
| Question | Why it matters |
|---|
| What exactly is being counted? | "Waste" could mean weight, volume, cost or number of items. |
| Compared with what? | A percentage without a base is meaningless. |
| Over what period? | A single day may not represent a term. |
| Who was included? | A survey of one class cannot speak for the school. |
| Who collected it, and why? | Purpose shapes what is measured and published. |
| What does it not show? | Data can reveal a pattern without explaining its cause. |
Misinformation and misleading content
| Pattern | Warning sign | Response |
|---|
| Fabricated source | A study or expert that cannot be found anywhere else. | Search the exact title and author; if it does not exist, do not use it. |
| False context | A real image or quotation attached to the wrong event. | Trace the earliest version and check the date and place. |
| Cherry-picked data | A trend line that starts at an unusual year. | Look for the full dataset and a longer time frame. |
| Impersonation | A website that mimics a familiar organisation. | Check the domain and find the organisation's official site. |
| Emotional urgency | "Share before this is deleted." | Slow down and verify before amplifying. |
| AI hallucination | Confident detail with no traceable source. | Ask for the source, then check it exists and says what is claimed. |
Source log
| Source and link | Author or organisation | Date | What it says | What it can show | What it cannot show | How I will use it |
|---|
| | | | | | |
| | | | | | |
| | | | | | |
Research ethics
- Do not collect personal information about students or staff without permission.
- Do not photograph or quote people without informed consent.
- Anonymise survey responses and store them securely.
- Do not publish a small group's responses in a way that identifies them.
- Respect copyright: link to sources rather than copying long extracts or images.
- Report findings honestly, including results that do not support the campaign.
Common errors
- First result bias. Using the first search result without checking who produced it.
- Source laundering. Citing an article that cites a study, without reading the study.
- Quote mining. Taking a sentence out of context to support a claim.
- False balance. Treating a fringe view as equal to a well-evidenced consensus.
- Link rot. Failing to record the title, author and date, so the source cannot be found again.
With the evidence checked, move to drafting the campaign.