Accuracy and Ethics
Explanation carries a special responsibility: people use it to make decisions. A clear explanation of a false claim is more dangerous than a muddled one, because it is more convincing. This page sets out the ethical standards that should sit behind every expository speech.
Accuracy
| Standard | What it means | Practice |
|---|---|---|
| Verification | Claims are checked against reliable sources. | Keep a source log; check the original, not just the summary. |
| Precision | Numbers and terms mean what you say they mean. | Define measures; avoid rounding that changes the point. |
| Context | Evidence is not stripped of its conditions. | Say where, when and for whom the finding applies. |
| Currency | Information is up to date. | Check the date; note when a field is changing. |
| Correction | Errors are acknowledged and fixed. | Keep a record; correct publicly if the error was public. |
Fact, inference and opinion
Audiences need to know what kind of claim they are hearing. A fact can be checked. An inference is a conclusion drawn from evidence. An opinion is a judgement or preference. Mixing them without signalling is a form of misleading.
| Claim | Type | How to signal it |
|---|---|---|
| “The river rose 1.2 metres.” | Fact | Give the source and date. |
| “The levee is likely to fail.” | Inference | “The evidence suggests ...” |
| “The levee should be rebuilt.” | Opinion or recommendation | “In my view ...” or “The case for ...” |
| “Everyone knows the levee is unsafe.” | Unsupported generalisation | Replace with evidence or cut. |
Uncertainty
Science and public knowledge are full of uncertainty. Pretending otherwise is not clarity; it is distortion. Say what is known, what is likely and what is unknown.
- “The evidence is strong that ...”
- “It is likely that ..., although the exact figure is debated.”
- “We do not yet know ...; here is how researchers are trying to find out.”
- “This finding applies to ..., not necessarily to ...”
Uncertainty does not make a speech weak. It makes it honest, and it gives the audience a more accurate picture.
Attribution
Credit the people who produced the knowledge. Attribution is fair to them and useful to the audience, who may want to check the source.
| Material | How to attribute |
|---|---|
| Statistic | Name the organisation and date: “The Australian Bureau of Statistics reported in 2024 that ...” |
| Quotation | Name the speaker and context: “As the author wrote in ...” |
| Image or graph | Credit the creator and source on the slide. |
| Idea | Name the thinker or tradition if it is distinctive. |
| Personal story | Use only with consent; say whose story it is. |
Consent and privacy
Explanations often use examples from real life. Those examples can expose people. Ask before using someone's story, especially if it involves health, family, legal trouble, immigration status, sexuality, gender identity or trauma. Remove identifying details unless the person has agreed to be identified.
- Do not read private messages or documents aloud.
- Do not use a photograph of someone without permission.
- Do not present a person as a representative of a whole group.
- Respect a request to withdraw a story.
- Be careful with data that could re-identify a small group.
Honest visuals
A graph can be technically accurate and still mislead. Check the axis, the baseline, the scale and the comparison. Do not use a visual to imply a causal relationship the data cannot support. If you crop an image, say what was cropped and why.
Contested topics
Some explanations touch on matters people disagree about: climate policy, vaccination, history, economics, religion. Explaining a contested topic does not require false balance, but it does require fairness.
| Practice | Why |
|---|---|
| Represent the strongest version of each serious position. | Weak versions are easy to knock down and unfair to the audience. |
| Distinguish scientific consensus from political debate. | They are different kinds of disagreement. |
| Name the evidence and its limits. | The audience can judge for itself. |
| Do not manufacture doubt where evidence is strong. | False balance misleads. |
| Do not present a fringe view as equivalent to a well-evidenced one. | Equivalence is a claim about evidence. |
Correcting errors
If you discover an error after a speech, correct it. If the speech was recorded or published, correct it in the same place. A short, direct correction protects trust more than silence.
“I was wrong about that” is a complete sentence. It does not require a long defence. Say what was wrong, what is right and what changed.
Activity: ethical audit
- Take a draft explanation.
- Highlight every factual claim and write its source.
- Mark each claim as fact, inference or opinion.
- Find one place where uncertainty should be acknowledged.
- Check every visual for a misleading scale or missing label.
- Check any personal story for consent and identifying detail.
- Revise one passage to make it more accurate or fair.
Common errors
- Confusing confidence with accuracy. A polished delivery can still be wrong.
- Hiding uncertainty. It usually backfires when the audience learns more.
- Using a statistic without a source. The audience cannot check it.
- Treating correlation as causation. It is one of the most common explanations errors.
- Using a real story without consent. Privacy matters more than a vivid example.
- False balance. Not every view has equal evidence.
Next, study delivering an explanation to carry these standards into performance. The building an expository speech page brings the unit together.