Validity and reliability
A claim is a statement that someone may or may not accept. Evidence is the facts, numbers, examples or expert views that give a reason to accept it. A strong claim needs strong evidence. Two words tell us how strong.
Validity: is it about THIS claim?
Evidence is valid when it really measures or shows what the claim says. A cricket score does not tell us anything about school start times, so it is not valid for that claim. Valid evidence is also relevant and enough: it should point straight at the claim.
Reliability: can we trust it?
Evidence is reliable when the source is trustworthy and the result would be the same if someone checked it again. A large survey done carefully, or a study repeated by other experts, is reliable. One person's opinion or an advert is weak, because it may be one-sided or hard to repeat.
| Valid? | Reliable? | |
|---|---|---|
| Doctors' sleep study | Yes | Yes |
| Friend's opinion | Yes (on topic) | Weak |
| Cricket score | No (off topic) | May be accurate, but useless here |
| Pillow advert | Partly | No (it wants to sell) |
Evidence must pass both tests. A trusted source can still be off topic, and an on-topic piece can still be weak.
Choosing evidence
Different kinds of evidence are useful. Numbers (statistics) show how big something is. Expert views give knowledge. Examples make an idea real. Experiments test cause and effect. Personal stories show feelings but are weak proof alone.
A quick checklist
- Relevance: does it speak to my exact claim?
- Source: who made it? Do they know the topic?
- Bias: do they gain if I believe it (like an advertiser)?
- Sample: how many people or cases? Do they fairly stand for everyone? 2,000 students from many schools beats 5 friends.
- Date: is it recent enough for this topic?
- Agreement: do other good sources say the same?
Using evidence in writing
Follow this pattern: claim, evidence, explanation. State the claim, give the evidence with its source, then say in one sentence why it supports the claim. Use at least two different kinds of evidence. If good evidence is against you, say so and answer it fairly.
Quick percent check
Numbers need care: "150 of 200 students" is 150 ÷ 200 × 100 = 75%. Say how many were asked, not only the percent.
Try it yourself
Pick a claim such as "Our class should have a plant corner". Write five pieces of evidence, some strong and some weak or off topic. Tick each with V (valid) and R (reliable). Keep only those with both ticks. In the 3D, switch on the cricket score and the pillow advert and see how little they add.
Key formulas and definitions
- Valid = on topic and really shows the claim
- Reliable = trusted source, same result if checked again
- Good evidence = Valid AND Reliable
- Pattern: Claim > Evidence (with source) > Explanation
- Percent = (part ÷ whole) × 100
Worked examples
1. Claim: "Reading daily improves marks." Which is better evidence: (a) "My cousin reads and tops his class" or (b) a study of 3,000 students over 2 years? Why?
(b). Both are on topic, but (b) is more reliable: large sample, long time, and it can be checked again. (a) is one case, so it is weak.
2. Claim: "Eating more vegetables keeps people healthy." Rate: (1) a vegetable seller's poster, (2) a hospital study on 10,000 adults, (3) the price of tomatoes today.
(1) On topic but biased: the seller gains, so weak. (2) Valid and reliable: strong. (3) Not valid: the price says nothing about health. Use (2) and perhaps a second expert source.
3. A survey asks 80 students, and 60 say they like the new timetable. What percent like it? Is it enough proof that "most students in the school like it"?
60 ÷ 80 × 100 = 75%. For a school of 1,200, 80 students may be too few, and if all 80 were from one class the sample is not fair. So it is a start, but not strong. Ask students from every class.
4. Write one claim-evidence-explanation line for "Cycling to school is good for the city".
Claim: Cycling to school is good for the city. Evidence: A city traffic report shows that when 20% of pupils cycle, car trips near schools fall by 15%. Explanation: Fewer cars mean less jam and less smoke.
Common mistakes
- Using evidence that is true but off topic (a cricket score for a school-time claim). That is not valid.
- Trusting a source that gains from your belief, such as an advert, without a second check.
- Treating one personal story as proof for everyone.
- Giving a percent without saying how many were asked, which hides a tiny sample.