The one-sentence truth: In GP, evidence is not just information you can quote. It is information whose quality, relevance and limits you can explain.
Many essays contain examples but little evidence evaluation. A student mentions a study, statistic or case and assumes the paragraph is now supported. Yet GP requires a second question: how much should this evidence actually make us believe?
Use the Top 100 General Paper Vocabulary Words for JC1 as the core inventory, and keep the broader JC1–JC2 vocabulary guide as the conceptual parent.
Core evidence vocabulary
- empirical — based on observation or measurement.
- anecdotal — based mainly on individual stories or cases.
- credible — worthy of belief because the source or reasoning is trustworthy.
- reliable — consistent or dependable.
- valid — genuinely measures or supports what it claims to.
- representative — reasonably reflects the wider group.
- verifiable — capable of being checked independently.
- corroborate — strengthen a claim with additional evidence.
- bias — a systematic tendency that can distort evidence or judgement.
- causality — a relationship where one factor contributes to producing another.
Credible does not mean true
A source can be credible and still be wrong. Credibility concerns whether the source deserves serious consideration; truth concerns whether the claim is actually correct. Strong GP writing keeps those ideas separate.
Reliable does not mean valid
A measurement can be consistent and still measure the wrong thing. A survey may reliably reproduce the same result while asking a leading question. Reliability is consistency; validity is whether the measure or conclusion actually fits the concept being studied.
Representative evidence
One striking case can illustrate possibility without showing prevalence. Useful language includes representative sample, isolated example, exceptional case, broader pattern, generalisable finding.
“This case proves that the policy fails” is usually too strong. “This case illustrates one pathway through which the policy can fail” is often more defensible.
Corroboration
Evidence becomes stronger when independent sources point in the same direction. Useful phrases include corroborating evidence, independently verified, consistent with, converging evidence, lends support to.
Bias vocabulary
- selection bias
- confirmation bias
- survivorship bias
- publication bias
- sampling bias
- conflict of interest
- motivated reasoning
You do not need to force technical labels into every essay. Use them when they clarify why the evidence might systematically mislead.
Evidence strength vocabulary
- suggests
- indicates
- supports
- corroborates
- demonstrates
- fails to establish
- remains inconclusive
- is consistent with
- should be interpreted cautiously
The verbs create an evidence ladder. Do not write demonstrates when the evidence merely suggests.
Before-and-after evidence evaluation
Weak: “A survey showed that young people are unhappy, so social media is harmful.”
Stronger: “The survey offers evidence of an association between heavy social-media use and lower self-reported wellbeing, but self-selection and reverse causality limit the strength of any causal conclusion.”
Weak: “One country banned the product and pollution fell, proving the ban works.”
Stronger: “The case is consistent with the claim that regulation can reduce pollution, although changes in enforcement, consumer behaviour and economic activity must also be considered.”
Source credibility
When evaluating a source, consider expertise, access to evidence, incentives, transparency, methodology and whether the claim can be independently checked. Authority alone should not end the analysis.
Evidence relevance
Even good evidence can be irrelevant. A study from one country may not transfer cleanly to another. Evidence about short-term behaviour may not answer a long-term question. Ask whether the evidence actually addresses the claim you are making.
Paper 1 use
Do not simply place an example after a claim. Explain what the evidence shows, why it is credible enough to matter, and what it still cannot establish. This is where evidence becomes reasoning.
Paper 2 use
Comprehension often depends on noticing evidential force. “May indicate” is weaker than “demonstrates”. “Anecdotal” carries a different evidential weight from “empirical”. Preserve these distinctions in paraphrase and synthesis.
A credibility drill
Take one statistic from an article. Ask five questions: who produced it, how was it measured, who was included, what alternative explanation exists, and what conclusion would be too strong? Then rewrite the claim at the appropriate evidence strength.
The deeper lesson
Evidence does not remove judgement. It disciplines judgement. A GP student becomes stronger when the question changes from “Do I have an example?” to “What exactly does this evidence allow me to claim?”
Return to the JC1 Top 100 GP Vocabulary Hero, especially the evidence, causation and qualification families. Together they form the language of intellectual honesty.
Phase 4 Expansion: Evidence Has Dimensions
Evidence is not strong or weak in one simple sense. A source can be credible but irrelevant, a sample representative but poorly measured, a statistic accurate but unable to establish causation. GP evaluation improves when students separate these dimensions instead of asking only whether evidence is “good”.
The five-part evidence test
- Relevance: does it answer this claim?
- Credibility: is the source worth taking seriously?
- Validity: does the evidence measure or support what it claims to?
- Representativeness: can the result travel beyond the observed cases?
- Causal strength: does it show association, contribution or actual causation?
Source incentives matter
Expertise is valuable, but expertise does not remove incentives. Governments, firms, activists, academics and journalists may all possess useful evidence while also having reasons to frame it in particular ways. A mature evaluation asks about methodology and transparency rather than dismissing evidence merely because a source has interests.
Numbers are not automatically stronger than stories
Statistics are useful for prevalence and scale. Anecdotes are useful for illustrating mechanisms and lived consequences. The mistake is using one person’s experience to establish frequency, or using a population average to pretend no individual variation exists.
Triangulation
Confidence rises when different kinds of evidence converge: surveys, behavioural data, case studies, historical comparison or experimental evidence. This is triangulation. It is especially valuable in GP topics where no single study can capture the whole system.
Paper 1 evidence sentence model
Do not stop at the example. Try: evidence → what it supports → limitation → implication. “The case suggests that congestion pricing can alter travel behaviour, although transfer to other cities depends on transport alternatives and enforcement capacity. It therefore supports the mechanism without proving universal effectiveness.”
Paper 2 evidence language
Watch how writers signal support: illustrates, suggests, is consistent with, lends support to, corroborates, establishes. These verbs are part of meaning. Preserving them protects the evidential force of a paraphrase.
The evidence audit
- What precise claim is this evidence supporting?
- What kind of evidence is it?
- What population and context does it cover?
- What alternative explanation remains?
- What is the strongest conclusion I can responsibly draw?
What mastery looks like
A mature GP student no longer asks merely, “Do I have a statistic?” The student asks what the evidence establishes, how far it travels, which uncertainty remains and whether independent evidence points in the same direction.
Continue with Evaluation Vocabulary for GP. Evidence tells you what is supported; evaluation tells you how much that support should matter to the final judgement.