Did you know that a beautifully cited sentence can still be unsupported by the source beside it? Students sometimes find a credible website, paste its link into a report and assume the evidence problem is solved. Yet the source may discuss the same general topic without establishing the specific conclusion the student has written.
Checking source support means comparing the exact claim in your sentence with what the original source actually reports—its observations, dates, quantities, conditions and limitations. A source supports a claim when the relevant information genuinely warrants that claim, not merely when it contains similar vocabulary. Read the passage yourself, narrow the wording when necessary and attribute the evidence accurately.
A reference is not automatically evidence
Suppose a student writes, “Our school reduced all canteen waste by 20% after introducing compost bins.” Beneath the sentence sits a link to a school garden log. The log records that a class collected 20 kilograms of food scraps for composting during April. That is worthwhile information, but it does not establish a percentage reduction, whether all canteen waste was measured or whether the bins caused any change. The link points to a relevant document; it does not validate the full sentence.
This distinction is easy to miss because source credibility and claim support answer different questions. A reliable school record can be an excellent source for the quantity it measured and a poor source for a claim about something it never measured. Likewise, a correct quotation can be placed beside a conclusion that reaches too far. Good research writing checks the relationship between the two.
Worked example: twenty kilograms is not twenty per cent
Here is an invented classroom illustration, not data from a real school. A fictional recycling team logs 8 kg of scraps collected during one week and 12 kg during the next. Their record shows 8 + 12 = 20 kg collected across those two weeks. The second week’s collected amount is 12 − 8 = 4 kg higher; relative to the first week’s 8 kg, the change in collection is (4 ÷ 8) × 100 = 50%. None of these calculations gives the percentage of total canteen waste reduced.
Why not? To estimate a percentage reduction in total canteen waste, we would need comparable measurements of total waste before and after an intervention, a clear definition of waste and an account of other factors that might have changed. Knowing that 20 kg entered compost bins supplies neither the total before nor the total after. More compost collection could even occur while total waste stays unchanged.
A defensible sentence is modest: “In the class’s two-week record, students collected 20 kg of food scraps for composting.” If a student wants to say why collection increased, that becomes a new question requiring additional evidence. The small act of narrowing a sentence is not weak writing. It is responsible writing.
Diagnose which part of the source check is missing
- Topic match without claim match: the student says, “Both are about recycling,” but cannot point to the sentence supporting the exact quantity or outcome.
- Number and denominator confusion: the student turns 20 kg into 20% or assumes a sample count represents all cases.
- Scope creep: the source reports one class or fortnight, but the draft says “all students,” “always” or “across Singapore.”
- Cause without comparison: the source documents a change and the student automatically credits an intervention without checking plausible alternatives.
- Attribution without verification: the reference exists, but the student has not read the page, passage, chart labels or surrounding caveats.
Ask the learner to underline the precise claim in their draft, highlight the actual supporting passage in the source and explain the connection aloud. If they cannot connect the two in a sentence, the problem may be interpretation, calculation, inference or simply an absent source. Each deserves a different correction.
A four-pass source-support routine
Pass 1 — extract the claim before opening a search tab
Write the claim as a plain testable sentence. Circle words that make it stronger: “all,” “proved,” “caused,” “doubled,” “best” or “significantly.” Ask what observation would have to be true for those words to be justified. Searching for a general topic can produce many related pages; searching for the evidence your exact claim needs is more efficient.
Pass 2 — inspect the original evidence and its units
Read beyond a search snippet or someone else’s summary. Find the relevant passage, table or figure; note who made the observation, when it was made, what was counted and whether the values refer to a sample or a whole population. A graph showing 20 households cannot automatically describe 200 households. A single measurement cannot establish a trend.
Pass 3 — test the inference
Now ask: does the evidence support this precise wording? Is a reported association being presented as a cause? Is a statement about one place stretched to every place? Have “percentage increase” and “percentage points” been confused? If a needed detail is missing, either gather better evidence or rewrite the claim at the level the source warrants.
Pass 4 — attribute and independently recheck
Record the source details and use the citation style requested by the school. Distinguish direct quotations from paraphrases; both need appropriate attribution. The Purdue Online Writing Lab’s guidance on paraphrasing also advises checking a paraphrase against the original meaning. Then close the source and explain what the evidence supports from memory before reopening it to confirm.
Two short practice questions
Question A: A student report shows that 18 of 30 survey respondents prefer the library’s quiet corner. Can the writer say “60% of everyone in the school prefers the quiet corner”? No. The calculation 18 ÷ 30 × 100 = 60% describes these 30 respondents. It is not automatically the proportion for the whole school. A defensible description says “60% of the 30 respondents,” followed by a note about how respondents were selected.
Question B: A science note says seedlings in one measured tray were 2 cm taller after a week. Can the writer say “The new light caused plants everywhere to grow 2 cm more”? No. The measured difference is limited to that tray and period, and cause requires a suitable comparison. Strong claims need stronger designs. These examples are teaching scenarios rather than published findings.
What students, parents and teachers can do this week
- Student: choose one paragraph from an existing assignment; beside each factual or numerical claim, write the source line that supports it, or mark “needs evidence.” Rewrite one overconfident sentence.
- Parent: ask “What exactly does this source show?” rather than “Did you add your references?” Praise an accurate qualification even if the final claim sounds less dramatic.
- Teacher or tutor: model a source that is on-topic yet insufficient. Ask students to correct the claim, not merely hunt for another link. Check understanding with a new claim–source pair the learner has not seen.
To observe progress, compare the first and second drafts: are quantities preserved, are sources traceable, and can the student state one limitation without prompting? Repeat with a different subject a few days later. The meaningful improvement is independent judgement, not the number of hyperlinks in the bibliography.
Limitations: a careful citation cannot repair every evidence problem
Some original sources are inaccurate, biased, inaccessible or superseded. A classroom exercise may simplify uncertainty for a beginner; a serious research claim can require multiple independent sources, study design appraisal or expert guidance. A source may genuinely support an observation while leaving causal explanations unresolved. The responsible move is to name that boundary, not to pretend complete certainty.
For broader navigation, see the eduKate Education library, the Core Aim learning hub and the earlier guide on source provenance. The emphasis on diagnosing a gap, explaining a principle, practising progressively and independently checking a new example also runs through the Clementi Mathematics learning reference.
Frequently asked questions
Is a reputable website enough to prove my claim?
No. Reputation is part of assessing a source, but your sentence still must be supported by the particular data or reasoning on that page. Check both the source and the inference.
Can I change the wording of a source in my own words?
Yes, with proper attribution and faithful meaning. A paraphrase must not turn “may be associated with” into “certainly causes,” or silently change a limited sample into a universal result.
What if two sources disagree?
Check definitions, dates, methods, sample sizes and the precise question each source answers. Disagreement can indicate different contexts rather than dishonesty. Ask for guidance when the evidence genuinely conflicts.
How many sources do students need?
There is no universal number. Follow the assignment requirements and select sources that directly support the claims. Two weakly related links do not automatically beat one directly relevant primary record.
A useful ending: make the sentence earn its evidence
A learner who can say “That source supports the smaller claim, but not the bigger one” has learned something valuable: intellectual care. Start with one real sentence, read its evidence closely, repair the scope and then try the method independently on another subject. Good education teaches not only how to find information, but also how to know what that information permits us to say.