A crystal that looks clean is not automatically a pure substance. A recovered mass is not an identity test, and a narrow melting interval is not a licence to advertise a numerical purity percentage. This G3 SEC Chemistry guide teaches the quieter but more powerful habit: decide exactly what the evidence establishes before choosing the words of the conclusion.
This chapter follows the salt preparation and batch-accounting sequence. It takes a new question: what may a laboratory record of melting behaviour tell us about a candidate solid? The worked data are fictional and designed for reading and reasoning, not for unsupervised practical work. Examination candidates should follow their registered syllabus and teachers’ laboratory safety instructions.
Quick summary: separate recovery, identity and purity; read the complete interval; check reference conditions; apply a documented instrument correction; compare candidates; describe uncertainty; and finish with an evidence-matched statement rather than an absolute claim.
Follow the earlier salt-preparation guide, harvest-accounting guide and cross-subject evidence hearing. The PSLE learner hub supports the underlying observation and measurement habits. Official controlling references are SEAB’s 2027 G3 syllabus directory and the 2027 K324 Chemistry syllabus, checked 6 November 2026.
Identity is not harvest mass
The collection record answers how much material was recovered. It cannot identify the compound and cannot quantify contamination. A tray containing 15 grams of crystals can be less pure than a tray containing 12 grams. Both statements can be true because mass and composition are different quantities.
First separate three questions on paper: What quantity was recovered? Which substance is probably present? What evidence bears on purity? A confident answer should not use the same measurement to answer all three.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Why melting point belongs in an identity investigation
A given reference melting temperature can help identify a solid. The useful evidence is the observed physical behaviour compared with reference information under suitable conditions. A match supports a candidate; it does not necessarily prove that no other candidate shares a similar property.
For a fictional material X, suppose the reference is 80.0 °C. A measured interval of 79.8–80.3 °C is consistent with X. The words ‘consistent with’ are deliberate. If another candidate has a similar value, another suitable observation may be required.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Understand the two ends of a reported interval
An observation of first visible melting is not always identical to the temperature when the final solid disappears. A proper record may include both endpoints. Their subtraction gives the measured interval width, while their position relative to a reference helps interpretation.
Suppose melting begins at 79.7 °C and finishes at 80.4 °C. The width is 0.7 °C. A student who reports a single exact ‘melting point of 80.4 °C’ without explaining what that number represents has lost information.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
A narrow interval is useful evidence, not a purity certificate
When conditions are comparable, a narrow melting interval close to a reference can be compatible with a relatively pure crystalline specimen. It does not establish a numerical purity percentage. Instrument effects, specimen state and other impurities can complicate the connection.
A school poster saying ‘0.5 °C interval means 99.5% pure’ invents a conversion formula that the data never supplied. Replace it with a qualitative, evidence-limited sentence and explain what further authorised testing would be needed for a quantitative certificate.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Do not confuse purity with identity
One substance might give a relatively narrow interval but still be a different substance from the intended product. Conversely, an intended substance can contain impurities that alter its observed melting behaviour. Identity and purity are related but separate conclusions.
A statement such as ‘because the graph is flat, the sample must be X’ skips the reference comparison. A statement such as ‘because the reference matches, the sample is guaranteed pure’ skips the impurity question. Build both inferences explicitly.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
The official assessment context
SEAB lists G3 Chemistry as K324 for the 2027 SEC. The syllabus explicitly includes using supplied melting-point and boiling-point data to deduce identities and purity. Candidates sit Paper 1 multiple choice (40 marks, 30%), Paper 2 structured/free response (80 marks, 50%), and Paper 3 practical (40 marks, 20%).
These facts are subject-specific. G3 Combined Science routes have distinct codes K326, K327 and K328. The example here uses K324; a combined Science candidate should follow their own registered science syllabus and not assume every stand-alone Chemistry practical requirement applies.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Why sample condition belongs in the record
Crystals freshly removed from liquid may retain solvent or solution. That can make the collected mass unreliable as a dry-harvest value and complicate the interpretation of a specimen. The observation record should identify the relevant state rather than concealing it.
The absence of a drying note does not prove the sample is wet, but it prevents a confident claim that the mass was measured under verified dry conditions. Distinguish missing evidence from positive evidence of an error.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Instrument correction has a direction
A thermometer that reads 0.6 °C higher than the accepted value has a positive bias. Subtract 0.6 °C from its reading when the documented correction applies. If it reads lower, the sign of the correction reverses. State the meaning of the correction before calculating.
An observed interval 80.8–81.4 °C becomes 80.2–80.8 °C after subtracting 0.6 °C from both endpoints. The interval width remains 0.6 °C. Correcting only the midpoint would leave an internally inconsistent record.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Accuracy differs from precision
Three closely grouped readings show repeatability under the recorded conditions. They do not independently demonstrate that the instrument reads the correct temperature. A systematic offset may shift all three results together.
Results of 81.6, 81.6 and 81.7 °C are close together. If the instrument is known to read 1.5 °C too high, the corrected series is 80.1, 80.1 and 80.2 °C. Explain both properties separately rather than calling the original readings ‘perfect’.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Similar reference values create ambiguity
The closest number is not necessarily a defensible identity. If two candidate reference points lie within the uncertainty of a measurement, identifying one uniquely would require more information. A correct answer sometimes states that the test cannot discriminate.
Candidate A is referenced at 81.0 °C, candidate B at 81.4 °C, and the instrument uncertainty is about ±0.4 °C. A report of 81.2 °C should not be treated as an exact fingerprint. Select an independent appropriate test rather than guessing.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Preserve observations before writing explanations
A rigorous record separates the raw reading, any authorised correction, the corrected data, the reference, and the interpretation. That order allows another person to follow the reasoning and find the point where a numerical or logical mistake was made.
‘It is pure, 80 degrees’ is too compressed to check. Better: ‘The corrected interval is 79.8–80.3 °C, close to X’s 80.0 °C reference. This is consistent with X under the stated conditions; exact purity has not been established.’
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Don’t invent a relationship between melting range and concentration
Some questions provide an explicit model relating a measured property to composition. If so, use that model. Otherwise, an observed interval width cannot be converted into percentage contamination by intuition alone. Do not supply a formula merely to make the answer look mathematical.
If the interval spans 4.0 °C, the result may deserve investigation. It does not follow that the sample contains 4% impurity, or that removing a quarter of its mass would narrow the range by one degree. Those are unsupported models.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
A second measurement must answer the remaining question
Once a first test suggests identity, ask which uncertainty remains. A repeat melting measurement can investigate repeatability. An authorised chemical test can address a different component of identity. A mass measurement cannot resolve a compositional ambiguity by itself.
The next test should be chosen because it can distinguish candidate explanations. ‘Do more experiments’ is vague; ‘repeat the comparison with a verified thermometer and independently checked sample condition’ is targeted and relates to a specified problem.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Read the graph, not your expectation
On a temperature-time graph, temperature belongs to the vertical axis only if labelled that way. A near-level segment may be relevant to a phase transition under the described conditions, but it could also result from instrument or recording choices. Read captions and axes before interpreting shape.
A graph rising from 65 °C to near 80 °C before a level region supports describing a change near 80 °C in an idealised case. It does not show crystal identity alone. Distinguish the observed shape from a causal explanation.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Why a late correction changes the conclusion
If a teacher verifies that a recorded value was wrong, the corrected value replaces it for the current inference. This does not mean rewriting history without trace. A careful record preserves the fact that the earlier version was superseded.
A table that once gave 81.1 °C may be corrected to 80.1 °C after a validated +1.0 °C bias is applied. The identity paragraph must be checked again. A student who updates the table but retains the original conclusion has not completed the repair.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Precision in numbers needs precision in language
Report temperatures with appropriate units and a resolution supported by the data. Do not write a long string of calculator digits and suggest that the final decimal places were observed. The wording of the conclusion should be at least as careful as the arithmetic.
For example, the difference 81.4 − 80.8 = 0.6 °C is easy. The harder step is explaining what that width means, and what it cannot mean, for the proposed identity and purity claim.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Planning the examination response
For ‘deduce’, give the supported inference and cite the relevant reference. For ‘explain’, connect observation to an appropriate principle. For ‘evaluate’, discuss strengths, named limitations and a targeted improvement. These are different instructions, and good candidates adjust their answer length and structure accordingly.
Read the final line of the question before beginning. If asked why the evidence is insufficient for a purity claim, another paragraph about how salt crystals form does not address the command. Use the required evidence directly.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Avoid generic evaluations
‘Human error’ and ‘be more careful’ are too broad to explain why a melting interval might be unreliable. The problem may involve an unverified instrument offset, different observed endpoints, the sample’s unrecorded state or a comparison made under inconsistent conditions.
Name the likely issue, its effect on the data or inference, and a specific improvement. A useful answer might say: ‘Because the thermometer has an uncorrected positive bias, all observed temperatures could be too high. Apply its documented calibration before comparing them with the reference.’
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Connect the earlier Chemistry and Mathematics work
Vol 0849 taught the clean preparation of a soluble salt; Vol 0850 explained recovered masses and remaining liquor; Vol 0852 asked what the school could claim. The present volume changes the purpose of the evidence. It addresses identification from physical-property observations, not production throughput.
A student who carries the old mass percentage into a new purity question is solving the previous task. Recognising when to switch models is part of upper-secondary examination control.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Explain a conclusion aloud before writing it
A concise oral explanation can expose an incomplete chain: ‘We corrected the instrument reading; compared the interval with the reference; and found the result consistent with candidate X. We cannot certify exact purity from this observation.’ Every clause has a distinct task.
Practise again after changing the supplied reference value. If the explanation stays identical although the evidence now points elsewhere, the method has become a memorised script instead of a transferable reasoning tool.
Try this examination check: write down the actual observation before explaining it, then identify the exact sentence in your answer that depends on the reference or correction. Ask whether a different given figure would change your conclusion. If it would, show the dependency plainly. If the result would remain the same, explain why rather than relying on a remembered answer. This habit makes the reasoning visible to a marker and makes the method reusable with unfamiliar data.
Twenty worked decision clinics for a stronger G3 answer
1. Recovery versus identity
A tray yields 16.0 g while another yields 12.0 g. The heavier tray does not necessarily contain the more pure or more correctly identified substance. State what the masses prove and what physical-property or chemical evidence is still missing.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
2. Range width
Melting is observed from 79.6 to 80.2 °C. Subtract the lower endpoint from the upper: 0.6 °C. Do not use the upper endpoint as the entire answer. Explain the relevance of the interval and the reference.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
3. Positive thermometer bias
The instrument reads 0.7 °C high, and the raw value is 81.0 °C. Subtract the positive bias to obtain 80.3 °C. Explain why adding it would move the result further from the correct value.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
4. Corrected interval
A sample melts across 81.0–81.8 °C on a thermometer with +0.8 °C bias. The corrected interval is 80.2–81.0 °C. Width remains 0.8 °C. Show the correction to both endpoints.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
5. Ambiguous candidate references
Candidate C melts at 80.0 °C; D at 80.3 °C. A measurement at 80.2 °C with stated uncertainty is not necessarily uniquely identifying. The best answer recognises overlap and requests a discriminating observation.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
6. An attractive numerical invention
A 1.0 °C interval does not imply a 99% purity. No quantitative model has been given. Distinguish a useful qualitative consistency check from an unsupported numerical certification.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
7. Outlier handling
Three trials give 79.9, 80.0 and 83.1 °C. Investigate the third reading and its conditions. It cannot simply be deleted because it raises the average. Record how a verified method fault would change interpretation.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
8. Sample still wet
A mass is labelled ‘dry crop’ but the preparation log never confirms drying. The right response is to question whether the label is justified, not to assert that it was definitely wet.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
9. Endpoint mismatch
The student’s record describes first softening, while the reference lists complete melting. The definitions may differ. Compare equivalent measurement conditions before concluding that the sample differs from the reference.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
10. A missing axis unit
A temperature graph has values near 80 but no stated scale unit. Add °C if supported by the instructions. A graph without a unit creates avoidable ambiguity even when its shape looks familiar.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
11. Relative wording
‘Consistent with the stated reference’ is defensible when the corrected interval is close. ‘Guaranteed 100% pure’ is not supported by a single interval. Rewrite a claim without losing the meaningful observation.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
12. Independent verification
If a plausible identity remains uncertain, specify a separate suitable authorised check. Merely repeating the same uncalibrated instrument does not independently remove systematic error.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
13. Correct source control
An old laboratory note is replaced by a teacher-verified correction. Show which figure now governs the comparison. Do not silently combine original and corrected figures into an average.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
14. Pure does not mean more mass
A smaller recovered crop can be more pure than a larger one. A heavier sample cannot be treated as a purity index without appropriate composition evidence.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
15. Boiling-point comparison
For a liquid, use the supplied boiling-point reference under the stated pressure rather than using a solid’s melting-point comparison. Avoid treating two physical properties as the same measurement.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
16. A competing explanation
A shifted observed interval may reflect a different substance, a specimen impurity, an instrument bias or inconsistent conditions. Name the uncertainty that is actually supported instead of claiming one cause is certain.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
17. Claim strength
A sentence ‘the tests suggest the expected compound’ communicates the limit of the record. A sentence ‘the tests prove absolute safety for consumption’ adds a claim no classroom melting-point comparison can justify.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
18. Transfer to unseen data
Create a new reference point at 92 °C and a measured interval at 91.7–92.3 °C. Reuse the same reasoning route rather than recalling an answer with 80 °C. Transfer is demonstrated when method, not memory of a number, survives.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
19. Improve evaluation
Replace ‘do it carefully’ with ‘verify the temperature correction on the instrument used, document the sample condition and repeat under comparable conditions’. The improvement now addresses named sources of uncertainty.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
20. Complete the evidence chain
State measurement, correction, reference comparison, inference and limitation in order. If any link is missing, the paragraph may still sound scientific but cannot be independently checked.
Write a three-part response for this case: first, identify the exact numerical or observed evidence; second, make the strongest conclusion it warrants; third, name one limit or next check that matters. Do not add facts that the question has not supplied. For a full-paper rehearsal, time the answer, then mark scientific accuracy, mathematical handling and clarity separately. Repair the part that failed before repeating another task.
How to practise from foundations to examination control
Start with observation versus inference. Cover the answers and explain why a heavier crop does not establish purity. Then calculate two interval widths and correct two thermometer biases. Once those steps are stable, combine them into a comparison of two candidate substances. End the session with one written evaluation. The sequence matters because a student cannot rescue an uncorrected reading with elegant language, and cannot rescue an unsupported identity with more arithmetic.
After a day, revisit the same concepts using changed figures. After a week, use one unfamiliar table and ask the student to produce a single paragraph without help. Identify whether the remaining weakness is reading, temperature arithmetic, chemical inference or English expression. The Mathematics learning route supports numbers and units; the English route supports evidence and qualifications. These are educational pathways, not a claim that the site offers Secondary Science tuition.
What to remember
A careful G3 Chemistry learner never confuses a clean-looking specimen with a certified pure substance. They identify the question, present the observed data, apply any legitimate correction, compare with reference information, and use language that matches the evidence. That combination is useful for K324 data-based and practical reasoning and for scientific communication beyond the examination.
