Originally published 7 June 2017 as a Punggol PSLE English tuition page. Rebuilt in 2026 as a noindexed guide to turning English mistakes into specific repairs. Dated programme claims, score promises, old contact details and unrelated travel imagery have been retired.

Quick answer: an English error log is useful only if it changes what the student does next. Do not record merely “Q6 wrong” or “grammar mistake”. Record the first failure, classify the mechanism, choose the smallest useful repair, retest on a different task after a delay, and note whether the same failure returns. The loop is attempt → locate → classify → repair → vary → delay → retest → update.

This page is intentionally noindex. Its reader job is narrow: how to build and use a PSLE English error log without turning it into a scrapbook of red marks.

An error log is not a list of failures

A long list of mistakes can make a student feel busy without making the next attempt better. The log earns its place only when it answers three questions:

• What failed first?
• What repair should happen next?
• What later task will show whether the repair transferred?

The unit of analysis is therefore not the worksheet question. It is the learning mechanism beneath the question.

Start with the original attempt

Keep enough of the student’s original work to reconstruct what happened. A clean model answer can hide the path that produced the error.

• the question or prompt;
• the student’s first answer;
• teacher or tutor feedback;
• the correction;
• how much help the correction required.

This lets the learner compare states rather than simply replace the wrong answer with the right one.

Find the first failure, not the final red mark

A wrong comprehension answer might begin with a misread question. A weak composition sentence might begin with a broken story plan. A grammar error might be correct in isolation but fail to transfer during writing.

Trace backward until the work was last secure.

question demand → evidence/idea → relationship → language choice → final answer.

The earliest broken link is usually the better repair target.

Use a small set of error classes

Do not create twenty categories. The classification should be precise enough to guide action and simple enough that the student can actually use it.

Separate observed, interpreted and unresolved

A strong log does not pretend the first explanation is certain.

• Observed: three answers copy long sections of the passage.
• Interpreted: evidence selection or question-demand reading may be weak.
• Unresolved: whether the student understands the passage but cannot select, or does not understand the relevant section at all.

The next task should distinguish those possibilities.

Comprehension errors need a chain diagnosis

For comprehension, record where this chain broke:

understand question → locate evidence → connect clues → infer within bounds → express precisely.

• If the question was misunderstood, more passage practice may not help.
• If evidence was found but the inference was too broad, practise evidence boundaries.
• If the reasoning is correct but the answer is vague, repair expression.

Grammar errors need a transfer check

Do not assume a grammar error means the rule is unknown.

• Can the student explain the rule?
• Can they correct an isolated sentence?
• Can they use it in a paragraph?
• Does the rule survive independent composition?

If isolated exercises are strong but writing is weak, the problem is transfer or attention, not necessarily knowledge.

Vocabulary errors should record the difference that was missed

“Wrong word” is too broad. Record whether the failure involved:

• meaning;
• word class;
• connotation;
• strength;
• collocation;
• register;
• retrieval under pressure.

A useful repair might compare two near-synonyms and then require both in new contexts rather than memorising another definition.

Composition errors should be ordered by leverage

Do not begin by correcting every adjective if the story does not answer the prompt.

A practical order is:

1. prompt relevance;
2. cause and consequence;
3. character motive or central idea;
4. paragraph function and pacing;
5. sentence clarity;
6. vocabulary precision;
7. grammar, spelling and punctuation.

The error log should lead the student to the highest-leverage repair first.

One entry should fit on one line

The detail belongs in the work itself. The log should be compact enough to scan for recurrence.

Frequency matters more than drama

A spectacular one-off mistake may be less important than a quiet error that appears every week.

• How often does the error recur?
• Does it affect several English components?
• Does it block high-value work?
• Is it changing after repair?

Prioritise patterns, not memorable red marks.

Use the smallest repair capable of changing the pattern

If the student misreads question demand, do not assign another complete paper. Practise five demand statements. If tense control collapses only in paragraphs, do not repeat an entire grammar chapter. Repair paragraph transfer.

Small repairs make cause and effect easier to see.

Immediate correction is not the finish line

Students often look “fixed” immediately after an explanation because the cue is still active.

Use three checks:

• near transfer: a similar but changed item;
• delay: the same mechanism several days later;
• farther transfer: the mechanism inside a different English task.

An error should eventually leave the log

The log should not grow forever. Mark a pattern as provisionally repaired when:

• it disappears across several independent tasks;
• the student can explain the repair;
• the student can detect the risk themselves;
• performance survives a delay.

Keep watching, but remove it from the active priority list.

Do not make the log punitive

The point is not to catalogue everything the student did badly. A useful log should create agency.

Write entries in neutral language:

• “evidence selection unstable” rather than “careless”;
• “prompt relevance missed” rather than “bad writer”;
• “tense transfer not yet stable” rather than “weak grammar”.

The language should describe the work, not the child’s identity.

Use the log before revision, not only after marking

Before a revision session, scan the active patterns and choose one or two high-leverage targets. Before an examination practice, ask the student which errors are most likely to recur and what checking cue will catch them.

The log becomes useful when it changes anticipation, not only post-mortem analysis.

Historical classroom context

The original 2017 article showed tutors marking student work, correcting mistakes and customising materials. That is the durable idea worth retaining. The rebuilt version adds the missing return path: marking should produce a narrow repair and the repair should later be tested in independent work.

A compact PSLE English error-log cycle

keep original attempt → find first failure → classify → record one-line entry → smallest repair → changed item → delay → transfer → close or reprioritise.

What parents, tutors and students can measure

• Are entries describing mechanisms rather than question numbers?
• Are the same error classes recurring?
• Does each active entry have a specific repair?
• Is the repair tested on a different task?
• Does it survive after a delay?
• Can the student increasingly predict and catch their own recurring errors?
• Is the active log getting shorter or more precise?

What not to conclude

• A long error log is not evidence of better revision.
• A corrected answer does not prove transfer.
• “Careless” is usually too vague to be a useful category.
• Every error does not deserve equal practice time.
• The log should describe work, not define the learner.
• The best error log changes what happens on the next independent attempt.

Related PSLE English routes

• What a PSLE English Tutor Should Diagnose Before Adding More Practice
• How to Make PSLE Composition Practice Produce Better Writing
• PSLE Comprehension Open-Ended — Find the First Weak Link