How to Check a PSLE Science Result for Internal Consistency Before You Explain It

Wait, What? Sometimes the Best Science Move Is Not to Explain the Result Yet

A results table says a cup contained 80 g of water at the start and 74 g at the end. A learner calculates that 16 g was lost, then writes a detailed evaporation explanation.

The explanation may be scientifically sensible. The result being explained is not. The change from 80 g to 74 g is 6 g, not 16 g.

Once the wrong derived value enters the reasoning chain, a good mechanism can be attached to a bad result. This is why strong PSLE Science learners need a short checkpoint between reading the result and explaining the result.

Before asking why something happened, first ask whether the quantities, labels, directions and conditions you are about to explain agree with one another.

Quick Answer

Check the scientific object, quantity, unit, starting value, final value, direction of change, any calculated value and the stated experimental condition. Compare the same result across the table, graph, diagram or written description. If those pieces disagree, locate the mismatch before building a causal explanation.

The checking chain is: IDENTIFY THE RESULT → TRACE IT TO THE ORIGINAL EVIDENCE → CHECK UNITS AND LABELS → CHECK START/END OR REFERENCE VALUES → CHECK DIRECTION → RECOMPUTE DERIVED VALUES → CHECK THE CONDITION → ONLY THEN EXPLAIN.

Owned PSLE Science Learning Job

This guide owns a narrow learner job: checking that a result is internally consistent with its own evidence before explaining it. It does not own scientific measurement generally, graph reading generally, or the scientific mechanisms used in the examples.

It is also not a licence to reject unexpected data merely because they surprise you. A surprising result can be real. Internal inconsistency means that two parts of the supplied or derived record cannot both be right as currently interpreted.

The Current PSLE Science Frame

For examination from 2026, SEAB assesses the 2023 Primary Science syllabus and includes interpreting and analysing information, evaluating observations and information, and communicating explanations and reasoning. Those jobs depend on reading evidence accurately before a conclusion or explanation is built.

The consistency checks below are an educational scaffold. They are not an official marking rubric and they do not imply that every examination question contains an error for students to hunt.

Result, Calculation and Explanation Are Three Different Layers

Layer	Example	Main check
Observed or recorded result	Mass fell from 80 g to 74 g.	Did I copy the values, object and unit correctly?
Calculated or derived value	6 g was lost.	Does the calculation follow from the original readings?
Scientific explanation	Water left the liquid surface by evaporation under the stated conditions.	Does the mechanism explain the verified result?

Skipping directly from the first layer to the third can hide a calculation error. Starting from an incorrect derived value can produce an elegant explanation of something that the data never showed.

The Eight-Point Internal Consistency Check

Object: Am I reading the correct set-up, specimen or time point?
Quantity: Is this mass, temperature, time, height, number, volume or another quantity?
Unit: Do the units match the quantity and comparison?
Reference: What starting value, control value or comparison point is being used?
Direction: Did the quantity increase, decrease or remain unchanged?
Derived value: If I calculated change, difference or rate, can I reproduce it?
Representation: Does the statement agree with the table, graph, diagram or written condition?
Condition: Does the explanation refer to the condition that actually changed?

Worked Example 1: Change Does Not Equal Final Value

Two seedlings are measured over one week.

Seedling	Starting height	Final height
P	8 cm	14 cm
Q	12 cm	16 cm

A learner says, “Q grew more because Q is taller at the end.” Check the result before explaining.

P increased by 6 cm. Q increased by 4 cm. Q has the larger final height, but P has the larger increase. The word grew more must be tied to the change, not simply to the final state.

Only after the comparison is corrected should the learner consider a scientific explanation for why the growth amounts differed, and only if the question supplies conditions that can support such an explanation.

Worked Example 2: A Unit Slip Can Reverse the Story

A table records one quantity as 0.8 g and another as 650 mg. A learner sees 650 and says the second value is much larger.

The numbers are not directly comparable until the units are aligned. 0.8 g is 800 mg, so the first value is larger.

This example is not about advanced conversion tricks. It is about refusing to let the visual size of a numeral replace the scientific quantity it represents.

Worked Example 3: The Graph and the Sentence Must Refer to the Same Object

A graph has two lines. P rises from 20 to 35 units. Q falls from 35 to 30 units. A learner correctly reads the rising line but writes, “Q increased.”

The scientific reasoning may be fine for P, but the object label has drifted. Before explaining, point to the line, trace it to the legend and restate the result with the correct object.

This kind of error can survive sophisticated thinking because the learner knows the relationship but attaches it to the wrong receiver. Object identity is therefore part of result consistency.

Worked Example 4: Smaller Number, Bigger Effect

Two identical objects travel the same distance under different conditions. P takes 4 seconds and Q takes 7 seconds. A learner writes, “Q was faster because 7 is greater than 4.”

The recorded result is not inconsistent. The interpretation is. For the same distance, shorter time indicates greater speed. The learner must translate the measured quantity into the process meaning before explaining it.

This is why internal checking includes the meaning of direction: bigger measurement does not always mean bigger underlying effect.

Worked Example 5: A Result Can Be Surprising Without Being Inconsistent

Three repeated measurements are 12.1 cm, 12.0 cm and 14.7 cm. The third is surprising because it differs strongly from the first two.

Do not delete it simply because it is awkward. First check whether the unit, object, instrument reading and transcription are correct. If they are, the result remains part of the evidence and may indicate variation, an unusual trial or a method issue that deserves investigation.

Internal consistency is not “make all results look alike”. It is “make sure each recorded and derived claim matches the evidence it is supposed to represent”.

Worked Example 6: A Calculation Can Be Correct but the Scientific Comparison Still Wrong

Set-up A changes from 30°C to 20°C over 5 minutes. Set-up B changes from 50°C to 35°C over 15 minutes. A learner correctly calculates changes of 10°C and 15°C, then says B cooled faster because it changed by more degrees.

The subtraction is correct. The rate comparison is not yet fair because the time intervals differ. If the question asks which cooled faster on average, the size of change must be considered together with time.

This illustrates an important principle: a correct calculation can still answer the wrong scientific question.

Representation Cross-Check

When the same result appears in more than one form, translate between them before explaining.

Table → verbal statement: “P increased from 4 to 7 units.”
Graph → table check: does the plotted point match the listed value?
Diagram → written condition: is the switch shown open when the sentence says closed?
Calculation → source values: can the derived change be traced back to the original readings?

If two representations appear to disagree, do not immediately choose the one that fits your expected concept. Re-check labels, scale, legend and time points first.

When the Question Itself Looks Inconsistent

In ordinary practice materials, printing, scanning or answer-key errors can occur. A missing unit, cut-off graph label or impossible arithmetic relation can make a question hard for the wrong reason.

Do not declare a question defective simply because the result is unexpected. Show the exact conflict: “The table says 40 g at the start and 35 g at the end, but the key says 15 g was lost.” That is checkable.

During the actual PSLE, follow the examination instructions and work from the paper as given. This guide does not invent a special procedure for challenging an examination item.

Earliest Weak-Link Diagnosis

Failure signature	Earliest weak link	Repair
You explain the wrong line on a graph.	Object/legend tracking	Trace the object label before reading the trend.
You compare 0.8 g with 650 mg by numeral size.	Unit meaning	Align units before comparison.
You confuse final amount with amount changed.	Reference value	Preserve starting and final values separately.
Your subtraction is right but your rate claim is wrong.	Question/quantity alignment	Identify whether the task asks amount, change or rate.
You reject an odd repeated result automatically.	Evidence handling	Check transcription and method before deciding what the result means.
Your explanation uses a condition that did not change.	Condition binding	Return to the set-up and identify the actual comparison.

Misconception Repair: “Scientifically Possible” Is Not the Same as “Internally Consistent”

A result can be scientifically unusual and still be internally consistent with the record. Conversely, a familiar-looking result can be internally inconsistent because a value was copied from the wrong row.

This guide therefore avoids the shortcut “Does this look realistic?” Primary learners should not reject evidence merely because it conflicts with expectation. Check the data structure first. Then use scientific knowledge to interpret the verified result.

A 30-Second Result Check

Before writing a long explanation, ask:

Which set-up and quantity am I talking about?
What are the units?
What is the reference or starting value?
Did it go up, down or stay the same?
If I calculated something, can I reproduce it?
Does my sentence match the graph or table?
Am I explaining the condition that actually differed?

This is not meant to become a ritual that slows every simple question. With practice, most checks become quick. Use the fuller version when the data are dense, the result is surprising or your explanation depends on a derived value.

Practice Sequence

Find one deliberately planted copying error in a short table.
Find one unit mismatch before comparing two values.
Distinguish final value from change using two different starting points.
Check a correct calculation that answers the wrong quantity.
Interpret an unusual repeated result without deleting it automatically.
Explain a verified result only after the consistency check passes.

Unfamiliar Transfer Challenge

A fictional sensor records “response units” for two materials. Material X starts at 15 and ends at 21. Material Y starts at 23 and ends at 27. A student says Y changed more because its final reading is higher.

You do not need to know what the sensor measures to check the arithmetic structure. X changed by 6 units; Y by 4. The scientific meaning of the sensor may still require supplied information, but the internal comparison can already be repaired.

This shows why some reasoning can transfer even when the scientific object is unfamiliar: preserve identity, reference and quantity before adding the mechanism.

Delayed Independent Return Test

Several days later, give a fresh question containing a table, a derived change and a written conclusion. Include one subtle mismatch. Do not tell the learner whether the problem is in the data reading, calculation or explanation.

The learner should locate the first mismatch, repair only that layer, then rebuild the explanation if needed. If every mistake triggers a complete restart, the checking process is not yet selective.

Result-Checking Receipt

Correct object?
Correct quantity?
Correct unit?
Correct starting/reference value?
Correct direction of change?
Correct calculation?
Correct representation?
Correct changed condition?
Only then: correct mechanism and outcome?

Parent and Tutor Teaching Guide

When a child’s explanation seems wrong, ask them to point to the result they are explaining before supplying the correct mechanism. Sometimes the Science concept is sound and the problem entered earlier through a copied value or reversed comparison.

Use deliberately inconsistent practice examples sparingly. Ask the learner to name the exact conflict rather than simply saying “something is wrong”. This develops auditable checking rather than suspicion.

When a result is merely surprising, model scientific restraint: verify the record, consider measurement and method, and keep the observation visible. Do not teach children that inconvenient data should be discarded.

Useful Internal Routes

Authoritative References and Evidence Boundary

The examples are original and simplified. The consistency protocol is a learning/checking scaffold, not a formal scientific error-detection algorithm or an official examination requirement. Real measurement uncertainty and experimental variation can be more complex than Primary-level examples.

The Quiet Return

A scientific explanation deserves a trustworthy result to explain.

So before you reach for the chapter keyword, look once more at the object, number, unit, reference and condition. Make sure the evidence says what you think it says. Then let the mechanism do its real job: explain the result that is actually there.

Small Group Tutorials