A reading can fall by half while the detector still records a signal
Imagine that a school science worksheet prints four readings from a radiation detector: 164, 84, 44 and 24 counts per minute. A student sees the numbers and declares, “The source halves every minute.” Another notices that 84 is not exactly half of 164 and rejects the idea of half-life. Neither response has asked the important question: What was the detector measuring before the source was included? If the background contribution is four counts per minute, the corrected values are 160, 80, 40 and 20. Those corrected readings do halve. The uncorrected readings do not.
This is the next skill after understanding the electromagnetic spectrum in Learner’s Guide 0275. Gamma radiation belongs to that spectrum, but radioactivity begins inside an unstable nucleus. The distinction matters. Not every electromagnetic wave is nuclear radiation, and not every form of nuclear radiation is an electromagnetic wave. A good answer therefore starts with the right physical model, then handles the numbers and ends with a statement the evidence actually supports.
You will learn to distinguish radiation from radioactive material, interpret half-life correctly, use nuclear notation, separate measured count rate from background, and solve structured and data-based questions. Every numerical situation in this guide is an original teaching example, not an official SEC question or an instruction to handle radioactive materials.
The verified 2027 SEC examination context
The SEAB 2027 G3 syllabus directory, updated 26 October 2026, lists K323 for G3 Physics, with 6091 given as the previous code. The official G3 Physics K323 syllabus includes Radioactivity as Topic 20. It expects knowledge of the atom and isotopes, random spontaneous decay, alpha, beta-minus and gamma radiation, nuclear equations, half-life, background radiation, uses and hazards. The prescribed understanding is broader than a single formula: you must read information, explain changes and judge what a measurement permits you to claim.
K323 requires three examination papers: Paper 1 is 40 multiple-choice questions in one hour, worth 30%; Paper 2 is 80 marks in one hour 45 minutes, worth 50%; Paper 3 is a practical paper of 40 marks in one hour 50 minutes, worth 20%. Paper 2 includes compulsory structured work and a choice between two questions in its final section. Data-based interpretation and clearly explained reasoning matter. These details describe G3 Physics K323, not the different arrangements for Combined Science or for English and Mathematics.
For a coordinated route through examination preparation, use the complete EMS learner’s guide, the Science Hub, and the PSLE Science Learning Guide when an earlier concept such as fair measurement or graph reading needs rebuilding. Those links are aids to learning, not substitutes for the official syllabus.
Start inside the nucleus, not with a detector
An atom has a small positively charged nucleus made of protons and neutrons, with electrons outside. The proton number, Z, tells you how many protons are in the nucleus and therefore which element it is. The nucleon number, A, tells you the total number of protons and neutrons. The number of neutrons is A minus Z. A nuclide symbol places A at the upper left of the element symbol and Z at the lower left. In plain text we can read carbon-14 as ¹⁴₆C: it has six protons and eight neutrons.
Isotopes are atoms of the same element with the same proton number but different neutron numbers. They do not become different elements merely because they have different masses. The isotope description is a statement about nuclear composition, not a prediction that every isotope must be unstable. Some isotopes are stable; some have unstable nuclei that can decay.
The exam habit is to write down what each number means before calculating. If a symbol is presented as ²³₁₁Na, the neutron count is 23 − 11 = 12. If the question then asks for the identity of an atom with 11 protons, changing the number of neutrons does not change the element. This basic discipline prevents later mistakes in alpha and beta-decay equations.
Random for one nucleus; patterned for a large sample
A radioactive nucleus may decay at an unpredictable time. The process is described as random and spontaneous. Random means we cannot say precisely which nucleus will decay next or the exact moment it will happen. Spontaneous means the decay is a property of the unstable nucleus; you do not have to strike, heat or shine a lamp on it to make ordinary radioactive decay begin.
A large collection of unstable nuclei nevertheless shows a dependable overall pattern. Over each half-life, approximately half of the nuclei that have not yet decayed will decay. That sentence is carefully worded. It does not promise that exactly half will decay in every small sample, nor that an individual nucleus lives for one half-life and then disappears on command. Statistical variation is part of real measurement.
An examination question may ask why two repeat measurements are not identical even with unchanged conditions. A useful explanation mentions the random nature of nuclear decay and counting fluctuations. “The stopwatch must be broken” is not a defensible default. Nor should a small difference in counts be treated as proof that the half-life itself has changed.
Three emissions, three different kinds of explanation
Alpha (α) radiation consists of helium nuclei, with two protons and two neutrons. It is positively charged and relatively strongly ionising but has low penetration. Beta-minus (β⁻) radiation consists of fast electrons emitted from the nucleus during a nuclear change; it has a negative charge and intermediate penetration. Gamma (γ) is high-energy electromagnetic radiation; it has no electric charge and is more penetrating, with shielding reducing intensity rather than promising a magical complete stop.
Do not confuse ionising power with penetrating power. A particle that produces dense ionisation along a short track need not travel far through matter. In a school explanation, alpha radiation can be stopped by paper or the outer dead layer of skin; beta radiation needs more substantial shielding; gamma radiation is difficult to attenuate. The exact amount of shielding depends on energy and material and should not be guessed from a classroom diagram.
The distinction between irradiation and contamination matters too. Being exposed to radiation from an external source is not the same as radioactive material being deposited on or inside an object. Never design a home experiment around collecting, opening or moving radioactive material. The worked count data in this guide are entirely hypothetical.
Nuclear equations are a conservation problem
In nuclear notation, the upper numbers on both sides of a valid equation must balance, and so must the lower numbers. For alpha decay, the nucleus loses a helium nucleus with A = 4 and Z = 2. The daughter nucleus therefore has A lower by four and Z lower by two. A familiar example is ²³⁸₉₂U → ²³⁴₉₀Th + ⁴₂He. Check both totals: 234 + 4 = 238 and 90 + 2 = 92.
For beta-minus decay, a neutron changes into a proton and a beta-minus particle is emitted. The daughter’s nucleon number stays the same, while its proton number rises by one. For example, ¹⁴₆C → ¹⁴₇N + ⁰₋₁e. The lower numbers balance because 7 + (−1) = 6. A gamma emission changes the energy state of the nucleus but does not by itself change A or Z. More complete particle descriptions exist beyond the level required by this calculation method.
Always do the conservation check after naming the emission. If the daughter gains one proton but loses no nucleons, your equation is not alpha decay. If A changes during a pure gamma emission, you have confused energy release with particle loss. The two-number check is quicker than trying to remember a different rule for every element.
What half-life means—and what it does not mean
The half-life is the time taken for the number of undecayed nuclei in a large radioactive sample to fall to half its initial number. Under consistent conditions, radioactive activity also falls to half, so activity-time or appropriately corrected count-rate graphs can be used to estimate it. The relationship can be expressed as remaining fraction = (1/2) raised to the number of half-lives. This is a compact way to record repeated halving, not a demand that every question use indices.
Suppose a teaching model begins with 800 undecayed nuclei and uses a half-life of three hours. After three hours the expected remainder is 400. After six hours it is 200. After nine hours it is 100. The time for each new halving remains three hours; the number lost during each interval becomes smaller. That is why the curve bends towards zero rather than following a straight line down.
Never say that a sample reaches exactly zero after several half-lives. Half-life produces successive fractions: half, quarter, eighth, sixteenth and so on. In practice a detector may eventually report readings close to its background or below its useful detection capability. That measurement limit is not evidence that every nucleus has vanished.
A table is often clearer than a hurried formula
For many K323 questions, first create two rows in your working: time and amount remaining. If the half-life is four minutes and the initial activity is 320 Bq, the expected activity at 0, 4, 8, 12 and 16 minutes is 320, 160, 80, 40 and 20 Bq. Writing this progression makes a mistake visible immediately.
Suppose the question instead gives an initial activity of 480 Bq and says the final value is 60 Bq. Move from 480 to 240 to 120 to 60. That is three half-lives. If the elapsed time was 15 days, each half-life lasted five days. You are solving a counting problem before you are solving a timing problem.
Activity is measured in becquerels, Bq, where one Bq means one nuclear decay per second. A detector’s count rate, such as counts per minute, is not automatically the same as the activity of the radioactive material. A detector may miss events or see only part of the emitted radiation. In a classroom table, use the specific quantity and units supplied by the question.
Background counts must come out before halving
Radiation detectors can register background events from natural and environmental sources even when the particular sample being studied is absent. A measured detector count rate therefore includes background contributions. If the problem supplies a mean background rate, the usual teaching calculation is corrected source count rate = measured count rate − background count rate. Do the subtraction before looking for a halving pattern.
Example: a detector records 196 counts/min with a source present, and the average background is 16 counts/min. The corrected source rate is 180 counts/min. Later the total detector reading is 106 counts/min. The corrected rate is 90 counts/min. That is a halving of the source contribution even though the total did not halve: 106 is not half of 196.
Another example: total detector readings of 510, 270, 150 and 90 counts/min are recorded at equal times; background is 30 counts/min. The corrected sequence is 480, 240, 120, 60. Each interval is one half-life. If the time between readings is seven minutes, the half-life is seven minutes. If you forget background subtraction, the graph appears to decay more slowly at late times and you risk the wrong answer.
A decay graph contains several stories at once
On a graph of corrected activity or count rate against time, the vertical axis describes how much radioactive activity remains, and the horizontal axis shows elapsed time. Start at one point on the curve, find half its vertical value, move across to the curve, and then down to the time axis. The horizontal difference is one half-life. Repeat this from another pair of values to check consistency.
A common error is to use half the reading on an uncorrected graph. If the graph levels towards a stable horizontal background at 12 counts/min, half of a 212-count/min total is not the correct next target. First subtract 12: the source contributes 200. After one half-life the source contributes 100 and the detector reads about 112 counts/min. The point of the graph is to separate signal from baseline.
Do not infer an exact half-life to more decimal places than the plot can support. A student can often read a broad estimate from a curve, then state what graph resolution limits it. Good physics explains uncertainty without treating all conclusions as equally uncertain. A clear pattern remains useful even when individual counts fluctuate.
Use the command word to choose the kind of answer
State calls for a fact, such as “Gamma radiation is an electromagnetic wave.” Describe calls for what is observed: the recorded count rate decreases as time increases and approaches background. Explain asks why the pattern occurs: fewer unstable nuclei remain to decay, while the decay of an individual nucleus remains random. Calculate demands operations, working, numbers and units. Evaluate needs a judgement tied to the limits and strength of the evidence.
If a question says, “Suggest why the experimental count rates at two equal intervals are not exactly in the ratio 2:1,” answer with random variations in recorded decays and sensible measurement issues. Do not claim that radioactive decay “slows because the substance is running out of energy.” The rate is lower because fewer undecayed nuclei remain. Your reasoning should respect the model rather than inventing a mechanical cause.
This is also where reading skill matters. In English comprehension you learn to separate what a text says from what you merely suppose. The same habit makes scientific explanation more accurate. Our English guide 0277 on evidence and effect offers a useful parallel: description, inference and interpretation must not be mistaken for one another.
An original structured worked question
A fictional study sheet describes a detector used under supervised conditions and supplies these readings at times 0, 6, 12 and 18 minutes: 340, 180, 100 and 60 counts/min. A separate background reading is 20 counts/min. The question asks for the half-life, a predicted detector reading at 24 minutes, and one reason why a real experiment might not produce exact numbers.
First correct the readings: 320, 160, 80 and 40 counts/min. The sample contribution halves in every six-minute interval, so the half-life is six minutes. The predicted corrected count rate at 24 minutes is 20 counts/min. Add background back because the question asks for the detector reading, not the corrected source rate. The predicted total is 40 counts/min.
For the explanation, note that radioactive decay is random and a detector records discrete events. Real short-interval counts will fluctuate around an expected pattern. The response must not pretend that the experimental device necessarily gives the exact mathematical progression, nor that the background source can be switched off by changing the sample.
A short scientific argument needs careful boundaries
Consider the claim, “The detector count rate fell by half, so the radiation is harmless.” The premise is a numerical change. The conclusion is a biological risk judgement. That leap is invalid: the information does not specify the radiation type, exposure conditions, intensity, location or relevant protective measures. A lower count rate by itself does not establish safety.
Likewise, “Gamma rays are electromagnetic waves, therefore every wireless signal is radioactive” confuses membership in one broad physical category with sharing all its properties. Gamma rays and ordinary radio signals are both electromagnetic radiation, but radioactivity is the spontaneous nuclear process that can emit gamma rays. You cannot infer unstable nuclei in a communication device from the mere presence of radio waves.
In a strong answer, identify the accurate observation, the unsupported inference, and the information needed to decide further. That habit will matter in data-based science questions and in everyday interpretation of technical reporting. For the public-facing communication task, continue to the next EMS guide after finishing the physics questions here.
Practice clinic: twenty-four questions that separate the errors
Try each question before reading its solution. Cover the answer, say what you expect physically, then do the arithmetic. When you get an answer wrong, label the cause: nuclear notation, wrong quantity, background omission, half-life counting, graph reading or claim beyond the evidence. These categories tell you what to repair; a raw percentage score does not.
1. A nuclide has proton number 15 and nucleon number 31. How many neutrons?
The neutron count is 31 − 15 = 16. Proton number tells you the element and nucleon number counts both nuclear particles; do not subtract electrons or use 31 as the proton number. Check: If your answer is 15 or 31, label the two numbers on the symbol before trying another isotope.
2. Two atoms have Z = 8 but nucleon numbers 16 and 18. Are they isotopes?
Yes. They share eight protons and differ in the number of neutrons. Their neutron counts are eight and ten. The different mass numbers do not make them different chemical elements. Check: State both the common quantity and the changed quantity; merely saying “different mass” is incomplete.
3. A source has 1,280 Bq initially and a half-life of two hours. Activity after six hours?
Six hours contains three half-lives. Halve 1,280 to 640, then 320, then 160 Bq. Check using a fraction of one eighth of the original, which is also 160. Check: Write the unit Bq; a numerical answer without a unit does not communicate the physical quantity.
4. A hypothetical sample changes from 600 to 75 expected undecayed nuclei. How many half-lives?
The sequence is 600, 300, 150, 75. Three halvings have occurred. If you divide 600 by 75 and see eight, remember that eight is 2 cubed, not eight half-lives. Check: Do not confuse a reduction factor with the number of repeated halving intervals.
5. The elapsed time in Question 4 was twelve days. Find the half-life.
Divide the elapsed twelve days by three halvings. The half-life is four days. The arithmetic becomes simple only after you have counted the repeated halvings correctly. Check: Express the answer as a time; 4 Bq or 4 counts/min would be dimensionally wrong.
6. Background is 14 counts/min; total is 174 counts/min. Find corrected rate.
Subtract background from the measured total: 174 − 14 = 160 counts/min. This represents the source contribution under the assumptions supplied by the question. Check: If you simply halve 174 in the next step, you will carry the background error through the whole problem.
7. With the same background, total later becomes 94 counts/min. Has the source rate halved?
The later corrected rate is 94 − 14 = 80 counts/min. It is half of the earlier corrected rate of 160. The observed totals are not in the ratio two to one. Check: Always compare like with like: corrected to corrected or total to total after modelling the background.
8. Corrected count rate is 240 counts/min, then 30 counts/min after nine minutes. Half-life?
The source rate halves from 240 to 120, then 60, then 30. Three half-lives take nine minutes, so one lasts three minutes. Check: Draw the halving chain instead of trying to subtract the rates and divide by time.
9. Measured rate is 50 and background is 10 counts/min. Predict total after one half-life.
The corrected source rate is 40 counts/min. After one half-life it is 20 counts/min. Adding the background produces a predicted detector reading of 30 counts/min. Check: The total is not 25 counts/min. This question tests which quantity obeys the halving model.
10. A corrected source rate falls from 90 to about 45 in five hours, then to about 22 in another five. Meaning?
The estimate is a half-life of about five hours. The 22 rather than 22.5 may reflect rounded data or random counting variation. Report an estimate appropriate to the evidence. Check: Do not claim the second interval proves a different half-life simply because the data were rounded.
11. A detector still shows 12 counts/min long after a nearby source contribution becomes very small. Explain.
It may be recording background radiation. The total measured signal can tend towards background while the corrected source contribution approaches zero within the limits of measurement. Check: A non-zero total does not prove that the particular source decays forever at a fixed 12 counts/min.
12. A student says alpha radiation is the most penetrating because it ionises most strongly. Correct?
No. Ionising ability and penetrating power are distinct properties. Alpha particles ionise strongly along short paths yet are readily stopped by matter; gamma radiation is more penetrating. Check: State both properties explicitly to show that you know why the proposed reasoning fails.
13. In alpha decay, A = 226 and Z = 88 initially. Daughter numbers?
Subtract four nucleons and two protons: A = 222, Z = 86. The emitted alpha particle accounts for the missing four and two in the balance. Check: You do not need to memorise the daughter name to show the crucial nuclear-number calculation.
14. In beta-minus decay, A = 32 and Z = 15 initially. Daughter numbers?
The daughter keeps nucleon number 32 while proton number increases to 16. The emitted electron carries the negative charge in the nuclear-number balance. Check: If you subtract one from A, you have mixed up beta-minus and alpha nuclear changes.
15. Does pure gamma emission change A or Z?
No. Gamma emission releases energy from the nucleus without ejecting a proton or neutron. The same nuclide can move from an excited nuclear state to a lower energy state. Check: A powerful electromagnetic photon is not a nuclear particle with a non-zero nucleon number.
16. What does 25 Bq mean?
It means an activity of 25 nuclear decays per second. It does not mean that a particular detector necessarily registers 25 counts per second, because detector response and geometry affect counts. Check: When an exam mixes Bq with counts/min, keep the units distinct until an explicit conversion is justified.
17. A graph uses time in minutes and corrected rate in counts/min. What does a half-life read-off measure?
It measures the horizontal time interval needed for the corrected rate to fall from one chosen value to half that value. The axes and units are part of the answer. Check: Do not read a vertical difference as the half-life; that would have units of counts/min.
18. A curve falls from 400 to 100 corrected counts/min in eight minutes. Half-life?
That is a reduction by a factor of four, or two halvings: 400 to 200 to 100. Each half-life lasts four minutes. Check: A factor-of-four reduction is not four half-lives, and the final value alone does not tell you the time.
19. A source has been through four half-lives. What expected fraction remains?
After four successive halvings, the fraction is 1/16, or 6.25% of the original. The fraction decayed is therefore 15/16, or 93.75%. Check: Keep the words “remains” and “decayed” separate; they are complements, not synonyms.
20. The activity has fallen to one quarter. How many half-lives have passed?
One quarter is 1/2 times 1/2. Therefore two half-lives have passed. Without the elapsed time, you cannot deduce the duration of one half-life. Check: A complete answer respects which quantity the question did and did not provide.
21. Can you predict exactly which one of ten unstable nuclei decays next?
No. Radioactive decay is random at the individual nucleus level. A large sample shows statistically predictable changes, but neither that pattern nor a half-life identifies one particular next nucleus. Check: The difference between a population model and an individual prediction is a frequent exam reasoning point.
22. A student claims that moving a source farther away shortens its nuclear half-life. Evaluate.
Changing detector distance may change the observed intensity or count rate, but it does not change the intrinsic half-life of ordinary spontaneous nuclear decay. Geometry and nuclear transformation are different causes. Check: Ask whether the claim concerns the source itself or how much of its radiation the detector receives.
23. Why take a background reading rather than assume zero?
Because the environment may contribute measured events. An independent background estimate lets you separate that baseline from the source contribution and interpret later low readings more honestly. Check: Background is not an inconvenient mistake to ignore: it is part of the measured system.
24. Is a source safe to handle because it has a short half-life?
That conclusion is unsupported. Safe management depends on the substance, radiation, activity, route and duration of exposure, shielding, and authorised controls. A half-life alone cannot establish a safe handling decision. Check: For real materials follow trained supervision and school safety rules; the exercise data here involve no real source.
How to write a full-credit explanation from a graph
A graph question might give you four marks across parts, rather than asking for one isolated value. Suppose you must describe the trend, estimate a half-life, explain the reason and state a limitation. Begin by describing the downward curve and identifying the plotted variable. Then show the two plotted values and the horizontal time difference that support your half-life. Explain that the number of undecayed nuclei decreases, so fewer decays are expected per unit time. Finally state the limitation: readings fluctuate and background or graph resolution may affect the estimate.
Avoid replacing these four tasks with one long paragraph saying that “radiation fades out with age.” That sentence gives no read-off, no units and no adequate cause. Strong answers are not longer because they use grander vocabulary; they are longer only where the marks demand separate evidence. The Examination Craft route can help you practise this relationship between command word, working and clarity.
If you are stuck, reduce the question to three lines: the observed numbers, the model or equation, and a conclusion in the unit requested. For half-life, the model is successive halving of the undecayed sample or corrected count rate, not a constant number of decays each minute. For background, the correction is subtraction from a measured total. For interpretation, the most important step is often naming what the data cannot show.
A seven-day revision path that builds independent control
Day one: reconstruct the atom, isotopes and the two nuclear numbers without a textbook. Day two: compare alpha, beta-minus and gamma by physical nature, charge and penetration, then check the differences with short questions. Day three: balance ten nuclear equations using A and Z separately. Day four: solve repeated-halving questions without a calculator where the values are powers of two, so the idea is secure before the arithmetic becomes harder.
Day five: move to detector questions that include background radiation. Draw a corrected-rate table and explain why total readings do not halve exactly. Day six: answer graph-based and data-based questions under a short time limit, including at least one with a curved graph and one with an estimated background. Day seven: mark your attempts by error cause, revisit only the weakest cause, and finish with a fresh mixed problem rather than rereading everything.
This sequence connects back to the PSLE Learning Guide for the earlier habit of reading measurements carefully and forward to the SEC requirement to justify a decision. It also prevents a familiar problem: doing the same easy calculation ten times and mistaking fluency with one format for readiness for an unfamiliar question.
Finish with a three-part reliability test
Before you turn a Physics answer into your final response, ask three questions. Physical meaning: have I described an unstable nucleus, a type of radiation, a detector reading or an exposure risk, and have I kept those ideas separate? Mathematical meaning: have I counted half-lives correctly, used the corrected quantity and kept the unit? Evidence meaning: does my conclusion follow from the information supplied, or have I silently assumed that a measurement proves more than it does?
If all three answers are sound, the work is becoming examination-ready. Follow the next guide for an integrated EMS public-information audit, then use the subject owners to repair any remaining uncertainty: the Science Hub for concepts, the Mathematics Hub for graph and rate calculations, and the Secondary English Learning Hub for precise claims. The examination is not a contest of confidence. It is a test of what you can demonstrate.
