Secondary 1 G3 Science investigations become clearer when students can explain what they changed, what they measured and what their results actually support. This Punggol guide teaches 100 words and terms for measurement, experiments and evidence, with short meanings, original examples and a check against common confusion. Use it to distinguish accuracy from precision, choose sensible units, identify variables, interpret graphs and write conclusions that stay within the evidence. Begin with the language in the practical or data question your school is currently teaching.
This is an original eduKate learning selection, not an official MOE or SEAB Top 100 list or a compulsory Secondary 1 word checklist. It develops the investigation strand of our broader Secondary 1 G3 Science vocabulary guide. MOE’s syllabus covers lower secondary, and schools determine their teaching sequence, so some supporting terms here may be introduced later or used as extension. The tasks below use invented data for explanation; they are not measurements from Punggol or instructions to reproduce school practicals at home.
- For: Secondary 1 students taking Science at G3 and parents helping them understand practical work and data questions.
- Included: exactly 100 distinct entries in ten sets, with a meaning, an original example and a caution about misuse.
- Most useful distinctions: observation and inference; independent and dependent variables; accuracy and precision; correlation and causation.
- Worked skills: units, displacement, controlled comparisons, repeat readings, means, unusual results, graph gradients and bounded conclusions.
- How to study: learn a few relevant terms, explain one worked example, then answer a changed question independently.
- School boundary: follow the teacher’s current scope and practical instructions; Secondary 1 does not have a national SEAB Science examination.
- Parent support: locate the particular confusion, discuss it without blame and check whether the student can use the correction in a new task.
Choose your next learning step
eduKate Punggol · Secondary 1 G3 Science measurement, experiments and evidence
Start with the words in today’s lesson, or choose a route below.
Open the chapter index · Continue to the wider subject guide
Chapter index
Vocabulary, meaning and worked examples
Practice, study plans and next steps
CHAPTER 1 OF 9 · Back to contents
1. Understand the purpose of this investigation vocabulary
A student can perform a calculation correctly and still reach a weak scientific conclusion. The numbers may come from an unfair comparison, the instrument may have an offset, or the answer may claim more than the observations establish. Investigation vocabulary gives students a way to identify those differences. It helps them ask whether the difficulty concerns measurement, design, interpretation or communication.
The MOE G2/G3 Lower Secondary Science syllabus, updated April 2024, includes inquiry, measurement, suitable units, accuracy, precision, zero error and parallax error. This article builds a teaching route around that foundation. It also introduces supporting language for discussing data carefully. Inclusion here does not mean every term is a separate assessed requirement for every Secondary 1 class.
Use a question from the current school chapter to choose the first set. If a child cannot explain the axes of a graph, begin with the graph group. If the child writes “repeat the experiment” for every improvement, begin with measurement quality and ask what the repetition would actually check. A correct improvement should address a named weakness.
The SEAB SEC information page concerns the later national qualification introduced from 2027. It does not define a national Secondary 1 practical paper. For a future SEC examination, use the requirements for the student’s actual examination year when available. For today’s work, the useful target is an independent explanation of a manageable investigation.
CHAPTER 2 OF 9 · Back to contents
2. The complete 100-term list
Each row is one learning item. Technical phrases such as “controlled variable” stay together because their combined meaning matters. The check column identifies a mistake to avoid; it is not an additional experiment to perform. Definitions are deliberately short, and the worked sections show how the distinctions function in a complete answer.
Questions, observations and explanations
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 1 | Investigation question | The specific question a study is designed to answer. | Does changing ramp height affect the trolley’s travel time? | Name a measurable outcome, not just a broad topic. |
| 2 | Observation | Something noticed directly or recorded using an instrument. | The pointer moved from 2 to 5 on the scale. | An explanation of why it moved is an inference. |
| 3 | Inference | An interpretation drawn from observations and relevant knowledge. | A wet surface suggests liquid reached it, but not necessarily rain. | Several explanations may fit the same observation. |
| 4 | Hypothesis | A proposed explanation or relationship that can be tested. | A higher ramp may produce a shorter travel time. | A hypothesis is not a result already established. |
| 5 | Prediction | A reasoned expectation about what will happen under stated conditions. | The trolley should arrive sooner when released from the higher position. | State the conditions; do not promise certainty. |
| 6 | Evidence | Information used to assess or support an explanation or claim. | Recorded travel times provide evidence about the proposed relationship. | A confident opinion alone is not measurement evidence. |
| 7 | Claim | A statement about what is true or what happened. | The group claims that the new arrangement reduced travel time. | A claim needs suitable evidence and reasoning. |
| 8 | Reasoning | The explanation connecting evidence to a claim or conclusion. | Similar loads were compared, so mass is less likely to explain the difference. | Listing results does not explain their relevance. |
| 9 | Conclusion | An answer to the investigation question supported by the findings. | Within the tested range, greater ramp height was associated with shorter times. | Keep the conclusion within the study’s conditions. |
| 10 | Limitation | A feature that restricts what the study can establish. | Only one model of trolley was tested. | State what the limitation affects, not just “human error”. |
Variables and fair comparisons
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 11 | Independent variable | The factor deliberately changed in a controlled investigation. | The group changes the ramp’s height between sets of trials. | Identify the factor, not the measured outcome. |
| 12 | Dependent variable | The outcome measured to examine the effect of a change. | Travel time over the marked distance is measured. | It is not every quantity mentioned in the question. |
| 13 | Controlled variable | A relevant factor kept constant during a comparison. | The same trolley and marked travel distance are used. | Explain how the factor will be kept constant. |
| 14 | Control setup | A comparison setup used to help isolate an investigated effect. | A test of a treatment includes a comparable untreated sample. | A control setup differs from a controlled variable. |
| 15 | Baseline | A starting or reference measurement used for comparison. | The group records the sensor reading before the change. | A baseline is not automatically a separate control setup. |
| 16 | Fair comparison | A comparison designed to limit relevant alternative explanations. | Only the planned factor changes while other important conditions are matched. | Changing two important factors weakens the comparison. |
| 17 | Test range | The span of independent-variable values included in the investigation. | The supplied data cover masses from 0 to 150 g. | Results may not hold beyond the tested range. |
| 18 | Interval | The difference between successive selected values or measurements. | Values of 10, 20 and 30 cm use 10 cm intervals. | Check whether the intervals actually remain equal. |
| 19 | Trial | One run of the investigation under specified conditions. | The second trial uses the same setting as the first. | A new setting and a repeat trial serve different purposes. |
| 20 | Sample | The material, objects or individuals selected for study. | Five leaves are selected from a larger collection. | A convenient sample may not represent the whole collection. |
Quantities, values and units
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 21 | Physical quantity | A measurable property expressed using a value and a reference such as a unit. | Length and time are different physical quantities. | Name what is measured before choosing an instrument. |
| 22 | Numerical value | The number part of a quantity expressed in a stated unit. | In 35 cm, the numerical value is 35. | Changing the unit changes the numerical value. |
| 23 | Unit | An agreed reference used to express a quantity. | The length is 35 centimetres, not simply 35. | A bare number can leave a measurement incomplete. |
| 24 | SI | The International System of Units used for consistent scientific measurement. | The metre, kilogram and second are SI base units. | Common laboratory units are not all SI base units. |
| 25 | Base unit | A unit for a base quantity in the chosen measurement system. | The SI base unit of mass is the kilogram. | The gram is useful, but is not the SI base unit. |
| 26 | Derived unit | A unit formed from units of other quantities. | Metres per second expresses speed using length and time. | Derived does not mean less valid or less useful. |
| 27 | Length | Distance measured along a specified direction or path. | Measure the strip between the two marked endpoints. | A bent route is not its straight-line separation. |
| 28 | Mass | The amount of matter in an object, measured in units such as kilograms. | A balance gives the sample’s mass as 45 g. | Mass and gravitational weight are different quantities. |
| 29 | Time interval | The duration between a defined starting event and ending event. | Start and finish markers define the travel time being measured. | A time of day is not the same as duration. |
| 30 | Temperature | A quantity indicating thermal condition, commonly described as how hot or cold. | The thermometer reports the sample temperature as 28°C. | Temperature is not the amount of thermal energy present. |
Volume, scale readings and instrument detail
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 31 | Volume | The amount of three-dimensional space occupied or enclosed. | The solid displaces 11 cm³ of water in the example. | Do not confuse volume with mass. |
| 32 | Capacity | The volume that a container is designed or able to hold. | A 250 mL container need not currently contain 250 mL. | Capacity and current contents are different. |
| 33 | Area | The size of a surface, expressed in square units. | A 3 cm by 4 cm rectangle has area 12 cm². | Area units are squared; volume units are cubed. |
| 34 | Density | Mass divided by volume for the object or substance considered. | A 33 g object occupying 11 cm³ has density 3 g/cm³. | A greater mass alone does not establish greater density. |
| 35 | Displacement | The volume of liquid moved aside by an immersed object in this context. | The reading rises from 28 mL to 39 mL. | Use the difference, subject to the method’s conditions. |
| 36 | Meniscus | The curved surface of a liquid near the container walls. | For a clear water sample, read the bottom of the curve as instructed. | The correct reading convention depends on the liquid and instrument. |
| 37 | Graduated scale | A set of marked divisions used to indicate measured values. | The cylinder has numbered marks with smaller divisions between them. | Count spaces between marks, not just printed lines. |
| 38 | Scale division | The value represented by the space between adjacent scale marks. | Ten equal spaces between 20 and 30 mL each represent 1 mL. | The labelled numbers may be several divisions apart. |
| 39 | Resolution | The smallest change that an instrument’s indication can meaningfully distinguish. | A display changing in 0.1 g steps cannot display 0.01 g changes. | More displayed digits do not guarantee greater accuracy. |
| 40 | Measurement range | The span of values an instrument can measure as specified. | A 100 mL cylinder cannot directly hold a 150 mL sample. | Choose a suitable range as well as useful divisions. |
Apparatus and instrument checks
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 41 | Apparatus | The equipment used in an investigation. | The apparatus list includes a balance and a measuring cylinder. | Listing equipment does not explain the measurement method. |
| 42 | Metre rule | A graduated rule used to measure lengths up to about one metre. | Read the marked endpoints rather than assuming the physical edge is zero. | Check the zero position and the viewing angle. |
| 43 | Measuring cylinder | A graduated container used to measure liquid volume. | The pupil reads the meniscus while the cylinder stands upright. | A beaker’s rough markings may be less suitable. |
| 44 | Electronic balance | An instrument that indicates mass electronically. | The display is checked with the empty pan before use. | A stable display can still have an offset. |
| 45 | Stopwatch | An instrument used to measure elapsed time. | The timer runs between the agreed start and finish events. | Human response can affect very short timings. |
| 46 | Thermometer | An instrument for measuring temperature. | The reading is recorded after the instrument has responded appropriately. | Follow instructions on placement and reading time. |
| 47 | Ammeter | An instrument for measuring electric current. | A teacher’s circuit diagram places the ammeter in series. | Use only the teacher-approved circuit and settings. |
| 48 | Voltmeter | An instrument for measuring potential difference between two points. | The diagram shows a voltmeter connected across a component. | Current and potential difference are different quantities. |
| 49 | Calibration | Establishing how readings relate to reference values and using that relation for measurements. | A laboratory compares balance indications with suitable reference masses. | Calibration is not merely changing the display to zero. |
| 50 | Zero error | A non-zero indication when the measured quantity should be zero. | An empty balance reads +2 g before an object is added. | Do not assume the same correction without checking the instrument. |
Measurement quality and uncertainty
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 51 | Accuracy | Closeness of a measured value to the true value, assessed using suitable reference information. | A reading near a trusted 100 g reference is relatively accurate. | Tightly grouped results can still be inaccurate. |
| 52 | Precision | Closeness of repeated measured values to one another under specified conditions. | Readings of 103.0, 103.1 and 103.2 g are closely grouped. | Precision does not show closeness to the true value. |
| 53 | Repeatability | Agreement of measurements under the same specified conditions over a short period. | One operator repeats readings with the same apparatus and method. | State which conditions remain the same. |
| 54 | Reproducibility | Agreement of measurements when specified conditions such as location, operator and equipment differ. | Different laboratories obtain comparable measurements using documented arrangements. | Changing conditions must be recorded, not ignored. |
| 55 | Measurement uncertainty | A quantified indication of the spread of values reasonably attributed to a measured quantity. | An uncertainty statement communicates limits in the measurement result. | It is not simply a mistake or an unknown error. |
| 56 | Random variation | Unpredictable differences among repeated measurements or outcomes. | Manual stopping times vary slightly between otherwise similar runs. | Averaging does not remove every source of error. |
| 57 | Systematic error | An error component that remains constant or changes predictably under repeated conditions. | A persistent +2 g offset shifts the readings upward. | Repeated readings can reproduce the same error. |
| 58 | Parallax error | A reading error caused by viewing a scale from an unsuitable angle. | A tilted view makes the meniscus appear against the wrong mark. | Use the instructed eye position, not a guessed correction. |
| 59 | Instrumental drift | A gradual change in an instrument’s indication unrelated to the intended measured change. | The baseline slowly shifts during an otherwise unchanged check. | A changing display need not mean the sample changed. |
| 60 | Reliability | Dependability of evidence or a method when relevant checks are considered. | Documented repeats and checks increase confidence in a result. | Say which checks help; reliability is not automatic truth. |
Recording and summarising data
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 61 | Qualitative data | Descriptions of qualities or categories rather than numerical measurements. | The sample is described as cloudy rather than clear. | Qualitative observations can still be useful evidence. |
| 62 | Quantitative data | Information expressed as counts or numerical measurements. | The recorded length is 14.2 cm. | State the quantity and unit where appropriate. |
| 63 | Raw data | The original recorded observations or readings before processing. | The notebook retains all four individual timings. | Do not replace inconvenient readings with preferred values. |
| 64 | Processed data | Information obtained by calculations or other operations on raw data. | A mean calculated from repeated timings is processed data. | Keep enough detail to check the calculation. |
| 65 | Data table | An organised arrangement of related observations or measurements. | Each row records one setting and its repeated results. | A table still needs meaningful labels and units. |
| 66 | Heading | A label explaining what a table column or row contains. | The column is headed “Time / s”. | A heading such as “Results” alone is too vague. |
| 67 | Repeated reading | Another measurement of the same quantity under stated conditions. | The length is read again without changing the object. | Copying the first number is not a repeat measurement. |
| 68 | Mean | The sum of numerical values divided by the number of values. | The mean of 12.2, 12.4 and 12.3 s is 12.3 s. | A mean can be strongly affected by an unusual value. |
| 69 | Median | The middle value when data are ordered, or the mean of the two middle values. | The median of 3, 5 and 9 is 5. | Order the values before finding the middle. |
| 70 | Outlier | A value unusually distant from the other data or the expected pattern. | A timing of 18.7 s stands apart from values near 12.3 s. | Investigate it; unusual does not automatically mean wrong. |
Graph vocabulary
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 71 | Horizontal axis | The left-to-right reference line on a conventional graph. | The chosen mass is placed on the horizontal axis. | Follow the question’s instructions about axis placement. |
| 72 | Vertical axis | The up-and-down reference line on a conventional graph. | The measured extension is placed on the vertical axis. | Label the measured quantity and its unit. |
| 73 | Origin | The point where the coordinate values are both zero on the usual axes. | A direct-proportion line passes through the origin. | Not every graph or relationship should pass through it. |
| 74 | Graph scale | The numerical values assigned to distances along an axis. | Each large square represents 50 g on the mass axis. | Equal spacing must match equal increments on a linear axis. |
| 75 | Coordinate | A value locating a point relative to a graph’s axes. | The point has horizontal coordinate 50 g and vertical coordinate 2 cm. | Check which quantity belongs to each axis. |
| 76 | Data point | A plotted representation of a pair of recorded values. | The result at 100 g and 4 cm becomes one data point. | A plotted point represents a measurement, not certainty. |
| 77 | Line of best fit | A line selected to represent the overall relationship in plotted data. | A straight line summarises readings that follow an approximately linear pattern. | It need not pass through every point or the origin. |
| 78 | Trend | An overall direction or pattern in how values change. | Extension increases as the selected mass increases in the supplied data. | Describe the trend before proposing a cause. |
| 79 | Gradient | Change in the vertical quantity divided by change in the horizontal quantity. | A 4 cm rise over 100 g gives 0.04 cm/g. | Use axis quantities and units, not only square counts. |
| 80 | Extrapolation | An estimate beyond the range of measured data using an assumed relationship. | Extending a pattern measured to 150 g to predict 300 g is extrapolation. | The relationship may change beyond the tested range. |
Interpreting relationships
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 81 | Correlation | An association in which variables tend to change together. | More umbrellas and wetter paths occur together in an imagined observation. | Association alone does not establish cause. |
| 82 | Causation | A relationship in which one factor contributes to producing a change. | A controlled comparison can help test whether the changed factor affects the outcome. | Consider other explanations before asserting causation. |
| 83 | Relationship | A connection between quantities, features or events. | The graph shows a relationship between mass and extension. | Specify the connection instead of saying only “related”. |
| 84 | Pattern | A recognisable arrangement or regularity in data. | The readings rise by similar amounts at each equal interval. | A pattern can include variation and exceptions. |
| 85 | Direct proportion | A relationship with a constant ratio between two quantities. | In the model data, doubling mass doubles extension. | The corresponding straight line passes through the origin. |
| 86 | Inverse proportion | A relationship in which the product of two quantities stays constant. | For a fixed distance, doubling constant speed halves travel time. | Any decreasing graph is not automatically inverse proportion. |
| 87 | Interpolation | An estimate within the range of measured values using the observed relationship. | Estimating extension at 75 g lies between measurements at 50 and 100 g. | An estimate remains dependent on the assumed pattern. |
| 88 | Anomaly | An observation or feature that does not fit the expected pattern. | One setting produces a result unlike the nearby settings. | Check the observation and the expectation. |
| 89 | Validity | How well the design or evidence addresses the intended question. | Timing a different route does not validly compare the two methods for the same route. | A repeatable measurement can address the wrong question. |
| 90 | Generalisation | A broader statement extended beyond the particular cases observed. | A finding about one trolley is applied to all trolley designs. | The extension may require additional evidence. |
Methods, practical judgement and reporting
| No. | Term | Meaning | Example | Check |
|---|---|---|---|---|
| 91 | Method | The overall approach used to investigate a question. | Compare measured outcomes at several settings while controlling relevant factors. | A method needs a clear connection to the question. |
| 92 | Procedure | The specific steps followed within the method. | The record states when timing began and ended. | “Do the experiment” is not a reproducible procedure. |
| 93 | Risk assessment | Consideration of possible harm and how it will be controlled. | The teacher identifies which handling precautions the practical requires. | Use the school’s assessment rather than inventing home procedures. |
| 94 | Hazard | Something with the potential to cause harm. | A damaged glass container is a hazard in a laboratory. | A hazard and the likelihood of harm are different. |
| 95 | Risk | The likelihood and possible severity of harm in the circumstances. | The risk depends on how damaged glass could be encountered or handled. | Risk changes when conditions and precautions change. |
| 96 | Precaution | An action taken to reduce a specified risk or measurement problem. | Reporting damaged apparatus allows the teacher to arrange safe replacement. | State the problem addressed by the precaution. |
| 97 | Observation record | A written, drawn or recorded account of what was observed. | The notebook separates measured values from later explanations. | Record uncertainty or missing observations honestly. |
| 98 | Supported explanation | An account that links relevant evidence with appropriate scientific reasoning. | The answer uses the readings and controlled conditions to explain its conclusion. | A longer answer is not necessarily better supported. |
| 99 | Model | A simplified representation used to describe, explain or predict a system. | A straight-line model represents the pattern within the measured range. | A model has limits and is not the whole system. |
| 100 | Ethical practice | Responsible and honest conduct in collecting, using and reporting evidence. | The group records an unexpected result and acknowledges help received. | Never invent results to match a prediction. |
The measurement distinctions are informed by the JCGM International Vocabulary of Metrology definition of accuracy and its separate account of precision. They are explained here in student-friendly language. The practical question is always what a result can tell us, which checks have been made and what remains uncertain.
CHAPTER 3 OF 9 · Back to contents
3. Measure the intended quantity and keep the units meaningful
Worked example 1: quantity, instrument and unit
An invented worksheet asks for the mass of a sample, the volume it occupies and the time it takes to travel between two marks. These require three different measurements. The balance is used for mass, a suitable volume method for volume, and a timing instrument for duration. A ruler reading cannot answer a mass question just because it produces a number.
The NIST guide to SI units identifies kilogram, metre, second and kelvin as the base units for mass, length, time and thermodynamic temperature respectively. Students may appropriately record small masses in grams and classroom temperatures in degrees Celsius. Distinguish “a suitable unit for this task” from “the SI base unit”. Also preserve letter case: mL means millilitres, whereas ML means megalitres.
Worked example 2: displacement and density
In an invented measurement record, the water reading rises from 28 mL to 39 mL after a suitable solid is completely immersed. The solid’s volume is 39 − 28 = 11 mL, equivalent to 11 cm³. Its measured mass is 33 g, so density = 33 ÷ 11 = 3 g/cm³. NIST’s volume guide explains cubic volume units and the relation between litres and cubic measures.
The method depends on conditions: the object must be fully immersed, and dissolving, water absorption, trapped bubbles or splashing could affect the interpretation. The answer should not mechanically subtract any two readings supplied. Ask what changed between them. These figures teach the reasoning; use the teacher’s chosen apparatus and instructions for actual practical work.
A final unit check catches many avoidable mistakes. A density answer needs mass per volume, not grams alone. A duration needs seconds or another stated time unit, not a clock time. Extra calculator digits do not create better measurements. Record and round to the precision requested by the teacher or justified by the supplied information.
CHAPTER 4 OF 9 · Back to contents
4. Design a comparison that answers one clear question
Worked example 3: name the variables and explain the controls
Imagine a teacher-led trolley investigation asking how ramp height affects travel time over a fixed marked section. Ramp height is the independent variable. Travel time is the dependent variable. Relevant controls include the trolley, marked distance, track surface, release position and release method. Naming “trolley” is only a start; explain that the same trolley is used so differences between models do not become an alternative explanation.
The timing rule also needs to be consistent. If one trial starts when the trolley is released and another starts when it reaches a later mark, the recorded durations describe different events. Even identical stopwatches cannot repair that design problem. Specify the start and finish events before comparing the results.
Now imagine the group raises the ramp and changes to a lighter trolley at the same time. A shorter time cannot confidently be attributed to the height change alone. There are two deliberate differences. Repeating this mixed comparison ten times may make its result consistent, but does not isolate the contribution of either change.
A control setup is not the same as a controlled variable
A controlled variable is a factor kept constant. A control setup is a comparison arrangement, such as an otherwise similar sample without the treatment being investigated. Some school investigations need a separate control setup; others compare several values of one variable without using a distinct untreated setup. Do not add “use a control” to every answer without identifying what comparison it provides.
A strong planning answer is specific enough for someone to understand the comparison. “Keep everything the same” is too vague because the independent variable must change. “Keep the trolley and measured travel distance the same, but change the ramp height” communicates the intended design. The teacher may identify further conditions that matter in the actual apparatus.
CHAPTER 5 OF 9 · Back to contents
5. Distinguish accuracy, precision and dependable evidence
Worked example 4: closely grouped readings can be wrong
Suppose a reference object has a trusted mass of 100.0 g for this exercise. Balance A gives 103.0, 103.1 and 103.2 g. Balance B gives 99.9, 100.0 and 100.1 g. Both sets are closely grouped, but B’s readings lie much closer to the reference. A’s consistency therefore does not establish its accuracy.
A child who writes “A is accurate because the readings are almost the same” has confused closeness to one another with closeness to the reference. Repair the distinction by asking two separate questions: “How much do the repeated values vary?” and “How close are they to the reference?” The first concerns precision; the second concerns accuracy.
Worked example 5: repetition does not remove a constant offset
An empty balance shows +2 g. If appropriate checks establish a constant +2 g offset over the relevant working range, a displayed 58 g corresponds to a corrected value of 56 g. Record the raw reading, the correction and its reason. Do not quietly replace the observation or assume that every instrument error is a constant offset.
The JCGM account of calibration distinguishes calibration from adjustment. Setting an empty balance to zero is not by itself a complete calibration. For school work, follow the instrument instructions and the teacher’s correction method. The important idea is that checking a reference can reveal something that repeated readings of an unknown object cannot.
Repeatability, reproducibility and uncertainty
Repeatability conditions keep the procedure, operator, measuring system and other specified conditions the same over a short period. Reproducibility conditions include specified changes such as different locations, operators and measuring systems. In a school discussion, say what was changed and what remained the same instead of treating the two words as interchangeable labels.
Reliability is a broader classroom term for dependability. Explain which evidence makes a result more dependable: sensible repeats, instrument checks, a clear method or agreement under a relevant comparison. No single repeated value proves that the method answers the intended question. Measurement uncertainty communicates a quantified spread of values associated with the result; it is not a confession that the student was careless.
CHAPTER 6 OF 9 · Back to contents
6. Process data without hiding inconvenient results
Worked example 6: a mean and an unusual reading
Three invented timings are 12.2, 12.4 and 12.3 s. Their sum is 36.9 s and their mean is 36.9 ÷ 3 = 12.3 s. A fourth recorded timing is 18.7 s. Including all four gives (12.2 + 12.4 + 12.3 + 18.7) ÷ 4 = 13.9 s. The difference between these means shows why an unusual observation deserves attention.
It does not tell us why 18.7 s occurred. It might reflect a recording problem, a timing difference, a changed condition or a real outcome. NIST’s guidance on outliers recommends investigating unusual values rather than simply deleting them. Preserve the original data and discuss any exclusion with the teacher, giving a clear reason. Removing a result merely because it spoils the expected pattern is not a valid correction.
Ordered values are 12.2, 12.3, 12.4 and 18.7 s, so the median is (12.3 + 12.4) ÷ 2 = 12.35 s. The median describes the middle differently from the mean. It does not make the unusual observation disappear or explain its cause. Use the summary the question requests, and discuss the data’s features when relevant.
Worked example 7: graph scale, gradient and proportion
Consider invented model data with masses of 0, 50, 100 and 150 g and corresponding extensions of 0, 2, 4 and 6 cm. Place mass on the horizontal axis and extension on the vertical axis, label both units and use evenly spaced values on each linear scale. The points follow a straight line through the origin.
The gradient between the points at 50 g and 150 g is (6 − 2) ÷ (150 − 50) = 0.04 cm/g. The constant extension-to-mass ratio for the positive values, together with the origin, is consistent with direct proportion in this model. A line that rises while crossing the vertical axis above zero would not express direct proportion, even if it were perfectly straight.
Interpolating at 75 g gives 3 cm if the line represents the relationship between the measured points. Extending the same line to 300 g predicts 12 cm, but this is extrapolation beyond the supplied range. A real object might behave differently there. A neat graph does not authorise an unlimited prediction or a physical test beyond the teacher’s approved conditions.
CHAPTER 7 OF 9 · Back to contents
7. Explain what the evidence supports
Worked example 8: association is not enough for a cause
Imagine an observation record showing more umbrellas on days when paths are wetter. That is an association. It does not establish that umbrellas cause the wet paths. Rain is a plausible factor connected to both observations. The purpose of the example is to practise alternative explanations before claiming causation.
A controlled intervention can strengthen a causal argument by deliberately changing a factor while limiting other explanations. Even then, describe the tested conditions and the remaining limitations. “Our findings support the idea that…” is often more accurate than “We proved this is always true.” A conclusion should answer the question without expanding one small investigation into a universal law.
Write an answer with a visible reasoning chain
A useful answer identifies the relevant result, links it to the comparison and explains the conclusion. For the trolley example, a student might write, “At the greater height, the recorded time was shorter. The same trolley and marked distance were used, so these factors do not explain the difference between those trials.” Further reasoning should use the concepts taught in that chapter, rather than add unfamiliar terminology for effect.
Separate an improvement to measurement from an improvement to design. A clearer start marker could reduce ambiguity in timing. More independent trials could reveal variation. A different comparison setup might address a confounding factor. “Repeat more times” does not perform all three jobs.
Practical judgement belongs in the same reasoning chain. Name the particular hazard, follow the teacher’s precautions and report damaged or unsuitable equipment. These vocabulary examples support understanding of a supervised lesson. Home study can use diagrams and supplied data without recreating electrical, heating or chemical procedures.
CHAPTER 8 OF 9 · Back to contents
8. Check understanding with original practice questions
Answer before looking at the explanations. These are learning exercises, not SEAB specimens or a school marking scheme. The numerical data are invented. Where a question asks for a limitation or an improvement, explain exactly what it affects.
Questions
- A student changes the width of a paper strip and measures how much it bends under the same load. Identify the independent variable, dependent variable and one relevant controlled variable.
- A measuring cylinder has five equal spaces between marks labelled 30 mL and 40 mL. What volume does each space represent?
- A suitable immersed solid changes the water reading from 42 mL to 57 mL. Its mass is 45 g. Find its volume and density.
- A trusted reference is 50.0 g. Repeated readings are 53.1, 53.0 and 53.1 g. Explain what the results suggest about precision and accuracy.
- Find the mean of 8.1, 8.4 and 8.7 s. State why the mean alone cannot show whether the instrument had a constant offset.
- A graph shows (20 g, 1 cm) and (100 g, 5 cm) on a straight best-fit line. Find the gradient in cm/g.
- Data were collected between 20 g and 100 g. Is a prediction at 60 g interpolation or extrapolation? What about a prediction at 140 g?
- Four readings are close together and a fifth is far away. Explain why “delete the fifth reading” is not an adequate first response.
- Two groups obtain similar measurements but use different rooms, operators and instruments. Which measurement-quality idea is relevant, and what should they document?
- A pupil repeats an unfair comparison with two changed factors twenty times. Explain why the repetition does not isolate the effect of one factor.
Answers and explanations
- Width is independent; the amount of bending is dependent. Relevant controls include the material, length, thickness, load and arrangement of supports. State how the chosen control stays the same. The exact set depends on the supplied apparatus.
- 2 mL per space. The 10 mL difference is divided by five equal spaces: 10 ÷ 5 = 2. Count the intervals, not merely the number of lines.
- 15 cm³ and 3 g/cm³. Volume = 57 − 42 = 15 mL = 15 cm³. Density = 45 ÷ 15 = 3. The calculation assumes the displacement method is suitable and the recorded rise represents the object’s volume.
- The readings are closely grouped but about 3 g above the reference. They suggest good precision under those conditions but poor accuracy relative to the reference. Repetition has not removed the offset.
- 8.4 s. The sum is 25.2 s and 25.2 ÷ 3 = 8.4 s. A common offset could affect all three readings and remain in their mean.
- 0.05 cm/g. Gradient = (5 − 1) ÷ (100 − 20) = 4 ÷ 80 = 0.05. The units follow the vertical quantity divided by the horizontal quantity.
- 60 g is interpolation; 140 g is extrapolation. The distinction concerns whether the prediction lies inside or outside the measured range. Both rely on a suitable assumed relationship.
- Investigate and retain the original record. Check the method, conditions, transcription and instrument. The unusual value may reveal an error or a genuine difference. Any exclusion needs a justified, recorded reason.
- Reproducibility is relevant. Record the changed and unchanged conditions, measurement procedure and results. Similar numbers without clear information about the methods are harder to interpret.
- Two factors remain changed together. Repeats may show that the mixed comparison gives consistent results, but cannot identify which change produced the effect. Redesign the comparison to separate the factors.
Use mistakes diagnostically. An incorrect unit conversion needs a different repair from a confused variable. A calculation slip does not necessarily mean the scientific idea is missing. Ask the student to explain the reasoning aloud, locate the first uncertain step and attempt a changed example after correction.
CHAPTER 9 OF 9 · Back to contents
9. Build a manageable learning routine and choose useful support
A small routine for a busy school week
Select four or five terms from the current topic. On the first session, connect each term to a diagram or a recent question. On the next, cover the definitions and explain the distinctions aloud. Later, answer one question with changed numbers or conditions. Finally, revisit one earlier error without looking at its correction. Adjust the pace to the child’s workload; the number of sessions is a suggested routine, not a prescribed learning schedule.
For a student falling behind, begin with a single distinction such as mass versus volume or independent versus dependent variable. For a student maintaining progress, mix a few older terms into the current chapter. For a student ready to extend, ask what evidence would weaken the conclusion or what conditions an extrapolation assumes. The next step should match the difficulty visible in the work.
Punggol parents do not need specialist equipment to help. A printed scale, an invented data table or a diagram is sufficient for a short discussion. Ask, “What exactly was measured?”, “Which value is the reference?” or “What else might explain the result?” When the science is unclear, note the question for the school teacher rather than turn the discussion into an argument about who should know the answer.
What a useful three-student lesson could do
In a proposed three-pax tutorial, each student first identifies the variables in a short investigation independently. The tutor checks whether the difficulties concern terminology, the diagram or the experimental design. During discussion, one learner proposes a conclusion, another identifies supporting evidence and the third checks a limitation. Rotate these roles, then give each learner a fresh individual question.
Parents can ask how the tutor records misconceptions, checks independent understanding and connects the lesson to current school work. Confirm the provider’s subject provision, grouping and timetable directly. The lesson model above describes a useful teaching approach; it does not assert current availability or promise a particular result.
Does my child need to memorise all 100 entries?
No. Use the list as a reference, beginning with language needed now. Some measurement terms have precise professional meanings that a young learner will develop gradually. A student who can distinguish accuracy from precision in a new example has made more useful progress than one who can recite both definitions but applies them interchangeably.
Does a repeatable result prove that an experiment is valid?
No. A method can repeatedly measure the wrong quantity or compare conditions unfairly. Check whether the design addresses the question and whether alternative explanations have been considered. Consistency is valuable evidence about one aspect of the work; it is not a substitute for an appropriate design.
Should every unusual result be ignored when calculating a mean?
No. Keep the record and investigate the reason. Follow the question’s instructions or the teacher’s justified treatment of the data. If an exclusion is appropriate, make it visible and explain it. Scientific honesty includes acknowledging when the available information does not establish why a result differs.
How can we tell whether the vocabulary work is helping?
Look for better choices: the student identifies the measured quantity, includes the unit, selects a relevant control, explains what repeats check and limits a conclusion appropriately. Compare independent attempts at similar tasks over time. The goal is a clearer method of thinking and communicating, rather than a larger collection of copied definitions.
Primary sources and scope
Source check: 16 November 2026. Official references establish the curriculum context, units and measurement terminology. The numbered selection, situations, data, questions and routines are original teaching material. The technical references clarify meanings; they do not add compulsory content to a school’s Secondary 1 programme.
- MOE G2/G3 Lower Secondary Science syllabus, updated April 2024 — school curriculum foundation.
- SEAB Secondary Education Certificate — later national examination context.
- NIST SI units and SI units: volume — quantities, symbols and volume relationships.
- JCGM International Vocabulary of Metrology: accuracy, precision, repeatability conditions and reproducibility conditions.
- JCGM measurement references: uncertainty and calibration.
- NIST Engineering Statistics Handbook: detection of outliers — investigating unusual observations.

