Working Scientifically
KS2SC-KS2-D001
Practical scientific methods, processes and skills taught through and integrated with substantive science content. In Lower KS2 (Y3-4), pupils broaden enquiry skills including fair tests, standard measurements, data recording and drawing conclusions. In Upper KS2 (Y5-6), pupils plan enquiries, control variables, use increasing precision, and evaluate evidence. Must not be taught as a separate strand.
National Curriculum context
Working scientifically at KS2 underpins all science learning and requires pupils to use a range of scientific methods and processes, progressing from the exploratory approaches of KS1 to systematic, fair-test investigations. Pupils develop the ability to plan different types of scientific enquiry — including observing over time, pattern seeking, classifying, comparative and fair testing, and researching using secondary sources — selecting the most appropriate method for the question being asked. The statutory curriculum requires pupils to recognise that scientific evidence is used to support or refute ideas, to take measurements using standard units, to record and present data in a variety of graphical forms, and to draw conclusions consistent with their evidence. By the end of KS2, pupils should be able to report and present findings from enquiries using appropriate scientific language and representations.
10
Concepts
4
Clusters
7
Prerequisites
10
With difficulty levels
Lesson Clusters
Ask scientific questions and choose the right type of enquiry
introduction CuratedSelecting appropriate enquiry type (including fair and comparative testing) is the gateway skill for all KS2 scientific investigation. Co_teach_hints link C001 and C002 directly.
Plan investigations by identifying and controlling variables
practice CuratedVariable control and precision/repeat measurement are the two features that distinguish KS2 fair testing from KS1 simple testing. They are always taught together in investigation planning.
Measure systematically and record data in a variety of formats
practice CuratedSystematic measurement, data classification/presentation, and complex data recording form a progression from basic to advanced data-handling skills. Co_teach_hints link C001, C004, and C009 together.
Draw conclusions, evaluate evidence, and communicate scientific findings
practice CuratedDrawing conclusions (C005), scientific reporting (C006), and causal relationship/evidence evaluation (C010) are the three sense-making and communication skills that close every enquiry cycle at KS2.
Access and Inclusion
5 of 10 concepts have identified access barriers.
Barrier types in this domain
Recommended support strategies
Prerequisites
Concepts from other domains that pupils should know before this domain.
Concepts (10)
Relevant Questioning and Enquiry Selection
process AI FacilitatedSC-KS2-C001
The ability to ask relevant scientific questions and select the most appropriate type of scientific enquiry to answer them (observation over time, pattern seeking, classifying, fair testing, secondary research). Develops from KS1 simple questioning to active enquiry design.
Teaching guidance
Introduce the five types of scientific enquiry explicitly: observation over time, pattern seeking, classifying and grouping, comparative and fair testing, and researching using secondary sources. Present pupils with a question and ask them to decide which type of enquiry would best answer it. Use a 'question sorter' display where pupils categorise questions by enquiry type. Model asking relevant questions by connecting to the content being studied — for example, during a rocks topic ask 'Which rock is the most permeable?' (fair test) versus 'How are fossils formed?' (secondary research).
Common misconceptions
Pupils often believe that all scientific questions must be answered by carrying out an experiment. They need explicit teaching that observation over time, classification, pattern seeking and secondary research are equally valid enquiry types. Some pupils confuse 'relevant' questions with 'interesting' questions — relevant questions are those that can be investigated scientifically.
Difficulty levels
Asking simple questions about a scientific topic and suggesting one way to find the answer, with teacher support.
Example task
We are learning about magnets. Think of a question we could investigate. How could we find the answer?
Model response: Which materials are magnetic? We could test different materials with a magnet and see which ones stick.
Asking relevant scientific questions and beginning to identify which type of enquiry (fair test, observation, classification, research) would best answer them.
Example task
Here are three questions: 'Which material is the best insulator?', 'How does a tadpole change into a frog?', 'How can we group these rocks?' Which type of enquiry would you use for each?
Model response: Best insulator — fair test (change the material, measure the temperature). Tadpole to frog — observation over time (watch and record changes over weeks). Grouping rocks — classifying and grouping (look at properties and sort them).
Independently asking relevant scientific questions and selecting the most appropriate type of enquiry from the five types (observation over time, pattern seeking, classifying, fair testing, secondary research).
Example task
We are studying sound. Write three different questions about sound and for each one explain which type of enquiry you would use and why.
Model response: 1. 'Does the length of a ruler hanging over a desk affect the pitch of the sound it makes?' — Fair test, because I can change one variable (length) and measure the pitch. 2. 'Is there a pattern between the size of an instrument and the pitch of its sound?' — Pattern seeking, because I am looking for a relationship. 3. 'How do different animals hear sound?' — Secondary research, because I cannot investigate this practically.
Evaluating questions for investigability, refining vague questions into precise testable ones, and justifying enquiry type selection with reference to the nature of the question.
Example task
A pupil asks 'Is exercise good for you?' Explain why this question is hard to investigate as written. Rewrite it as a question that could be answered by a specific type of enquiry.
Model response: The question is too vague — 'good for you' is not measurable and 'exercise' is not specific. A better question: 'Does 5 minutes of star jumps increase a person's heart rate?' This can be answered by a fair test — measure heart rate before and after, keeping the person, time of day and number of jumps the same. Or 'Is there a pattern between the number of minutes of exercise and heart rate increase?' for pattern seeking. Good scientific questions need a measurable outcome and a clearly defined variable.
Delivery rationale
Science fair test concept — requires physical apparatus and variable control, but AI can structure the enquiry sequence.
Access barriers (2)
KS2 working scientifically involves planning and conducting investigations with multiple steps: question, prediction, method, variables identification, data collection, recording, analysis, conclusion. This is a 7-8 step procedure.
Planning investigations requires children to design their own method — deciding what to measure, how to measure it, and what to keep the same. This is a high-level open-ended task demanding both scientific understanding and executive function.
Fair and Comparative Testing
process AI FacilitatedSC-KS2-C002
Setting up simple practical enquiries as fair tests, recognising and controlling variables. Understanding that a fair test changes only one variable at a time while keeping others constant, enabling valid comparisons.
Teaching guidance
Begin with comparative tests (e.g., 'Which surface does the car travel furthest on?') before introducing fair tests with controlled variables. Use a planning framework that explicitly asks: 'What will I change? What will I measure? What will I keep the same?' Provide practical contexts such as testing which paper towel is most absorbent or which material is best for insulating a cup of warm water. Encourage pupils to predict outcomes before testing and to explain why the test must be fair to get reliable results.
Common misconceptions
Pupils frequently change more than one variable at a time without recognising this invalidates the comparison. Some pupils believe that a test is 'fair' simply because everyone in the group gets a turn, rather than understanding it as controlling variables. Pupils may also confuse prediction with guessing — a prediction should be based on existing knowledge or prior observations.
Difficulty levels
Carrying out a simple comparison of two options to find which is better for a purpose, keeping one condition the same with teacher support.
Example task
Test which of these two surfaces a toy car rolls furthest on. Make sure you start the car from the same place each time.
Model response: I pushed the car from the top of the ramp each time. On the smooth floor it went 95cm, on the carpet it went 42cm. The smooth floor let the car go further.
Setting up a fair test by identifying what to change, what to measure and what to keep the same, with some independence.
Example task
Plan a fair test to find out which paper towel is the most absorbent. Name your variables.
Model response: I will change: the brand of paper towel. I will measure: how much water each towel absorbs (by weighing it wet). I will keep the same: the size of each piece, the amount of water, and the soaking time. This makes it fair because the only thing different is the towel brand.
Planning and carrying out a fair test with controlled variables, making repeat measurements, and drawing conclusions from the results.
Example task
Plan and carry out a fair test to find out which material is the best thermal insulator. Wrap cups of warm water in different materials and measure the temperature over 20 minutes.
Model response: Independent variable: insulating material (wool, foil, bubble wrap, no insulation). Dependent variable: temperature after 20 minutes. Control variables: same volume of water, same starting temperature, same cup type, same room conditions. I tested each one three times and calculated the mean. Wool kept the water warmest (52°C mean), then bubble wrap (48°C), then foil (44°C), then no insulation (38°C). Wool is the best insulator because it trapped the most heat.
Designing a fair test for a complex question, evaluating the reliability of results, identifying anomalies, and suggesting improvements.
Example task
You tested which material was the best insulator but one of your results seemed very different from the others. What should you do? How could you improve the investigation?
Model response: The anomalous result could be due to an error — maybe the thermometer was not read at the right time, or the material was not wrapped the same way. I should check if the result is truly anomalous by comparing it with my other repeat measurements. If it is very different, I should discard it and use the remaining results to calculate the mean. To improve: I could use data loggers for more precise temperature readings, ensure all materials have the same thickness, and increase from three to five repeats for more reliable averages.
Delivery rationale
Science fair test concept — requires physical apparatus and variable control, but AI can structure the enquiry sequence.
Access barriers (1)
Extended scientific investigation requires sustained focus across multiple phases: setting up, observing, recording, and analysing — often over 30+ minutes. Children with ADHD may lose focus between phases.
Systematic Measurement with Standard Units
skill AI FacilitatedSC-KS2-C003
Making systematic and careful observations using standard units (e.g., degrees Celsius, metres, grams). Using a range of equipment including thermometers and data loggers. Increasing accuracy over Lower KS2.
Teaching guidance
Teach pupils to select appropriate equipment for the measurement being taken — rulers for length, thermometers for temperature, measuring cylinders for liquid volume, balances for mass. Practise reading scales with different intervals (ones, twos, fives, tens). Introduce data loggers for temperature and light investigations, demonstrating how they capture readings automatically over time. Link measurement to maths by reinforcing standard units (cm, m, g, kg, ml, l, °C) and by practising estimation before measuring.
Common misconceptions
Pupils commonly misread scales by counting lines rather than intervals, leading to systematic errors. Some pupils believe that digital instruments are always more accurate than analogue ones. Children may not understand that reading a thermometer requires waiting for the liquid to stop moving before recording the value.
Difficulty levels
Making simple measurements using standard units with teacher support, such as reading a ruler in centimetres.
Example task
Measure the length of this leaf using a ruler. Give your answer in centimetres.
Model response: The leaf is 8cm long.
Using appropriate equipment (rulers, thermometers, measuring cylinders) to take measurements in standard units with increasing accuracy.
Example task
Measure the temperature of this cup of water using a thermometer. Wait until the reading is steady before you record it.
Model response: I placed the thermometer in the water and waited for the red liquid to stop moving. The temperature is 42°C.
Selecting the most appropriate equipment for different measurements, reading scales accurately including those with intervals of 2, 5 or 10, and recording with correct units.
Example task
You need to measure the mass of a rock, the temperature of water and the volume of liquid in a beaker. Which equipment would you use for each? Take the measurements.
Model response: Mass: balance scales — the rock is 145g. Temperature: thermometer — the water is 23°C. Volume: measuring cylinder — there is 75ml of liquid (I read it at the bottom of the curved surface at eye level). I chose each piece of equipment because it measures the right quantity in standard units.
Explaining the importance of measurement precision, choosing equipment with appropriate sensitivity, and using data loggers for continuous measurements.
Example task
We want to measure how quickly a cup of water cools over one hour. Should we use a thermometer or a data logger? Why? What would a data logger show that a thermometer cannot?
Model response: A data logger would be better because it takes readings automatically at regular intervals (every 30 seconds, for example) without anyone needing to be there. It records every measurement precisely and we can download the data as a table or graph. A thermometer would only give us readings when someone looks at it, and humans might read it slightly differently each time. The data logger would show the exact pattern of cooling — whether it cools at a steady rate or faster at first and slower later. A thermometer could miss that detail if we only read it every 5 minutes.
Delivery rationale
Science skill involving measurement/practical work — AI structures, facilitator supervises.
Access barriers (2)
Recording and presenting scientific data requires drawing tables, charts, labelled diagrams, and writing conclusions. The recording demand can dominate the session if the child has fine motor difficulties.
Scientific recording at KS2 involves substantial writing: copying table headings, recording observations, writing up methods and conclusions. When science is the learning objective, the handwriting load should be minimised through templates.
Data Classification and Presentation
skill AI FacilitatedSC-KS2-C004
Gathering, recording, classifying and presenting data in a variety of ways: simple scientific language, labelled diagrams, classification keys, bar charts, and tables. Moving from recording towards analytical presentation.
Teaching guidance
Teach each data presentation format explicitly: tables with headings and units, bar charts with labelled axes, and classification keys. Use sorting and grouping activities with real specimens (leaves, rocks, minibeasts) before introducing branching databases or keys. Model how to construct a bar chart step by step — choosing a scale, labelling axes, drawing bars accurately. Encourage pupils to select the most appropriate format for their data type (categories → bar chart, yes/no questions → classification key).
Common misconceptions
Pupils often confuse bar charts with pictograms or line graphs and may draw bars that do not start from the baseline. Some pupils think a table is simply writing numbers in any arrangement — they need to understand that tables have clear headings, units and consistent formatting. Children may also believe that more elaborate presentation (colour, decoration) makes data 'better'.
Difficulty levels
Recording results in a simple table with given headings, using scientific vocabulary with support.
Example task
Fill in this table to show which materials are magnetic and which are not. The headings are: Material | Magnetic? (yes/no)
Model response: Iron nail: yes. Wooden block: no. Copper coin: no. Steel paper clip: yes. Plastic ruler: no.
Creating simple tables with own headings, recording results accurately, and presenting data as a bar chart with labelled axes.
Example task
We measured how far a toy car rolled on four different surfaces. Design a table for the results and draw a bar chart.
Model response: Table with headings: Surface | Distance rolled (cm). Rows: carpet — 32cm, sandpaper — 28cm, wood — 78cm, tile — 85cm. Bar chart with title, labelled axes (Surface on x-axis, Distance in cm on y-axis), correct scale and four bars drawn accurately.
Choosing the most appropriate way to present data (table, bar chart, classification key, labelled diagram) depending on the data type, and using the presentation to identify patterns.
Example task
We tested how temperature affects the time for sugar to dissolve. Our data is: 20°C — 180s, 40°C — 95s, 60°C — 52s, 80°C — 28s. Choose the best way to present this data. What pattern does it show?
Model response: A line graph is best because both variables are continuous numbers. X-axis: temperature (°C), Y-axis: time to dissolve (seconds). The graph shows a clear pattern: as temperature increases, the dissolving time decreases. Sugar dissolves faster in hotter water. The decrease is steeper between 20°C and 40°C than between 60°C and 80°C.
Selecting and creating the most effective data presentation for the audience and purpose, interpreting graphs to draw conclusions and identify anomalies.
Example task
Two groups investigated the same question about dissolving rate and temperature but got slightly different results. How would you present both sets of data to compare them? How would you decide which is more reliable?
Model response: I would plot both sets of results on the same line graph using different colours, so we can see where they agree and where they differ. Points that are close together show the results are reliable. If one point is far from the general pattern and the other group's results, it is likely an anomaly caused by measurement error. The group with more consistent results (closer to a smooth curve) probably has more reliable data. We could also calculate the mean of both sets for each temperature to get an even more reliable picture.
Delivery rationale
Science skill involving measurement/practical work — AI structures, facilitator supervises.
Drawing Conclusions and Predictions
process AI FacilitatedSC-KS2-C005
Using results to draw simple conclusions, make predictions for new values, suggest improvements and raise further questions. Iterative development of scientific understanding through evidence-based reasoning.
Teaching guidance
Scaffold conclusion-writing with sentence starters: 'Our results show that...', 'This happened because...', 'If we changed X, I predict that...' Teach the difference between describing results (what happened) and explaining them (why it happened). Introduce prediction for new values by asking pupils to extend patterns — for example, 'If we tested a fourth surface, what do you think would happen based on our results?' Encourage pupils to identify when results are unexpected and to suggest reasons or further tests.
Common misconceptions
Pupils often restate their prediction as their conclusion regardless of what the data actually shows. They may confuse correlation with causation — for example, concluding that one thing caused another simply because they occurred together. Some pupils think that if their prediction was wrong, the investigation has 'failed', rather than understanding that disconfirming evidence is equally valuable in science.
Difficulty levels
Describing what happened in an investigation using simple language, stating whether the prediction was correct.
Example task
We predicted that the red car would roll furthest. It actually rolled the shortest distance. What happened?
Model response: My prediction was wrong. The blue car rolled the furthest, not the red one.
Drawing a simple conclusion from results, using evidence to answer the original question, and making a simple prediction for an untested value.
Example task
We tested how the number of elastic bands on a ball launcher affected the distance it fired a ball: 1 band — 30cm, 2 bands — 55cm, 3 bands — 78cm. What is your conclusion? Predict the distance for 4 bands.
Model response: More elastic bands make the ball go further because there is more force. I predict 4 bands would fire the ball about 100cm because the distance goes up by about 25cm each time.
Drawing evidence-based conclusions, identifying patterns in results, making predictions for new values, and suggesting improvements to the investigation.
Example task
We investigated how the height a ball is dropped from affects how high it bounces. Results: 50cm drop — 28cm bounce, 100cm — 54cm, 150cm — 80cm. Draw a conclusion, predict the bounce height from 200cm, and suggest how we could improve.
Model response: The higher the drop height, the higher the bounce. The bounce height is approximately 54% of the drop height. I predict a 200cm drop would give about 108cm bounce. To improve: we could use a slow-motion camera for more accurate bounce measurements, do more repeats for each height, and test more drop heights to confirm the pattern. We could also try different surfaces to see if the relationship changes.
Evaluating conclusions critically, distinguishing between what the data shows and what it does not, recognising the limits of the investigation, and raising further questions.
Example task
Our investigation showed that sugar dissolves faster in hotter water. Can we conclude that all solids dissolve faster in hotter water? Why or why not?
Model response: No, we can only conclude that sugar dissolves faster in hotter water because that is the only solid we tested. Other solids might behave differently — some substances actually become less soluble at higher temperatures. To make a broader conclusion, we would need to test many different solids (salt, bicarbonate of soda, chalk) at different temperatures. Scientific conclusions should not go beyond what the evidence supports. A good follow-up question would be: 'Do all soluble substances dissolve faster in hotter water, or only some?'
Delivery rationale
Science process concept — enquiry methodology benefits from structured AI guidance with facilitator.
Access barriers (2)
Drawing conclusions and reporting findings requires formal scientific language: 'The results show...', 'This suggests that...', 'The evidence indicates...'. This register is linguistically demanding for children with SLCN.
Scientific conclusion-writing requires self-generated explanatory text linking observations to scientific concepts. There is no single correct phrasing, which makes this task paralysing for children who need structured response formats.
Scientific Reporting and Communication
skill AI DirectSC-KS2-C006
Reporting on findings from enquiries through oral and written explanations, displays and presentations. Using scientific language and illustrations to communicate findings to different audiences.
Teaching guidance
Provide opportunities for both oral and written reporting. Use 'science conferences' where groups present findings to the class, encouraging questions from the audience. Teach the structure of a scientific report: question, method, results, conclusion. Model the use of scientific vocabulary in explanations and provide word banks. Encourage pupils to use labelled diagrams and data displays as part of their communication, not just written prose.
Common misconceptions
Pupils often think that writing more equals better science communication. They need to understand that clarity and precision matter more than length. Some pupils believe that scientific reports should only contain text, not realising that diagrams, tables and charts are essential components of scientific communication.
Difficulty levels
Telling the class or a partner what they found out in a simple investigation, using some scientific words.
Example task
Tell your partner what we found out about which material was the most waterproof.
Model response: We tested four materials by pouring water on them. Plastic was the most waterproof because no water went through. Paper was the worst because it soaked up the water.
Writing a simple scientific report with sections for question, method, results and conclusion, using scientific vocabulary.
Example task
Write a report about our investigation into which ball bounces highest. Include: question, what we did, results and conclusion.
Model response: Question: Which type of ball bounces highest from 1 metre? Method: We dropped each ball from 1m and measured the bounce height. We did three drops for each. Results: Table showing tennis ball — 58cm average, rubber ball — 72cm, ping pong ball — 65cm. Conclusion: The rubber ball bounced highest because rubber is very elastic and returns to its shape quickly.
Reporting findings using appropriate scientific language, illustrations and data displays, presenting to different audiences through oral and written formats.
Example task
Prepare a presentation of our electricity investigation for a school assembly. You need to explain what we discovered about conductors and insulators so younger children can understand.
Model response: We wanted to find out which materials let electricity flow through them. We built a test circuit with a gap and tested different materials. Materials that completed the circuit and lit the bulb are called conductors — metals like copper, iron and aluminium. Materials that did not complete the circuit are called insulators — plastic, wood and rubber. That is why electrical wires have metal inside (conductor) covered in plastic outside (insulator to keep us safe). Diagram showing the test circuit with labelled parts.
Producing a comprehensive scientific report that includes method evaluation, discusses reliability, and suggests further questions, using precise scientific language throughout.
Example task
Write a full scientific report on our parachute investigation. Include an evaluation of our method and suggestions for further work.
Model response: Title: Does canopy size affect parachute fall time? Hypothesis: Larger canopies will fall more slowly due to greater air resistance. Method: Made parachutes from plastic bags (25cm², 50cm², 75cm², 100cm²), same string length and mass. Dropped from 2m, timed with stopwatch, 3 repeats each. Results: Table and line graph showing fall time increases with canopy size. Conclusion: Larger canopies create more air resistance, slowing the fall. Evaluation: Timing with stopwatches introduced human error (reaction time ~0.3s). Could improve with light gates. The parachutes sometimes drifted sideways, affecting consistency. Further work: test different canopy shapes (circle vs square), different materials, or different masses to see if the pattern still holds.
Delivery rationale
Science data/analysis skill — graph interpretation and data handling are digitally deliverable.
Variable Control in Enquiry Planning
process AI FacilitatedSC-KS2-C007
Planning scientific enquiries by recognising and controlling variables where necessary. Understanding which variables to keep constant, which to change, and which to measure. Required for valid experimental design.
Teaching guidance
Use a structured planning template that requires pupils to explicitly name the independent variable (what I will change), the dependent variable (what I will measure), and the control variables (what I will keep the same). Provide practice in identifying variables across different contexts before asking pupils to plan independently. Use 'spot the error' activities where pupils critique a flawed investigation plan and identify uncontrolled variables. Connect variable control to the concept of reliability — explaining why uncontrolled variables make results untrustworthy.
Common misconceptions
Pupils often confuse the independent and dependent variables, or believe that 'keeping things the same' refers to getting the same results rather than controlling conditions. Some pupils think that controlling variables means nothing should change at all, not understanding that the independent variable must be deliberately changed. Upper KS2 pupils may struggle to identify hidden variables that they have not considered.
Difficulty levels
Identifying what was changed and what was measured in a given investigation, with teacher support.
Example task
We changed the height of the ramp and measured how far the car rolled. What was the thing we changed? What did we measure?
Model response: We changed the height of the ramp. We measured how far the car rolled.
Naming the independent variable (what to change), dependent variable (what to measure) and control variables (what to keep the same) for a given investigation.
Example task
We are testing whether the type of surface affects how quickly an ice cube melts. Name the three types of variables.
Model response: Independent variable (what I change): the type of surface (metal, wood, plastic). Dependent variable (what I measure): the time it takes for the ice cube to melt. Control variables (what I keep the same): the size of ice cube, the room temperature, the starting temperature of each surface.
Planning an investigation by explicitly identifying and controlling all relevant variables, explaining why each control variable matters for the validity of results.
Example task
Plan a fair test to find out whether the thickness of insulation affects how quickly a cup of hot water cools. Identify all variables and explain why each control variable must be kept the same.
Model response: Independent: thickness of insulation (1 layer, 2 layers, 3 layers, no insulation). Dependent: temperature of water after 15 minutes. Controls: same volume of water (keeps heat capacity the same), same starting temperature (so all start equal), same material for insulation (so we only test thickness, not material), same cup type (different cups might retain heat differently), same room temperature. Each control matters because if any were different, I could not be sure whether the thickness or the other factor caused the change.
Evaluating the variable control in an investigation design, identifying uncontrolled variables that could affect results, and explaining how this limits the conclusions.
Example task
A group tested whether exercise increases heart rate. They had five pupils do star jumps — but each pupil did a different number, some were already tired from PE, and they measured heart rate at different times after stopping. What problems can you identify?
Model response: There are several uncontrolled variables. Different numbers of star jumps means the exercise intensity was not the same — this should have been standardised. Some pupils being tired from PE means their resting heart rate was already elevated — they should all have rested first. Measuring at different times after stopping means some hearts had longer to recover. To fix this: all pupils should do the same number of jumps, all should rest for 5 minutes before starting, and heart rate should be measured at the same time after stopping (e.g. immediately and then at 1 minute). Without controlling these variables, we cannot be sure the results are valid.
Delivery rationale
Science process concept — enquiry methodology benefits from structured AI guidance with facilitator.
Precision and Repeat Measurements
skill AI DirectSC-KS2-C008
Taking measurements with increasing accuracy and precision using a range of scientific equipment. Understanding the value of repeat readings to improve reliability and identify anomalies.
Teaching guidance
Demonstrate the importance of repeat measurements by having different groups measure the same thing and comparing their results. Discuss why readings differ and how taking multiple measurements and calculating a mean improves reliability. Teach pupils to identify and discard anomalous results before calculating averages. Use contexts where precision matters — such as measuring the time for a pendulum swing or the distance a ball rolls — to motivate the need for careful, repeated measurement.
Common misconceptions
Pupils often confuse accuracy (closeness to the true value) with precision (consistency of repeat readings). Some pupils believe that taking more decimal places automatically makes a measurement more precise. Children may think that if their result differs from a classmate's, one of them must be 'wrong', rather than understanding natural variation in measurement.
Difficulty levels
Making a measurement carefully and recording it, understanding that taking care leads to better results.
Example task
Measure the length of this pencil. Make sure you read the ruler carefully and start from zero.
Model response: The pencil is 17.5cm long. I made sure to line up the end with the zero mark and read the ruler at eye level.
Taking the same measurement more than once and understanding that repeat measurements can improve reliability.
Example task
Measure how long it takes for a marble to roll down this ramp. Do it three times. Why might you get slightly different results each time?
Model response: Attempt 1: 2.3 seconds. Attempt 2: 2.5 seconds. Attempt 3: 2.4 seconds. The results are slightly different each time because of small differences in when I start and stop the stopwatch (reaction time) and small differences in exactly where I release the marble.
Taking repeat measurements, calculating the mean, identifying anomalous results, and explaining why repeat readings improve reliability.
Example task
You measured the time for a pendulum to make 10 swings. Your results: 12.3s, 12.5s, 12.4s, 15.1s, 12.3s. Calculate the mean. Is there an anomaly?
Model response: The fourth result (15.1s) is an anomaly — it is much higher than the others, probably because of a timing error. I should discard it and calculate the mean from the remaining four: (12.3 + 12.5 + 12.4 + 12.3) ÷ 4 = 12.375s, which rounds to 12.4s. Repeat measurements improve reliability because random errors tend to cancel out when averaged, giving a result closer to the true value.
Explaining the difference between accuracy and precision, selecting equipment to match the required precision, and evaluating whether results are reliable enough to support a conclusion.
Example task
Two groups tested how temperature affects dissolving time. Group A got: 180s, 95s, 52s, 28s (one test each). Group B got: 178s, 182s, 176s (three tests at 20°C only) with a mean of 178.7s. Which group has more reliable data for 20°C? Which has more useful data overall?
Model response: Group B is more reliable for 20°C because they repeated the measurement three times and the results are consistent (within 6 seconds of each other), which shows good precision. Group A tested more temperatures but only once each, so any single measurement could be inaccurate without repeats to check. However, Group A's data is more useful overall because it shows the pattern across temperatures. Ideally, you would combine both approaches — test at multiple temperatures with repeat measurements at each. Precision means consistent results; accuracy means close to the true value.
Delivery rationale
Science data/analysis skill — graph interpretation and data handling are digitally deliverable.
Complex Data Recording
skill AI FacilitatedSC-KS2-C009
Recording data and results of increasing complexity using scatter graphs, line graphs, bar charts, classification keys, tables and scientific diagrams. Selecting the most appropriate representation for different types of data.
Teaching guidance
Introduce scatter graphs and line graphs as tools for continuous data, distinguishing them from bar charts used for categorical data. Teach pupils to plot points accurately and, where appropriate, draw lines of best fit. Provide opportunities to select the most suitable graph type for different investigations. Use ICT tools (spreadsheets, graphing software) alongside hand-drawn graphs. Model how to read information from graphs, including interpolating between data points and identifying trends.
Common misconceptions
Pupils often join scatter graph points dot-to-dot rather than drawing a line of best fit. They may confuse line graphs (for continuous data) with bar charts (for categorical data), using the wrong type for their data. Some pupils believe that a graph must always go through the origin (0,0), not understanding that this depends on the data.
Difficulty levels
Recording data in a simple table or drawing with teacher-provided headings, using standard formats.
Example task
Record the results of our magnet test in this table. For each material, write whether it was magnetic or non-magnetic.
Model response: Table completed with correct results for each material tested, with consistent formatting (yes/no or magnetic/non-magnetic).
Choosing between a table, bar chart or classification key to record data, and constructing the chosen format with appropriate labels.
Example task
We surveyed the minibeasts in two habitats. Record our findings using the most appropriate format.
Model response: A table with habitats as columns and minibeast types as rows, showing tallied counts. Or a paired bar chart comparing counts in each habitat. Axes labelled, title included, scale appropriate.
Selecting the most appropriate data recording method for different data types (categorical vs continuous), constructing scatter graphs and line graphs for continuous data, and interpreting the patterns shown.
Example task
We measured the height of a plant every 3 days for a month. What type of graph should you draw? Create it and describe the pattern.
Model response: A line graph because both variables are continuous (time in days, height in cm). Points plotted accurately with a smooth curve or line of best fit. The graph shows rapid growth in the first two weeks, then the growth rate slowed down. By day 27, the plant was hardly growing at all. The pattern shows growth that starts fast then levels off.
Using multiple data representations to communicate complex findings, drawing and interpreting lines of best fit on scatter graphs, and using ICT tools alongside hand-drawn methods.
Example task
Present the results of our investigation into how arm span relates to height. We measured 25 pupils. Choose the best graph type and explain what the data shows.
Model response: A scatter graph is best because we are looking for a relationship between two continuous measurements. Each pupil is a point (x = arm span, y = height). The line of best fit shows a positive correlation — as arm span increases, height tends to increase too. But the points do not all sit exactly on the line, showing natural variation. I could use a spreadsheet to plot this and calculate the correlation. This tells us arm span and height are related but not identical — some people have proportionally longer arms or legs.
Delivery rationale
Science skill involving measurement/practical work — AI structures, facilitator supervises.
Causal Relationships and Evidence Evaluation
process AI FacilitatedSC-KS2-C010
Identifying causal relationships in data. Evaluating degree of trust in results. Identifying scientific evidence that supports or refutes ideas and arguments. Distinguishing opinion from fact in secondary sources.
Teaching guidance
Use investigations with clear causal links (e.g., increasing temperature speeds up dissolving) to model the difference between correlation and causation. Teach pupils to evaluate the trustworthiness of results by considering sample size, whether variables were controlled, and whether results are repeatable. Introduce the idea that scientific claims require evidence by examining real examples of scientific debate. Use newspaper or media science claims as stimuli for critical evaluation, asking 'What is the evidence? Is it strong enough?'
Common misconceptions
Pupils commonly assume that correlation implies causation — for example, that because ice cream sales and sunburn both increase in summer, ice cream causes sunburn. They may also believe that a single investigation provides definitive proof, rather than understanding that scientific conclusions become stronger with repeated, independent evidence. Some pupils struggle to distinguish between scientific evidence and personal opinion.
Difficulty levels
Identifying a simple cause and effect in an investigation — 'this happened because of that' — with teacher support.
Example task
The ice melted faster in the warm room than in the cold room. What caused the ice to melt faster?
Model response: The warm room made the ice melt faster because it was hotter.
Identifying a causal relationship in results and beginning to evaluate whether evidence is strong enough to support a claim.
Example task
We found that plants near the window grew taller than plants far from the window. Can we say that light caused the difference? What else might have caused it?
Model response: Light probably caused the difference because the plant near the window got more light. But the window might also be warmer because of the sun, or there might be a radiator nearby. So it could be the temperature, not just the light. We would need to control temperature to be sure it was the light that made the difference.
Identifying causal relationships in investigation data, distinguishing between correlation and causation, and evaluating the trustworthiness of results.
Example task
A news article says 'Children who eat breakfast do better in school tests.' Does this mean breakfast causes better test results? What other explanations could there be?
Model response: It shows a correlation — children who eat breakfast tend to do better — but it does not prove causation. Other factors could explain it: families that provide breakfast might also provide more support with homework, have a calmer morning routine, or have higher incomes (which are linked to academic outcomes). The breakfast itself might help (providing energy for concentration), but we cannot be certain from this data alone. To prove causation, you would need a controlled experiment where the only difference is whether children eat breakfast.
Evaluating scientific evidence critically, distinguishing between fact and opinion in secondary sources, and assessing the degree of trust in results based on methodology.
Example task
Two websites make claims about magnets: Site A says 'Magnets can cure headaches' and cites one person's experience. Site B says 'A study of 200 people found no evidence that magnets reduce headache pain compared to a placebo.' Which should you trust more? Why?
Model response: Site B is more trustworthy because it is based on a study with 200 people (large sample size) and used a placebo (control), which is how scientists test whether something really works. Site A is based on one person's experience — this is anecdotal evidence. One person might have felt better for many reasons (the headache was going away anyway, the placebo effect). Scientific claims need to be tested with controlled studies and large samples before we can trust them. I would look at who conducted the study and whether it was published in a scientific journal.
Delivery rationale
Science process concept — enquiry methodology benefits from structured AI guidance with facilitator.
Access barriers (2)
Using classification keys requires following branching decision sequences: 'Does it have legs? Yes → Does it have 6 legs? Yes → It is an insect.' Each branch is a conditional step that must be followed in order.
KS2 animal classification extends to invertebrate groups (insects, arachnids, crustaceans, molluscs, annelids) plus classification keys and dichotomous keys. The taxonomic vocabulary load is substantial.