Cambridge A Level Information Technology (ICT): Modelling (9626)

This study guide covers topic 9, Modelling, of Cambridge International AS & A Level Information Technology (9626). It follows the Cambridge International AS & A Level Information Technology 9626 syllabus for examination in 2025, 2026 and 2027 (version 3), section 9.1 Modelling and simulations. Topic 9 is an AS Level topic, so it is part of both the AS Level and the full A Level. The syllabus bases Paper 1 (Theory) questions on sections 1–11 and Paper 2 (Practical) tasks on sections 8–11, so you can meet modelling in both papers. Paper 4 (Advanced Practical) may also include practical tasks from sections 8–10 within a problem-solving context.

Modelling uses the spreadsheet skills from topic 8, and A Level topic 17 (Data analysis and visualisation) builds on it. Course hub: Cambridge A Level ICT. Printable checklist: 9626 checklist. Not sure where your gaps are? Try a free 10-minute diagnostic.

What this topic covers

Section	What you must be able to do	Stage
9.1	Use what-if analysis: predict the result of changing data, and change data to model different scenarios	AS
9.1	Use goal seek	AS
9.1	Describe uses of what-if analysis: financial forecasting, population growth, climate change, weather systems, queue management, traffic flow, construction	AS
9.1	Describe the characteristics of modelling software	AS
9.1	Explain the need for computer models	AS
9.1	Evaluate the effectiveness of spreadsheet models	AS
9.1	Explain the use of a model to create and run simulations: natural disaster planning, pilot training, learning to drive a car, nuclear science research	AS

Every outcome can be asked about in writing in Paper 1. Because Paper 2 tasks are based on sections 8–11, you should also be able to build a model, change its data and use goal seek in practical work.

Models and simulations

A computer model is a representation of a real system using data (the variables) and rules (formulas or relationships between the variables). A spreadsheet that works out a shop’s profit from its prices, costs and sales is a model of that shop’s finances.

A simulation is what you get when you run a model to see how the system behaves, often over time or under conditions you choose. You might step a population model forward year by year, or feed random customer arrivals into a queue model.

Key vocabulary:

• Variable – a value you can change, such as a price or a growth rate.
• Rule – a formula linking the variables, such as profit = sales × (price − cost) − fixed costs.
• Assumption – a simplification built into the model, such as “the growth rate stays the same every year”.
• Scenario – one set of input values, such as “best case” or “price rise”.

What-if analysis

What-if analysis means changing one or more inputs to a model and seeing how the outputs change. Because the spreadsheet recalculates automatically, you can try many scenarios in seconds. The syllabus lists two parts:

• Predict the result of changing data – before you change an input, say what you expect to happen, then check.
• Change data to model different scenarios – change inputs to compare options such as “raise prices” against “cut costs”.

Worked example 1: a café’s monthly profit

A café owner sets up this model.

	A	B
1	Price per cup (£)	3.20
2	Cost per cup (£)	1.10
3	Fixed costs per month (£)	1890
4	Cups sold per month	1000
5	Profit (£)	=B4*(B1-B2)-B3

Step 1 – current profit. Each cup earns 3.20 − 1.10 = £2.10 towards the fixed costs. Profit = 1000 × 2.10 − 1890 = £210.

Step 2 – predict, then change. The owner asks: “What if I sell only 800 cups?” Prediction: profit falls by 200 × 2.10 = £420. Change B4 to 800: profit = 800 × 2.10 − 1890 = −£210, a loss. The prediction was right.

Step 3 – a second scenario. Reset B4 to 1000 and change B1 to 3.50. Each cup now earns £2.40, so profit = 1000 × 2.40 − 1890 = £510.

Notice that every input sits in its own labelled cell and the formula refers to those cells. If the price were typed into the formula itself, each scenario would mean editing the formula, which is slow and easy to get wrong.

Goal seek

What-if analysis changes the inputs and reads the output. Goal seek works the other way round: you choose the output you want and the software finds the input value that produces it. A goal seek tool needs three things:

1. the cell that holds the result (it must contain a formula);
2. the target value you want that cell to reach;
3. the input cell the software is allowed to change.

The software then tries input values until the result matches the target.

Worked example 2: break-even for the café

The owner wants to know how many cups give a profit of exactly zero.

• Result cell: B5. Target value: 0. Cell to change: B4.
• Goal seek returns 900. Check: 900 × 2.10 − 1890 = 0.

At the higher price of £3.50, goal seek returns 1890 ÷ 2.40 = 787.5. Goal seek has no idea that half a cup cannot be sold, so you must interpret the answer: the café needs 788 cups to cover its costs.

Worked example 3: a population model

A planner models a town of 50,000 people growing at 2% a year. The growth rate is stored in a cell named Rate (2%).

	A	B
1	Year	Population
2	0	50000
3	1	=B2*(1+Rate)
4	2	=B3*(1+Rate)

The formula in B3 is filled down to B7 (year 5). Because Rate is a named cell, every copied formula still refers to it.

• Year 1: 50,000 × 1.02 = 51,000. Year 2: 52,020. Year 5: 55,204 (to the nearest person).
• Goal seek: “What yearly growth rate gives 60,000 people after 5 years?” Result cell B7, target 60000, cell to change Rate. Goal seek returns about 3.71% (to 3 s.f.).

Uses of what-if analysis

The syllabus names seven uses. For each, know what is varied and what is predicted.

Use	Typical inputs changed	What the model predicts
Financial forecasting	prices, sales, costs, interest rates	profit, cash flow, loan repayments
Population growth	birth, death and migration rates	future population; demand for schools, housing
Climate change	greenhouse gas emissions, land use	temperature and sea-level trends
Weather systems	air pressure, temperature, humidity, wind readings	the weather over the coming days
Queue management	arrival rate, service time, number of servers	waiting times and queue lengths
Traffic flow	traffic light timings, number of lanes, vehicle numbers	congestion and journey times
Construction	materials, loads, dimensions, costs	whether a structure is safe, and what it costs

Worked example 4: queue management

A post office expects 72 customers an hour at its busiest time. One clerk serves 15 customers an hour on average.

• Clerks needed = 72 ÷ 15 = 4.8, so the model recommends 5 clerks.
• What if arrivals rise to 96 an hour? 96 ÷ 15 = 6.4, so 7 clerks.
• What if new equipment lets each clerk serve 18 an hour? 72 ÷ 18 = exactly 4 clerks.

The last result shows a limit of a simple model. With 4 clerks working flat out, any short burst of arrivals creates a queue that cannot clear, because real customers do not arrive at an even rate. A simulation that generates random arrivals would show this; the average-rate spreadsheet does not.

Characteristics of modelling software

Modelling software (including a spreadsheet used as a model) typically lets you:

• store variables in cells or fields and change them easily;
• define rules and formulas that link the variables;
• recalculate automatically when an input changes;
• carry out what-if analysis and goal seek;
• set up and save several scenarios and compare their results;
• step the model through time intervals (days, years) to show change over time;
• generate random values so that a simulation reflects chance events;
• display results as graphs, charts or animated visual output, so trends are easy to see;
• import real data so the model can be checked against what actually happened.

The need for computer models

Computer models are used because, compared with testing the real thing, they can be:

• Cheaper – testing a new traffic layout on screen costs far less than rebuilding a junction.
• Safer – no one is put at risk while a dangerous situation is studied.
• Faster – years of population change can be calculated in seconds.
• Repeatable – the same scenario can be run again with one variable changed, to see its effect alone.
• Possible at all – you cannot run a real experiment on the future climate or on an earthquake.

They also let decision-makers compare many options before committing money or resources.

The effectiveness of spreadsheet models

A spreadsheet model is only as good as its rules, its data and its assumptions. When asked to evaluate one, weigh both sides.

Strengths

• Formulas recalculate instantly, so what-if analysis and goal seek are quick.
• Widely available and familiar; non-specialists can follow the layout.
• Results can be shown as charts.
• Easy to update when new data arrives.

Weaknesses

• Simplified rules: real systems depend on more variables than a spreadsheet includes.
• Fixed assumptions: a constant growth rate or average arrival rate may not hold.
• Input data quality: inaccurate data gives inaccurate predictions.
• Formula errors: one wrong cell reference can spread through the whole model.
• Human behaviour and rare events are hard to express as formulas.
• Spreadsheets are poor at visual, real-time or three-dimensional simulation.

A good evaluation ends with a judgement: for example, a spreadsheet is effective for a café’s monthly forecast but too limited for a full weather system.

Using a model to create and run simulations

Simulation	What is simulated	Benefits	Drawbacks
Natural disaster planning	floods, earthquakes, storms; how people and services respond	emergency plans and evacuation routes tested without danger; resources placed in advance	real disasters may not follow the model; depends on good data about the area
Pilot training	aircraft behaviour, weather, instrument readings, faults	emergencies such as engine failure practised safely; no fuel used; scenarios repeated	high cost of simulator equipment; trainees know they are not in real danger
Learning to drive a car	roads, other vehicles, hazards, weather	hazards practised without risk to the learner or the public; useful for first lessons	cannot fully reproduce the feel of a real car; may cause motion sickness
Nuclear science research	reactions, reactor behaviour, effects of changing conditions	no exposure to radiation; experiments impossible in reality can be run	results depend on the accuracy of the scientific rules in the model

In each case the model holds the rules (how the aircraft responds to its controls, how floodwater spreads) and the simulation runs those rules under conditions the user chooses.

Common errors

• Confusing what-if analysis with goal seek. What-if: change inputs, read the output. Goal seek: set the output, find an input.
• Leaving a model’s answer unrounded when the context needs whole units (787.5 cups, 6.4 clerks).
• Typing values into formulas instead of referencing input cells, so scenarios cannot be changed easily.
• Writing “a model is cheaper” without saying cheaper than what. Compare with the real-life alternative.
• Using “model” and “simulation” as if they mean the same thing.
• Giving only advantages when the question asks you to evaluate.

Next steps

Condense this with the Modelling revision notes, then test yourself on the Modelling practice questions. For the sensors and control systems behind weather data, see Monitoring and Control. For the hardware used to run large simulations, see Hardware and Software.

Official syllabus

Cambridge International AS & A Level Information Technology 9626, syllabus for examination in 2025, 2026 and 2027, version 3, Cambridge University Press & Assessment. Topic 9: Modelling (9.1 Modelling and simulations).