Data science and analytics: formula sheet

Key formulas

The 5 Vs of big data

Volume + Velocity + Variety + Veracity + Value

Learn them as a checklist. For each V, state what it means and give a scenario example.

Insight to action chain

Data → Analysis → Insight → Decision → Performance improvement

Use this to show value. Data alone creates no benefit until it changes a decision.

Cost-benefit test for data projects

Expected benefits from better decisions > Cost of collecting, storing, analysing and securing data

This is the link between the Value V and a business case. Costs include skills, systems and compliance.

Descriptive

What happened?

Past and current data. Reports, dashboards, KPIs, summary statistics.

Diagnostic

Why did it happen?

Drill-down, comparison, correlation, root cause and variance investigation.

Predictive

What is likely to happen?

Forecasts and probabilities from models. Outputs are estimates, not certainties.

Prescriptive

What should we do?

Recommends actions using optimisation, simulation and decision rules, within constraints.

Value and complexity

Descriptive → Diagnostic → Predictive → Prescriptive

Insight and decision value usually rise along this order, and so do data and skill needs. This is a general pattern, not a strict rule.

Quick revision

• Big data is commonly described by features such as volume, velocity, variety and veracity; know what each means with an example.
• Descriptive analytics shows what happened; diagnostic explains why.
• Predictive analytics estimates what is likely to happen; prescriptive suggests what action to take.
• Data mining looks for patterns in large data sets; machine learning lets systems improve from data without being explicitly programmed for each case.
• AI can automate decisions, but its output depends on the data and design behind it.
• Dashboards should show a few relevant KPIs clearly, linked to objectives and to the user's decisions.
• Good visualisation helps managers see trends and exceptions quickly, but poor design can mislead.
• Data quality means accuracy, completeness, timeliness and consistency; poor quality leads to poor decisions.
• Governance sets who owns data, who can access it and how it is kept secure and used.
• Ethical issues include privacy, consent, bias in algorithms and transparency of decisions.
• Technology can speed up reporting and widen the range of measures, but it adds cost, security risk and a need for skills.
• Always apply points to the scenario and finish with a clear, justified recommendation.

Common mistakes

• Listing the 5 Vs with textbook definitions only. Fix: Add a scenario example to every V. Marks go to application, and professional skills marks reward relevance.
• Saying big data simply means a lot of data. Fix: Explain that volume is one of five characteristics. Velocity, variety and veracity matter just as much, and big data differs from traditional data in structure, source and speed.
• Mixing up predictive and prescriptive analytics. Fix: Predictive estimates what will probably happen. Prescriptive recommends the action to take. If the answer is a forecast, it is predictive. If it is a choice, it is prescriptive.
• Giving textbook definitions with no link to the scenario. Fix: After each definition, add one example using the company's products, customers or processes.
• Treating data mining, machine learning and AI as the same thing. Fix: Define each in one line: mining finds patterns, machine learning learns and predicts, AI is the broad field. Say they overlap when relevant.
• Writing a generic list of benefits that ignores the scenario. Fix: Name the business problem and data from the case in every point. Use the company's products, customers or processes.
• Describing visualisation in general terms with no link to the scenario. Fix: Name the company, its users and its KPIs in every point. Ask yourself what this means for this business.
• Designing a dashboard with dozens of measures. Fix: Limit each dashboard to the few measures tied to strategy and to the user's decisions. Explain that overload hides what matters.
• Listing generic risks without using the scenario. Fix: Tie every point to a named fact, such as the data source or the decision affected.
• Only describing problems and giving no controls. Fix: Pair each risk with a practical safeguard and who is responsible for it.

Exam tips

• Always apply the 5 Vs to the scenario. A bare list earns few marks.
• Read the requirement verb. 'Evaluate' needs benefits, risks and a judgement. 'Explain' needs clear definitions with examples.
• Link every data point to a performance measure, such as a KPI, target or forecast, to show commercial awareness.
• Mention ethics and regulation when personal data is in the scenario. It is a common marker expectation.
• Keep paragraphs short and use headed points. Clear structure earns communication skills marks.
• Always tie each type to the scenario's decision. Generic definitions earn few marks.
• Use the four questions as a structure, then add data needs, benefits and limits to reach the depth the requirement asks for.
• When asked to evaluate or recommend, give a balanced view and a clear conclusion, including which type to adopt first.