Skip to content

Actuarial Statistics · Purpose and function of data analysis

Purpose and Aims of Data Analysis in Actuarial Work

Updated 11 October 2026 · Fact-checked

Actuaries analyse data to turn raw records into decisions. The main aims are to describe what happened, infer facts about a wider population, predict future outcomes, and support choices such as pricing, reserving and risk control. In exams, name the aim first, then match the method to it.

Understand Purpose and Aims of Data Analysis

Data analysis is the process of collecting, checking, summarising and modelling data so that it answers a question. Actuaries do not analyse data for its own sake. Every analysis starts with a business problem: What premium should we charge? How many claims will we pay next year? Is this mortality rate different from the last one?

There are four broad aims. Describing means summarising the data you have, using averages, spreads and charts. Inferring means using a sample to draw conclusions about a larger population, with an allowance for uncertainty. Predicting means using patterns in past data to estimate future or unseen outcomes. Supporting decisions means turning the results into a recommendation, such as a premium, a reserve or a reinsurance choice.

These aims build on each other. You usually describe first, to understand the data and spot problems. Inference and prediction come next. The decision comes last, and it must account for uncertainty, cost and business limits.

The aim decides the method. A claims summary table is description. A confidence interval for mean claim size is inference. A model for next year's claim count is prediction. Recommending a rate increase is decision support. Examiners often give a short scenario and ask which aim applies, or why an analysis is needed.

Good actuarial analysis also needs data that is fit for purpose. Poor, biased or incomplete data leads to wrong answers, however clever the model. So the purpose of the analysis also shapes what data you collect and how much checking it needs.

How to solve Purpose and Aims of Data Analysis questions

Use this method for any question that asks why data is analysed or which aim an analysis serves.

  1. 1Read the scenario and find the business question being asked.
  2. 2Decide whether the task is to describe, infer, predict or support a decision. Some tasks involve more than one.
  3. 3State the aim in one clear sentence, using the scenario's own context (for example, motor claims or policyholder mortality).
  4. 4Name the kind of output that serves the aim, such as a summary table, an interval, a fitted model or a recommendation.
  5. 5Link the output to the decision it supports, for example pricing or reserving.
  6. 6Mention one limit: data quality, sample size, or uncertainty in the result.
  7. 7Check that your answer uses the context given and not only general statements.

Quickest way: Four-word match

When to use it: Use for multiple-choice questions and short written parts where you must identify the aim quickly.

  1. Look for the verb in the question: summarise or display means describe.
  2. Generalise from a sample or test a claim means infer.
  3. Forecast, estimate future or score new cases means predict.
  4. Choose, price, set or recommend means support a decision.
  5. If two options seem right, pick the one that matches the final goal in the scenario.

Common mistakes in Purpose and Aims of Data Analysis

  • Treating description and inference as the same thing.

    Both use data summaries, so they look alike.

    Fix: Ask whether the statement is only about the data in hand (description) or about a wider population (inference).

  • Giving generic answers with no insurance or finance context.

    Students memorise the list of aims without applying it.

    Fix: Tie each aim to a concrete example, such as claim frequency, mortality or reserves.

  • Saying the purpose of analysis is just to build a model.

    Modelling feels like the main actuarial skill.

    Fix: Remember the model serves a question. State the decision or question first, then the model.

  • Ignoring data quality when stating the aims.

    Students focus on methods and forget that results depend on inputs.

    Fix: Add one line that the analysis is only as reliable as the data, and that checks are needed before use.

  • Claiming a prediction is certain.

    Forecasts are written as single numbers.

    Fix: Say predictions carry uncertainty and may be given with intervals or scenarios.

Worked examples

Example 1

An insurer holds five years of motor claims records. State four aims for which it might analyse this data, giving one example of each.

Show the solution
  1. Describing: summarise claim counts and average claim size by year and vehicle type.
  2. Inferring: use the sample to estimate the true mean claim size for all policyholders, with a confidence interval.
  3. Predicting: fit a model to forecast the number of claims next year.
  4. Supporting decisions: use the forecast and cost estimates to set premium rates or decide how much reinsurance to buy.

Answer: The four aims are describing, inferring, predicting and supporting decisions, each matched to a motor insurance example as above.

Example 2

A life insurer finds that the average age at death in its annuity portfolio last year was higher than in the previous year. A manager asks whether this proves longevity is improving for all annuitants in India. Explain which aims are involved and comment.

Show the solution
  1. The comparison of the two averages is description. It only reports what happened in this portfolio.
  2. The manager's question about all annuitants in India asks for inference, because it goes beyond the data held.
  3. Inference needs a test or interval that allows for random variation, and a check that the portfolio is representative of the wider population.
  4. Annuitants are often wealthier and healthier than the general population, so the portfolio may be biased.
  5. A decision such as changing annuity rates would need prediction of future mortality as well.

Answer: The higher average is a description of the portfolio only. It does not prove national longevity improvement. Inference with a test, representativeness checks and prediction would be needed before supporting a pricing decision.

Exam tips

  • Start answers with the aim in a single sentence, then give examples. This secures marks quickly.
  • Always use the context in the question. Generic lists score less than applied points.
  • For 'discuss' questions, mention data quality and uncertainty as limits of any analysis.
  • In multiple-choice questions, check whether the scenario talks about the sample only or the wider population.
  • Link analysis to a decision at the end, such as pricing, reserving or risk management.

Practice questions from Purpose and function of data analysis

Purpose and Aims of Data Analysis: frequently asked questions

Why do actuaries use data analysis?

Actuaries use it to measure and manage financial risk. Analysis shows past experience, estimates uncertain quantities and supports decisions on pricing, reserving and capital.

What are the main aims of data analysis?

The main aims are to describe the data, infer facts about a population, predict future outcomes and support decisions. Exam answers should name the aim and give a context example.

Is this topic only theory with no calculations?

Mostly yes. It is usually tested through short written answers and multiple-choice questions. Later chapters apply the aims with calculations.

How is description different from inference?

Description reports on the data you hold. Inference uses that data to make statements about a wider population, and it must allow for sampling uncertainty.