What is data analytics?

The four types of analytics

Data analytics is often framed as four progressively more advanced types. Descriptive analytics summarises what happened, using aggregates and reports. Diagnostic analytics investigates why, drilling into data to find causes and relationships. Predictive analytics estimates what is likely to happen next, using statistical models or machine learning. Prescriptive analytics goes furthest, recommending actions and weighing their likely outcomes. Each level builds on the last and answers a more demanding question, generally at greater analytical complexity.

How analytics is done

Analytics combines data preparation, statistical methods, and visualisation, often supported by dashboards and reporting tools. Descriptive and diagnostic work leans on aggregation, segmentation, and clear visual summaries.

Predictive and prescriptive work draws on modelling, forecasting, and optimisation. As the questions become more forward-looking, the role of statistical assumptions and model validation grows — a prediction is only as trustworthy as the data and assumptions behind it.

Analytics versus analysis

In everyday use, data analysis and data analytics overlap heavily. A common convention treats analysis as the close examination of a particular dataset to explain what happened, and analytics as the broader discipline that also encompasses systems, processes, and forward-looking prediction across an organisation or research programme. Both fall under the umbrella of data science.

Analytics in research

In research and evaluation, analytics supports monitoring, forecasting, and decision-making — for example tracking the progress of a study or projecting resource needs. As with all data work, conclusions depend on data quality and on honest treatment of uncertainty. Predictive analytics in particular can mislead if a model is applied outside the conditions it was built for, so validation, documentation, and clear communication of limitations are essential.