Data Analytics Concepts and Fundamentals
The analytics ladder
Data analytics turns raw data into decisions. Distinguish data (raw recorded values), information (data processed into meaning, such as a monthly total) and knowledge (information combined with experience to guide action). Analytic work is usually described in four levels. Descriptive analytics answers "what happened" (totals, trends, dashboards). Diagnostic asks "why did it happen" (drill-downs, correlation, root-cause analysis). Predictive estimates "what is likely to happen" (forecasting, regression, scoring). Prescriptive recommends "what should we do" (optimisation, simulation, decision rules).
Types and scales of data
Data is qualitative/categorical or quantitative/numeric. Quantitative data is discrete (countable, e.g. number of orders) or continuous (measurable on a scale, e.g. delivery time). Measurement scales matter because they limit which statistics are valid: nominal (labels only, e.g. region), ordinal (ordered but gaps not equal, e.g. satisfaction 1–5), interval (equal gaps, arbitrary zero, e.g. temperature in °C), ratio (equal gaps, true zero, e.g. revenue).
Also know population versus sample: a population parameter is a fixed but usually unknown value; a sample statistic is an estimate that varies from sample to sample. Data may be structured (tables), semi-structured (JSON, logs) or unstructured (text, images), and primary (collected for your purpose) or secondary (already collected by someone else).
Traps
Questions love mixing up predictive and prescriptive; freely averaging ordinal ratings; calling a sample mean a parameter; assuming numeric-looking codes such as postcodes or IDs are quantitative; and treating interval data as ratio (20 °C is not "twice as warm" as 10 °C).
Rule of thumb
Ask whether zero means "none of the thing": if yes, ratio; if zero is just a point on the scale, interval; if only order matters, ordinal.