Classifying data correctly is the first step before choosing any statistical test or chart.
Numerical data divides further into discrete counts and continuous measurements.
Discrete numbers count indivisible units, such as the number of cars parked in a driveway.
Continuous numbers can take any real value along an interval, such as height or temperature.
Categorical data divides further into nominal categories and ordinal categories.
Nominal categories possess no natural order, such as colours, product names, or departments.
Ordinal categories carry an intrinsic ranking, such as small, medium, and large shirt sizes.
Even though ordinal groups have an order, the mathematical distance between them is rarely uniform.
The gap between strongly agree and agree is not necessarily equal to the gap between agree and neutral.
A widespread trap is storing categories as numbers and computing an average over them.
Assigning postcode numbers or telephone area codes does not make those values numerical data.
Calculating the average postal code produces a number with zero physical or mathematical meaning.
Visualising them also differs: categories use bar charts while continuous numbers use histograms.
Converting categories into numbers for computation requires careful encoding rather than plain arithmetic.
Always verify whether differences between values reflect true measurable distance.