Types of Data
Not all data is the same. The type of data you collect determines which statistical methods you can use, which charts make sense, and what conclusions you can draw.
Qualitative vs Quantitative
The broadest division. Qualitative data describes categories (colors, names). Quantitative data represents numbers you can count or measure.
Nominal Data
Categories with no inherent order. You can count them, but you cannot rank or average them.
- Blood types — A, B, AB, O
- Eye color — brown, blue, green
- Gender — labels, not rankings
- Nationality — categories, not numbers
Ordinal Data
Categories with a natural order, but the gaps between ranks are not necessarily equal. You know which is “more,” but not by how much.
- Satisfaction — poor, fair, good, excellent
- Education — high school, bachelor's, PhD
- Pain scale — 1 to 10, unequal gaps
Interval Data
Equal spacing between values, but no true zero. The difference is meaningful, but ratios are not. 40°C is not “twice as hot” as 20°C.
Ratio Data
Equal intervals plus a true zero. All math operations are valid. $80k is truly twice $40k because $0 means no income.
Measurement Scales
Each level adds more mathematical power:
- Nominal — categories only, count & mode
- Ordinal — + order, median & percentiles
- Interval — + equal spacing, mean & std dev
- Ratio — + true zero, all operations valid
Discrete vs Continuous
Discrete data takes countable values (25 students, not 25.7). Continuous data can be any value in a range (height: 170.3 cm).
Structured vs Unstructured
Structured data fits in tables with rows and columns. Unstructured data — text, images, audio — has no predefined format.
How We Gather Data
Three primary methods, each with different strengths:
- Surveys — self-reported, efficient but biased
- Experiments — controlled, can prove causation
- Observational — natural, but correlation only
Matching Data to Methods
The data type dictates the analysis. Using the wrong method for your data type produces invalid results.
What you learned
Data types form a hierarchy — nominal, ordinal, interval, ratio — each unlocking more mathematical operations. Always classify your data first, then choose your tools.