Nominal Data
What is Nominal Data?
Nominal data consists of categories or labels with no inherent order. 'Male' is not greater than 'Female' - they're just different categories. You can only count frequencies.
Real-World Examples
Nominal data appears everywhere: ZIP codes, product categories, marital status, brand names. The key is there's no meaningful way to rank or order these categories.
How to Analyze Nominal Data
You can count nominal data and find the mode (most frequent category). You cannot calculate means or medians because there's no numeric value or order to work with.
Ordinal Data
What is Ordinal Data?
Ordinal data has categories with a meaningful order, but the gaps between categories may not be equal. 'Excellent' is better than 'Good,' but we can't say how much better.
Real-World Examples
Survey responses, rankings, education levels, and any scale with ordered categories but unequal or unknown intervals. The gap between 'Agree' and 'Strongly Agree' isn't necessarily the same as between 'Neutral' and 'Agree.'
How to Analyze Ordinal Data
Ordinal data adds median to your toolkit. Some researchers calculate means on Likert scales (treating 1-5 as numbers), but this assumes equal intervals—a debated practice.
Nominal vs Ordinal
The key difference: Can you rank the categories? If 'yes,' it's ordinal. If the categories are just different names with no logical order, it's nominal.
Interval Data
What is Interval Data?
Interval data has equal spacing between values, but zero is arbitrary—not a true absence. The difference between 20°C and 40°C equals 40°C to 60°C. But 0°C doesn't mean 'no temperature.'
Real-World Examples
The classic examples are temperature scales (Celsius/Fahrenheit). 0°C is just the freezing point of water—not 'no heat.' You can't say 40°C is 'twice as hot' as 20°C.
How to Analyze Interval Data
Interval data unlocks most statistical operations: mean, standard deviation, t-tests, ANOVA. The one limitation: you can't form meaningful ratios because zero is arbitrary.
Why Zero Matters
The 'zero test': Does zero mean complete absence? 0°C doesn't mean no temperature (molecules still move). But 0 kg means no mass at all. This distinction determines whether ratios are meaningful.
Ratio Data
What is Ratio Data?
Ratio data has equal intervals AND a true zero point. This means ratios make sense: $40 is truly twice $20. Zero dollars means no money at all.
Real-World Examples
Most physical measurements are ratio: height, weight, distance, time, speed, money. Kelvin temperature (unlike Celsius) is ratio because 0K is absolute zero—true absence of thermal energy.
Full Statistical Power
Ratio data supports every statistical operation. You can calculate percentage change, create ratios, use geometric means, and apply any test. It's the most flexible data type.
Interval vs Ratio
The key distinction: Can you make meaningful ratio statements? 'Person A earns twice as much as Person B' makes sense (ratio). 'Today is twice as hot as yesterday' doesn't (interval).
Decision Guide
Data Type Hierarchy
Data types form a hierarchy. Ratio data can be analyzed like interval, ordinal, or nominal. But nominal data can only be counted—you can't calculate a mean on categories.
Identification Flowchart
Three questions determine data type: (1) Can values be ordered? (2) Are intervals between values equal? (3) Does zero mean complete absence? Answer these in order.
At-a-Glance Summary
| Type | Order | Equal Gap | True 0 | Example |
|---|---|---|---|---|
| Nominal | ✗ | ✗ | ✗ | Colors |
| Ordinal | ✓ | ✗ | ✗ | Rankings |
| Interval | ✓ | ✓ | ✗ | °C / °F |
| Ratio | ✓ | ✓ | ✓ | Weight |
Each data type builds on the previous: Ordinal adds order to Nominal, Interval adds equal spacing to Ordinal, Ratio adds true zero to Interval.
Real-World Examples
Apply your knowledge to practical scenarios
Customer Satisfaction Survey
A company collects customer feedback with the following variables: Customer ID, Gender, Age, Income, Satisfaction Rating (1-5 stars), Product Category, and Purchase Amount.
The Temperature Trap
A weather report says 'Today at 30°C is 50% warmer than yesterday's 20°C.' Is this statement correct?
Key Insights
- Always check if zero means 'complete absence' before making ratio statements.
Clinical Trial Data Classification
A pharmaceutical study tracks: Patient ID, Treatment Group (Drug A/B/Placebo), Pain Level (None/Mild/Moderate/Severe), Blood Pressure (mmHg), Temperature (°F), Recovery Time (days), and Outcome (Recovered/Not Recovered).
Data Types Mastery Quiz
Test your understanding • Passing score: 70%