Unit 1 · Find your way around
Levels of measurement in SPSS: nominal, ordinal and scale
How to choose Nominal, Ordinal or Scale in the SPSS Measure column, with examples and practice questions.
Quick answer: How do you set the level of measurement in SPSS?
- Go to Variable View and find the Measure column.
- Choose Nominal for unordered categories (such as course).
- Choose Ordinal for ordered categories (such as a 1–5 rating).
- Choose Scale for numbers with equal intervals (such as age or exam score).
What are levels of measurement?
The level of measurement describes what the numbers in a variable actually mean. Nominal numbers are labels for categories with no order. Ordinal numbers show an order, but the gaps between them aren’t guaranteed to be equal. Interval and ratio numbers have equal gaps; SPSS calls both of these Scale. Choosing the right level tells you which statistics and tests are meaningful.
When to use it
- Setting the Measure column in Variable View.
- Choosing a statistic: modes for nominal, medians for ordinal, means for Scale.
- Choosing a test: categorical outcomes need chi-square tests; Scale outcomes usually need t-tests, ANOVA or regression.
How it works
Ask two questions. Does the order of the numbers mean anything? If not, it’s nominal. If it does, are the steps between values equal in size? If not, or you can’t be sure, it’s ordinal. If they are, it’s Scale. Ratio data (such as reaction time) also have a true zero, so you can say one value is twice another; interval data (such as temperature in °C) don’t.
Worked example
Course (1 = Psychology, 2 = Nursing) is nominal. A 1–5 satisfaction rating is ordinal. Age in years is Scale (ratio). Many researchers treat a total score made by adding several Likert items as Scale.
Key terms
- Nominal
- Unordered categories, such as eye colour.
- Ordinal
- Ordered categories with uneven or unknown gaps, such as rankings.
- Interval
- Equal gaps, no true zero, such as °C.
- Ratio
- Equal gaps and a true zero, such as weight. SPSS treats interval and ratio both as Scale.
Common mistakes
- Averaging nominal codes: a “mean course” of 1.8 means nothing.
- Leaving every variable as SPSS’s default instead of setting Measure yourself.
- Treating a single Likert item as precise Scale data without saying why.
How to do it in SPSS
The SPSS syntax
Prefer typing to clicking? Open File › New › Syntax, paste this in, select it and choose Run › Selection. Every dialog's Paste button writes the same kind of code.
* Set the level of measurement (the Measure column).
VARIABLE LEVEL gender course (NOMINAL)
/age score (SCALE).Practice data
Download the file used in this lesson and run it yourself in SPSS: you should get the same results as the screens above.
- students.savStudent survey: gender, age, course, exam score and attention check (N = 100)
Now practise it interactively →
Reading is only half of it. Tap through the real SPSS menus, read the output and pick the right write-up in 5 quick questions, with instant feedback on every answer.
This lesson’s practice is free. No sign-up needed to start.