Skip to content
Getting startedNot reviewed

Understand variable types and measurement levels

Review detected numeric, text, categorical, ordinal, and scale-like variables before analysis.

When to use this

Use this immediately after upload and whenever coding changes.

Why it matters

Statistical procedures depend on both stored values and their substantive measurement meaning.

Step-by-step instructions

1. Open the dataset and select Variables.
2. Compare names and labels with the questionnaire.
3. Review detected data type and measurement level.
4. Mark identifiers and demographics.
5. Link indicators to approved constructs only when justified.
6. Confirm or correct classifications and record notes.

Expected outputs

Reviewed variable metadata linked to one immutable dataset version.

How to interpret the results

A numeric storage type does not automatically mean a variable is continuous. Treat Likert items according to the planned analysis and measurement rationale.

Common mistakes

Do not infer independent or dependent roles from column order. Do not treat identifier codes as measurements.

Was this article helpful?

Your response helps the methodology team improve this documentation.