How to Use Balance Tables
Check whether pre-treatment variables are similar across randomized arms.
The Short Version
Check whether pre-treatment variables are similar across randomized arms. A practical way to begin is to fetch or load responses, then open Results. Finish by making sure you export the balance CSV or journal tables. The full walkthrough below explains each part in order.
Is This Guide for You?
Researchers preparing reviewer-facing diagnostics for survey experiments.
Follow These Steps
Work through these 6 steps at your own pace. The wording is intentionally practical, and you can return to the checklist at the end when you are ready to review your work.
- Fetch or load responses.
- Open Results.
- Open Balance and diagnostics below the main findings.
- Select the balance variable you want to inspect.
- Read means or category summaries by arm.
- Export the balance CSV or journal tables.
What This Helps You Accomplish
Balance tables are a simple way to show that random assignment did not create obvious pretreatment differences across arms.
A Quick Confidence Check
- Use only pre-treatment variables.
- Treat p-values as diagnostics, not proof of validity.
- Report any meaningful imbalance transparently.
- Run balance after applying exclusions.
Common Questions
Can I use this guide if I am new to this?
Yes. Researchers preparing reviewer-facing diagnostics for survey experiments. Follow the steps in order, start with a small test, and use the final checklist before you field or report anything important.
What is the simplest way to get started?
Begin by fetch or load responses. Next, open Results. You do not need to perfect every setting before running a small preview or pilot.
How do I know when I am ready?
Use the confidence check above. In particular: Use only pre-treatment variables. Treat p-values as diagnostics, not proof of validity. Report any meaningful imbalance transparently. Run balance after applying exclusions. When the decision is consequential, keep your study documentation and ask a qualified colleague or methods reviewer to examine the design as well.
Related Guides
Research artifact: sample data, codebook, and scriptsResearch Pathways
Product names and fielding platforms mentioned in this guide belong to their respective owners. Conjoint Survey is an independent academic research tool and is not affiliated with, sponsored by, or endorsed by those services.