Free Calculator

MaxDiff Sample Size Calculator

Calculate the minimum sample size for MaxDiff (Best-Worst Scaling) studies based on the number of items, set size, and tasks.

Study Design Parameters

How many features, messages, or concepts in your MaxDiff study (typically 10–30)

How many items appear in each best-worst choice task (typically 4–5)

Number of best-worst sets each respondent completes (typically 8–15)

⚠️ Each item appears only 2.7× per respondent. At least 3 appearances are needed for stable estimates. Increase tasks or reduce items.

Minimum Sample

188

Statistical floor

Recommended Sample

226

With design buffer

With Recruitment Buffer

260

+15% for dropouts

Item appearances per respondent

With this design, each item appears 2.7× per respondent. Increase tasks to improve stability.

Want the methodology behind this?

Read our full guide on maxdiff analysis, covering study design, interpretation, and common mistakes.

Read the guide