Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Statistics›Trimmed Mean Test (Yuen's Test)
Regression model

Trimmed Mean Test (Yuen's Test)

Two-Sample Trimmed Mean Test (Yuen's Test) · Also known as: Yuen's test, Yuen-Welch test, robust mean comparison, kırpılmış ortalama testi

The trimmed mean test compares two groups using trimmed means, which discard a fixed proportion of the most extreme observations in each tail before averaging. Introduced by Karen K. Yuen in 1974, it is a robust alternative to the classical t-test when the data are non-normal or contain outliers and the population variances are unequal.

ScholarGate
  1. Regression model
  2. v1
  3. 2 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Trimmed Mean Test
MAD EstimationMann-Whitney U testPermutation TestWelch t-testBootstrap InferenceWinsorized Estimation

When to use it

Use the trimmed mean test to compare two independent groups on a continuous outcome when normality is doubtful or outliers are present, especially if the group variances differ. It needs independent observations and a reasonable sample (at least about 20 per group), and it assumes the trimmed data still represent the phenomenon of interest. It is a strong default when a plain t-test would be distorted by heavy tails but you still want to reason about a mean-like centre.

Strengths & limitations

Strengths
  • Resistant to outliers and heavy tails because the most extreme observations are removed before comparison.
  • Handles unequal population variances through Yuen-Welch degrees of freedom, unlike the pooled t-test.
  • Keeps a familiar mean-difference interpretation while gaining robustness, and is simple to compute and report.
Limitations
  • Discarding tail observations sacrifices some information and statistical efficiency when the data really are normal.
  • The result depends on the chosen trimming proportion, which must be fixed in advance rather than tuned to the data.
  • Unreliable in very small samples (below about 10 per group) and when the outlier share is so high that even trimming cannot stabilise the centre.

Frequently asked

How much should I trim?

Trimming 10-20% from each tail is the common default and works well across many distributions. The key rule is to fix the proportion before looking at the test result, since choosing it to get a significant outcome invalidates the test.

How is this different from the Mann-Whitney U test?

Both are robust to outliers, but the trimmed mean test compares trimmed means and keeps a mean-like, interpretable difference, whereas Mann-Whitney compares stochastic dominance through ranks. Use the trimmed mean test when you want to reason about a central tendency on the original scale.

What happens if there are too many outliers?

When the outlier share climbs above roughly 25%, even trimming cannot fully stabilise the centre. You should either raise the trimming proportion or switch to a median-based or MAD-based comparison.

Can I use it on small samples?

It needs a reasonable sample, about 20 per group, to be reliable. Below roughly 10 observations per group the trimmed mean test becomes untrustworthy and a permutation test is the better choice.

Sources

  1. Yuen, K. K. (1974). The Two-Sample Trimmed t for Unequal Population Variances. Biometrika, 61(1), 165-170. DOI: 10.1093/biomet/61.1.165 ↗
  2. Wilcox, R. R. (2012). Introduction to Robust Estimation and Hypothesis Testing (3rd ed.). Academic Press. ISBN: 978-0123869838

How to cite this page

ScholarGate. (2026, June 1). Two-Sample Trimmed Mean Test (Yuen's Test). ScholarGate. https://scholargate.app/en/statistics/trimmed-mean-test

Related methods

MAD EstimationMann-Whitney U testPermutation TestWelch t-test

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • MAD EstimationStatistics↔ compare
  • Mann-Whitney U testStatistics↔ compare
  • Permutation TestStatistics↔ compare
  • Welch t-testStatistics↔ compare
Compare side by side →

Referenced by

Bootstrap InferencePermutation TestWinsorized Estimation

Similar methods

Robust independent samples t-testRobust one-sample t-testRobust paired samples t-testRobust Mann-Whitney U testRobust ANOVARobust one-way ANOVAIndependent t-testWelch t-test

Related reference concepts

Rank-Based MethodsMultivariate Analysis of VarianceData Distribution and NormalityNonparametric StatisticsMeasures of Central TendencyPermutation Tests

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Trimmed Mean Test (Two-Sample Trimmed Mean Test (Yuen's Test)). Retrieved 2026-07-21 from https://scholargate.app/en/statistics/trimmed-mean-test · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Karen K. Yuen
Year
1974
Type
Robust two-group comparison
Estimator
Trimmed mean with Winsorized variance
Outcome
continuous
DefaultTrim
10-20% per tail
MinSample
20
Related methods
MAD EstimationMann-Whitney U testPermutation TestWelch t-test
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account