Trimmed Mean Calculator

Calculate a robust average by trimming extreme values from both ends of your sorted dataset. Ideal for noisy data, outlier resistance, and fair summary statistics.

Calculator

10 means trim 10% from the low end and 10% from the high end.

Trim count = n × (trim%/100), then apply selected rule.

Trimmed Mean Calculator Guide: Formula, Steps, and Practical Use

What is a trimmed mean?

A trimmed mean is an average computed after removing a fixed percentage of the smallest and largest values in a dataset. It is a robust measure of central tendency designed to reduce the influence of extreme observations (outliers). Instead of letting one unusually high or low number dominate the average, trimmed mean focuses on the “middle bulk” of the data.

For example, if you use a 10% trim per tail, you remove 10% of values from the low end and 10% from the high end after sorting the data. Then you calculate the mean of what remains.

Why use trimmed mean instead of ordinary mean?

The arithmetic mean is sensitive to extreme values. In real datasets—exam scores, income values, process measurements, user behavior logs, or sensor data—rare spikes or anomalies can drag the mean away from the true “typical” level.

Trimmed mean formula

Let the sorted sample be x(1) ≤ x(2) ≤ ... ≤ x(n). Let p be trim fraction per tail (for 10%, p = 0.10). Define k = rule(n × p), where rule may be floor, round, or ceil based on analysis convention.

The trimmed mean is:

TM = (1 / (n − 2k)) × Σ x(i), for i from k+1 to n−k.

This means the first k values and last k values are removed before averaging.

How to calculate trimmed mean step by step

  1. Collect your data points.
  2. Sort numbers from smallest to largest.
  3. Choose trim percentage per tail (e.g., 5%, 10%, 20%).
  4. Compute trimmed count per tail, k.
  5. Remove lowest k and highest k values.
  6. Average the remaining values.
Tip: Most workflows use floor for k to avoid over-trimming in smaller samples.

Worked examples

Example 1 (with outlier): Data = 3, 4, 5, 6, 7, 8, 100. A regular mean is pulled upward by 100. With a 10% trim and n=7, k=floor(0.7)=0, so trimming does not remove any value. In small samples, you may need higher trim or a different rounding method. With 20% trim, k=floor(1.4)=1, remove 3 and 100, and average 4, 5, 6, 7, 8.

Example 2 (larger sample): Suppose n=30 and trim is 10%. Then k=floor(3)=3. Remove the 3 lowest and 3 highest values, and compute the mean of the remaining 24 observations.

How to choose trim percentage

There is no one-size-fits-all trim level. The best choice depends on data quality, sample size, and how frequently outliers occur.

For small samples, check how many values are actually removed. If k=0, your chosen trim may have no effect.

Trimmed mean vs mean vs median

Statistic Outlier Sensitivity Uses all values? Best use case
Mean High Yes Symmetric, clean distributions
Median Very low No (position-based) Skewed data, robust center
Trimmed Mean Moderate to low Mostly Robust averaging with more efficiency than median

Where trimmed mean is commonly used

Best practices for reliable results

Trimmed mean calculator FAQ

Is trimmed mean the same as winsorized mean?
No. Trimming removes extreme values. Winsorizing replaces extremes with boundary values before averaging.

Can trim percentage be 50%?
No. At 50% per tail, nothing remains to average. Practical upper bounds are well below 50%.

Does this calculator support decimals and negatives?
Yes. You can paste integers, decimals, and negative numbers separated by commas, spaces, or line breaks.

What if my dataset contains text?
Non-numeric tokens are ignored. For best accuracy, clean your data before calculating.

Use this page whenever you need a robust, practical average that resists outliers while preserving most of your data signal. The trimmed mean is one of the most useful statistics for real-world datasets where perfect normality is rare.