Trimmed Mean Calculator

Trimmed Mean Calculator

Calculate a more resistant average by removing the lowest and highest values before finding the mean of the remaining dataset.

What a trimmed mean does

A trimmed mean removes the same amount from both ends of a sorted dataset, then averages what remains. This helps reduce the influence of extreme outliers without ignoring the main pattern in the data.

For example, a 10% trimmed mean removes the lowest 10% and highest 10% of values before calculating the average.

Separate values with commas, spaces, or new lines
%

How to Use This Calculator

  1. Paste your dataset: Enter numbers separated by commas, spaces, or line breaks.
  2. Choose a trim method: Use percentage per tail for common trimmed means, or count per tail if you know exactly how many low and high values to remove.
  3. Calculate the result: The tool sorts the dataset, removes equal values from both ends, then averages the values that remain.
  4. Compare the outputs: Review the trimmed mean next to the original mean and median to understand how much the extremes affected the average.

Trimmed Mean Formula

A trimmed mean is calculated by sorting the dataset, removing the same number of values from the low and high ends, and taking the arithmetic mean of the remaining values.

A trimmed mean calculator calculates the average after removing a fixed percentage of the highest and lowest values. It reduces the impact of outliers and produces a more robust central value. For example, a 10% trimmed mean removes the top 10% and bottom 10% before averaging the remaining data.

Trimmed Mean = Sum of Retained Values / Number of Retained Values

If you choose a percentage, this calculator uses the floor of n x trim percentage to decide how many values to remove from each tail. That keeps the trim balanced and avoids removing partial observations.

When a Trimmed Mean Is Useful

A trimmed mean is helpful when your dataset has a few unusually high or low observations that may distort the ordinary average. It is often used in statistics, surveys, judged competitions, finance, and measurement data where one extreme value should not control the result.

  • Use it for outlier resistance: It keeps most observations while reducing the pull of extremes.
  • Use it for fair comparisons: It can make averages more stable when sample values vary widely.
  • Do not over-trim: Removing too much data can hide meaningful variation and make the result less representative.

How to Choose the Right Trim Percentage

The best trim level depends on how noisy your data is and how many values you can afford to remove. If you already have a result, jump to how to read the trimmed mean or when not to trim.

5% Trim

A light trim works well when the dataset is fairly clean and you only want mild protection from one or two unusual values.

10% Trim

This is a practical default for many real-world samples because it reduces the pull of extremes while still keeping most observations in the average.

20% Trim

A heavier trim can help with very skewed or messy data, but it removes much more information and should be used deliberately.

Use count-per-tail when a grading rule, lab protocol, or judging method says exactly how many low and high values must be removed from each side.

How to Read the Trimmed Mean Result

The most useful part of this calculator is not just the trimmed mean itself, but how it compares with the ordinary mean and the median. Looking at all three helps you see whether extreme values are changing the story your dataset tells.

  • If the trimmed mean is close to the original mean: Outliers are probably not having much influence.
  • If the trimmed mean is much lower than the original mean: A few high values are likely inflating the standard average.
  • If the trimmed mean is much higher than the original mean: A few unusually low values are likely pulling the standard average downward.
  • If the trimmed mean and median are close together: The center of the distribution is probably more stable than the raw mean suggests.

This comparison is especially helpful when you need to decide what number to report in a summary, chart, or analysis note.

When Not to Use a Trimmed Mean

A trimmed mean is powerful, but it is not always the right statistic. In some cases, removing the tails can hide information that matters more than the average.

  • Do not trim tiny datasets too aggressively: With a small sample, even one removed value can change the result a lot.
  • Do not trim when extremes are the point: Maximum wait times, safety failures, income inequality, and rare events may require keeping the full range of values visible.
  • Do not substitute it for a required method: If a class, research paper, or reporting standard calls for the ordinary mean, median, or weighted average, follow that rule.

A good next step is to report the trim rule clearly, such as 10% trimmed mean with one value removed from each tail, so readers know exactly how the result was produced.

Trimmed Mean Example

Step Example Result
Original data 12, 14, 15, 16, 18, 19, 21, 23, 24, 120 10 values
10% trim per tail Remove 1 lowest and 1 highest value Remove 12 and 120
Average remaining values 14, 15, 16, 18, 19, 21, 23, 24 Trimmed mean = 18.75

Interesting Fact

Trimmed-average thinking shows up in judged sports, not just statistics classes. World Aquatics explains that in artistic swimming scoring, the highest and lowest awards are removed and the three remaining marks are used, which helps reduce the effect of one unusually harsh or generous judge. In other words, 2 of the 5 awards are discarded before the routine score is calculated. Source: World Aquatics, Sport Terminology: Understanding Artistic Swimming.

Frequently Asked Questions

What is a trimmed mean in a dataset?

A trimmed mean is a statistic that finds an average after you sort a dataset and remove a fixed cutoff from each tail. It is useful when a few outlier values or unusual observations would otherwise pull the result too far away from the center of the data distribution.

What does a 10% trimmed mean remove from a sample?

A 10% trimmed mean removes the lowest 10% and highest 10% of values from the sample, then averages the middle 80%. In a calculator, that percentage works as the trim setting for each tail, so the exact cutoff depends on the sample size and the sorted data.

Is a trimmed mean the same as median or another average?

No. The median is the middle value in a sorted dataset, while the trimmed mean is an average based on many retained observations after the extremes are removed. In analysis, the trimmed mean often acts as a middle-ground statistic between the ordinary mean and the median because it still uses more of the distribution.

Can I use negative numbers as input?

Yes. This calculator accepts negative numbers because the trimmed mean formula relies on sorting and averaging, not on roots or logarithms. As long as your input values are valid numbers, the output is simply the trimmed average of the remaining data after the low and high tails are removed.

How much should I trim for analysis?

A 5% to 20% trim is common when you want a robust result without discarding too much data. The best percentage depends on your dataset, whether you expect outliers, and how much of each tail you want to remove before interpreting the final average. For a small sample, even a small cutoff can remove several important observations.

Disclaimer: This trimmed mean calculator is an educational tool for statistical estimates. Review your dataset, trim settings, and calculation requirements before using the result for research, finance, grading, or other important decisions.

Last updated: April 23, 2026