Home / Tools / Percentile Calculator
Percentile Calculator
A percentile calculator turns a column of numbers into position measures: the value below which a given share of the data falls. Paste your data and read the full percentile table, from the 1st to the 99th plus any custom percentile, or flip it around and ask what percentile rank a particular value holds. It uses R's default type-7 rule (the same as Excel's PERCENTILE.INC), with a type-6 toggle and matching R code.
I want to find the value at a given percentile
-
How these numbers are computed
mean(x <= v), the same number R's ecdf(x)(v) returns.
✓ No data leaves your browser
✓ Verified against R's quantile(type = 7), quantile(type = 6) and ecdf()
✓ Free, no sign-up
Percentiles, both directions
A percentile is a position, not an average. Reading it two ways answers most real questions about where a number stands.
The 90th percentile is the value at or below which 90% of the data falls. It marks a cut-off: a lab's reference range, a service-level target, a grading boundary. Because it ignores how extreme the extremes are, one huge outlier barely moves it.
Turn it around: a score of 85, where does it stand? Its percentile rank is the share of the data at or below it. This is how test reports say "you scored higher than 72% of test takers," computed straight from the empirical distribution.
Type 7 is R's and Excel's default and interpolates between the two nearest order statistics. Type 6 (Minitab, SPSS) shifts percentiles slightly toward the tails. They match at the median and diverge most at the 1st and 99th, especially for small samples.
What each measure means
| Measure | What it tells you | In R |
|---|---|---|
| Percentile (type 7) | Value at or below which a given share of the data lies, R and Excel default. | quantile(x, p, type = 7) |
| Percentile (type 6) | Same idea, Minitab and SPSS interpolation, slightly further into the tails. | quantile(x, p, type = 6) |
| Quartiles | The 25th, 50th and 75th percentiles: the boundaries of a boxplot. | quantile(x) |
| Median | The 50th percentile, the middle value, robust to outliers. | median(x) |
| IQR | Q3 minus Q1, the width of the middle 50% of the data. | IQR(x) |
| Percentile rank | Share of the data at or below a value, the reverse of a percentile. | ecdf(x)(v) |
The calculator above reports the full percentile grid from this table for your data, in both type 7 and type 6, and runs the percentile rank for any value you enter.
Frequently asked questions
How do I calculate a percentile from a list of numbers?
Paste your values into the box, separated by commas, spaces, tabs or new lines. The calculator sorts them and reports the value at each percentile from the 1st to the 99th, plus any custom percentile you type. Switch to the other mode and it runs the reverse: give it a value and it returns that value's percentile rank, the share of the data at or below it.
What is the difference between a percentile and a percentile rank?
A percentile answers "what value sits at the 90th percentile?" and returns a number in your data's units. A percentile rank answers the reverse, "this value of 85, what percentile is it?" and returns a percentage. The 90th percentile is the value at or below which 90% of the data falls; the percentile rank of 85 is the percentage of values at or below 85. This tool does both directions.
Which method does this use, type 7 or type 6?
By default it uses R's type-7 quantiles, the same linear interpolation as quantile(x) with no type argument, which also matches Excel's PERCENTILE.INC and QUARTILE.INC. The toggle switches to type-6, used by Minitab and older SPSS, which places percentiles a little further into the tails. The two agree at the median and differ most near the 1st and 99th percentiles, especially for small samples.
What does the 90th percentile actually mean?
The 90th percentile is the value at or below which about 90% of the data lies, so only about 10% of values are higher. If a test score of 94 is the 90th percentile, 90% of test takers scored 94 or below. It is a position measure, not an average, and it is far more robust to a few extreme values than the mean.
Does this calculator match R and Excel?
Yes. Every value is checked against R's quantile() for both type 7 and type 6, and the percentile rank against ecdf(), across more than four hundred cases including ties, negatives, huge magnitudes and samples as small as one. Type 7 also matches Excel's PERCENTILE.INC. Excel's older PERCENTILE.EXC uses a different rule, close to type 6, and can differ near the tails.
Are the minimum and maximum the 0th and 100th percentiles?
Yes. Under type 7 and type 6 the 0th percentile is the minimum and the 100th percentile is the maximum, because the interpolation is anchored at the smallest and largest observed values. That is why the table runs from 1 to 99: the endpoints are just the extremes, which you can read as the min and max.