Deciles slice a data set into ten equal-population bands — the language of income distribution, exam performance bands and benchmark reports everywhere. This decile calculator takes your numbers, sorts them, and returns all nine cut points, D1 through D9, using the same linear-interpolation percentile method as spreadsheet PERCENTILE functions, with each decile’s position shown so you can audit the arithmetic instead of trusting it.
How Deciles Are Computed
A decile is a percentile in tens: D1 is the 10th percentile, D5 the 50th (the median), D9 the 90th. This calculator uses the linear-interpolation method — the one Excel's PERCENTILE.INC and most statistics packages implement: for percentile p, the rank position is p × (n − 1) in the sorted data, interpolating between neighbours when the position falls between two values. The default data set (5, 8, 12, 15, 20, 22, 25, 30, 35, 40; n = 10) gives D1 = 7.70 (position 0.9, nine-tenths of the way from 5 to 8), D5 = 21.00, and D9 = 35.50, with the full run D1–D9 = 7.70, 11.20, 14.10, 18.00, 21.00, 23.20, 26.50, 31.00, 35.50.
Reading the Nine Numbers
Each decile is a border crossing. Ten percent of the data lies at or below D1; ten percent lies above D9; and the spread between D1 and D9 — the "middle 80 percent" — is a robust summary of dispersion that ignores outliers at both ends, which is precisely why income researchers quote decile ranges instead of min-to-max spans. The ratio D9/D1 is the classic inequality shorthand (how many times the top tenth's threshold exceeds the bottom tenth's), and D5 splits any distribution into its two halves. If your D9 sits far above D8 while D1 through D7 march evenly, the data is telling you its top tail is stretched — a story the mean alone would summarise into oblivion.
Deciles vs Quartiles vs Percentiles
One family, three granularities. Quartiles cut at 25/50/75 (Q1 = D2.5's neighbourhood, exactly the 25th percentile); deciles cut every 10; percentiles every 1. They are fully interconvertible: D7 is the 70th percentile, Q3 is the 75th, and this calculator's method matches the site's percentile tool so the numbers agree across pages. Which granularity to report is audience, not mathematics — exam boards publish deciles because ten bands are explainable, actuaries publish percentiles because their readers want the resolution, and a wise analyst computes at percentile fineness and reports at decile coarseness.
The Method Warning, Honestly Stated
There is no single God-given decile algorithm: textbooks teach at least four ranking conventions (nearest-rank, (n+1) positions, linear interpolation, and others), and for small samples they disagree in the decimals. This page uses linear interpolation on (n − 1) — the spreadsheet and NIST-consistent choice — and says so on the label, which is the professional habit: quote the method with the number. For large samples the methods converge to the same deciles, because with hundreds of points the neighbours being interpolated are close together; for n under about 20, treat the second decimal as method, not truth.
Where Deciles Earn Their Keep
Income and wealth distribution (the decile share tables behind every inequality debate), standardised testing (stanines and decile bands for reporting cohorts), growth and development benchmarks, housing-market reporting (deciles of prices by region), and any dashboard where "top decile" is the club being measured. In each, the decile's virtue is the same: it is a rank-based summary, immune to the unit of measurement and to the freak values at the extremes, so a billionaire entering the sample moves the mean violently and the deciles barely at all. Marketing borrowed the idea wholesale — the "top decile" of customers by lifetime value gets its own retention programme in most consumer businesses — and so did hospital benchmarking, where decile bands of waiting times survive the one catastrophic month that would poison an average.
Common Mistakes
Confusing "the first decile" (the bottom 10 percent group) with D1 the cut point — reports use both senses; know which yours means. Computing deciles on unsorted data by hand. Mixing methods between compared data sets, then debating differences the method manufactured. And reading D5 as the mean: it is the median, and in skewed data the gap between them is itself the finding.
Related Tools
The distribution bench continues in the Statistics Calculators hub — the Percentile Rank Calculator answers the reverse question (where does one score sit?), and the Sum of Squares Calculator feeds the variance family.
Frequently Asked Questions
What is a decile?
One of nine cut points dividing sorted data into ten equal-population parts. D1 is the 10th percentile, D5 the median, D9 the 90th percentile. Ten percent of the data falls below D1 and ten percent above D9.
How do you calculate deciles?
Sort the data, then for decile k find the value at position (k/10) × (n − 1), interpolating between the neighbouring values when the position is fractional — the method spreadsheets use for percentiles. For 5, 8, 12, 15, 20, 22, 25, 30, 35, 40, D1 = 7.7 and D9 = 35.5.
Is D5 the same as the median?
Yes — the fifth decile is the 50th percentile, the median. In the example data set D5 = 21, halfway between the 5th and 6th sorted values (20 and 22).
What is the difference between deciles and percentiles?
Granularity only. Deciles mark every tenth (10th, 20th, … 90th percentile); percentiles mark every hundredth. D7 and the 70th percentile are the same cut point.
Why do different calculators give slightly different deciles?
Several ranking conventions exist (nearest-rank, (n+1) position, linear interpolation). They differ in the decimals for small samples and converge for large ones. This calculator uses linear interpolation, the spreadsheet standard, and labels it.
What does the D9/D1 ratio mean?
It compares the threshold of the top tenth with the threshold of the bottom tenth — a standard inequality measure in income research. A D9/D1 of 5 says the top decile starts at five times the value where the bottom decile ends.