Median of five numbers calculator

Definition

For a finite sample of real numbers x₁, …, xₙ with odd n, the sample median is the middle value after ordering: if x₍₁₎ ≤ x₍₂₎ ≤ … ≤ x₍ₙ₎ are the order statistics, the median equals x₍₍ₙ₊₁₎/₂₎. For n = 5 that is x₍₃₎ — the third element of the ascending list. This calculator always uses exactly five fields a, b, c, d, e: it collects the values, sorts them ascending, and returns the element at index 2 (zero-based), matching the page script and the historical the page formula path.

The median is a measure of location (central tendency). Unlike the arithmetic mean it does not sum observations: it depends only on rank order. For the same five values 1, 2, 3, 4, 100 the median is 3 while the mean is 22. For odd n the definition is unique; for even n the usual convention averages the two central values — this form has no such path because n is always 5.

In the literature the median is the empirical 50th percentile (second quartile Q₂). NIST/SEMATECH e-Handbook lists the median beside the mean among location measures; MathWorld (Statistical Median) and Wikipedia (Median) give the same sort-based construction. Tukey’s Exploratory Data Analysis treats the median as the basic center description in exploratory work precisely because it resists outliers.

Historical and conceptual context

The idea of a middle value after ordering is older than the modern phrase “order statistics.” In the nineteenth century Francis Galton popularized percentiles and rank-based summaries; the median as the 50th percentile entered descriptive statistics alongside Quetelet’s mean. In the twentieth century the theory of order statistics (David & Nagaraja, Order Statistics) systematized the distributions of x₍ₖ₎, and the median became a standard location estimator with a high breakdown point.

John W. Tukey’s EDA (1977) made the median and the box-and-whisker plot everyday tools: the box center is the median, and the whiskers describe spread without assuming normality. In robust statistics the median is the canonical example of an estimator with a 50% breakdown point: more than half the observations must be corrupted before the result can be driven arbitrarily far — unlike the mean, which a single infinite value destroys.

Conceptually this calculator does not compute quartiles Q₁/Q₃, the interquartile range IQR, MAD (median absolute deviation), or variance. It also does not accept a variable-length list. An open list belongs on the neighboring median calculator; the mean belongs on the arithmetic-mean calculator. Here the goal is precise and didactic: five fields, one middle, an explicit mean comparison on the same numbers.

Formal properties

Input: five finite real numbers (a, b, c, d, e). Output: median m = x₍₃₎ after sorting. Always x₍₁₎ ≤ m ≤ x₍₅₎. If all values are equal, m equals that common value. Under a shift, m(xᵢ + c) = m(xᵢ) + c; under positive scaling λ > 0, m(λx) = λ m(x).

Robustness: changing one or two extreme observations (at n = 5) cannot move the median outside the interval fixed by the remaining three — the finite-sample breakdown fraction is floor((n−1)/2)/n = 2/5, and asymptotically the median reaches 50%. The arithmetic mean has breakdown 0. L¹ estimation: the median minimizes the sum of absolute deviations Σ|xᵢ − t| in t (for odd n the minimizer is x₍₃₎).

Field order does not affect the result because sorting is total. Zeros and negatives are allowed. Duplicates are allowed: for 2.5, 2.5, 7, 2.5, 9 the sorted list is 2.5, 2.5, 2.5, 7, 9, so the median is 2.5.

Full derivation

Step 1. Read five fields and parse them as real numbers (comma and period decimals are equivalent). An empty field blocks the result — the form requires a full set of five.

Step 2. Build L = [a, b, c, d, e] and sort ascending: L′ = sort(L) with numeric comparison (x − y), not lexicographic.

Step 3. Median m := L′[2] (zero-based) = x₍₃₎. That is the only primary result of this calculator.

Step 4 (expanded panel). Mean μ = (a+b+c+d+e)/5; min = L′[0]; max = L′[4]; range R = max − min; sum S = 5μ. These quantities do not replace the median — they support comparison and checking.

Correctness: a finite totally ordered chain of odd length always has a unique middle element. The browser sorts floating-point numbers; for typical school and engineering inputs the result matches hand calculation.

Worked path: 1, 5, 3, 4, 2 → L′ = [1, 2, 3, 4, 5] → m = 3. Robustness path: 1, 2, 3, 4, 100 → L′ = [1, 2, 3, 4, 100] → m = 3, μ = 22.

Special cases

All equal (10, 10, 10, 10, 10): median 10, range 0, mean 10. One outlier (1, 2, 3, 4, 100): median 3, mean 22. Permuted order (5, 1, 4, 2, 3): still 3. Fractions (−2, 0, 4, 1.5, 8): sorted −2, 0, 1.5, 4, 8, median 1.5. Middle ties (2.5, 2.5, 7, 2.5, 9): median 2.5.

This calculator does not handle n ≠ 5. It does not average two central values. It does not drop outliers by a filter — the median resists extremes by rank construction. Invalid text in a field yields a parse error message.

Numerical notes

Sorting five numbers is exact for finite floating-point arithmetic values in the range users typically enter. Avoid mixing thousands separators with a decimal comma in one token. Integer results format without trailing zeros; fractions keep needed precision.

The number-line diagram is an illustration, not a separate mathematical model. When all values are equal the axis expands visually so points remain readable, while the numeric median stays the common value.

Related concepts and sources

Related: arithmetic mean (sensitive to extremes), variable-length list median (even n), min/max and range (extreme order statistics), quartiles and IQR, MAD. In robust estimation the median is the starting point for Huber M-estimators and trimmed means.

Sources NIST/SEMATECH e-Handbook of Statistical Methods — Measures of Location; Wikipedia — Median / Order statistic; MathWorld — Statistical Median; John W. Tukey, Exploratory Data Analysis (Addison-Wesley, 1977); H. A. David & H. N. Nagaraja, Order Statistics (Wiley). Educational tool only — not a substitute for full statistical analysis.