statistics

chi squared distribution

The distribution of a sum of squared Gaussian variables; the reference curve for goodness of fit.
8 related articles

Description

The chi-squared distribution is a probability distribution that arises naturally when you sum the squares of independent, standard normal (Gaussian) random variables. If you have k such variables, the sum of their squares follows a chi-squared distribution with k degrees of freedom. Its shape is skewed for small k and becomes more symmetric as k grows, eventually resembling a normal distribution.

History

Introduced by Karl Pearson in 1900 as part of his work on goodness-of-fit tests, specifically the chi-squared test.

How it works

The distribution is defined by a single parameter, the degrees of freedom (k). Its probability density function involves a gamma function and is zero for negative values, since squares are always non-negative. The mean equals k and the variance equals 2k. It is the sampling distribution of the sample variance for normally distributed data, which is why it underpins many statistical tests.

In practice

Used in hypothesis testing (e.g., chi-squared test for independence), in constructing confidence intervals for variance, and in model fitting to assess how well observed data match expected distributions.

Fun fact

The chi-squared distribution is a special case of the gamma distribution, with shape parameter k/2 and scale parameter 2.

Concept graph →

Related articles · 8

q-bio.PE2026-07-261 minarXiv:2607.23728express
When a Disease Wave Decides Its Own Speed
A disease outbreak is like a wildfire: the fire runs through dry grass, and its speed depends on wind and humidity. But an infection has its own 'wind'—how often people infect each other. When this rate changes along with the fraction of already infected, the epidemic wave can behave differently.
Cosmology2026-07-205 minarXiv:2607.17942
Gaussian Mask Falls: How 4 610 Universes Changed Cosmological Statistics
4 610 universes — that's exactly how many cosmological models astrophysicists needed to show that the familiar Gaussian mask no longer works for next-generation data. Future sky surveys, such as the Euclid mission and the Vera C.
math.OC2026-07-151 minarXiv:2607.13701express
How to Keep a Problem from Ballooning: A Fast Method for Machine Learning
The standard approach to a certain class of machine learning problems is like a spring that's been stretched and then forgotten: standard methods artificially inflate the problem's size, turning computations into a black hole for memory and time.
High Energy2026-06-102 minarXiv:2606.12205express
The Hidden Spin of Black Holes
Two black holes spiraling toward each other resemble figure skaters in a pair spin: each rotates about its own axis, and together they weave a shared dance. This collective twist, the so-called effective spin, can be measured by astronomers from the gravitational waves rippling outward from the pair, like ripples on a pond.
Statistical Mech2026-05-092 minarXiv:2605.08967express
When Disorder Gives Birth to Order
Spread a handful of coins on the table. Each coin is a possible outcome: heads, tails, or something else. If we know how chaotic this set is (its entropy), which combinations of probabilities will we encounter more often? It turns out the answer depends on how much chaos we allow.
Quantum Physics2026-05-053 minarXiv:2605.03807express
Almost Orthogonal Worlds
Toss a dozen coins on a table. The number of heads-and-tails combinations is just over a thousand. For two dozen coins, it's already a million. The number grows as 2 to the power of the number of coins, and that's fast. But quantum systems are far richer. Each particle here is not a coin with two sides but an arrow that can point in any direction.
HEP Theory2026-02-042 minarXiv:2602.04949express
Complexity on Graphs: How Quantum Walks Speed Up Time
A quantum walk is like a drunk person who turns randomly at every intersection. Except instead of streets, there are graph nodes, and instead of steps, there are quantum transitions. Scientists have discovered that such a walk has a hidden measure — Krylov complexity.
HEP Theory2026-02-021 minarXiv:2602.02682express
Universe Without a Beginning: Quantum Cosmology Confirms Hawking's Idea
The birth of the Universe is usually associated with the Big Bang — a point of infinite density. But at that point, the laws of physics break down, which bothers scientists. Stephen Hawking and James Hartle back in the 1980s proposed a different view: time has no boundary, like the surface of a globe.