Definition 1: The chi-square distribution with k degrees of freedom, abbreviated χ²(k), has probability density function

k does not have to be an integer and can be any positive real number.

Click here for more technical details about the chi-square distribution, including proofs of some of the propositions described below. Except for the proof of Corollary 2 knowledge of calculus will be required.

Observation: The chi-square distribution is the gamma distribution where α = k/2 and β = 2.

Property 1: The χ²(k) distribution has mean k and variance 2k

Observation: The key statistical properties of the chi-square distribution are:

• Mean = k
• Median ≈ k(1–2/(9k))^3
• Mode = max (k – 1, 0)
• Range = [0.∞)
• Variance = 2k
• Skewness = 
• Kurtosis = 12/k

The following are the graphs of the pdf with degrees of freedom df = 5 and 10. As df grows larger the fat part of the curve shifts to the right and becomes more like the graph of a normal distribution.

Figure 1 – Chart of chi-square distributions

Theorem 1: Suppose x has standard normal distribution N(0, 1) and let x₁, …, x_kbe k independent sample values of x, then the random variable

has a chi-square distribution χ²(k).

Corollary 1:

1. If x has distribution N(0, 1) then x² has distribution χ²(1)
2. If x ~ N(μ, σ) and z = (x–μ)/σ then over repeated samples z² has distribution χ²(1)
3. If x₁, …, x_kare independent observations from a normal population with normal distribution N(μ,σ) and for each i, z = (x–μ)/σ , then the following random variable has a χ²(k) distribution
   

Property 2: If x and y are independent and x has distribution χ²(m) and y has distribution χ²(n), then x + y has distribution χ²(m + n)

Theorem 2: If x is drawn from a normally distributed population N(μ,σ) then for samples of size n the sample variance s² has distribution

Corollary 2: s² is an unbiased, consistent estimator of the population variance

Corollary 3: If x is drawn from a normally distributed population N(μ, σ), then for samples of size n the random variable  has a χ²(n–1), distribution

Property 3: The mean of the sample variance s² is σ² and the variance is 

Proof: This can be seen from the proof of Corollary 2.

Excel Functions: Excel provides the following functions:

CHIDIST(x, df) = the probability that the chi-square distribution with df degrees of freedom is ≥ x; i.e. 1 – F(x) where F is the cumulative chi-square distribution function.

CHIINV(α, df) = the value x such that CHIDIST(x, df) = 1 – α; i.e. the value x such that the right tail of the chi-square distribution with area α occurs at x. This means that F(x) = 1 – α, where F is the cumulative chi-square distribution function.

With Excel 2010/2013 there are a number of new functions (CHISQ.DIST, CHISQ.INV, CHISQ.DIST.RT and CHISQ.INV.RT) that provide equivalent functionality to CHIDIST and CHIINV, but whose syntax is more consistent with other distribution functions. These functions are described in Built-in Statistical Functions.

In Excel 2010 CHISQ.DIST(x, df, TRUE) is the cumulative distribution function for the chi-square distribution with df degrees of freedom, i.e. 1 – CHIDIST(x, df), and CHISQ.DIST(x, df, FALSE) is the pdf for the chi-square distribution.

Real Statistics Functions: The Real Statistics Resource Pack provides the following functions.

CHISQ_DIST(x, df, cum) = GAMMA.DIST(x, df/2, 2, cum) = GAMMADIST(x, df/2, 2, cum)

CHISQ_INV(p, df) = GAMMA.INV(p, df/2, 2) = GAMMAINV(p, df/2, 2)

These functions provide better estimates of the chi-square distribution when df is not an integer. The first function is also useful in providing an estimate of the pdf for versions of Excel prior to Excel 2010, where CHISQ.DIST(x, df, FALSE) is not available.

The Real Statistics Resource also provides the following functions:

CHISQ_DIST_RT(x, df) = 1 – CHISQ_DIST(x, df, TRUE)

CHISQ_INV_RT(p, df) = 1 – CHISQ_INV(p, df)

Example 1: Suppose we take samples of size 10 from a population with normal distribution N(0, 2). Find the mean and variance of the sample distribution of s².

By Property 3