Basic Properties
The following properties are Properties 3 and 4 from Order Statistics from a Continuous Population.
Property 1: If the population follows the uniform distribution on the interval (0,1), the kth order statistic has a beta distribution Bet(k, n-k+1).
Property 2: If the population follows the uniform distribution on the interval (0,1), the mean and variance of the kth order statistic are
E.g. the mean of the 3rd order statistic from a sample of size 10 is 3/11 = .2727. We get the same value when using the formula =ORDER_MEAN(3,10,,”uniform”,0,1). See Order Statistics from a Continuous Population.
Observation: For a sample of size n from a population with a uniform distribution on (0,1), the distribution of the kth order statistic can be expressed in Excel by
fk(x) = BETA.DIST(x,k,n-k+1,FALSE)
Fk(x) = BETA.DIST(x,k,n-k+1,TRUE)
For any continuous distribution, the cdf F(x) is an increasing function on the interval (0,1). Thus, if x1, …, xn is a sample from this distribution in increasing order, then F(x1), …, F(xn) is a sample from the uniform distribution on (0,1) and the kth order statistic from the uniform distribution is F(x(k)).
Generalization
Property 3: For any distribution with cdf F(x)
Fk(x) = BETA.DIST(F(x),k,n-k+1,TRUE)
Proof: F(x) has a uniform distribution on (0,1) and so
F(x)∼ Bet(k, n-k+1)
and the result follows from Property 1.
Observation: For example, the kth order statistic for the gamma distribution Gamma(α, β) could be calculated in Excel by the formula
Fk(x) = BETA.DIST(GAMMA.DIST(x,α,β,TRUE),k,n-k+1,TRUE)
Joint Distribution
Property 4: For a uniform distribution on (0,1) where x < y
Proof: By Property 3 of Joint and Range Order Statistics from a Continuous Population
Property 5: For a uniform distribution on (0,1)
Proof: Click here for the proof using calculus.
Observation: Note that the distribution of x(k) – x(j) only depends on the difference k – j. Thus, for example, the distribution of x(5) – x(2) is the same as the distribution of x(7) – x(4).
Based on Property 5, we see that the mean and variance of x(k) – x(j) are
Also, note that
Thus
Solving for cov(x(k),x(j)), we get the covariance between two order statistics.
References
Border, K. C. (2016) Lecture 15: Order statistics; conditional expectation. Caltech
https://healy.econ.ohio-state.edu/kcb/Ma103/Notes/Lecture15.pdf
Ma D. (2010) The distribution of the order statistics. A Blog on probability and statistics
https://probabilityandstats.wordpress.com/2010/02/20/the-distributions-of-the-order-statistics/
Ma, D. (2010) The order statistics and the uniform distribution. A Blog on probability and statistics
https://probabilityandstats.wordpress.com/2010/02/21/the-order-statistics-and-the-uniform-distribution/