Sampling distribution

From Wikipedia, the free encyclopedia

In statistics a sampling distribution or finite-sample distribution is the probability
distribution of a given statistic based on a random sample. Sampling distributions are
important in statistics because they provide a major simplification on the route to
statistical inference. More specifically, they allow analytical considerations to be based
on the sampling distribution of a statistic, rather than on the joint probability distribution
of all the individual sample values.
Contents
Introduction
! Standard error
" #$amples
% Statistical inference
& 'eferences
( #$ternal links
Introduction
)he sampling distribution of a statistic is the distribution of that statistic, considered as
a random variable, when derived from a random sample of si*e n. It may be considered as
the distribution of the statistic for all possible samples from the same population of a
given si*e. )he sampling distribution depends on the underlying distribution of the
population, the statistic being considered, the sampling procedure employed, and the
sample si*e used. )here is often considerable interest in whether the sampling distribution
can be appro$imated by an asymptotic distribution, which corresponds to the limiting
case as n .
For e$ample, consider a normal population with mean and variance +. ,ssume we
repeatedly take samples of a given si*e from this population and calculate the arithmetic
mean for each sample - this statistic is called the sample mean. #ach sample has its
own average value, and the distribution of these averages is called the .sampling
distribution of the sample mean.. )his distribution is normal /n is the sample
si*e0 since the underlying population is normal, although sampling distributions may also
often be close to normal even when the population distribution is not /see central limit
theorem0. ,n alternative to the sample mean is the sample median. When calculated from
the same population, it has a different sampling distribution to that of the mean and is
generally not normal /but it may be close for large sample si*es0.
)he mean of a sample from a population having a normal distribution is an e$ample of a
simple statistic taken from one of the simplest statistical populations. For other statistics
and other populations the formulas are more complicated, and often they don1t e$ist in
closed2form. In such cases the sampling distributions may be appro$imated through
Monte23arlo simulations
454p. !5
, bootstrap methods, or asymptotic distribution theory.
Standard error
)he standard deviation of the sampling distribution of a statistic is referred to as the
standard error of that 6uantity. For the case where the statistic is the sample mean, and
samples are uncorrelated, the standard error is7
where is the standard deviation of the population distribution of that 6uantity and n is
the si*e /number of items0 in the sample.
,n important implication of this formula is that the sample si*e must be 6uadrupled
/multiplied by %0 to achieve half /8!0 the measurement error. When designing statistical
studies where cost is a factor, this may have a role in understanding cost2benefit tradeoffs.
Examples
Population Statistic Sampling distribution
9ormal7
Sample mean
from samples of
si*e n
:ernoulli7
Sample
proportion of
.successful
trials.
)wo
independent
normal
populations7
a
nd
;ifference
between sample
means,
,ny
absolutely
continuous
distribution F
with density
Median
from a sample
of si*e n < !k =
, where sample
is ordered
to
,ny
distribution
with
distribution
function F
Ma$imum
from a random
sample of si*e n
Statistical inference
In the theory of statistical inference, the idea of a sufficient statistic provides the basis of
choosing a statistic /as a function of the sample data points0 in such a way that no
information is lost by replacing the full probabilistic description of the sample with the
sampling distribution of the selected statistic.
In fre6uentist inference, for e$ample in the development of a statistical hypothesis test or
a confidence interval, the availability of the sampling distribution of a statistic /or an
appro$imation to this in the form of an asymptotic distribution0 can allow the ready
formulation of such procedures, whereas the development of procedures starting from the
joint distribution of the sample would be less straightforward.
In :ayesian inference, when the sampling distribution of a statistic is available, one can
consider replacing the final outcome of such procedures, specifically the conditional
distributions of any unknown 6uantities given the sample data, by the conditional
distributions of any unknown 6uantities given selected sample statistics. Such a
procedure would involve the sampling distribution of the statistics. )he results would be
identical provided the statistics chosen are jointly sufficient statistics.
)his article needs additional citations for verification. >lease help improve this
article by adding citations to reliable sources. ?nsourced material may be
challenged and removed. (May 2014)
References
. Mooney, 3hristopher @. /AAA0. Monte arlo simulation. )housand Baks,
3alif.7 Sage. IS:9 ACDEDE"A&A%"A.
Merberg, ,. and S.F. Miller /!EED0. .)he Sample ;istribution of the Median..
ourse !otes for Math 1"2# Mathemati$al %tatisti$s, on the web at
http788web.williams.edu8Mathematics8sjmiller8publicGhtml8:rown3lasses8(!8Ha
ndouts8Median)hmE%.pdf, pgs 2A.
External links
Ienerate sampling distributions in #$cel
Mathemati$a demonstration showing the sampling distribution of various
statistics /e.g. J & +0 for a normal population
[hide]
v
t
e
Statistics
[sho]
!escriptive statistics
[sho]
!ata collection
[sho]
Statistical inference
[sho]
Correlation
Regression anal"sis
[sho]
Categorical # $ultivariate # %ime-series # Survival anal"sis
[sho]
&pplications
3ategory
>ortal
Butline
Inde$
3ategories7
Statistical theory
'avigation menu
3reate account
Kog in
,rticle
)alk
'ead
#dit
Liew history
Main page
3ontents
Featured content
3urrent events
'andom article
;onate to Wikipedia
Wikimedia Shop
Interaction
Help
,bout Wikipedia
3ommunity portal
'ecent changes
3ontact page
%ools
What links here
'elated changes
?pload file
Special pages
>ermanent link
>age information
Wikidata item
3ite this page
Print#export
3reate a book
;ownload as >;F
>rintable version
(anguages
;eutsch
#spaMol
NOPQR
:asa Sunda
#dit links
)his page was last modified on A Fune !E% at EA7&E.
)e$t is available under the 3reative 3ommons ,ttribution2Share,like KicenseS
additional terms may apply. :y using this site, you agree to the )erms of ?se and
>rivacy >olicy. WikipediaT is a registered trademark of the Wikimedia
Foundation, Inc., a non2profit organi*ati