Probstat/notes/sample means and sample variances
- This is part of probstat
Consider a certain distribution. The mean of the distribution is the expected value of a random variable sample from the distribution. I.e.,
.
![{\displaystyle \mu =E[X]}](https://wikimedia.org/api/rest_v1/media/math/render/svg/4b8d5d31fd9dc7cd9849f1a07a72e82420448133)
Also recall that the variance of the distribution is
![{\displaystyle \sigma ^{2}=Var(X)=E[(X-\mu )^{2}]=E[X^{2}]-E[X]^{2}.}](https://wikimedia.org/api/rest_v1/media/math/render/svg/413cc5ad0d25f8ab364154479aee3f44d26e3cc0)
And finally, the standard deviation is .
เนื้อหา
Sample Statistics
Suppose that you take samples independently from this distribution. (Note that are random variables.)
Sample means
The statistic
is called a sample mean. Since are random variables, the mean is also a random variable.
We hope that approximates well. We can compute:
![{\displaystyle E[{\bar {X}}]=E\left[{\frac {1}{n}}\sum _{i=1}^{n}X_{i}\right]={\frac {1}{n}}E\left[\sum _{i=1}^{n}X_{i}\right]={\frac {1}{n}}\sum _{i=1}^{n}E[X_{i}]={\frac {1}{n}}n\mu =\mu }](https://wikimedia.org/api/rest_v1/media/math/render/svg/2df57bd2c3c1cfc8a2937d14ac740a5627a5a58c)
and since are independent, we have that
Sample variances and sample standard deviations
We can also use the sample to estimate .
The statistic
is called a sample variance. The sample standard deviation is .
Note that the denominator is instead of .
We can show that .
We note that since and are independent, we have that
![{\displaystyle E[X_{i}X_{j}]=E[X_{i}]E[X_{j}]=\mu \cdot \mu =\mu ^{2}}](https://wikimedia.org/api/rest_v1/media/math/render/svg/3c180ef6fd6cf5f43cac88f3106d697eb57e4d44)
Let's deal with the middle term here.
Let's work on the third term which ends up being the same as the middle term.
![{\displaystyle E\left[\left(\sum _{j=1}^{n}X_{j}\right)^{2}\right]=E\left[\sum _{j=1}^{n}\sum _{k=1}^{n}X_{j}X_{k}\right]=\sum _{j=1}^{n}\sum _{k=1}^{n}E[X_{j}X_{k}]=nE[X^{2}]+n(n-1)\mu ^{2}.}](https://wikimedia.org/api/rest_v1/media/math/render/svg/bf78bddcae38881f34e4191e061ae76bce37c466)
Let's put everything together:
![{\displaystyle {\begin{array}{rcl}\mathrm {E} [S^{2}]&=&{\frac {1}{n-1}}\left(\sum _{i=1}^{n}E[X_{i}^{2}]-(2/n)\cdot \sum _{i=1}^{n}\sum _{j=1}^{n}E\left[X_{i}\cdot X_{j}\right]+(1/n)\cdot E\left[\left(\sum _{j=1}^{n}X_{j}\right)^{2}\right]\right)\\&=&{\frac {1}{n-1}}\left(nE[X^{2}]-(2/n)(nE[X^{2}]+n(n-1)\mu ^{2})+(1/n)(nE[X^{2}]+n(n-1)\mu ^{2})\right)\\&=&{\frac {1}{n-1}}\left(nE[X^{2}]-2E[X^{2}]-2(n-1)\mu ^{2}+E[X^{2}]+(n-1)\mu ^{2}\right)\\&=&{\frac {1}{n-1}}\left((n-1)E[X^{2}]-(n-1)\mu ^{2}\right)\\&=&E[X^{2}]-\mu ^{2}=\sigma ^{2}\\\end{array}}}](https://wikimedia.org/api/rest_v1/media/math/render/svg/d23da700b512a8307a14076e1ad12f1b64166190)
Summary
Sample means:
Sample variance:
Properties of sample means and sample variances
Distribution of sample means
While we know basic properties of sample means , if we want to perform other statistical calculation (i.e., computing confidence intervals or testing hypotheses), it is very useful to know the exact distribution of .
For a general population, it will be hard to deal the the distribution of exactly. However, if the population is normal, we are in a very good shape.
Recall the definition of :
Therefore, is a sum of independent normally distributed random variables. A nice property of normal random variables is that the sum of normally distributed random variables remains a normal random variable. Since a normal random variable is uniquely determined by its mean and variance, we have the following observation.
Distribution of sample means of normal populations.
If the population is normally distributed with mean and variance , the distribution of is normal with mean and variance .
Examples
Ex1. Suppose that the population has mean and variance . If you select a sample of size 20, what is the probability that the sample mean is greater than 17?
Solution:
The sample mean is normal with mean and variance . Therefore,
is unit normal.
Note that
We can look at the standard normal table and find out that , for a unit normal random variable Z. Thus, the probability
which is roughly 1%.
Ex2.
- To be added...
Why do we use normal distributions?
Normal random variables appear very often in our treatment of statistics. This is not just a coincidence. See limit theorems.
