Natalia Mora Ramírez

UNIT 2 – Sampling distributions

1. BASIC NOTIONS AND SAMPLING DISTRIBUTION DEFINITION

- Parameter: mathematical function defined about the numeric values relating to measurable
characteristics in a population. The parameter is a fixed (but unknown) number.

- Statistic: mathematical function defining numeric values related to measurable

characteristics in a sample. As each random sample from the same population could be
different, the statistic can be too. Because of this, it is considered a random variable.

We need some information about the parameters to estimate them through their own
statistics. This information is the sample statistics equal to the parameter. For instance, if
we can estimate the mean of a population, we use the mean of the sample.

As the statistic is considered a random variable it’ll have own probability distribution.

The Sampling Distribution of a statistic is the probability distribution of the values it could
take over all possible samples of the same number of observations drawn from the
population.

The concept of a sampling distribution is very important because it permits us:

- To select the best statistic when estimating a parameter based on the characteristics of its
probability distribution.

- To assign probabilities of error or a correct answer in the estimation of the parameter from
the statistic.

2. SAMPLING DISTRIBUTION OF THE SAMPLE MEAN. THE CENTRAL

LIMIT THEOREM.
Suppose that the sample mean X is the mean of a sample of n observations from a population
with mean X and variance 𝜕2X

The mean of the sampling distribution of the sample mean X is the population mean X

1
Natalia Mora Ramírez

Example: Suppose that we want to obtain the sampling distribution of the sample mean of the
age of a group (population) with 70 students. We take samples with two observations with
replacement. The age distribution is the following:

As there is replacement, the probabilities are constant:

40 20 10
P(X=19)=70 ; P(X=20)=70 ; P(X=21)=70

We can obtain the sampling distribution of the sample:

As the events are independent the probability of their intersection will be the product of each
one’s probabilities.
If we calculate the mean of the sample means:

2
Natalia Mora Ramírez

And afterwards we calculate the mean of the age of the entire population (N= 70 students)

Conclusion: The mean of the sample means coincides with the population mean (µ); therefore,
we can use the sample mean to estimate the population mean.
It is also of interest to determine how close the sample mean is likely to be to the population
mean. The answer will depend on the spread, or variance in the sampling distribution.
If the number of population members is very large compared to the number in the sample, then
the distributions of the individual sample members are independent of one another.
In this case, the variance of a sum is the sum of variances, we have:

Since each Xi has a variance of 𝜕2X , it follows that:

𝜕2X + 𝜕2X + 𝜕2X + ….+ 𝜕2X = n𝜕2X

Therefore, the variance of the sample mean is:

In conclusion, we can affirm that Sampling distribution of sample mean X has the standard
deviation:

Where:
n: the sample size
𝜕X: value of the population standard deviation
The standard deviation of the sample mean, or “standard error of sample mean” is an accuracy
indicator of the sample mean as an estimator of the population mean.

Let X denote the mean of a random sample of n observations from a population with mean and
variance:

E(Xi)= X ; Var(Xi)= 𝜕 2X i=1,…..,n;

3
Natalia Mora Ramírez

𝜕2
The sampling distribution of X has mean X and variance
𝑛

If the population distribution is normal, then the random variable X is normal:

EXAMPLE: Suppose that Annual salary increase percentage for the chief executive officers of
all mid-size corporations are normally distributed with mean 12.2% and standard deviation
3.6%. A random sample of 9 observations from this population of percentage salary increases is
taken. What is the probability that the sample mean will be less than 10%?
Solution: 0.0336

CENTRAL LIMIT THEOREM

If the population distribution is not normal but the sample size, n, is moderately large (n≥30),
the distribution of the sample mean is distributed as standard normal.