SURVEY SAMPLING AND SAMPLING DISTRIBUTION

 Population
The totality of observation with which we are concern is called population.
Example:
 Population of alphabets
 Population of students in university.
 Population of doctors in a hospital

The population is also called universe. The size of population is denoted by “N”.
 Sample
Any small part of population is called sample.

A sample is always taken to represent the characteristics of population. Sample size is

denoted by “n”.

 Parameter
The numerical quantities calculated for all population is called parameter.
Example:
 Population mean
 Population variance
 Population S.D
 Population proportion
Population parameters are usually constant.
 Statistic
The numerical quantities calculated only for sample is called statistic.
Example
 Sample mean
 Sample variance
 Sample S.D
 Sample Population
 Statistic are generally variance.
 Statistical Inference
The process of drawing conclusion about the population by using sample
information is called statistical inference.

 Sampling
It is a procedure by which we select the samples from the population.
Generally, we use the following two statistical techniques of sampling;
 Sampling with replacement method
 Sampling without replacement method

 Sampling With Replacement Method

If the selected unit is replaced to the population before selection the next unit then
such a technique of sampling is called sampling with replacement method.

Hence a sampling unit can be selected more than once in sampling with replacement
method.

If a population having N units out of which a sample of size n units select then we find all
possible samples by sampling with replacement method by using the following formula;

=(𝑵)𝑛

 Sampling Without Replacement Method

If the selected unit is not replaced to the population before selecting the next unit.
then such a technique of sampling is called sampling without replacement method. Hence
a sampling unit cannot be selecting more than once in sampling without replacement
method.

If a population having N units out of which n units have to selected by sampling

without replacement method. then we can find all possible samples by using the following
formula.

 By permutation
 By Combination
 Advantages of Sampling
The important advantages of sampling over the complete enumeration are given;

 Sampling saves a lot of time and energy than the complete enumeration.
 Sampling saves money as it is much cheaper to collect the data from a small sample then
from the whole population.
 Sampling provides information that is almost as accurate as that obtained from a complete
of enumeration.
 Sampling makes it possible to obtained more detailed information from each unit of the
sample as collecting data from a few units of the population.
 Sampling is the only method of collecting data in case of any infinite population.
 Sampling is essential to obtaining the data when the measurments process physically
damage or destroys the sampling unit under investigation.
 It is much easier to cover non-response error in sampling than the complete enumeration.
 Sampling is extensively used to obtained some of the cencus information.
 The most important advantage of sampling is that it provides a valid measure of reliability
for the estimator.
 Sampling Design
A sample design is a definite statistical plan concerned with the all principle test taken
in the selection of a sample and the estimation of procedure.

 Sampling Frame
A sampling frame is a complete list or a map that contains all the N sampling units in
a population.

 Sample Survey
The term survey has been defined as a mean of collecting information to meet a definite
need. The main steps in a sample survey are;

 Clearly state the objective of the survey.

 Define the population we wish the study as clearly as possible.
 Construct the sampling frame by clearly defining the sampling units.
 Choose an appropriate sample design and a proper sample size.
 Summarize and analyzed the data.

 Type of Sampling Design

There are two types of sample design;

 Probability sample design

 Non-probability sample design

 Probability Sampling Design

When each unit in a population has a known non-zero (not necessarily equal) prob
of its being included in the sample. Sampling is said to be probability sampling.

A probability sampling is also called Random Sampling. The major type of Prob
sampling are:

 Simple Random Sampling

 Stratified Random Sampling
 Cluster Sampling
 Systematic sampling

 Simple Random Sampling

The sampling is defined to be simple random sampling (SRS) if;

“Each unit in the population has an equall probability of being included in the sample.

each possible sample of the same size has an equall probability of being the sample
selected”

 Selection of Simple Random Sample

A simple random sample can be selected by the following method;

 Gold fish bowl method

 Using random number table
 Using mechanical method
 By using scientific calculator by using computer.
 Stratified Randome Sample
A sample of size n is defined to be a tartified random sample if it is selected from a
population which has been divided into a number of non-overlapping group or sub-
population, called strata, such that part of the sample is drawn at random from each stratmn.

Stated differently let a hetrogenus population containing N units then we divide these
hetrogenous N units into K subgroups N1, N2,…..NK in such a manner that N1, N2,…..NK
are homogenous called stratum. Now such the sampling technique is called Stratified
Random Sample.

 Allocation Method

 Equall allocation method.

 Proprtional allocation method
 Neymen allocation method.
 Optimum allocation method.

 Systhematic Sampling
A sample of size n is defined as to be a systematic random sample if it is obtained by
choosing one unit at random from the first k unit and thereafter selecting every Kth unit
after the N unit in the population have been serially numbered from 1-N or arranged in a
systhematic fashion. The letter K called the sampling interval.

i.e K= N/n in the systematic sampling.

 Cluster Sampling
A random sample is said to be a cluster sample if it consists of first selecting at random
group of indivisual unit called cluster (treated as sampling unit) into which a population
can be divided and then including in a sample either all the units from each of the choosen
cluster, or selecting a random sample of the units which the cluster comprises. Such a
sample is called cluster sample. Cluster sample required that the clusters should be as
internally dissimilar as possible and different cluster should be very similar.
 The procedure is called one stage cluster sampling. when all the units which each of the
qsample clusters comprises are included. if each of the sample cluster is sub sampled then the
sampling plan is called 2 stage cluster sampling. The plan is called multi stage cluster sampling.
when more than two stages are involved in taken the sampling. when the clusters related tpo
geographical areas then cluster sampling is also called Area sampling.

 Multi Stage Cluster Sampling

A sample is called a multi stage sample when it is selected in stages. the sampling unit at
each stage being sub sampled from the larger unit selected at the previous stage. here a
population is divided into a number of unit called first stage unit which are sub sampled. Each
of the selected second stage units is further divided into third stage units from which a sub
sampled is again selected and so on.

 Multi Phase Sampling

A sample is said to be two phase sample when certain items of information are collected
from all the units in a sample and other items of usually more detailed information are collected
from a subsample of the units corresponding to the original sample.

 Sequential Sampling
This is another method of sampling when the samples size is not fixed in advance but
sampling units are drawn one by one or in lots, and the decision is based on a definite rule
relating to the sampling unit themselves. That is we draw one unit at a time and after each draw
we make a decision whether to accept the lot or draw, whether to reject it or whether to continue
sampling.

 Non-Probability Sample
A non-prob sample or non-random sample is also called judgemental sample. The
important types of non-probability sample are;

 Purposive sample
 Quota Sample
  Purposive Sample
A purposive sample is a non-probability sampling in which the selection of the
sampling unit is based on the person’s expertise about the population. A purposive sample
is liable to bias to be introduce by the deliberate objective choice of the person who select
the sample. As the purposive sampling is not basedon probability theory, therefore there is
no objective method for measuring the reliability of the sample result and hence the
information gathers from such a sample cannot be made a basis of statistical Inference. it
is also called judggemental sampling.

 Quota Sampling
A quota sample is a type of judgemental sample. it is a sample, usually of human
being in which the information is called aulacted purposing from the segment of the
population (quota)

Example

 The quota of Men & Women

 The quota of Urban & Rurals.

 Errors in Sampling
There are two types of errors in sampling;
 Sampling error
 Non-Sampling

 Sampling Error
A sample being only a part of population cannot perfectly represent the population. No
matter how carefully the sample is selected. Thi results in a difference bbetween the values
of the sampling statistic and the true value of the corresponsding population. Such a
difference is called sampling error for the sample

 Non-Sampling Error
A part from the sampling error there are some errors which arrives because a sample
comprises only a portion of the population. These errors occurs at the stage of gathering or
collecting data, faulty of interview, inaccurate or non-response etc These errors are called
non-sampling errors.

 Sampling Distribution
It is defined as a probability distribution of the values of statistic such as mean, standard
deviation,, proportion etc computed from all possible samples of the same size which might
be selected with or without replacement from a population.

 Standard Error
The standard deviation of the sampling distribution of a sample statistic is called standard
error abrivated as S.E.

 Central Limit Theorem

The C.L.T is perhaps the most important theorem in all of statistical inference. it is
concerned with the means of large sample and provide the soultions when the shape of the
population distribution is unknown or highly skwed.

The cental limit theoram states that “if a variable X from a population has mean and
finite variance then the sampling distribution of sample mean approaches a normal
distribution with mean and variance S^2 /n as the sample size n approaches infinity”.