Basic concepts in sampling theory

1 Population: A group of individuals having same characteristics in same surrounding is known as population.
According to A. C. Rosander “A population is the totality of objects under consideration”. In short, group of all
objects which are coming under the definition of investigation unit. For example, the group of employees of a institute
shall be the population for every investigation related to the employees. Individuals of the population are known as a
unit or element in population. Hence, a group of individual or item under study is called population.
The total no of individual item of population is called as a population size of that population .The population size of
population is generally denoted by N i.e. the population size N may be finite or infinite.
Finite or Infinite Population: In a finite population the number of units is finite i.e. number of students in
a class, number of employees in a college. In a infinite population, the number of units are infinite, for example the
number of hair in head etc. and also if the units of a population are unable to count and the limits cannot be made then
the population is known as Infinite population, for example stars in sky, leaves in tree, number of viewers of a T.V.
serial, etc.
Real or Hypothetical Population : Population of a concrete subject is called real population. For example,
students in a university, employees of an institution. In a hypothetical population the subjects are not concrete,
whether they are hypothetical. For example, the population made by the number of Heads or Tails based on the trial of
a coin.
[Link] units: The elementary unit is a member of the population whose measurement on the variable of
interest is what we wish to examine. In other words the elementary unit is the ultimate unit to be sampled for the
purpose of the survey. It is distinguishable and identifiable.
3. Sampling unit: Before selecting a sample for a given population it is necessary that the population to be sample
must be subdivided into finite no of items, non-overlapping and distinguishable units .This units must cover the entire
population such unit are called sample units. Sampling unit may consist, one or more elementary units and each
elementary unit belongs to one and one sampling units. The sampling units must cover the entire population and they
must be distinct, unambiguous and non-overlapping in the sense that every element of the population belongs to one
and only one sampling unit. In a Socio economic survey, whether a family or a member of a family is to be the
ultimate sampling unit.
4. Sampling Frame: Before selecting a sample it is essential for purpose a sampling to prepare complete list of all
sampling units of the population the sample such a list should be include all the sampling units of population ,
these list sampling units are called sampling frames. Hence A list of all the sampling units belonging to population
to be studied with their identification particulars or a map showing the boundaries of sampling units is known as
sampling frame. A sampling frame represents the population under investigation and it is base for drawing sample.
As far as it should be up to date i.e. free from duplicates and omissions. For e,g A list of village ,a list of collages
Census or complete enumeration:
In census, we study about each and every unit of the population. Population means total units of investigation area.
Incensus, whole group related to investigation is investigated and the information‟s are collected, i.e. Census of
population of a country, Census of import and export, etc.
Census investigation is useful in following situations:
1. When a deep study to be performed;
2. When study area is limited;
3. When an adequate accuracy and reliability is desired;
4. When investigator have resources; and
5. When use of sampling method is tough and prohibited.
Merits and Demerits of Census
Merits:
1. Useful in Heterogeneity: This method is very appropriate when the units are heterogeneous from each other and hard
to be succeeded for sampling method.
2. Deep Study is Possible: Through census, deep study of the subject is possible so that the investigator can get the total
information of the variable of interest. He also knows the things or subjects which otherwise overlooked.
3. High Level of Accuracy: High level of accuracy is expected by following this method. Because of that the personnel
investigation is to be performed in investigation area therefore the results are accurate at higher degree.
4. Necessary in Some Situation: If the nature of investigation is like that the involvement of all units is necessary than
census is necessary i.e. census of population of a country.
Demerits
1. Useless in Case of Destructive Units: If the units are destructive type and been destroyed by examining, this method
is useless. For example, to study about the hardness of a chawk or quality of crackers or life of bulbs or tube lights.
2. More Time and Energy: This method is very time consuming and large number of persons required to complete the
process. This method consumes much of energy and hard work to perform the study.
3. More Expensive: Much of time and organization of big size is needed for a census investigation. A large number of
field investigators to be involved in this work and arrangement of their training is also needed. In all, in this
process large scale expenditure is needed.
4. Investigation Remains Incomplete: In this method time, money, organization and large number of field investigators
are required. Population is also large. Therefore, the investigation may remain incomplete due to weakness of
investigating team or time or lack of availability of resources. In that situation, the effort or expenses which have
already done that become useless.
5. Inconvenient: This method is very inconvenient because this needs a whole department to be established separately.
Problems related to management arise.
6. Not Possible for all: This method is not comfortable for all because this method can be used solely by powerful
person or organizations.
7. Not Possible in Every Situation: Census is not possible from many reasons in various situations, where the
investigation area is large and wide. In such cases contact to each and every unit is not possible.
8. Statistical Error: In this method we cannot have the knowledge of statistical error.
Need of Sampling:
If the population is infinite, census is not possible. Also, if the units are destroyed in the course of inspection, 100%
inspection though is not possible at all desirable. But even if the population is finite or the inspection is not
destructive, 100% inspection is not taken recourse too, because of the administrative, financial and time factor
related problems. So we take the help of sampling.
Advantages of sampling over census method :
[Link] cost: If data is collected from only small fraction of population expenditures smaller than if a complete
census is attempted with dirge population .we can get accurate result from samples that represent a population
[Link] speed (reduce time): Data from small fraction can be collected be maximum speed than complete census
analysis .when the information is urgently required then we consider only sample method.
3. Greater scope: There is generally a greater scope in a sample survey than in census. Some inquiries may require
highly trained personnel or specialized equipment for collection of data, then the census may be inconceivable
4. Reliability: For sampling survey we can employee person of high quality they can make careful supervision of field
work .a sample gives more accurate results than complete enumeration.
5. Details of information: Due to small sample every member of sample can be studied carefully and detailed
information can be obtained. In case of complete enumeration due to huge size detail study of each member is
not possible.
6. Only method: If population is too large or spread over a large geographical area for e.g. trees in jungle, the fishes in
the sea , it become very difficult to collect information of all items. In this case sampling is only method.
7. Destructive nature of certain test.: There are of destructive nature i.e. if quality of an article can be determined only
by destroying the article in the process of testing e.g. a) testing of crackers or explosives b) testing of an
electrical items such as light bulb or tube c) testing quality of milk d) testing blood of individual . If we use
census in this case none item will be available for use. So in such cases sampling is only method.
8. If population is hypothetical , sampling method is only scientific method of estimating the characteristics
(parameter) of population.
9. A sample survey gives data of better quality than a complete census, because in a sample survey it may be possible
to use better resources than complete census.
Non-Probability Sampling
In this method, the sample is selected with a definite purpose in view and the choice of the sampling units depends
entirely on the discretion and judgment of the investigator. While selecting a sample, investigator tries to include
each and every characteristics of population in sample.
Non-Probability Sampling scheme can be classified as:
1. Purposive Sampling: In this sampling the sample is selected with definite purpose in view and the choice of
sampling units depends entirely on the discretion of the surveyor. This sampling suffers from drawback of
favoritism and nepotism of the surveyor.
2. Judgment Sampling: In judgment sampling respondents are selected on the judgment of the surveyor with the hope
that they will meet requirements of the study. The underlying assumptions are that the respondent truly represents
the entire population. To find out the potential guide for the food and catering technology a researcher go to the
teachers of Hotel Management Department may be the example of judgment sampling.
3. Convenience Sampling: In convenience sampling method, a surveyor selects the sample at his/her own convenience,
often as the study is being conducted. Convenience sampling is based on assumption that the target population is
homogeneous and the individuals selected and interviewed yields similar information with regard to the
characteristics under study. In our example of the 10,000 university students, we were only interested in
achieving a sample size of 100 students who would take part in our research. As such, we would continue to invite
students to take part in the research until our sample size was reached. Since the aim of convenience sampling is
easy access, we may simply choose to stand at one of the main entrances to campus of the University where it
would be easy to invite the many students that pass by to take part in the research. Such a sampling is known as
convenience sampling.
4. Quota Sampling: If the cost of selected random samples in each stratum is very high in stratified sampling then the
sampling units are assigned in a quota (fixed number of units) in each stratum and the actual selection of units is
left at the decision of the surveyor. For example, if there were 6,000 male students (60% of the total) and 4,000
female students (40% of the total), our sample would need to be made up of 60% males and 40% females. If our
desired sample size was 100 students, this would mean our sample should include 60 male students and 40 female
students.
5. Snowball Sampling: where research participants recruit other members for the study. This method is particularly
useful when participants might be hard to find. For example, a study on working prostitutes or current heroin users.
For e.g. If researchers are conducting a study of individuals with rare diseases, it may be difficult to find these
individuals. However, if they can find just a few initial individuals to be in the study then they can ask them to
recruit further individuals they may know through a private support group or through some other means. For
example, there are no lists of drug users or prostitutes that a researcher could get access to, especially lists that
could be considered representative of the population of drug users or prostitutes. Another example is the sensitivity
of coming forward to take part in research is more acute in such research contexts. Individuals that are drug users
or prostitutes, for example, are likely to be less willing to identify themselves and take part in a piece of research
than many other social groups. However, since snowball sampling involves individuals recruiting other individuals
to take part in a piece of research, there may be common characteristics, traits and other social factors between
those individuals that help to break down some of the natural barriers that prevent such individuals from taking
part.
Random or Probability Sampling
The technique of random sampling is of fundamental importance in the application of Statistics. The estimation theory
is based on the assumption of random sampling. Probability sampling is the scientific method of selecting samples
accordingly to some laws of chance in which each unit in the population has some definite pre-assigned probability
of being selected in the sample. The following are the different types of probability sampling:
1. Where each unit has an equal chance of being selected
2. Sampling units have different probabilities of being selected
3. Probability of selection of a unit is proportional to the sample size.
 Non-Probability Sampling Methods
Probability Sampling Methods
Non-probability sampling is a
Probability Sampling is a sampling technique in
sampling technique in which the
which samples from a larger population are
Definition researcher selects samples based on
chosen using a method based on the theory of
the researcher‟s subjective judgment
probability.
rather than random selection.
Alternatively Known as Random sampling method. Non-random sampling method

Population selection The population is selected randomly. The population is selected arbitrarily.

Nature The research is conclusive. The research is exploratory.

Since the sampling method is
Since there is a method for deciding the sample,
arbitrary, the population
Sample the population demographics are conclusively
demographics representation is almost
represented.
always skewed.
Takes longer to conduct since the research This type of sampling method is quick
Time Taken design defines the selection parameters before since neither the sample or selection
the market research study begins. criteria of the sample are undefined.
This type of sampling is entirely
This type of sampling is entirely unbiased and
biased and hence the results are
Results hence the results are unbiased too and
biased too, rendering the research
conclusive.
speculative.
In probability sampling, there is an underlying
In non-probability sampling, the
hypothesis before the study begins and the
Hypothesis hypothesis is derived after conducting
objective of this method is to prove the
the research study.
hypothesis.
Designing a good questionnaire:
No survey can achieve success without well designed questionnaire. Unfortunately design of questionnaire has no
theoretical base. All researchers have to guide themselves to design a good draft of questionnaire born out of
experiences of other researchers past and present experiences.
Following are main steps for drafting good questionnaire:
1. Consulting with all stake holders related purpose of survey: The good questionnaire cannot be written by only one
person of not merely writing questions. To have a good questions which are relevant to extract information that
sticks to purpose of survey, we have to consult with experts and all stake holders in that field.
2. Decision regarding questions: We have to first decide regarding various questions to be incorporated. This decision
of selection of questions depends upon purpose of enquiry. Irrelevant and unnecessary questions should be
avoided.
3. Deciding types of questions: Next step of drafting the questionnaire is to decide in which way the question should be
asked? Questions can be asked in no. of ways like
a) simple alternative questions: These questions may be answered between two alternatives such as yes or no. for e.g.
Do you have a car? , Do you take drinks?
b) Multiple choice questions: These questions can be answered in multiple ways. For e.g. What is your marital status?
The answers are married , unmarried, widow , divorced.
c) Specific information question: Such questions are asked to obtain some specific information. For e.g. are you
married?, Are you employed?
d) open ended questions: Such question lave the door to open any answer. For e.g. How can the problem of
unemployment be solved?
[Link] regarding question wording: Another step in drafting good questionnaire is to decide about wording of
different questions. Question should be worded in such a manner that they must be easily understandable.
5. Decision regarding order of questions: Questions should be in a sequence or adjusted in such a manner there will be
not be no sudden jumps. One question should lead to and flow into another.
6. Pretesting of questionnaire: Before sending questionnaire to the respondents it must be properly tested.
.
Characters of Good questionnaire
1. Questionnaire Should deal with important topic to create interest among respondents.
2. It should seek data which cannot be obtained from other sources. Only relevant Questions should be asked.
[Link] should be as short as possible but should be comprehensive.
[Link] should be attractive.
[Link] should be represented in good psychological order proceeding from general to more specific responses.
6. The question should be simple, clear and precise. Its language should be very simple so respondents may easily
understand.
7. No personal questions should be asked from respondents.
8. It should avoid embarrassing questions.
9. Questions should be objective.
10. The questionnaire should be set in such a way that there may be cross examination of information supplied by
respondents.
11. There should not be questions related to past memory.
[Link] negative questions should be avoided.
Sampling and non-sampling error:
The errors involved in the collection, processing and analysis of data may be broadly classified under the following
two heads:
1. Sampling Error
2. Non-sampling Error
Sampling Error
The error which arises only in sample survey is termed as sampling error. This error arises because in sample survey a
part of the population is only studied. This is the reason why sampling error is absent in census. The main factors
of sampling error are:
1. Some of the bias is introduced by the use of defective sampling techniques for the selection of a sample
2. Substitution of a non-selected a convenient unit of the population in place of a selected unit to which the investigation
is difficult leads to some biases in the sample survey
3. Bias due to defective demarcation of sampling units, particularly in area/filed survey
4. Constant errors due to improper choice of the statistics for estimating the population parameters.
Non-Sampling Error
The non-sampling error arises at the stages of observation, ascertainment and processing of the data. This is the reason
why the non-sampling error presents in both the census and the sample survey. Non-sampling error can occur at
every stage of the planning or execution of census or sample survey. Non-sampling errors arise due to the following
factors:
1. Data specification being inadequate and inconsistent with respect to the objective of the study
2. Error due to location of the units and actual measurement of the characteristics
3. Error due to ill designed questionnaire
4. Lack of trained and qualified investigators and lack of adequate supervisory staff
5. Errors due to lack of correct responses furnished by the respondents. This error is called as response error it may be
arise due to following reasons:
a) Respondents may misunderstood questions and unintentionally furnish improper information.
b) Sometime respondents upgrade or downgrade his response due to prestige issue. For e.g. he may downgrade his age
of upgrade response related to income , education etc.
c) A person may give understatement of his salary of production and over state his expenses due to personal interest.
d) Sometimes due to biased interviewer, response of respondents may affect.
d) Questions related to pat memories creating problem to respondent to remembering and associating event
information.
6. Non-response biases occur if full information is not obtained on all the sampling units
7. If the objectives of the survey are not stated clearly, it may result in inclusion of the units which are not to be included
and exclusion of the units which are to be included in the sample
8. Due to error in various operations of data processing such as editing and coding of the responses, punching of cards,
tabulation and summarizing the observation made in the survey
9. The errors may be committed during presentation and printing the results of the survey.
Handling Non Response:
1. Give some incentives like lotteries, donations, charities and gifts. This use of incentives increase response rates is
based on principle of reciprocation.
2. Devising questionnaire acceptable to respondents.
3. Sending reminders to potential respondents.
[Link] flexible with time frames of respondents.
5. Assuring respondents that the data collected kept confidential
[Link] of post survey adjustments including imputation and weighting which reduces nonresponsive bias.
METHODS OF SELECTION OF A SAMPLE
The random sample obtained by a method of selection in which every item has an equal chance to be selected in
the sample. The random sample depends not only on method of selection but also on the sample size and nature of
population.
Simple Random Sampling
If a sample of n units is selected randomly from a population of size N, this method is known as simple random
sampling. As the name suggest, simple random sampling is a method in which the required number of elements
/units are selected simply by random method from the target population. One can select a simple random sample
by either of these two methods with replacement method and without replacement method.
Simple Random Sampling with Replacement (SRSWR)
When simple random samples are selected in the way that units which has been selected as sample unit is remixed
or replaced in the population before the selection of the next unit in the sample then the method is known as
simple random sampling with replacement. SRSWR is a method of selection of n units out of the N units one by
one such that at each stage of selection, each unit has an equal chance of being selected, i.e., 1/ N.
Characteristic under study: A property of each individual or item of population in which we interested is called a
characteristic under study .generally the characteristic under study denoted by letter Y.
Population values: The values of the characteristic of interest of the individual of items of the population under study
are called as population values are denoted by Y1, Y2……..YN
Parameter: Any function of a population values Y1, Y2……..YN is called as parameter. For e.g Population mean,
population total
Population Total: Let a population contains Y1, Y2……..YN values then total of all values is called as population
total.
Population Mean: Let a population contains Y1, Y2……..YN values then average of all values is called population
mean.
=
Some results for SRSWOR:
1. In SRSWOR , the probability of a specified unit being selected in sample at any given draw is equal to 1/N.
Proof: In simple random sampling method an equal probability of selection is assigned to each unit of the
population at the first draw.
Thus, in SRS from a population of N units, the probability of drawing any unit at the first draw is 1/N, the
probability of drawing any unit in the second draw from among the available N-1 is 1/(N-1) and so on.
Let, Er be the event that any specified unit is selected at the rth draw.
P (Er) = Prob.{A specific unit is not selected at any one of previous (r-1) draws and then selected at
the rth draw}

That means
P (Er)= P (E1)
2. The probability that a specified unit is selected in the sample of size n is n/N.
Proof: Since a specified unit can be selected in the sample of size n in n mutually exclusive ways, viz. it
can be selected in the sample at the rth draw (r=1, 2,…, n) and since the probability that it is selected at rth
draw is P (Er)=1/N, r=1,2,….,n where Er be the event that any specified unit is selected at the rth draw.
Therefore, the probability that a specified unit is included in the sample would be the sum of the
probabilities of inclusion in the sample at 1st draw, 2nd draw,…, nth draw. Thus, by addition theorem of
probability, we get

3. The possible numbers of sample of size n from a population of size N if sampling is done with replacement is
Nn.
Proof: The first unit can be drawn from N units in N ways. Similarly, second unit can also be drawn in N
ways because the first selected unit again mixed with the population. So on up to the selection of nth unit.
Thus, the total number of ways are
.
[Link] SRSWOR , the probability of drawing a sample of size „n‟ from a population of N units is or In SRSWOR
every one of the possible samples of size n has an equal probability of being selected as a sample.
Proof: If u1,u2,….,un are the n units selected in the sample, then the probability of their selection is
p(u1,u2,….,un)=p(u1)*****p(un)
Note that when the second unit is to be selected, then there are (n – 1) units left to be selected in the sample
from the population of (N – 1) units. Similarly, when the third unit is to be selected, then there are (n – 2)
units left to be selected in the sample from the population of (N – 2) units and so on.
p(u1)=n/N, p(u2)=(n-1)/(N-1)……. p(un)=1/(N-n+1)
Thus
p(u1,u2,….,un)=n/N*(n-1)/(N-1)*….*1/(N-n+1)

5. In SRSWR the probability of a specific unit included in sample is 1-(1-1/N)n.

Proof: We know that in SRSWR the probability of selecting specific unit at first draw =1/N
Therefore probability of not selecting specific unit at first draw=1-(1/N)
Similarly probability of not selecting specific unit at second draw=1-(1/N) and so on
probability of not selecting specific unit at n th draw=1-(1/N)
since all draws are independent by compound probability theorem , In SRSRWR
probability of not selecting specific unit in the sample =[1-(1/N)]n
so probability of selecting specific unit at in the sample =1-[1-(1/N)]n