Sampling &

Sampling Distribution
2
NTRODUCTION
of
In all the spheres life (such as
investigation and data analysis is risingBconomic, Social
statistical dataa:(i) Census Method, and (ii)
day by day. and Business) the need for statistical
There are two
relatingtothe entire field of
sample method, rather than collectinginvestigation Sample Method.
or units of Under
population
methods of collection of
census method, information
relatingto only.selected units is collected. information
about
Before
we make aall the
is
units collected;
of whereas under
willexplain some basic concepts related to them. detailed study population,
of both the information
SOME BASICCONCEPTS methods, we
Universe or Population : IIn
about which we obtain
investigation about which information.statistics,
A universeor or population means an
universe
is sought. For population means the entire aggregate
about the average knowledge
monthly expenditure of all example, if we want to field
of items
under
aggregate of 2,000 students will be termed as the 2,000 students of a collect information
inds () Finite and (ii)
Infinite. Universe or college, then the entire
of students or In a finite
teachers in a college. On the population, Population. A
population
number of items is definite can be of two
items e.g., number of other hand, an infinite such as, number
tree or number of hairsstars in the sky,
on the head. number of water drops in an ocean, population has infinite number of
(2) Sample:A part of number of leaves on a
the population is known as population is
a sample. In called sample. In other words, selected or sorted
0r tne purpose of fact, a sample is that part of the population whichunits from
sLudents of acollege who investigation. For example, if an investigator we select
selects 200 students from 2000
means Somerepresent
Thus, sample all of them, then these
200
units selected out of a students will be termed as a sample.
CENSUS AND population which represent it.
Ihere are two SAMPLEto METHODS
methods collect statistical data :
(1)Census Method
(2) Sample Method
1) Census Method
unit ofCenstheus method is that method in which information or data is
heir basisS.
Sone popul
This amethod
tion relating collected from
to the problem under investigation and conclusionseach
is also called as are
and every
drawn on
information Complete Enumeration Method. For
(like Monthly Expenditure, Average Height, Average Weight, example, suppose
etc.) is to be
 38 Sampling &Sampling
collected regarding 2000 students of a college. For that purpose if we collect data by
Distribution
and every student of the college then this method will be
called as Census method. In inquiring
this each
the whole college i.e., all 2000 students will be considered as a population and example,
levery student as an
individual will be called the unit of the population. Population in India is conducted after
years by using censusmethod.
every ten
Merits and Demerits of Census Method
Merits
(i) Reliable and Accurate Data : Data obtained by census method have more reliabilib..
accuracy because in this method data are collected by contacting each and every unit of the
universe.
(ii) Extensive Information: This method gives detailed information about each unit of a
universe. For example, Indian population census does not only provide the knowledge
about the number of persons butalso information about their age, occupation, income.
education, marital status, etc.
(iüi) Suitability : This method is more suitable for the population with limited scope and
diverse characteristics. Use of this method is also appropriate where intensive shudy ie
desired.
Demerits
(i) More Expensive : Census method is an expensive one. More money is needed for it as
information is collected from each unit of the population. This is why this method is used
by Government mostly for very important issues like Census, etc.
(ii) More Time: This mnethod involves much timne for data collection because data are
collected from each and every unit of the population. This results in delay in making
statistical inferences.
(iii) MoreLabour: This method of data collection also involves very much labour. For this the
enumerators in a large number are required.
(iv) Not Suitable for Specific Problems: This method is not suitable relating to certain specific
or items of the
problems and infinite population. For example, if the population is infinitenot
population are perishable or very complex type, then the census method is suitable.
(2) Sampling Method
Sampling method is that method in which data is collected from the sample of items selected
be made
from population and conclusions are drawn from them. For example, if a study is to collecting
regarding the monthly expenditure of 2000 students of acollege, then instead of students
information from each student of the college, if we collect information by selecting some
possiple
like 100, then this willbe called Sampling Method. On the basis of sampling method, it ishas thre
to study the monthly expenditure of all the students of the college. Sampling methodinferences
(ii) to make
main stages (i) to select a sample (ii) to collect information from it and
regarding the population.
Importance of Sampling Method
statistical inquiry.
In modern times sampling method is animportant and popular method of
Besides economic and business world, this method is widely used in dailytwo-three life. For exampeonly.
grains
housewife comes to know ofthe coating of the whole lot of rice by observing the same way,
Adoctor tests the blood of a patient by examining one or two dropssof blood only. In wheat, rice,
use like
we learn about the quality of a commodity while buying the items of daily quality controllerinspects
pulses, etc. by observing the sample or specimen. In factories, statistical aboutthe
knowledge
the quality of items by examining afew units produced. Ateacher gets the
smpling Sampling Distribution
39
ticac
ayfhis by putting questions
steachingmethod used.
toa few students. In
reality,there is scarcely any area
wheresampling is not
pt and |
Demeritssof Sampling Method
Merits
Merits
9 SavingofTime and Money : Sampling method is less expensive. It
labourbecause only afew units of the population are studied. saves money and
)Savingof Time : In sampling method, data can be collected more quickly as these are
obtained from some items of the universe. Thus much time is saved.
(ii)IntensiveStudy : Asi number of items is less in sampling method, they can be intensively
studied.
(iv) Organisational Convenience : In this method, research work can be organised and
executed more conveniently. More skilled and competent investigators can be
(v) More Reliable Results: If sample is selected in such a manner as it represents appointed.
totally the
universe, then the results derived from it will be more accurate and reliable.
(vi) More Scientific : Sampling method is more scientific because data can be inquired with
other samples.
(vi) Only Method : In some fields where inquiry by census method is impossible, then in such
ituation, sampling method alone is more appropriate. If the population is infinite or too
idespread or of perishable nature, then sampling method is used in such cases.
Demerits
aIess Accurate : Sampling method has less accuracy because rather than making inquiry
about each unit of the universe, partial inquiry or inquiry relating to some selected units
only is made.
6i) Wrong Conclusions: If method of selecting asample is not unbiased or proper caution has
not been taken, then results are definitely misleading.
(ii) Less Reliable : Compared to census method, there is more likelihood of the bias of the
investigator, which makes the results less reliable.
liv) Need of Specified Knowledge: This is a complex method as specialised knowledge is
required to select a sample.
(v) Not Suitable : If all units of a population are different from one another, then sampling
method willnot prove to be much useful.
Diference between Census and Sample Method
Ihe main difference between the census method and the sampling method are as follows :
whereas in
U Scope : In census method, all items relating to a universe are investigated
Sampling method only a few items are inquired.
money and labour
eost : Census method is expensive from the point of view of time,
whereas method economises on them.
Sampling whereas
(iüi) Field of method isusedin investigations with limited field
Investigation:Census
Sampling method is used for investigations with large field.
(iv) the population are heterogeneous
7: Census method is useful where units ofpopulation
Homogeneity
whereas sampling
) Type of
method proves more useful where
: In such fields where study of
units are homogenous.
each and every unit of the universe is
Universe
census method is more appropriate. On
population is
the contrary, when then
necessary,
or vast or liable to be destroyedas aresult of complete
infinite enumeration, sampling
method is considered to be more appropriate.
 40
Sampling &
Sampling
SAMPLING METHODS
The method of selecting a sample out of a given
Distribution
population is called sampling. In
sampling denotes the selection of a part of the aggregate other
obtaining information about the whole. Now adays, there are statistical material
various methods of with a view wodrs,of
from a population in accordance with various needs.
(A) Probability Sampling Methods : sample selecting
(1) Simple Random Sampling
(2) Stratified Random Sampling
(3) Systenmatic Random Sampling
(4) Multistage Random Sampling
(5) Cluster Sampling
(B) Non-Probability Sampling Methods :
(6) Judgement Sampling
(7)Quota Sampling
(8) Convenience Sampling
(9) Extensive Sampling
Sampling Methods

Probability Sampling Methods Non-Probability

Sampling Methods

Simple Stratified Systematic Multistage Cluster Judge Quota

Random Random Random Conve Extensive
Random Sampling ment
Sampling Sampling Sampling Sampling Sampling nience Sampling
Sampling Sampling
(A) Probability Sampling Methods
Probability sampling methods are such methods of selecting a sample from the population n
which all units of the universe are given equal chances of being included in the sample.
There are various variants of probability sampling methods, which are given below :
(1) Simple Random Sampling : Simple random sampling is that method in which each item
of the universe has an equal chance of being selected in the
included in the sample and which not, such decision is not madesample. Which item wl
by an investigator o u
willbut selection of the units is left on chance. According to random
two methods of selecting a random sample: sampling, there a
(i) Lottery Method : In this method, each unit of the population is named or numbered
which is marked on separate piece of paper. Such chits are folded and put nto s
urn or bag. Thereafter as many chits are made selected by some person as many units
are to be included in a sample.
(ii)) Tables of Random Numbers: Some experts has constructed random numbertables.
These tables help in selection of asample. Of all such various tables, Tippett'sTables
 &, Sampling Distribution
snpling

most famous and are in use. 41

are Tippett has constructed a
by using numbers as many as
aumber ation are written 41,600. In this method,four-digit
first of
table of 10,400
serialy. lhereafter by making use of
accordance with the size of the
all, all the items
help of
:Tippetts table sample,
can numbers are selected. The Tippett's tables, in
with the be made clear by an selection of asample
2952
An Extract of
6641 3992
Tippett's Table example:
9792 7979
3170 5524 4167 5911
9525 1545 1396
7203 4356 1300 2693 2370 7483
3408 2762 3563 6107
6913 7691
0560 5246 1112 9025 6008 8127
For example, suppose 12 units are to be chosen out of 5000
decide such units, firstly 5000 units will be serially ordered units. toWith Tippett's table, to
Tippett's table, 12 numbers will be chosen from the from1 5000 and then from
These 12 numbers are follows : beginning which are less than 5000.
2952 4167 4356 2370
3992 1545 1300 3408
3170 1396 2693 2762
The items of such serial numbers will be included in the
are less than 100, then 4 digit random numbers will be sample. If units of the population
numbers, and then such two digit numbers will be selected.made compact into two digit
Like as to select 6 units out of
60 units, the units with serial numbers.29, 39, 31, 41, 15 and 13 willbe inchuded in
the
sample.
Merits
0 There is nopossibility of personal prejudice in this method. In other words, this method is
free from personal bias.
(ü) Under this method, every unit of the universe gets the equal chance of being selected.
(i) The use of this method saves time, money and labour.
Demerits
9) t sample size is small, then sample is not adequately represented.
(u) lfuniverse is very small, then this method is not suitable.
rsome items of the universe are so important that their inclusion in the sample is very
essential, then this method will not be appropriate.
iis method will not be appropriate when population has units with diverse
(2) characteristics.
Stratified Random Sampling : This method is used when units of the universe are
heterogenesample
divided into differentous
rather than in accordance Under
stratahomogeneous. this method, first of all units of the population are
with their characteristics. Thereafter by using random
ssFoups
emleplcteidng,outon theo basisitemsof Arts,
students are to be
are selected from each stratum. For example, if 150 divided into three
of 1500 students college,then firstly the college students will be
of Commerce
and Science. Suppose there are 500, 700, 300 students
a

RsASSpuemcetsiveequally
,70 and30 instudents
three faculties
respectively willsample
and 10% be selected
is to by using then
be taken, on the
random basis of random sampling
sampling. Thus, this method
representation to each class or group and all the units of the
universe get equal
thance being
of
selected in the sample.
 Sampling &Sampling Distribution
42

Merits
likelihood of representation of units in this method.
(i) There is more possible under this methed
(ii) Comparative study on the basis of facts at different strata is
(iii)This method has more accuracy.
when tho
Demerits
has limited scope because this method can be adopted only
(i) This method known.
population and its different strata are population is not properly stratified.
of prejudice if the
(ii) There can be the possibility
small in size, it is difficult to stratify it.
population is too universe are
(iii)If the
Sampling: In this mnethod, all the items of the intervals, For
(3) Systematic Randomnumbered and then sample units are selected at equal would be
systematically arranged and
to be selected for asample, then 50 students
example, if 5 out of 50 students
are at random
One item of the first 10 would be selected
numbered and systematically
arranged.
selected number will be selected to frame a sample. If the
Subsequently, every 10th item from the would be 15th, 25th, 35th and 45th.
the subsequent numbers
item, then
first selected number is 5th
Merits
method. Samples can be easily obtained by it. accurate.
(i) It isa simple
little time in sample selection and results are almost
(ii) This method involves very
only
Demerits
not stand the equal chances of being selected because
each unit does
(i) In thismethod,selected on random sampling basis.
the first unit is characteristics, then results will not be appropriate.
different in through many stages,
(ii) If all the units are sampling procedure passes
Sampling : When or population is
(4)Multistage Random sampling. In this method, firstly the entire universe
selected on random sampling
then it is known as multi-stage
substages. From the each stage some units are
sampling again some
divided into stages or subdivided and on the basis of random example,
basis. Thereafter these units
are sub-division further and selection on. For
goes on with will be
sub-units are selected. Thus, this Education in Haryana State, first some districts tehsils,
Adult districts, some tehsils and
out of
for the purpose of a study regarding the selected
selected on random basis.
Thereafter out of
further out of the villages or towns, some
may be thus selected, the
be selected from whom
some villages or towns and out of the wards, some households will
neighbourhood, or wards problem at hand.
concerning the
inquiry willbe made
Merits universe orpopulation on regional basis.
This is the best method of studying a
where decisions on the basis of sample alone
(i) suitable for those problems
(ii) This method is
taken.
can not be
the level of accuracy which involves
Demerits
requires many tests to correctly estimate
(i) This methodand labour. predecided which does not seem
alot of time accuracy level is
this method, level of estimated
(ii) In
is divided into many groups caled
logical.
Sampling : In this method, simply the universe
random basis and then the clusters
are
(5)Cluster are selected on in pharmaceutical
and out of which afew clusters applied in industries like as
cluster method is usually
complete enumerated. This
Sampling& Sampling Distribution
machine produces medicines
industry, tablets
fewrandomly selected batches
inspection,a are
in the
batches of hundred each, then for
43

Non-Probability Sampling Methods examined.. quality

Non-probabilitysampling methods are those
thebasisofconvenience
c or judgement of the methods in which selection of units is
such methods, selection of units isinvestigator rather than on the basis of made on
chance. In
convenienceofthe investigator.
made in probabilityandor
accordance with the specific objectives
(6) Judgement Sampling: Under this method, the
excusivelyon the judgement of the investigator. In other selection of the sample items depends
judgementinthe choice and includes those items in the sample
words, the investigator exercises his
theuniversee with regardto the characteristics under study. For which he thinks
example, are most
if asample of 20typical
studentsof
istobe selected from a class of 80 students for. analysing the spending habits of the 10 students, the
investigator would select 20 students, who in his opinion are representative of the class.
Merits
i) This method is less expensive.
i) This method is very simple and easy.
a This method is useful in those fields where almost similar units exist or some units are too
important to be left out of the sample.
Demerits
prejudice in this method.
() There is greaterchance of the investigator's own
reliable.
(ü) This method is not very accurate and according
Sampling:In this method, the investigators are assigned definite quotasquota. The
(7) Quota to obtain the required number to fill in each
to some criteria. They are instructed personal
individuals (i.e., sample items) to collect information on their
nvestigators select the or a part of the whole quota is not available or
When all
Juagements within the quotas. responds. Quota sampling is a type
supplementing new
Pproachable, the quota is completed by
of judgement sampling.
included.
Merits
chance ofimportant units being theunits of the quotas
() In this method, there is greater
organised in this method on
account of
sical inquiry is mnore
being fixed.
Demerits remain. method. sample
i) Possibility of prejudice shall sampling errorin this sampling,thechoice ofthe
according
(ii) There is greater likelihood of type of non-probability
investigator obtain asampleon thebasis
(8) Convenience Sampling:: Inthis investigator. The teachersconvenientlyregarding his
is left t convenience ofthepublisherselectssomefeedback fromthem This
to his completely to Forthe
example,
unreliable.
a book prospectus and gets unscientificand forsample
of the convenience. college simple butismethod is appropriate available
list of the teachers from theexpensiveand
more This units isnot
tpubl
nethiodcatioresults
n. enumerators. the
This method is less the defined orlistof population
on clearly
in more dependence populationis not asthecollectionis
selection where
Dt sample units the universe
are not or themselves.
clearin sample size is takenalmost
outfor
as big
whichdata
method, those unitsareleft
(9) this
tself 1ikeExt90%ensithe
ve section of thep
Sampling : Inpopulation. Only