Sampling Distribution

Population: By population, we mean the collection of individuals or of

their attributes or individuals obtained as results of some experiment or
operation.
Example- Collection of weights of the University Students.
Types of Population :
i) Existent Population
ii) Hypothetical Population
iii) Finite Population
iv) Infinite Population
Sample : A part of the population selected according to some rule is
known as a sample.
Size of Sample: The number of members contained in a sample is
called sample size.
Sampling : The Process of drawing or selecting a sample is known as
sampling.
Need of Sampling: In our daily life we come across persons making an
assessment of the population though sample. For Example: the
housewife tests a small quantity of nice to see if it has been well
cooked or not. In most of cases it is neither practical nor desirable to
collect the data for each and every member of the population,
particularly when the population is either large or infinite or it is a
hypothetical population, as this collection is times suming , too
expensive and so on.
Types of sampling :
i) Random sampling
ii) Simple sampling
iii) Purpose sampling
iv) Mixed or stratified sampling
v) Multi stage sampling
vi) Quota sampling
vii) Cluster or Block sampling
viii) Systematic or quasi-random sample
Techniques of Random sampling :
i) Lottery system
a) Chit Method
b) Card Method
c) Lottery System
ii) Arranging the numbers of the population is some order and selecting
individuals in a sequence.
iii) Random number method
The Process of sampling is based on two important principles of
statistics.
i) Law of statistical regularity
ii) Law of inertia of large number.
Parameter :
It is a statistical measure based on all the units of a population.
e.g: Population mean   , Population s.d   , Proportion of defective in
the population are parameters.
Statistic : It is a statistical measure based only on all the units selected
in a sample.
e.g: Sample mean x  , sample s.d (s) proportion of defectives in the
sample are statistic.
Sampling distribution of statistic :
we know the value of a statistic may vary from sample to sample. Let t1 ,
t 2 , t3 ......... tk be the values of statistic for the K. samples. Each of
these values occurs with a definite probability. Thus we can
construct a table showing the set of values t1 , t2 , t3 ......... tk of t with
their respective probabilities. This probability distribution of t is
known as the sampling distribution of t.
Central limit theorem :
(i) If x be the mean of a rendom sample of sizer drawn from a
population having mean  and standard deviation  , then the
sampling distribution of the sample mean x is approximately a
normal distribution with mean  and S.D= S.E of x provided the
sample size n is sufficiently large.
(ii) If p be the proportion of defectives in a random sample of size n
drawn from a population having the proportion defectives P, then
the sampling distribution of the sample proportion of defectives p is
approximately a normal distribution with mean p and S.D = S.E of
p provided the sample size n is sufficiently large n  30
Standard Error (S.E) of Statistic :
The S.E of a statistic is the s.d. of the sampling distribution of the
statistic. Thus S.E of sample mean x is the s.d of the sampling
distribution of sample mean. It will play a very important role in the
sampling theory. S.E is used a tool in tests of hypothesis or test of
significance. It gives an idea of the reliability and precision of a
sample. It helps in determining the confidence limits within which
the parameters are expected to lie.
Formula of S.E :
6
(i) For sample mean x
n

pq
(ii) For proportion p,  : Q 1  P
r


(iii) For s.d =
2n

(iv) For x  x  =
1 2
 12
n1

 22
n2

P1Q1 PQ
(v) For P1  P2    2 2
n1 n2

When a sample is drawn without replacement.

 N n
(i) For mean = .
n N 1

PQ N  n
(ii) For Proportion = .
n N 1
 n N n
If  0.05 then finite population correction factor is used.
N N 1

P-1: A sample of size 36 is drawn from a finite population

consisting of 101 units. If the population S.D is 12.6, find the
standard error of sample mean when the sample is drawn (i) with
replacement (ii) without replacement.
Ans: Here n= 36,  = 12.6, N= 101
 12.6
  S.E. for x =   2 .1
n 6

 N n
(ii) S.E for x = .
n N 1

12.6 101  36
=
6 101  1

65
= 2.1
100
= 2.1  0.806
= 1.6926
 1.69
Estimation :
The principles of sampling theory are used in statistical inference which is defined by
W.A Spurand C.P Boninias, “the process by which we draw a conclusion about some
measure of a population based on a sample value. The measure might be a variable, such
as the mean, s.d etc. The purpose of sampling is to estimate some characteristics for the
population from which the sample selected”.
One of the important problems of statistical inference is the estimation of population
parameters from the corresponding sample statistics.
Statistical Estimation:
It is the technique of estimating the parameters of population from that of sample.
e.g: If we are to find the approximate weight of dust in a bag of wheat, we draw carefully
a sample of wheat from the middle portion of the bag and weight it. After clearing this
sample, it is weighted again to find the amount of dust in the sample. This weight of dust
in the sample multiplied by the ratio of the wheat in the bag to the weight of the sample
gives an approximate of dust in the bag of wheat.
Estimator: The statistic by which estimate the population parameter is called estimator.
Properties of good estimator :
There are the following properties be the proportion of a good estimator viz.
i) Sufficiency (i) Efficiency (iii) Urbaneness (iv) Consistency
Unbiase and Biased Estimator : A statistic is called unbiased estimator of the
corresponding parameter, if the mean of the sampling distribution is equal to the
corresponding population parameter. Otherwise the statistics is called biased estimator of
the corresponding parameter.
Point Estimator : If an estimate of a population parameter is given by a single point then
the estimator is called a point estimator.
Interval estimate : If an estimate of a population parameter be given by two distinct
numbers between which the parameter may be considered to lie, then the estimate is called
an interval estimate of the parameter.
As the value of a point estimate fluctuates from sample to sample, then interval estimates
are preferable to point estimates. Also the interval estimate indicates the accuracy of an
estimate. A statement of error of an estimate is called its reliability.

Confidence Interval : In an interval estimation of the population parameter  , if we find

two quantities t1 and t2 based on sample observations drawn from the population such that
the unknown parameter  is called in the interval [t1, t2] in a specified percentage of cases
then this interval is called a confidence interval for the parameter  .
The lower limit t1 and the upper limit t2 of this interval are called confidence limits of  .
* If the sampling distribution of the sample mean x of a random sample of size n n  30
drawn from population having mean  and s.d  , it is approximated a normal
distribution with mean  and s.d = S.E of x =  | n
* For small samples (n< 30) drawn from a normal population, t-distribution is used to find
the confidence interval of the population mean is  is not known.
For large random samples, confidence limit :
1. Confidence limits for  :
 i) 95% confidence limits x 1.96 S .E of x  
ii) 99% confidence limits x  2.58 ( Do)
2. Four 1   2 :
(i) 95% Confidence limits x  x   1.96 S.E of
1 2 x1  x 2 
(ii) 99% confidence limits x  x   2.58 Do 
1 2

3. For Population Proportion p:

(i) 95% Confidence limits p  1.96 S E of  
(ii) 99% confidence limits p  2.58 Do 
4. For p1  p2 :
(i) 95% confidence limits p1  p2  1.96 S . E of p1  p2 
(ii) 99% confidence limits  p1  p2   2.58 Do 
P-1: A random sample of the height of 100 students takes from a large
population of student in a university having s.d of 0.75. It has an
average height of 5.6 ft. find (i) 95% and (ii) 99% confidence limits for
the average height of all the students of the university.
x  5 .6
  0.75
Ans: Here, n  100
 0.75
S .E x    0.075
n 10

(i) confidence limits = 

x 1.96 S . E of x 
= 5.6  1.96  0.075

= 5.6  0.147
= 5.453 and 5.747
(ii) Confidence limits = 
x  2.58 S . E of x 
= 5.6  2.58  0.075

= 5.6  0.1935
= 5.4065 and 5.7935
P-2 : In a large consignment of arranges a random sample of 500 arranges revealed that 65
arranges were bad find 95% confidence limit of bad arranges in %.
 65
Ans : Here, P   0.13
500
q = 1 – 0.13 = 0.87
PQ q
S.E of p  
n n
0.13  0.87
=
500
= 0.015
95% confidence limit of p = p  1.96 S .E of  
= 0.13  1.96  0.015
= 0.13  0.0294
= 0.1594 and 0.01004
= 15.94 % and 10.04 %
P-3: 10 life insurance policies in a sample of 200 taken out of 50.00 were found to be
insured for less than Tk 5,000. How many policies can be reasonably expected to be
insured for less than Tk. 5000 in the whole lot at 95% confidence level.
Ans : Here n = 200
10
p= = 0.05
200
q= 1-0.05 = 0.95
pq 0.05  95
S.E of p =   0.01541
n 200
Confidence limits for p= p  1.96 S .E of p 
= 0.05  1.96  0.01541
= 0.05  0.0302036
= 0.080 2036 and 0.0197964
Required number of policy lies between
0.0197964  5000 and 0.0802036  5000
i.e. 989.82 and 4010.18
i.e 990 and 4010