Sampling Theory & Sampling Distribution

Presented by:
MOHAMMAD FAROOQ
ACADEMIC COORDINATOR
MATHEMATICS
 LEARNING OBJECTIVES
 Differentiate between a population
parameter and a sample statistic.
 Describe different Sampling methods
 Types of error
 Describe the properties of the sampling

distribution of the sample mean.

 The Central Limit Theorem and its

applications
 POPULATION
 Population—consists of all items of
interest in a statistical problem.
 Population is described by parameters.
 A parameter is a constant, whose value may be
unknown.
 Only one population.
 SAMPLE
 Sample—a subset of the population.
 Sample Statistic is calculated from sample
and used to make inferences about the
population.
 Sample is described by statistics.
 A statistic is a random variable whose value
depends on the chosen random sample.
 Can draw multiple random samples of size n.
 BIASED SAMPLE
 the tendency of a sample statistic to
systematically over- or underestimate a
population parameter.
 in which all individuals, or instances, were

not equally likely to have been selected. 

 if it systematically favors some outcomes

over others
 SAMPLING METHODS
 Simple random sample is a sample of n
observations which has the same
probability of being selected from the
population as any other sample of n
observations.
 Most statistical methods presume

simple random samples.

 However, in some situations other

sampling methods have an advantage

over simple random samples.
Systematic Random Sampling
 The items or individuals of the population are arranged in
some order. A random starting point is selected and then
every kth member of the population is selected for the
sample.

EXAMPLE
A population consists of 845 employees of Fraooq Industries. A
sample of 52 employees is to be selected from that population.
First, k is calculated as the population size divided by the sample
size. For Farooq Industries, we would select every 16th
(845/52) employee list. If k is not a whole number, then round
down. Random sampling is used in the selection of the first
name. Then, select every 16th name on the list thereafter.
 Stratified Random Sampling
 A population is first divided into subgroups, called
strata, and a sample is selected from each stratum.
Useful when a population can be clearly divided in
groups based on some characteristics
 Divide the population into mutually exclusive and
collectively exhaustive groups, called strata.
 Randomly select observations from each stratum, which are
proportional to the stratum’s size.
 Advantages:
 Guarantees that the each population subdivision is

represented in the sample.

 Parameter estimates have greater precision than those

estimated from simple random sampling.

 Stratified Random Sampling
 EXAMPLE :Suppose we want to study the advertising expenditures
for the 352 largest companies in the United States to determine
whether firms with high returns on equity (a measure of profitability)
spent more of each sales dollar on advertising than firms with a low
return or deficit.
 To make sure that the sample of 50 companies is a fair
representation of the 352 companies, the companies are grouped on
percent return on equity and a sample proportional to the relative
size of the group is randomly selected.
 CLUSTER SAMPLING
 A population is divided into clusters using naturally occurring
geographic or other boundaries. Then, clusters are randomly
selected and a sample is collected by randomly selecting from
each cluster.
 Divide population into mutually exclusive and collectively

exhaustive groups, called clusters.

 Randomly select clusters.

 Sample every observation in those randomly selected

clusters.
 Advantages and disadvantages:
 Less expensive than other sampling methods.
 Less precision than simple random sampling or stratified
sampling.
 Useful when clusters occur naturally in the population
 CLUSTER SAMPLING

EXAMPLE :Suppose you want to determine the views of
residents in Oregon about state and federal environmental
protection policies.
 Cluster sampling can be used by subdividing the state into
small units—either counties or regions, select at random say
4 regions, then take samples of the residents in each of these
regions and interview them. (Note that this is a combination
of cluster sampling and simple random sampling.)
 Stratified versus Cluster Sampling
 Stratified Sampling  Cluster Sampling

 Sample consists  Sample consists

of elements from of elements from
each group. the selected
groups.
 Preferred when
the objective is to  Preferred when
increase the objective is
precision. to reduce costs.
 QUOTA SAMPLING
 A sampling method of gathering representative
data from a group. As opposed to random
sampling, quota sampling requires that
representative individuals are chosen out of a
specific subgroup. For example, a researcher
might ask for a sample of 100 females, or 100
individuals between the ages of 20-30
 Quota sampling is based on the researcher’s

judgment and is considered a 

non-probability sampling technique.
 QUOTA SAMPLING
Advantages:
 Easy to administer.
 Fast to create and complete.
 Inexpensive.
 Takes into account population proportions, if desired.
 Can be used if probability sampling techniques are
not possible.
Disadvantages:
 Selection is not random.
 Selection bias poses a problem
Non-Sampling Error
Non-sampling error – Error associated with collecting
and analyzing the data
 Coverage error
◦ Wrong population definition
◦ Flawed sampling frame
◦ Interviewer or management error in following sampling frame
 Response error
◦ Badly worded question results in invalid or incorrect response
◦ Interviewer bias changes response
 Non-response error
◦ Respondent refuses to take survey or is away
◦ Respondent refuses to answer certain questions
 Processing errors
◦ Error in data entry or recording of responses
 Analysis errors
◦ Inappropriate analytical techniques, weighting or
imputation are applied
Sampling Error
Sampling error – Error associated with failing to interview the entire
population
 Sampling error is known after the data are collected by calculating
the Margin of Error and confidence intervals

 Surveys don’t have a Margin of Error, questions do

 Power analyses use estimates of the parameters involved in

calculating the margin of error

 It is common to see sample sizes of 400 and 1000 for surveys

(these are associated with 5% and 3% margins of error)

 In most cases the size of the population being sampled from is

irrelevant

 The margin of error should be calculated using the size of the

subgroups sampled
Sampling Distribution of
the Sample Mean
A large tank of fish from a hatchery is being
delivered to the lake. We want to know the
average length of the fish in the tank. Instead
of measuring all the fish, we randomly
sample some of them and use the sample
mean to estimate the population mean .
 Sampling Distribution
 Sampling Distribution of the Mean X
◦ Each random sample of size n drawn from
the population provides an estimate of m
the sample mean. X
◦ The sampling distribution of the sample
mean is a probability distribution
consisting of all possible sample means of
a given sample size selected from a
population.
 Sample with replacement
 When we sample with replacement, the two
sample values are independent. Practically,
this means that what we get on the first one
doesn't affect what we get on the second.
Mathematically, this means that the
covariance between the two is zero.
 Sample mean Sample standard deviation
x   
x 
n
 Sample without replacement
 In sampling without replacement, the two
sample values aren't independent. Practically,
this means that what we got on the for the
first one affects what we can get for the
second one. Mathematically, this means that
the covariance between the two isn't zero
 Sample mean Sample standard deviation
x   x 
  N n 
 
n  N  1 
  ACTIVITY. 1 (Pumpkin Weights)
 The population is the weight of six pumpkins (in
pounds) displayed in a carnival "guess the weight"
game booth. You are asked to guess the average
weight of the six pumpkins by taking a random
sample without replacement from the population.
Pumpkin A B C D E F
Weight
19 14 15 9 10 17
(in
pounds)
 a. Calculate the population mean μ.
 b.  Calculate the population variance Var  X    2

 C. Obtain the sampling distribution of the

sample mean for a sample size of 2 when one
samples without replacement.
Sampling Distribution of the sample
mean for a sample size of 2
Sample weight X P( X) Sample weight X P (X )
1) A, B 19,14 16.5 1/15 9) B, F 14, 17 15.5 1/15
2) A, C 19,15 17.0 10) C, D 15, 9 12.0
3) A, D 19, 9 14.0 11) C, E 15, 10 12.5
4) A, E 19,10 14.5 12) C, F 15, 17 16.0
5) A, F 19,17 18.0 13) D, E 9, 10 9.5
6) B, C 14,15 14.5 14) D, F 9, 17 13.0
7) B, D 14, 9 11.5 15) E, F 10, 17 13.5
8) B, E 14,10 12
Sampling Distribution for the sample mean
when the sample size is 5
Sample Weight X P(X )
1) A, B, C, D, E 19, 14, 15, 9, 10 13.4 1/6
2) A, B, C, D, F 19, 14, 15, 9, 17 14.8
3) A, B, C, E, F 19, 14, 15, 10, 17 15.0
4) A, B, D, E, F 19, 14, 9, 10, 17 13.8
5) A, C, D, E, F 19, 15, 9, 10, 17 14.0
6) B, C, D, E, F 14, 15, 9, 10, 17 13.0
The Expected Value and Standard Deviation
of the Sample Mean

 Expected Value
 The expected value of X,

E X  
 The expected value of the sample mean,

 
E X E X  
 Variance of X
Var  X    2

 Standard Deviation
 of X SD  X    2  

 Where n is the sample

 of
X SD  X   size. Also known as the
n
Standard Error of the
Mean.
The Central Limit Theorem
 If all samples of a particular size are
selected from any population, the sampling
distribution of the sample mean is
approximately a normal distribution. This
approximation improves with larger
samples.
 The mean of the sampling distribution

equal to μ and the variance equal to σ2/n.

Graphical presentation of CLT
 Application of CLT
 For any population X with expected value m and
standard deviation s, the sampling distribution of
X will be approximately normal if the sample size
n is sufficiently large.
 If the shape is known to be non-normal, but the
sample contains at least 30 (n > 30)
observations, the central limit theorem
guarantees the sampling distribution of the mean
follows a normal distribution
 As before, if X is approximately
X 
normal, then we can transform it to Z 
 n
 Sampling from a
Normal Distribution
 For any sample size n, the sampling
distribution of X is normal if the
population X from which the sample is
drawn is normally distributed.
 If X is normal, then we can transform it
into the standard normal random variable
as: For a distribution of
For a sampling
distribution. the values of x.

Z
   X 
X E X
Z
x E X

x
SD  X   n SD  X  
 THE t DISTRIBUTION
DEFINITION
 The t distribution is a theoretical probability
distribution. It is symmetrical, bell-shaped, and
similar to the standard normal curve. It differs from
the standard normal curve, however, in that it has
an additional parameter, called degrees of
freedom, which changes its shape.
 More about the t Distribution
 The t distribution is similar to the standard normal
distribution (i.e., z Distribution) in that it is bell-shaped.
However, the height of the t distribution differs
depending on the sample size. By doing so, the
distribution becomes more conservative for smaller
sample sizes to account for some error that may occur
from estimating σ with s from a small sample.
As n approaches infinity (∞) the t distribution approaches
the standard normal distribution.

 The height of the t distribution is determined by the

number of degrees of freedom.
 df = n−1. (distribution based on one group)
 Degrees of freedom (fig)
 The effect of df on the t distribution is
illustrated in the four t distributions below.
 Note that the smaller the df, the flatter the

shape of the distribution, resulting in greater

area in the tails of the distribution.
 Student t- distribution
We knows the significant of the difference
between the mean of the sample and the
mean of the population
 t = X- 

/n
 Sampling Distribution of the
Differences of Means
Mean Standard
Value Variance Deviation

Population or Treatment 1 1 12 1

Population or Treatment 2 2 22 2

Sample Standard
Size Mean Variance Deviation
2
Population or Treatment 1 n1 x1 s
1
s1
Population or Treatment 2 n2 x2 s 2
2
s2
Sampling Distribution for Comparing
Two Means
If the random samples on which x1 and x 2
are based are selected independently of
one another, then
1.  x x   x   x  1   2
1 2 1 2

  2 2
2.  2
x1  x 2   
2
x1
2
x2 1 2
n1 n 2
 
2 2
x x  
1 2
1 2
n1 n 2
Sampling Distribution for Comparing
Two Means
If n1 and n2 are both large or the population
distributions are (at least approximately)
normal, then the distribution of
x1  x 2  (1   2 )
z
 
2 2

1 2
n1 n 2
Is described (at least approximately) by the
standard normal (z) distribution.
 Two-Sample t Test for Comparing Two
Population Means
If n1 and n2 are both large or if the population distributions
are normal and when the two random samples are
independently selected, the standardized variable
x1  x 2  (1   2 )
t
2 2
s s
1
 2
n1 n 2
Has approximately a t distribution with
 THANKS TO
Business Statistics: Communicating with
Numbers
By Sanjiv Jaggia and Alison Kelly

and
Prof. Dr. David Crystal
Rochester Institute of Technology
Dept of Mathematics and Statistics