SAMPLING & ESTIMATION

Main Issues
 Universe/Population
 Sampling Frame
 Sampling Unit
 Sample Size
 Budgetary Constraints
 Sampling Procedure
 Universe/Population
 CENSUS STUDY
 Sample
 Sampling Unit
 Sampling Frame: representation of the elements of the target
population. Examples of a sampling frame include the telephone book,
an association directory listing the firms in an industry, a customer
database, a mailing list on a database purchased from a commercial
organisation, a city directory, or a map. If a list cannot be compiled,
then at least some directions for identifying the target population
should be specified, such as random-digit dialling procedures in
telephone surveys.
 Sample Size
 Budgetary Constraints
 Sampling Procedure
Criteria of Sampling Design
Cost of
collecting &
analyzing Data

Minimise cost of sampling

Cost of
incorrect
inferences
Systematic bias &
Leads to
Sampling error

Systematic bias – Inherent in the System

Design Errors: Selection error, Sampling frame error, Measurement scale error
Administering Error: Questioning error, Recording error
Response Error: Data error (intentional/ unintentional)
Non response Error: Failure to contact all members, Incomplete responses
Random/Sampling error – Random variation, controllable by sample size
difference between measure obtained from the sample and the true measure of
the population
 Sampling Methods
A.Non-random/Non-probability-based sampling: relies
on the personal judgement of the researcher rather
than on chance to select sample elements.
• Convenience sampling: selection of sampling units is
left primarily to the interviewer. Often, respondents
are selected because they happen to be in the right
place at the right time. Examples: (1) use of students
and members of social organisations, (2) street
interviews without qualifying the respondents, (3)
some forms of email and Internet survey, (4) tear-out
questionnaires included in a newspaper or magazine.
• Judgmental sampling: elements are selected based on
the judgement of the researcher because he/she
believes that they are representative of the population
of interest or are otherwise appropriate. Examples: (1)
test markets selected to determine the potential of a
new product, (2) purchase engineers selected in
industrial marketing research because they are
considered to be representative of the company, (3)
product testing with individuals who may be
particularly fussy or who hold extremely high
expectations, (4) expert witnesses used in court.
Quota sampling: two-stage restricted judgemental sampling that is
used extensively in street interviewing.
• The first stage consists of developing control characteristics, or
quotas, of population elements such as age or gender. To develop
these quotas, the researcher lists relevant control characteristics
and determines the distribution of these characteristics in the
target population, such as Males 49%, Females 51% (resulting in
490 men and 510 women being selected in a sample of 1,000
respondents). Often, the quotas are assigned so that the
proportion of the sample elements possessing the control
characteristics is the same as the proportion of population
elements with these characteristics. In other words, the quotas
ensure that the composition of the sample is the same as the
composition of the population with respect to the characteristics
of interest.
• In the second stage, sample elements are selected based on
convenience or judgement.
• Snowball sampling: an initial group of respondents is selected
who possess the desired characteristics of the target
population. After being interviewed, these respondents are
asked to identify others who belong to the target population.
Subsequent respondents are selected based on the referrals.
By obtaining referrals from referrals, this process may be
carried out in waves, thus leading to a snowballing effect. The
main objective of snowball sampling is to estimate
characteristics that are rare in the wider population.
• Examples: users of particular government or social services,
such as parents who use nurseries or child minders, whose
names cannot be revealed; special census groups, such as
widowed males under 35; and members of a scattered
minority ethnic group; Industrial buyer using some special
equipment or technology;
B. Random/Probability- based sampling
1. Simple random sampling
 Each element/item has equal chance of getting included in a
sample. Randomness.
 Sampling with/without replacement
 Random number table, pseudo-random number generator.
2. Stratified Sampling
 Each stratum is a homogeneous group and different from
other strata.
 Random selection from each stratum, proportionately.
3. Cluster sampling
 Least or no variation among clusters.
 Clusters are selected randomly for further
analysis.
 Area sampling in geographical clusters.
 Multi-stage sampling as a special case.
4. Systematic sampling
 Elements selected at a uniform interval.
 Selection evenly spread, less cost & time, more
convenient.
 the sample is chosen by selecting a random starting
point and then picking every i th element in succession
from the sampling frame.
 The sampling interval, i, is determined by dividing the
population size N by the sample size n and rounding to
the nearest whole number. For example, there are
100,000 elements in the population and a sample of
1,000 is desired. In this case, the sampling interval, i, is
100. A random number between 1 and 100 is selected.
If, for example, this number is 23, the sample consists
of elements 23, 123, 223, 323, 423, 523, and so on.
Sample Size Determination:
 2z2
n 2
for mean
D
p (1  p ) z 2
n 2
for proportion
D
D  Level of precision
z is associated with Confidence Interval
 SAMPLING DISTRIBUTION
• Sampling Distribution: Distribution of a sample
statistics, usually mean.
• Standard error( ): Standard deviation of the
sampling distribution.
• Mean of sampling distribution( ) of means, taking
all possible samples exhaustively, approaches to
population mean (µ), particularly for normal
population distribution.
• As sample size increases, standard error decreases.
Assuming Normal Population Distribution

n = Sample size
Central Limit Theorem:

Irrespective of shape of population distribution, sampling

distribution approaches to normal, as sample size increases.
Point Estimate
Interval Estimate.
 Confidence Level:
 Level of significance, α
 Probability that is associated with an interval
estimate (1- α), of any population parameter.
 Higher confidence level => Wider confidence
interval
Estimation of mean from large sample(usually n> 30):
As sample size is large, sampling distribution of
mean is normal.
1. Compute from either known or estimated

2. Get Z value from standard normal distribution table

corresponding to confidence level (1- α).
3. The confidence interval
Estimation of means from small samples(n<30):
t-distribution:
 Applicable for smaller sample size.
 Unimodal and almost like a bell shape.
 Flatter than normal.
 Larger the sample size less flatter the distribution shape and
closer to normal.
 Value of t varies with d.f.i.e.(n-1) as the distribution shape
changes.
Step 1. Compute ( ) as usual
Step 2. Get t value from t- distribution table corresponding to
(n- 1) as d.f. and (1- confidence level) as the area under curve.
Step 3. ± t is the confidence interval/limit.
 Two sided Confidence
Case Interval (CI)

Population standard deviation, σ 𝜎

𝑥 ± 𝑍𝛼/2
known 𝑛

Population Sample size n > 30 𝑠

𝑥 ± 𝑍𝛼/2
standard 𝑛
deviation, σ
unknown
Sample size n ≤ 30 𝑠
𝑥 ± 𝑡𝛼,𝑛−1
2 𝑛
Example 1: A sample of size 20 was collected
and the sample mean and standard deviation
are estimated as 9.8525 and 0.0965. Find 95%
two-sided CI for the mean.
Example 2:
A manufacturer produces piston rings for an automobile engine. It is known that ring diameter is approximately
normally distributed and as a standard deviation σ = 0.001 mm. A random sample of 15 rings has a mean diameter
of = 74.036 mm. Construct a 90% two-sided CI.
• Example 3: The life in hours of a light bulb is
known to be approximately normally distributed
with standard deviation of 25 hours. A random
sample of 40 bulbs has a mean life of 1014 hours.
1. Construct a 95% two-sided CI on the mean life.
2. Construct a 95% one-sided lower CI of the mean life.

One-sided confidence interval: Appropriate lower or upper

confidence limit are found by replacing
𝑍𝛼/2 by 𝑍𝛼 and 𝑡𝛼,𝑛−1 by 𝑡𝛼,𝑛−1
2
• Example 4: The following result shows the
investigation of the haemoglobin level of hockey
players (in g/dl).
15.3 16.0 14.4 16.2 16.2
14.9 15.7 14.6 15.3 17.7
16.0 15.0 15.7 16.2 14.7
14.8 14.6 15.6 14.5 15.2

a) Find the 90% two-sided CI on the mean 15.43684211

0.83413996
haemoglobin level.
b) Also construct 90% Upper CI on the mean
haemoglobin level.
Confidence Interval on the Variance of a Normal Distribution

Confidence Intervals on a Population Proportion

Example: An automatic filling machine is used to fill bottles with
liquid detergent. A random sample of 20 bottles results in a
sample variance of fill volume of 0.0153. If the variance of fill
volume is too large, an unacceptable proportion of bottles will
be under- or overfilled. We will assume that the fill volume is
approximately normally distributed. Calculate 95% upper-
confidence interval for variance.

Therefore, at the 95% level of confidence, the data indicate that the process
standard deviation could be as large as 0.17
Example: In a random sample of 85 automobile engine
crankshaft bearings, 10 have a surface finish that is rougher than
the specifications allow. Therefore, a point estimate of the
proportion of bearings in the population that exceeds the
𝑥 10
roughness specification is 𝑝 = = = 0.12 . Compute 95%
𝑛 85
two-sided confidence interval for p.
Example:
The Salk polio vaccine experiment in 1954 focused on the effectiveness of the vaccine in combating paralytic
polio. Because it was felt that without a control group of children there would be no sound basis for
evaluating the efficacy of the Salk vaccine, the vaccine was administered to one group, and a placebo (visually
identical to the vaccine but known to have no effect) was administered to a second group. For ethical reasons,
and because it is suspected that knowledge of vaccine administration would affect subsequent diagnosis, the
experiment was conducted in double-blind fashion. That is, neither the subjects nor the administrators knew
who received the vaccine and who received placebo. The actual data for this experiment are as follows:
Placebo group: n = 201299 110 cases of polio observed
Vaccine group: n = 200745 33 cases of polio observed.
a. Find a 95% two-sided CI on the proportions of children in the two groups who contracted paralytic
polio.
b. What conclusions can you draw from the CI in part (a).