You are on page 1of 19

Chapter 4

Sampling Design
• A sample is a fractional part of the population
selected for investigation.
• Sampling is the process of selecting a number
of study units form a defined study
population.
• Population is defined as the entire collection
of all persons, objects or items concerning a
specific area of investigation or study.
Criteria for selecting sampling
procedure or STEPS
1. Objectives
Define objectives of survey in clear and concrete terms
Objectives commensurate with money, manpower and time
2. Population
What should be the population ?
(Population should be clearly defined)
3. Sampling units and frame
Decision must be taken before selecting sample
Sampling unit – Geographical one – state, district, village
Construction unit – House, flat
Individual
List of sampling units is called ‘Frame’ or ‘sampling frame’
It contains names of all items of universe (in case of finite universe)
4. Size of sample
The size of sample should neither large nor too small it should
be optimum.
5. Parameters of interest
Statistical constraints of population are called parameters eg;-
population mean, proportion, etc.
When we do census survey we get actual value of parameters.
In sample we get estimate of unknown population.
6. Data collection
No irrelevant information should be collected and no essential
information should be discarded.
Objective of the survey should be clear in the mind of surveyor.
7. Non-respondents
Data may not be collected for all the sampling units because of
practical difficulties.
Non-response tends to change the result.
The reason for non-respondent should be recorded by the
investigator.
8. Selection of proper sampling design
Researcher must decide.
9. Organizing field work
The success of survey depends on the reliable field work.
Should be efficient supervisory staff and trained personal for the
field work.
10. Pilot survey
Always helpful to try out on a small scale before going to field.
Called pilot survey/ pretest.
It gives better idea of practical problems and troubles.
11. Budgetary constraint
Cost consideration for size and type of sample.
Implications for sample design
• A sample design is a definite plan for obtaining a sample
form the given population.
• Technique or the procedure researcher would adopt in
selecting items
• Sample design is determined before data are collected
• Researcher must select sample design which should be
reliable and appropriate for his study.
Characteristics of good sample design
• Sample design must result in a truly representative sample
• Each member in population should get an opportunity of being
selected
• Absence/ sampling error of sampling error
• Economically viable
• Generalized and applicable
• Goad oriented
• Proportional
• Randomly selected
• Actual information provider
• Practical
Different types of sample design
I. Non-probability sampling
II. Probability sampling
I. Probability sampling
• Based on theory of probability
• Sample is representative of whole population
• It Involves random selection
• Every subject in a population has equal chance
to be selected
• Chances of sampling and systematic bias are
less
Types of probability sampling
1. Simple random sampling
(i) Lottery method
(ii) Use of table of random numbers
(iii) Use of computer
2. Stratified random sampling
3. Systematic random sampling
4. Cluster/ Multistage sampling
5. Sequential sampling
(i) Lottery method ( simple random
sampling)
• 2 essential requirement
– Population must be homogeneous
– Researcher must have list of all members/elements in
population (sampling frame)  sample drawn
• Oldest and most mechanical one
• Each number of population is allotted a unique
number
• Blind-folded researcher chose numbered tags from the
bowl.
– Replacement method (chosen numbers are replaced back)
– Non-replacement method
(ii) Table of random numbers

(iii) Use of computer


www.randomizer.org
Simple random sampling
Advantages Disadvantages
• Ease of assembling • Need all members of
• Equal opportunity to population
select • Lots of procedure to be
• Free from sampling performed
erros • Expensive and time-
consuming
2. Stratified random sampling
• Used for heterogeneous population
• Eg:- study anthropometric measurements of
students 100 kerala, tamilnadu, karnataka
3. Systematic random sampling
• Selection of every Kth case eg:- 10th person,
100th person from phone directory
• K= Number of subject in target population(N)/
size of samples(n)
• To choose 100 subject from population of 500
• 500/100 ie, every 5th person is selected
4. Cluster or multistage sampling
• When a simple random sampling cannot be
taken eg:- entire population of Asia
• One-stage cluster
• Two-stage cluster
• Multistage cluster
5. Sequential sampling
• Here sample size is not fixed.
• Eg:- association between smoking and lung
cancer
II. Types of Nonprobability sampling
1. Purposive sampling : Subjects are chosen to be
part of the sample with a specific purpose in
mind
2. Convenience sampling: Subjects are selected
because of their convenient accessibility and
proximity to the researcher
3. Consecutive sampling: Pick up all the available
subjects who meet the present inclusion and
exclusion criteria
4. Quota sampling: Equal or proportionate
representation of subjects from each quota
5. Snowball sampling
Snowball sampling
• Locating the initial
subject, and then taking
assistance from the
subject to identify
people with a similar
trait of interest

You might also like