SAMPLING

Lets Try Few Questions

Q.1 If you were asked to buy a good quality wheat, how will you do?
Options 1: You will take a han dful of wh eat and see if it h as good colour, size,
uniformity, quality etc.
Option 2: You will check each and ever y piece of w hea t fo r its colour, size,
uniformity, quality etc.
Option 3: Will buy the wheat without testing it for anything.

Q.2 You are outside an inox multiplex at Hyd era bad and thinking of watching a
movie. The probl em is you didn’t hear the ongoing movie na me a nd even actors ar e
also not known to you. How will you decide w hether to see or not to s ee th e
movie?
Option 1: You will enquire a ll the outgoing peopl es who have just wa tched th e
movie regarding review.
Option 2: You will get th e r eview from certain sel ected p eopl es who ha ve just
watched the movie.
Option 3: You will directly watch the movie and then decide if it was good or bad.
 SAMPLING

Introduction:
Sampling is a scientific tool to shortlist certain selected
respondents or objects for study from amongst the wide
arena, with a view to save Cost, Time and other Resources.

Definition:
“In the social science, it is not possible to collect data from
every respondent relevant to our study but only from some
fractional part is called Sampling.
- Davis S. Fox
 BASIC CONCEPTS IN SAMPLING

Universe/Population:
• All the units collectively related with the subject matter
about which an investigation is being conducted.
• Types: Finite Universe and Infinite Universe
Statistical Population:
• A statistical population is a set of entities concerning which
statistical inferences are to be drawn, often based on a
random sample taken from the population.
Sample:
• Sample is a portion of the population which is examined
with a view to estimating the characteristics of the
population.
 BASIC CONCEPTS IN SAMPLING

Sampling Frame:
• Sampling frame is the actual set of units from which a
sample will be drawn.
Sample Design:
• A sample design is a definite plan for obtaining a sample
from the sampling frame. Population
Sample
Frame
Sample
CHARACTERISTICS OF A GOOD SAMPLE

• Representative • Free from Bias

• Objective • Accurate
• Comprehensive • Economical
• Approachable • Good Size
• Feasible • Practical
 SAMPLING PROCEDURE

Define the Universe

Develop Sample Frame

Specify Sampling Unit

Select Sample Design

Determine Sample Size

Select the Sample

 SAMPLING

Advantages:
• Saves time, money and resources.
• More Effective
• Faster and Cheaper
• More Accurate
• More Comprehensive Information
Disadvantages:
• Less Accuracy
• Can’t be inferred to generalizations
• Misleading Conclusions
• Need for Specialized Knowledge
 SAMPLING METHODS
Two Types:
• Probability Sampling
• Non Probability Sampling

Probability Sampling Non Probability Sampling

• Simple Random Sampling • Convenience Sampling
• Systematic Sampling • Judgement Sampling
(Purposive Sampling)
• Stratified Random • Quota Sampling
Sampling (Purposive Sampling)
• Cluster Sampling • Panel Sampling

• Multi Stage Sampling • Snowball Sampling

• Area Sampling -
 PROBABILITY SAMPLING
a ) S i mple Ra n dom S a mp ling:
• Gives each unit an equal chance of getting selected.
• Omits Biases and Prejudices in Sample Selection
Typ es o f S i mple Ra n dom S a mpling :
• Lotter y Method
• Random Number Table
Ad vanta ges:
• Free from Bias
• Representativeness
• Ease of Sampling and Analysis
Di s advantages :
• Costly
• Requires the Availability of a current Universal Elements i.e. Population.
• Statistical Efficiency
• Administrative Difficulty
 PROBABILITY SAMPLING

b) S ys tematic S a mp ling:
• First element is selected at random and then subsequent elements at some
common inter vals are selected.
• Can be used if population is finite and elements can be arranged in some system
say alphabets, numerals etc.
Ad vanta ges:
• Simple and Convenient
• Gives Similar Results
• Independent
• Little chance of Biases
• Helps in Random Selection.
Di s advantages :
• High Sampling Error
• Possibility of selecting impracticable units
• Biased
• Not suitable for Large Population.
 PROBABILITY SAMPLING

c) Stratified Random Sampling:

• The universe to be sampled is divided into groups (Strata) having some
characteristics (Mutually Exclusive & All Inclusive i.e. Exhaustive)
• And then a simple random sampling is used from each group (Strata) to
select element.
Advantages:
• More Representative Sample
• Certainty
• Greater Precision
• Administrative Convenience.
Disadvantages:
• Needs more attention
• Time Consuming
• Complicated
• Expensive.
 PROBABILITY SAMPLING

d ) C l u s ter S a mpling:
• The universe is divided into some recognizable subgroups called Clusters.
• And then a simple random sampling is used for selecting certain number of
subgroups i.e. Clusters
• And lastly all the elements from the selected clusters constitute the sample.
Ad vanta ges:
• Cheap, Quick and Easy
• Larger Sample Size
• Convenient to obtain
• Cost Effective
Di s advantages :
• Less Representative
• High Sampling error
• Less efficient
• Sometimes leads to non appropriate Sample.
 PROBABILITY SAMPLING

e) Multi Stage Sampling:

• This is a modified Cluster sampling.
• As in Cluster sampling we select all the elements from certain
selected cluster, here we form the different stages of cluster elements
and then select the sample from different stages.
Advantages:
• Flexible
• Saves Time
• Administrative Efficiency
• Helps in survey of undeveloped areas
• Sampling of large units is possible.
Disadvantages:
• Large number of errors are possible
• Greater Variability
• Less Efficient.
 PROBABILITY SAMPLING

f) Area Sampling:
• Area Sampling is a form of multistage sampling in which
maps, rather than lists are used as sampling frame.
• If cluster sample selects geographical clusters then it is
better termed as Area Sampling.
Advantages:
• Convenience
• Cost
• Feasible.
Disadvantages:
• Similar Elements
• Sometimes Costly.
 NON PROBABILITY SAMPLING

a ) C o n venience S a mpling:
• Selection of the sample on the basis of convenience or accessibility.
• When population is difficult to locate or when some element can be better
sample then other due to some characteristics then convenience Sample is used.
• This sample might be suitable when total sample size is ver y small.
Ad vanta ges:
• Economical
• Proper Representation
• Avoid Irrelevant Items
• Intensive Study
• Accurate Results.
Di s advantages :
• Personal Bias
• No Equal Chance
• No degree of accuracy
• Unsuitable for large samples
 NON PROBABILITY SAMPLING

b) Judgement Sampling:
• Here Researcher applies his judgement in selecting the sample for
study.
• When population is difficult to locate or when some element can be
better sample then other due to some characteristics then Judgment
Sample is used.
• This sample might be suitable when total sample size is ver y small.
Advantages:
• Suitable for Small Sampling
• Studying unknown traits of population
• Solves ever yday business problems.
Disadvantages:
• Non Scientific
• No method to calculate sampling error.
 NON PROBABILITY SAMPLING
c) Quota Sampling:
• One of the most popular method in consumer surveys.
• Quota Sampling also uses the principle of stratification. Wherein groups
are formed on the basis of certain characteristics mostly demographic
identity.
• Then researcher defines the Quota from each group to be studied
weather proportionately or disproportionately.
• Selection of sample within Quota is vested with the Field Worker who is
actually involved in the data collection.
Advantages:
• Economical
• Administratively Convenient
• Minimum Memor y Errors
• Independent.
Disadvantages:
• Difficulty in calculating standard errors.
• Difficulty in obtaining representative sample.
• Hampers quality of work.
• Difficult to supervise and control.
 NON PROBABILITY SAMPLING

d) Panel Sampling:
• In panel sampling firstly on the basis of random sampling
participants or panel is identified.
• And then the same panel is asked for the information over a period
of time.
• Suitable for Longitudinal Studies.
Advantages:
• Saves Cost and Time
• Helps in measuring changes
• Helps in tracing shift in behaviour.

Disadvantages:
• Not representative.
• Members become conditioned.
• Difficult to preserve representative character of panel.
 NON PROBABILITY SAMPLING

e) Snow Ball Sampling:

• This is a special non probability method used when the desired
sample is not known or rare to identify.
• It relies on referrals from earlier element or respondent.

Advantages:
• Helps in identifying the perspective respondent or sample.
• Useful in Qualitative research.
• Needs little planning.
• Less costly.
Disadvantages:
• Biased
• Limited data structure
• Limited control
• Researcher has no idea of distribution.
 SAMPLING ERRORS

Meaning:
• Sampling error means generation of errors in the research process
due to faulty or deceased sample selection and or sampling
procedure.
Types of Sampling Errors
• Biased Sampling Errors
• Unbiased Sampling Errors
Causes for Sampling Errors:
• Faulty sample selection.
• Non suitable sampling method selection.
• Incorrect Sample size selection.
• Error due to substitution.
Preventing Sampling Errors:
• Increase Sample Size.
• Divide Population by doing Stratification.
 NON SAMPLING ERRORS
M ea n ing:
• Non Sampling error means generation of errors in the pre and or post sampling
procedure i.e. Research Planning, Data Collection, Data Measurement and Data
Analysis and Interpretation.
Typ es o f N o n S a mp ling E r rors:
• Frame Error
• Non Response Error
• Measurement Error
• Data Processing Error
• Data Analysis Error
C a u ses f o r N o n S a mpling E r rors:
• Errors due to Faulty planning and definition.
• Response Error.
• Non Response Error.
• Error in coverage.
• Compilation errors.
Pr eventing N o n S a mpling E r rors:
• Results should be thoroughly edited.
• All steps of processing must be well performed.
• The Trained Professionals must be employed in the research.
 SAMPLE SIZE CONSTRAINTS

Meaning:
• Generally larger the sample size higher is the quality of
research output, however organizational limitations like -
a) Cost;
b) Time Availability and
c) Resources Availability
poses some sort of Constraints leading to sample size
selection.
Determination of Sample Size:
a) Determination of Sample Size (n) when estimating the
Population Mean
b) Determination of Sample Size (n) when estimating the
Population Proportion
 SAMPLE SIZE CONSTRAINTS
a) Determination of Sample Size (n) when estimating the
Population Mean:
z6 2
n =
e
Where,
z- For Large Samples; 6- Standard Deviation & E- Sampling
Error
b) Determination of Sample Size (n) when estimating the
Population Proportion:
z 2 pq
n =
e2
SAMPLE SIZE BY MORGAN TABLE