Beruflich Dokumente
Kultur Dokumente
BY
Bhaskar Banerjee
METHODS
PROBABILITY SAMPLING
NON- PROBABILITY SAMPLING
PROBABILITY SAMPLING
Every unit in the sampling frame has an
equal or known chance of being
included in the sample.
( BOYD AND KRESS)
It includes
Simple Random Sampling
Stratified Random Sampling
Cluster Sampling
Systematic Sampling
Area Sampling
SIMPLE RANDOM SAMPLING
It is a sample generated by a process
that guarantees, in the long run, that
every possible sample of a given size
will be selected with known and equal
probability
Eg: drawing a card from a pack of
well shuffled deck.
Selection Method: Table of random
numbers, Lottery method
Used to sample a universe with
following properties
It is small
Satisfactory list of universe items exists.
Cost per interview is practically
independent of the location of the sample
items.
Other than a list of items, no other
universe information is available.
Eg.- survey of members of gym in New
Delhi
PROBLEMS
Cost
Availability of a Current Listing of
Universe Elements
Statistical Efficiency
Administrative difficulty
STRATIFIED RANDOM
SAMPLING
The universe to be sampled is divided
(or stratified) into groups that are
mutually exclusive and include all
items in the universe.
A simple random sample is then
chosen independently from each
group or stratum.
Example
STORE SIZE NUMBER OF PERCENTAGE
STRATUM STORES OF STORES
Medium
30,000 30
stores
PROPOTIONAL ALLOCATION.
Sample each stratum in proportion to its
relative weight in the universe as a
whole.
DISPROPORTIONAL ALLOCATION
Sample institutional universes (grocery
stores, manufacturers) rather than
human universes.
When the variability among observations
within a stratum is high, one samples
the stratum at a higher rate than for
strata with less internal variation.
LIMITATIONS
Greater complexity of both design
and analysis.
A separate list of items within each
stratum is required
CLUSTER SAMPLING
Sampling methods in which universe
elements are chosen in groups
Widely used in sampling of human
populations
Example
BLOCK HOUSES
1 X1,X2,X3,X4
2 X5,X6,X7,X8
3 X9,X10,X11,X12
4 X13,X14,X15,X16
COMPARISION WITH SIMPLE RANDOM
SAMPLING
More or less statistically efficient than
simple random sampling.
Depends upon the degree of intracluster
heterogeneity obtained
Lower relative cost of obtaining
observations
TYPES
SYSTEMATIC SAMPLING
AREA SAMPLING
SYSTEMATIC SAMPLING
Most widely used form of cluster
sampling
Example
To study students’
opinion towards SAMPLE IDENTIFICATION
food provided in
canteen, the 1 1,6,11…96
researcher decides
to sample 20 2 2,7,12…97
students out of
100.
3 3,8,13…98
One method
Draw a random
number between 1 4 4,9,14…99
to 5.
Possible samples 5 5,10.…100
SELECTING A SYSTEMATIC SAMPLE
Sampling interval=
Number of universe items
Desired sample size
LIMITATION
Cannot be used if the universe items are
large and scattered.
JUDGEMENT SAMPLING
Is selection of universe items by
mean of expert judgment
Specialists in the subject matter of
the survey choose what they believe
to be the best sample for that
particular study.
Eg: Sales managers might select the
sample of grocery store in a city they
regarded as representative
APPLICATION
May be used when the total sample
size is extremely small
LIMITATION
This approach has been found
empirically to produce unsatisfactory
results.
QUOTA SAMPLING
Is most widely used in consumer
survey
Uses principle of stratification
The researcher begins by
constructing strata
Sample sizes (called quota) are
established for each stratum
The sampling within strata may be
proportional or disproportional
Example
STRATUM QUOTA
Men, 18-34 50
Men, 35-49 50