2. ī§ To understand the terms Sample and sampling.
ī§ To understand the need for taking a sample
ī§ To be able to enumerate various uses of sampling.
ī§ To understand different types and methods of sampling- its
advantages, disadvantages , uses and examples.
ī§ To understand the basic types of sampling errors.
Seminar ObjectivesâĻâĻ.
3. ī§To know whether the rice has been cooked
properly or not only a few grains are
enoughâĻâĻâĻâĻ..
4. What is Sampling?
īąSampling is defined as the practice of
taking a small part of a large bulk to
represent the whole.
ī§According to Non Lin, âsampling design is a subset of
cases from the population chosen to represent it. By using
the subset, we can infer the characteristics of the
population.â
5. â A sampling method is a scientific and objective procedure
of selecting units from a population & providing a sample
that is expected to be representative of the population as a
whole.â
Sampling
6. Population is defined as The Entire Group under study.
Sometimes it is also called as the âUniverse.â
Population
Target population
A set of elements larger than or different
from the population sampled and to which
the researcher would like to generalize
study findings.
7. Population versus Sample
ī§ Population = Parameter (N-size, Îŧ-mean, Ī -SD)
ī§ Sample = Statistic (n-size, x- Mean, s-SD)
ī§ Statistic gives estimates about parameter.
8. Sample - characteristics
ī§ True representative of population.
ī§ Inference drawn refers only to defined population
ī§ Have its statistic almost equal to population parameter
ī§ Precision (sample size should be large)
ī§ Unbiased in character
ī§ Still difference will persist which may be due to the sampling error
9. Sampling Unit
ī§ Smallest element of the population from which sample can be selected.
ī§ Generally well defined and identifiable
ī§ E.g.
a) To select persons,
sampling unit = individual person
b) To select families,
sampling unit = a family
10. Sampling frame
ī§ A List of all the sampling units
from which sample is drawn.
ī§ Example:
īTelephone directory
ī List of students in a college
12. Sampling scheme
ī Method of selecting sampling units from sampling frame
īType of sampling
īProbability sampling
īNon-probability sampling
13. ī§ To gather data about the population in order
to make an inference that can be generalized
to the population
The purpose of samplingâĻ
14. Define the target population
Select a sampling frame
Conduct fieldwork
Determine if a probability or nonprobability
sampling method will be chosen
Plan procedure for selecting
sampling units
Determine sample size
Select actual sampling units
Stages in the
Selection
of a Sample
15. âĸProbability sampling â a method of sampling that uses random
selection so that all units/ cases in the population have an equal
probability of being chosen.
âĸ Non-probability sampling â does not involve random selection and
methods are not based on the rationale of probability theory.
TYPES OF SAMPLING
Sampling
Techniques
Probability
Non-
Probability
19. Features of Probability Sampling
ī§ Provides equal chance to every element of the population in
getting selected in the sample
ī§ The results of probability sampling are more accurate and
reliable.
ī§ It helps in the formulation of a true representative sample
by eliminating human biases
20. SIMPLE RANDOM SAMPLING
īļBasic , most commonly and easiest method
īļPrinciple â Laws of Chance
īļAll subsets of the frame are given an equal probability. Each element of the frame thus has an
equal probability of selection.
īļTwo approaches are used in drawing a sample:-
1. Lottery method
2. Use of random number table
21. Lottery method
Two types of lottery method :
1) Lottery method with replacement
- Possibility of selected twice or more
2) Lottery method without replacement
- More convenient and provide more precise sample
22. Lottery method
Suppose 30 patients are required to be put in trial out of 300 available patients
Note the serial number from 1 to 300 on a card & shuffle
Draw out one and note the number
reshuffle & draw the 2ndcard(Replace/or Donât replace the card drawn)
Repeat the process till 30 numbers are drawn
23. Random number table
īļ Use of randomly generated number tables
ī§ It provides use of random numbers specially designed for sampling
purposes
ī§ Such type of random table are mostly found at the end of statistical
textbooks
25. SIMPLE RANDOM SAMPLING
Every subset of a specified size n from the population
has an equal chance of being selected
26. Simple Random Sampling
Advantages:
ī§ Sample is representative of the population
ī§ No need of prior information about the population
ī§ Equal and independent chance of selection of every element
Disadvantages :
ī§ Not feasible in case of large population
ī§ Requires a lot of resources
ī§ More chances of misleading sample
27. SYSTEMATIC RANDOM SAMPLING
ī§ Simple and convenient to adopt
ī§ Equal sampling interval.
ī§ Only the first unit is picked randomly.
28. Steps for systematic random
sampling
ī§ Number the units in the population from 1 to N
ī§ Decide on the n (sample size)
ī§ k = N/n = the interval size
ī§ Randomly select first unit between 1 to k
ī§ Take every kth unit
29. Every member ( for example: every 20th person) is
selected from a list of all population members.
SYSTEMATIC RANDOM SAMPLING
30. Advantages:
ī§ The sample is evenly spread over
the entire population.
ī§ The sample is easy to select
ī§ Helps in organizing field work.
ī§ Cost effective
Disadvantages:
ī§ Difficult if population is heterogeneous.
ī§ Pattern or periodicity
ī§ Precision of estimate is less as compared to
SRS.
ī§ Sample may be biased if hidden periodicity in
population coincides with that of selection.
ī§ Each element does not get equal chance
ī§ Ignorance of all in between elements
31. ī§ The population is divided into two or more groups called strata,
according to some criterion, such as geographic location, grade level,
age, or income, and subsamples are randomly selected from each
strata.
Stratified Random Sampling
32. ī§ When the population is not homogeneous
ī§ Heterogeneous population is divided into homogeneous sections or strata
Stratified Random Sampling
33. ī§ Samples are selected from each strata
ī§ Simple random or systematic sampling
ī§ Example:
Entire population can be divided into several
socioeconomic groups or strata
Stratified Random Sampling
34. Alcohol Consumption among a Population:
Stratified Random Sampling
20 - 30 years old
(homogeneous within)
(alike)
30 - 40 years old
(homogeneous within)
(alike)
40 - 50 years old
(homogeneous within)
(alike)
Heterogeneous
(different)
between
Heterogeneous
(different)
between
Stratified by Age
35. Advantages:
1. Every unit in the stratum has equal chance of being selected .
2. Characteristic of each stratum can be estimated and comparisons made.
Disadvantages:
1. Requires accurate information on proportions of each stratum
2. Stratified lists costly to prepare.
Stratified Random Sampling
36. Cluster Sampling
īąUsed to cover large-spread area. The main reason for cluster sampling is
âcost efficiencyâ (economy and feasibility)
īą Cluster :
ī A group of population elements, constitutes the sampling unit, instead of a single
element of the population
ī Villages , wards , blocks , factories worker
īą Steps for cluster sampling
1) Divide population into clusters
2) Randomly sample clusters
38. Advantages
īąEasier to apply to larger Geographical area
īąSaves time and cost of travelling
īą Examples
īļWHO selected 30 Clusters for the study of national immunization
coverage under EPI.
Cluster Sampling
39. Disadvantages
ī§ Requires cluster/group-level information.
ī§ Higher sampling error than random sampling techniques (less
efficient)
ī§ Advised to take large number of small clusters rather than small
number of large clusters.
ī§ In School example take sections 11th A , 12th B as clusters instead of
Class 11th and 12th .
Cluster Sampling
40. STRATIFICATION Vs CLUSTERING
STRATIFICATION CLUSTERING
Divide population into groups different From
each others: Sexes, Race, ages
( Homogeneous)
Divide population into comparable groups
Cities , Villages ,Wards , slums
( Heterogeneous)
Sample randomly from each group Randomly sample some of the groups
Less error compared to simple random More error compared to simple random
More expensive to obtain stratification
information before sampling
Reduce costs to sample only some areas or
organization
41.
42. Probability Proportional to Size sampling
ī§ Each sampling unit has equal chance of getting included in probability or
random sampling
ī§ This chance can be different for every sampling unit
ī§ Bigger the size of sampling unit or cluster, higher is chance of being
included in the sample or highest number of samples from that cluster.
ī§ Proportional to size of subgroups / clusters.
43. Example = Selection of Delhi hospitals using PPS
ī§ Sampling interval is calculated = total population/sample size.
ī§ Eg = 8300/10=830, if there are 10 clusters and total population is 8300
ī§ A Random number ( 3 digits) between 1 and 830 is selected by SRS using
random number table.
ī§ If the random no. selected is 677.
ī§ The first cluster selected will be found from the cumulative population column
A.
Probability Proportional to Size sampling
44. ī§ Next add on the sampling interval and look for its location in the
cumulative frequency
ī§ 677+830 = 1507 ,Hospital C, and so on till 10 clusters
ī§ Depending upon the sample size, equal no. taken from each cluster.
For n = 400 take 40 from each.
Limitation :
ī§ We need a sampling frame i.e. Listing of all the units of the
population.
Probability Proportional to Size sampling
45. Hospital with large population, more than one cluster selected
Hospitals in Delhi Population of
female staff
Cumulative
population
Selected Clusters
A 700 700 1 677th location
B 450 1150
C 600 1750 2 677+830=1507
D 300 2050
E 1200 3250 3,4
F 100 3350
G 200 3550
H 1500 5050 5,6
I 500 5550
J 250 5800 7
K 2500 8300 8,9,10
Maximum
Clusters
selected
46. MULTI-STAGE SAMPLING
ī§ Is selected in various stages but only last sample is studied .
âĸ Employed in large country survey
âĸ Different type of sampling methods may be used at different stage
48. Advantages:
ī§ Good representative of population
ī§ Improvement of other sampling method
Disadvantages:
ī§ Difficult and complex method
MULTI-STAGE SAMPLING
49. MULTIPHASE SAMPLING
ī§ Used when additional information for screening is not available and
resource limited.
ī§ Collection of basic information from a larger sample..
ī§ More specific information from successive sub samples.
ī§ Sampling unit at each phase remains same structurally
50. E g:- In a tuberculosis survey(hypothetical):-
Phase 1:-Physical examination or Mantoux Test may be done in all cases of
the sample.
Phase 2 :-Chest x-ray may be done in Mantoux positive cases and in those
with clinical symptoms.
Phase 3 :- Sputum may be examined in x-ray positive cases only.
Hence it avoids the expenses.
MULTIPHASE SAMPLING
52. NON-PROBABILITY SAMPLING
ī§ Also called as judgment sampling.
ī§ Units in the population do not have an equal chance or opportunity of
being selected in the sample.
ī§ Believes in selecting the sample by choice and not by chance.
ī§ Suffers defects , like personal bias and sampling error.
ī§ Unscientific and less accurate method of sampling, hence it is only
occasionally used in research activities.
54. QUOTA SAMPLING
ī§ The population is first segmented into mutually exclusive sub-
groups, similar to stratified sampling.
ī§ Then judgment used to select subjects or units from each segment
based on a specified proportion till the specified no. of sampling
units have been achieved.
ī§ For example, an interviewer may be told to sample 200 females and
300 males between the age of 45 and 60.
ī§ It is this second step which makes the technique one of non-
probability sampling.
ī§ In quota sampling the selection of the sample is non-random.
ī§ Based on pre-specified quotas regarding demographics, attitudes,
behaviors, etc
55. ī§ Advantages
ī§ Contains specific subgroups in the proportions desired
ī§ Does not require a sampling frame.
ī§ Easy to manage, quick
ī§ Disadvantages
ī§ Dependent on subjective decisions
ī§ Representativeness cannot be assured.
ī§ only reflects population in terms of the quota, possibility of
bias in selection, no standard error
QUOTA SAMPLING
56. CONVENIENCE SAMPLING
ī§ Sometimes known as grab or opportunity
sampling or accidental or haphazard sampling.
ī§ Selection of whichever individuals are easiest to
reach
ī§ It is done at the âconvenienceâ of the researcher
ī§ This type of sampling is most useful for pilot
testing.
57. Advantage: A sample selected for ease of access, immediately known
population group and good response rate.
Disadvantage: cannot generalise findings (do not know what population group
the sample is representative of) so cannot move beyond describing the sample.
âĸProblems of reliability
âĸDo respondents represent the
target population ? â is a question.
âĸResults are not generalizable
Use results that are easy to get
CONVENIENCE SAMPLING
58. JUDGMENTAL SAMPLING OR
PURPOSIVE SAMPLING
ī§ - The researcher chooses the sample based on who they think would be
appropriate for the study. This is used primarily when there is a limited
number of people that have expertise in the area being researched
ī¨ Selected based on an experienced individualâs belief
ī§ Advantages
ī§ Based on the experienced personâs judgment
ī§ Disadvantages
ī§ Cannot measure the representativeness of the sample
59. ī§ Useful when a population is hidden or difficult to gain access to. The contact with an
initial group is used to make contact with others.
ī§ Respondents identify additional people to included in the study
ī§ The defined target market is small and unique
ī§ Compiling a list of sampling units is very difficult
ī§ Advantages
ī§ Identifying small, hard-to reach uniquely defined target population
ī§ Useful in qualitative research
ī§ access to difficult to reach populations (other methods may not yield any results).
ī§ Disadvantages
ī§ Bias can be present
ī§ Limited generalizability
ī§ not representative of the population and will result in a biased sample as it is self-
selecting.
SNOWBALL SAMPLING
60. Sequential Sampling
ī§ Used in clinical trials for rare cases or costly drug.
ī§ Sample size in not fixed.
Double Sampling
ī§ Used for a costly research
ī§ To reduce cost another variable is used.(closely related & cheap to study)
61. INTERPENETRATING
SUBSAMPLING(REPLICATED SAMPLING)
ī§ Used by The National Sample Survey Organization(NSSO)
ī§ Independent subsamples are drawn from the population by simple
random sampling without replacement.
ī§ Distinct sampling unit should provide the required sample size
62. ī§ A researcher has 10 research assistants each with his/her own equipment that they
use to measure time it takes for people to respond to a command
ī§ A simple random sample of 80 people are taken
ī§ Since the researcher believes the assistants will produce slightly biased measurements,
He divides 80 people into 10 subsamples of 8 people each.
ī§ Each of the 10 assistant is assigned to one subsample
For example:
63. Advantages:
ī§ Reduces work load
Disadvantages :
ī§ Rarely used in clinical studies.
ī§ Inter observer variation.
INTERPENETRATING
SUBSAMPLING(REPLICATED SAMPLING)
65. Sampling Error
īļ The error arises as a result of taking a sample from a population rather than
using the whole population
īļExample: Healthy population Hb level- 14.5g%
Estimated sample value â 13.5g%
īļ Factors :-
1.Sample size (inadequate)
2.Natural variability of individual readings (chance)
Unavoidable in sample study, can be minimized.
66. Type 1 error
ī§ The probability of finding a difference with our sample
compared to population, and there really isnât oneâĻ.
ī§ Known as the alpha error (or âtype 1 errorâ)
67. Type 2 error
ī§ The probability of not finding a difference that
actually exists between our sample compared to the
populationâĻ
ī§ Known as the beta error (or âtype 2 errorâ)
68. SUMMARY
ī§ Sampling is used in all kinds of survey and research everywhere in the world .
ī§ Because of lack of resources and feasibility the investigator is forced to collect
information from a sample rather than the entire population.
ī§ Better sampling technique gives the best and appropriate result .
69. References
ī§ Principles and Practice of Biostatistics- by Dr. J.V. Dixit (6th edition)
ī§ Medical statistics â Principles and Methods â by K.R. Sundaram, S.N. Dwivedi, V.
Sreenivas. ( 2009)
ī§ Methods in Biostatistics â by Dr. K.S. Negi (2nd edition)
ī§ Parkâs Textbook of Preventive and Social Medicine- by K. Park. (23rd Edition)