2.
Whenever we have to carry out a study we will get
exact results only if we study the whole population.
In reality to study a whole population is not
possible..
So we take appropriate sample from the population
that will represent the entire population
INTRODUCTION
3. Definitions:
POPULATION :- The group of individuals which have
something in common is called a population. This is also known as
total population or Target population or reference population or
Universe.
e.g.: a group of all patients suffering from lung cancer in India, all
patients with IHD in the world etc.
STUDY POPULATION:- Study population is a subset / part of total
population i.e. it is a collection of individuals from the total
population who are feasible to study.
e.g.: a group of all patients suffering from lung cancer who are
residing in Pune, all patients with IHD admitted to a particular
teaching hospital etc.
SAMPLE:- Sample is a subset of the study population, selected in a
such way that it is a representative of the total population.
e.g.: a group of all patients suffering from lung cancer who are
residing in Pune, attending OPD at SKNMC
4. POPULATION- all the
patients suffering from
lung cancer in INDIA
STUDY POPULATION- all
the patients suffering from
lung cancer in pune
SAMPLE- Patients attending
cancer opd at SKNMC
Relationship between Total population, study population
and sample in a study examining all patients suffering from
lung cancer
5.
Sampling Unit: It is an
individual member of the
population on which the
information is obtained.
Sampling unit may be an
individual or a household
or a family or a village/
state/country. etc..
Sampling Frame: A complete list of each and every individual
(sampling unit) in the study population is called the Sampling frame.
e.g.: If you want to draw a sample of adolescent girls from Wanowari area,
then the sampling frame is list of all adolescent girls residing in Wanowari
Area. Here each adolescent girl is called as a sampling unit
6.
Sampling is the
process of selecting a
small number of
individuals from a larger
defined target population
so that the information
gathered from the small
group can be generalized
to the target population.
What is SAMPLING ????
7.
Sampling is commonly used in our day today life.
e.g.: whenever we go to market to buy something, we
find the quality of product by taking sample of it.
Sampling is used in research in health related fields
Applications of SAMPLING
8.
Sampling enables us to know about a large population
Advantages of
SAMPLING
AT
REASONABLY
LOW COST
AT GREATER
SPEED
WITH
INCREASED
ACCURACY
TO DRAW
INFERENCES
ABOUT
POPULATION
9.
Classification of
Sampling Techniques
Probability Sampling
1. Simple random
sampling
2. Systematic random
sampling
3. Stratified random
sampling
4. Multi-stage sampling
5. Multi-phase sampling
6. Cluster sampling
10.
Involves random selection of elements in
which each element has a chance of being selected
Simple Random Sampling
Systematic Sampling
Stratified Sampling
Multi-stage sampling
Multi-phase sampling
Cluster Sampling
Probability Sampling
11.
Applied when the population is SMALL, HOMOGENOUS
and READILY AVAILABLE (eg.patients coming to hospital
or lying in the wards)
Used in experimental medicine or clinical trial.
SRS can be done by using
i. Lottery method
ii. Table of random numbers
Simple random sampling
12.
Lottery method
Prepare the sampling frame and assign numbers between 1
to N
Prepare N slips bearing numbers
from 1 to N
Mix all the slips
Select a slip
Record the number on the
slip
Continue the process till desired
sample size is achieved
13.
e.g.: In an area there are 500 families, how will you
select a sample of 15 families randomly to find out the
standard of living in that area ?
in this question N = 500 (which is a 3 digit number)
assign number from 1 to 500 to each of the
500 families
Example for table of random
numbers
14.
Select a page randomly from random number table
From this selected page, select a point randomly and from this
selected point consider 3 columns (since N contains 3 digits)
If the selected random number < N;
then the member of population corresponding to the random
number is included in the sample.
If a selected random number > N;
then divide the random number by N and the member of the
population corresponding to remainder is included in the sample.
Move down in the random number table from the last selected
random number till the desired sample size is achieved.
Table of random numbers
15.
4652 3819 8431 2150 2352 2472 0043 3488
9031 7617 1220 4129 7148 1943 4890 1749
2030 2327 7353 6007 9410 9179 2722 8445
0641 1489 0828 0385 8488 0422 7209 4950
Part of random number table…..
16. select a point randomly and from this selected point
consider 3 columns (since N contains 3 digits)
Now we start from the third row, the numbers are:
203 023 277 353 600 794 109 179
272 284 450 641 148 908 280
If selected Random No. < N,
then the member of population corresponding to the
random number is included in the sample:
203, 023, 277, 353, 109, 179, 272, 284, 148, 450, 280
Selected Random number > N are:
600, 794, 641, 908
then divide these random number by N and the member
of the population corresponding to remainder is
included in the sample
corresponding remainders are 100, 294, 141, 408 resp.
17.
So the selected 15 families will be
203 023 277 353 100 294 109 179
272 284 450 141 148 408 280
i.e.
203 023 277 353 100 294 109 179
272 284 450 141 148 408 280
18.
Applied when the population
is LARGE, SCATTERED and
HETEROGENOUS.
THIS METHOD IS POPULARLY
USED WHEN A COMPLETE LIST OF
POPULATION IS AVAILABLE
BEFORE A SAMPLE IS DRAWN.
THEN ARRANGE THE LIST IN ANY
ORDER. E.G.: ALPHABETICAL
ORDER OR NUMBERS ETC..
Systematic random
sampling
19.
Assign number to each individual or unit.
Choose a sample by taking every “K” th patient or unit
where “K” is sample interval.
In this method, first unit is selected by simple random
sampling method then remaining units are selected in a
definite sequence at equal interval.
HOW DO WE DO SYSTEMATIC
RANDOM SAMPLING???
K = TOTAL POPULATION / DESIRED SAMPLE SIZE
20.
If population is 5000
And sample size is 250
now the value of K will be
K = 5000/250 = 20
Here K = 20 is a sampling interval
e.g.: if we have chosen no. 8; then our sample will be
8, 28, 48, 68, 88, …..
EXAMPLE
K = 20
22.
Stratified random sampling is done when the population
is heterogeneous.
Here first heterogeneous group will be divided into
HOMOGENOUS GROUP called as STRATA ,with respect
to characteristic of interest. e.g.: age, sex, socioeconomic
status, etc.
The samples are drawn from each stratum at random in
proportion to its size.
Stratified random
sampling
24.
If we want to do study with respect to socioeconomic group in a
population.
The population can be divided into 4 strata
1. Higher class ( 20%)
2. Upper middle class ( 40%)
3. Lower middle class ( 30%)
4. Lower class (10%)
If in the given population, higher class strata is 20% of the total
population then 20% of the sample should be from higher class.
Similarly 40% , 30% and 10% of the sample should be from
upper middle class, lower middle class and lower class
respectively. and if the sample size of 50 is to be taken, then…..
Example of stratified random
sampling
25.
DETAILS
STRATA / SOCIOECONOMIC GROUP
HIGHER
CLASS
UPPER
MIDDLE
LOWER
MIDDLE
LOWER
CLASS
TOTAL
PERCENTAGE
OF
POPULATION
20% 40% 30% 10% 100%
DISTRIBUTION
OF SAMPLE
SIZE TO BE
SELECTED
10 20 15 05 50
26.
Multi stage sampling
This method of sampling is applied when the
population is large, scattered and not homogeneous.
used for large data surveys.
The sampling is carried out in stages (first stage unit,
second stage unit, and third stage unit and so on) using
suitable random sampling method.
27.
Example of multistage sampling
suppose we want to find out immunization status among children
under five years of age in Maharashtra. Here in this study
sampling unit is a child under five years of age. The sampling
may be carried out as shown in the table.
STAGE UNIT SAMPLE
FIRST DISTRICT
Sample of children
under 5 in Pune
SECOND TALUKA
Sample of children
under 5 at Lonawla
THIRD VILLAGE
Sample of children
under 5 at Kusgaon
28. Multi-stage Sample Design
state
state
state
state
state
District
District
District
District
village
village
village
village
29.
MULTI-PHASE SAMPLING
In this method, part of information is collected from
whole sample and part of information from sub-sample.
Example – “ TUBERCULOSIS SURVEY ”
1st PHASE - We first select a sample population. Then we
carry out a Montoux test on all the members in the
sample.
2nd PHASE - We carry out X-ray Chest only on those
members who had a positive Montoux test.
3rd PHASE - We carry out sputum test only on those people
who are X-ray positive.
30.
So, those who need sputum
examination will be small. Number of
members in the IInd and IIIrd phases
will be successively smaller and smaller
31.
CLUSTER SAMPLING
Method by which the population is divided into
groups (clusters), any of which can be considered a
representative sample.
For cluster sampling you must know the total
population of the area to be surveyed and the
population of the cities, towns and villages in the
area.
32.
Cluster sampling is used when the
population is heterogeneous, scattered
and when it is difficult, costly or
impossible to make a complete list of
population.
Cluster Sampling
33.
A very useful method for field
epidemiological research and administrators.
A special form of cluster sampling called “30
cluster sampling” has been recommended by
the WHO for field studies in assessing
Immunization coverage.
34.
In this a list of all villages for a given geographical
area is made.
30 clusters selected from these villages.
From each of the selected clusters, 7 subjects are
randomly chosen. Thus a sample of 30X7 = 210
subjects are chosen
35.
Example
e.g. We want to find how many children or
what %age of children in a particular area are
fully immunized. For this we use a sampling
technique called CLUSTER SAMPLING.
For this survey, we have to select 30
clusters. Each cluster containing 7 children in
the age group of 12-23 months
36.
How to draw clusters??
1. List all the villages included in the area to be studied.
2. List the individual population of each village.
3. Calculate and write in the cumulative populations as each village is
added.
To obtain a cumulative population you must add the population
of the next village to the combined total of all populations in preceding
villages.
The final cumulative population is the same as the total population
to be surveyed.
38. Distribution of the clusters
A
B
C
D
E
F
G
H
I
J
1600
220
3200
400
800
200
1200
200
1600
400
9820
1600
1820
5020
5420
6220
6420
7620
7820
9420
9820
Total population = 9820
Compute cumulated population
39. Distribution of the clusters
Then compute sampling INTERVAL :
K= 9820/ 30 = 327
Draw a random number ( between 1
and 327 with same number of digits as
sampling interval ), which should be just
less than or equal to sampling interval
so we have chosen :
example: 320
Start from the village including “320” and
draw the clusters adding the sampling
INTERVAL
A
B
C
D
E
F
G
H
I
J
1600
1820
5020
5420
6220
6420
7620
7820
9420
9820
I I I I
I
I I I I I I I I I I
I
I I
I
I I I I
I
I I I I I
I
40.
Identify the Village in which Cluster 1 is located. This is done by
locating the first village listed in which the cumulative population
equals or exceeds the random number (320)
Identify the village in which Cluster 2 is located. Use the formula
below.
The cumulative population listed for that village will equal or
exceed the number you calculate.
Random number + sampling interval = 320+327 = 647 =2nd Cluster.
So both cluster 1 and cluster 2 will be from Village A
Cluster 3 = 647+327 = 974 so cluster 3 is also from village A Thus we will
identify 30 clusters.
We will get 4 Clusters from village A, 1 from village B and so on.
41.
Drawing households and children
Go to the center of the village , choose direction
(random)
Number the houses in this direction (e.g. 21)
Draw random number (between 1 and 21) to identify
the first house to visit
From this house progress until finding the 7 children.
42.
Then from each cluster we will randomly
select 7 children in the age group of 12-23 months.
Thus we will get 210 children. Then from these 210
children, we will find out how many children are
fully immunized.
This we can find the %age of children who are fully
immunized (Immunization Coverage).
. Im
Im 100
210
Total no of children fully munized
munizationCoverage
43.
Q1: The study has been planned to find out the health
profile of batch A of IInd MBBS students. For this
study 5 students have to be selected from the batch
of 45 . How will you select students?
Sol.: As population is very small (i.e. 45 students) & all
are from IInd MBBS. So by Simple Random
Sampling Method, we have selected :
9, 42, 6, 12, 33
Journal Questions
44.
Q 2: Select a sample of 10 from a population of 300
female patients attending the MCH centre. Draw a
sample by using table of random number.
Sol.: N = 300 (So 300 is a 3 digit figure)
Now we will choose 3 digit no. from random
number table. So the numbers are:
696, 563, 033, 728, 522, 846, 438, 569, 899, 013
Now we will take the remainder of all the numbers
which are more than 300. So the actual nos. are:
96, 263, 33, 128, 222,246, 138, 269, 299, 13
45.
Q3: A study has been planned to find out sanitation facility in
village of 200 houses from Pune district, by doing a survey
of 10% houses from that villages. How will you select a
sample?
Sol: Sample Size = 10%
Population of houses = 200
So 20 houses have to survey
Sampling Interval = 200/20 = 10
1, 2, …., 6,…., 10
11, 12,…., 16, … , 20
…….
191, 192, …, 196,…,200
So the sample we have selected is 6 to 196
Thus the no. of houses to be chosen are: 6, 16, 26, 36, 46,56, 66,
76, 86, 96, 106, 116, 126, 136, 146, 156, 166, 176, 186, 196