Sample size and sampleing

Amna Khairy
 MBBS. MPH
Community Medicine Lecturer
EBM Instructor/Altababa Training Center
1/28/2021 1

 To Understand the interpretation of
important terms related to Sample size and
Sampling .
 To Be able to choose appropriate Sampling.
 To calculate sample size for cross sectional
studies .
1/28/2021 2

Cross Sectional study on :prevalence of
depression among student at university of
Khartoum 2017. sample size is 400 students,
how we are going to get a sample ?
1/28/2021 7

1/28/2021 8
Probability
(Random)
Non -
Probability

Probability
(Random) sampling
Simple
Systematic
Stratified
Cluster
Non probability
snowballing
Convenient
others
1/28/2021 9

 Each student is called sampling unit (states or
provinces, districts, communities,
households, individuals).
 List of all Units (students at university of
Khartoum) is called sampling frame.
1/28/2021 11

University of Khartoum Faculty of Medicine
1. Faculty of Medicine
2. Faculty of pharmacy
3. Faculty of Engineering
4. ………………….
5. …………………….
6. …………………..
7. …………………..
1. Amna Khairy
2. Marwa Mohammed
3. Asma Mohammed
4. ………………….
5. …………………….
6. …………………..
7. …………………………
…………………
1/28/2021 12

Simple random sampling
 A survey on “ knowledge attitude
and practice of hand hygiene
among doctors at Ibrahim Malik
hospital in 2017 “
 we have calculated the sample size
and we need 200 doctor, how we
are going to get a random sample
?
1/28/2021 13

Choose 200 random numbers from 1-400
HOW?
1/28/2021 15
1. MOHAMMED AHMED
2. SOMAIA HASSAN
3. FATIMA OMER
4. ………….
5. ………………
6. …………………..
7. ……….
8. …………….
9. ………………..
10.……………..
11.…………………..
12.……………..
13.………………….
14.………………
………………..……..
400……..

Simple random sampling
among doctors at Ibrahim Malik
hospital in 2017 “
 we have calculated the sample size
and we need 200 doctor, how we
are going to get a random sample
?
1/28/2021 17

 Sampling frame :all the doctors in Ibrahim
malik Hospital.(ep: 400 doctor)
 Sampling interval :400/200 = 2.
 Select a random number within the interval.
Ep: 2.
 2, then followed by 4, 6, 8..
1/28/2021 19

Stratified random sampling
among health care provider
(nurses , doctors and midwife)at
Ibrahim Malik hospital
 What is the best way of sampling?
1/28/2021 20

When the parent population is subdivided into
strata and marked variations are existing
between the strata or subgroups
1/28/2021 21

 sample size =300, total # of health care
providers=800.
 Within each strata use simple random
sampling to get your sample.
1/28/2021 22
Doctors
400
#of doctors in
the sample=
400/800*300 =
150
Nurses
300
# of nurses in
the
sample=300/8
00*300=113
Midwifes
100
# of midwife in
the
37
sample=

 Cross section study on: The length of hospital
stay and related factors for patients at public
Hospitals in Khartoum state2017 .
 The researcher used Multistage cluster
sampling to get the sample.
1/28/2021 23

Cross sectional study on Prevalence of
epilepsy among school children in Khartoum
State, Sudan.
1/28/2021 24

Hospital based cross-sectional study : Factors
associated with diabetic retinopathy among
Sudanese diabetic patients Makkah Eye Complex
(MEC) 2016.
1/28/2021 27

Probability
(Random) sampling
Simple
Systematic
Stratified
Cluster
Non probability
snowballing
Convenient
others
1/28/2021 30

Snow balling (non probability sampling)
1/28/2021 32

Sample is made up of people who are easy to
reach.
1/28/2021 34

 Example:
Research on Satisfaction of Medical students
with a Research Methodology course.
The researcher can interview only students who
have attended all the lectures.
1/28/2021 36

Researcher relies on his or her own judgment
when choosing members of population to
participate in the study.
1/28/2021 37

 Budget
 Study design
 Margin of error
 Confidence level
1/28/2021 38

Budget and Logistics .
1/28/2021 39

Study Design
Observational
Descriptive
Cross
sectional
surveys
Ecological
Analytical
Case control
studies
Cohort
studies
Cross
sectional
(analytical)
Experimental
Randomized
control Trials
Non-
Randomized
control trials
Quasi design
1/28/2021 40
Things that affect sample size

In across sectional study on : Awareness of
Doctors about Zika virus among Doctors at
Khartoum state 2017 , 60% of doctors
scored Average awareness about zika
virus.5% margin of error, sample size was
400 .
1/28/2021 41

Degree of error due to inadequate sample
size , Which you can tolerate later in your
result !
1/28/2021 42

 A researcher found
that 60% of doctors
in sample at
Khartoum state , are
aware about zika
virus, with 5% margin
of error
 55% - 65% of
doctors at Khartoum
state are aware about
zika virus.
 A researcher found
that 60% of doctors
in sample at
Khartoum state , are
aware about zika
virus, with 10 %
margin of error
 50% - 70% of doctors
at Khartoum state are
aware about zika
virus.
1/28/2021 43

In across sectional study among pregnant
ladies attending Al –ban jaded antenatal
care clinic in 2017 , prevalence of anemia
was 40%.
 5% margin of error and 95% confidence level
, sample size was 400 with 5% margin of
error ).
1/28/2021 44

confidence level refers to the percentage of
all possible samples that can be expected to
include the true population parameter.
1/28/2021 45

if we repeated this survey 100 times on the
same population , using different samples
with same size , 95 of the surveys will result
in 35%-45% prevalence .
1/28/2021 46
Higher Confidence level =larger
sample size is needed .

1/28/2021 47
how many is enough?

Cross sectional study : Factors influencing
voluntary counseling and testing for HIV among
students at University of Khartoum in 2018 .
 Margin of error =5%
 confidence level=95%
 Total number of student at University of
Khartoum.
1/28/2021 48

Cross Sectional survey on :
prevalence of anemia among pregnant ladies
attending Al –ban jaded antenatal care clinic
in 2018.
1/28/2021 51

 n= sample size
 = (z=1.96).
 p= the frequency of occurrence of an event =
prevalence of anemia among pregnant ladies
attending antenatal care from previous
studies )
 q= 1-p , e = Margin of Error (0.05)
1/28/2021 52

 From the previous study : p= 0.62
 Apply that to the equation ?
 Sample size = 362
1/28/2021 54

 If no previous frequency found , assume
50/50 chance for the event to occur .
 For example : There is 50/50 chance that a
pregnant lady will have anemia
 In the previous equation P=50% i.e. 0.5.
calculated Sample size =384.
1/28/2021 56

 always Take in consideration that not all
participants will respond to you .
 Assume response rate of 80%
 Non response = 20% so increase your sample
size by 20% of the original (calculated ) one
,So sample size =480
1/28/2021 64

 Try to use a random method where possible
and remember that random does not mean
‘haphazard’ or ‘arbitrary’.
 Obtain independent advice on sample size
usually be from either a trained statistician or
from a researcher in your field who has
longstanding experience
1/28/2021 65

 https://wwwn.cdc.gov/epiinfo/user-
guide/StatCalc/introduction.html
 http://epitools.ausvet.com.au/content.php?p
age=SampleSize
 http://www.stats.gla.ac.uk/steps/glossary/ba
sic_definitions.html
1/28/2021 66

https://www.cdc.gov/epiinfo/support/downloa
ds.html
1/28/2021 67

https://www.cdc.gov/epiinfo/support/tutorials
.html
1/28/2021 68

 Feel free to contact me at:
 amnakh22@gmail.com
1/28/2021 69

Sample size and sampleing

More Related Content

What's hot (20)

Similar to Sample size and sampleing (20)

More from Amna Khairy (10)

Recently uploaded (20)

Sample size and sampleing

Editor's Notes