population sampling rss6 2014
TRANSCRIPT
Population Sampling
Dr Fayssal M Farahat MD, MSc, PhD
Public Health Consultant
Infection Prevention and Control Department Associate Professor, Faculty of Medicine, Menoufia University, Egypt
Research Fellow, Oregon Health & Science University (OHSU), USA
Complete set of people
with a specified set
of characteristics SAMPLE
=
Subset
of
The
population
Clinical & demographic
Teenagers with asthma
Teenagers with asthma
living in Jeddah in 2013
Population
Study
subjects
Truth in the universe
Target
population
Findings in the study
Generalisability
Infer
Study
subjects
Truth in the universe
Target
population
Generalisability
design
Specify
clinical,
demographic,
& geographic
characteristics
Findings in the study
Specify accessible
population
and
approach
to select them
Study
subjects
Target
population
FIRST
Whether the sample differs from the population
Workers General
population
Study
subjects
Truth in the universe
Target
population
Findings in the study
SECOND
Validity of generalizing from study subjects to target population
Infer Association
bet HTN & CHD
in a sample
of Jeddah adults
Same
Association
Exists in
Saudi adults
IS IT .. Can Be .. • No sample is the exact mirror
image of the population.
• Select samples with acceptable
errors.
Representative Generalized
Inclusion criteria • Main characteristics of the target
population.
Clinical Demographic
Age, sex, Race Geographic
A 5-year trial of
calcium supplementation
for preventing osteoporosis
Demographic
Clinical
Geographic
Temporal
White females 50 – 60 ys
In good general health
Patients attending PHC Jeddah
Bet Jan 1 – Dec 31 of next year
Men Black female
HTN Paraplegia Metastatic lung dis
• Including alcoholics in the
osteoporosis study would expand
generalizability and allow to study
alcohol consumption as a cause of
demineralization.
• Exclude alcoholics to avoid a big
problem due to loss of follow-up.
• Exclusion in clinical trials is more specific
and may be mandated by ethical
considerations.
BE CAREFULL ..!
• EXCLUSION might threaten the validity of
generalizing the findings to the population.
Sampling ..
Sampling ..
Minimal cost
Maximum speed
Maximum accuracy
Impossible to examine the entire population
Terminology
• Sampling unit (element) – Subject under observation on which
information is collected • Example: children <5 years, hospital discharges,
health events…
• Sampling fraction – Ratio between sample size and population
size • Example: 100 out of 2000 (5%)
Terminology
• Sampling frame
– List of all the sampling units from which sample is drawn
• Lists: e.g. children < 5 years of age, households, health care units…
• Sampling technique – Method of selecting sampling units from
sampling frame • Randomly, convenience sample…
Samples Probabilty
Non Probabilty
Samples Probability
Non Probability
Probability
of selection
KNOWN
Probability
of selection
UNKNOWN
Non probability samples
Convenience
Quota
Snowball
Some elements of the population have no chance of selection “out of coverage”
ease of access
friend ….etc
Specific quota for
subgroup
Purposive Purpose
Convenience samples
Consecutive design
a practical approach for most clinical research projects
Entire accessible population over a long enough period
Avoid seasonal variations
Avoid changes over time
Generalisability
Gold Standard for
Probability Sampling
Each unit has specified
chance of selection
RANDOM
the choice of one subject will
not affect the chance of other
subjects being chosen
RANDOM
Generalisability
Sampling
RANDOM
Simple Random Sample عينة عشوائية بسيطة
Systematic Sample (متوالية)عينة عشوائية منتظمة
Stratified Random Sample عينة عشوائية طبقية
Cluster Sample عينة عشوائية عنقودية
Random Sample
Multistage sample عينة عشوائية متعددة المراحل
Simple Random Sample عينة عشوائية بسيطة
Ideal Bowel
Random number table
Computer-generated
Simple random sampling
57172 42088 70098 11333 26902 29959 43909 49607
33883 87680 28923 15659 09839 45817 89405 70743
77950 67344 10609 87119 15859 74577 42791 75889
11607 11596 01796 24498 17009 67119 00614 49529
56149 55678 38169 47228 49931 94303 67448 31286
80719 65101 77729 83949 83358 75230 56624 27549
93809 19505 82000 79068 45552 86776 48980 56684
40950 86216 48161 17646 24164 35513 94057 51834
12182 59744 65695 83710 41125 14291 74773 66391
13382 48076 73151 48724 35670 38453 63154 58116
38629 94576 48859 75654 17152 66516 78796 73099
60728 32063 12431 23898 23683 10853 04038 75246
01881 99056 46747 08846 01331 88163 74462 14551
23094 29831 95387 23917 07421 97869 88092 72201
15243 21100 48125 05243 16181 39641 36970 99522
53501 58431 68149 25405 23463 49168 02048 31522
07698 24181 01161 01527 17046 31460 91507 16050
22921 25930 79579 43488 13211 71120 91715 49881
68127 00501 37484 99278 28751 80855 02035 10910
55309 10713 36439 65660 72554 77021 46279 22705
92034 90892 69853 06175 61221 76825 18239 47687
50612 84077 41387 54107 09190 74305 68196 75634
81415 98504 32168 17822 49946 37545 47201 85224
38461 44528 30953 08633 08049 68698 08759 45611
07556 24587 88753 71626 64864 54986 38964 83534
60557 50031 75829 05622 30237 77795 41870 26300
Table of random numbers
1200
Students
100
EXCEL
Systematic Sample (متوالية)عينة عشوائية منتظمة
Random Sample
Select sample at regular
intervals based on sampling
fraction.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
46 47 48 49 50 51 52 53 54 55 ……..
• N = 1200, and n = 60
1200/60 = 20
• List persons from 1 to 1200
• Randomly select a number between 1 and 20 (ex : 8)
1st person selected = the 8th on the list
2nd person = 8 + 20 = the 28th
etc .....
Systematic Sample (متوالية)عينة عشوائية منتظمة
Systematic sampling
Stratified Random Sample عينة عشوائية طبقية
Random Sample
Stratified sampling
• When the sampling frame contains
clearly different categories (strata)
– Males and females
– Social classes
• What we do : – Classify population into internally
homogeneous subgroups (strata)
– Draw sample in each strata
Stratified sampling
Equal vs.
proportional
allocation.
Stratified sampling (example)
• N=12000 K=3
• N1=6000 N2=4000 N3=2000
• N=240
• n1= (240×6000)/12000 = 120
• n2= (240×4000)/12000 = 80
• n3= (240×2000)/12000 = 40
Cluster Sample عينة عشوائية عنقودية
Random Sample
–In selected clusters, all units or
proportion (sample) of units
included.
–All students in a classroom.
Cluster Sample عينة عشوائية عنقودية
Random Sample
Multistage sample عينة عشوائية متعددة المراحل
– 1rst stage : drawing regions
– 2nd stage : drawing city from each region.
– 3rd stage : drawing areas from each city.
– 4th stage: drawing houses from each area.
Multistage sample عينة عشوائية متعددة المراحل
Determine vaccination coverage in a country
Section 4
Section 5
Section 3
Section 2 Section 1
Simple Random Sample عينة عشوائية بسيطة
Systematic Sample (متوالية)عينة عشوائية منتظمة
Stratified Random Sample عينة عشوائية طبقية
Cluster Sample عينة عشوائية عنقودية
Random Sample
Multistage sample عينة عشوائية متعددة المراحل
• Random sample of the gallbladder surgery
patients.
• Reviewing hospital records of patients with
lung cancer from allover the country.
The use of random numbers is
generally preferable to using
systematic random.
Agree Dis-agree
The use of random numbers is
generally preferable to using
systematic random.
Agree
The regularity of selection can coincide by chance with some
unforeseen regularity in the presentation of the material for study –
Hospital appointments being made from patients
from certain practices on certain days of the week
2nd Advanced Course on Applied Medical Research and Biostatistics 22 – 24 March 2010 50
The Errors of Research
No study is free of errors
The goal is to maximize the validity
The best is to prevent errors from occurring (design & Implementation)
Errors can be addressed in the analysis
2nd Advanced Course on Applied Medical Research and Biostatistics 22 – 24 March 2010 51
Random Error
Wrong result due to chance
20% 18
19
21
22
28
12
Sample Size
precision
2nd Advanced Course on Applied Medical Research and Biostatistics 22 – 24 March 2010 52
Systematic Error
Wrong result due to BIAS
Sample (respondents) or
Measurement (unclear Q)
OR
Accuracy
Sample size
Response rate
= proportion of eligible persons who
agree to enter the study.
People difficult to reach.
People refused to enter.
…..
?
25%
• Acquire additional information on the non-
respondents.
or best
• Deal with non-response bias at the outset
Deal with non-response bias at the outset
• Series of repeated contacts (mail, telephone, home visit).
• Choosing a design that avoids invasive and uncomfortable tests.
• Using brochures and discussion to minimize the anxiety and discomfort.
• Providing incentives (reimbursing the costs of transportation and providing the results of the tests).
To anticipate ..
• Pre-test help to estimate the response rate
and how much to increase to get your
required sample.
• During the actual study, monitor the non-
response and find solutions to overcome
before continue to next sample.
Who will be included
Technique
of selection
How many ..?
Practical issues • Allow for drop-outs and non-consent
when planning sample size,
particularly when subjects are being
followed up for a long period of time.
• A pilot study may be necessary to
obtain suitable estimates.