Yekatit 12 Hospital Medical College

School of Public Health

Department of Epidemiology and Biostatistics
Lecture 3: Sampling and sample size
For MPH Student
BY
Dube Jara
Assistant Professor of Epidemiology
Email: [email protected]

February,
2025
Addis Ababa,
1 Ethiopia
 Outline
 Introduction
 Sampling
 Types of sampling
 probability and non probability sampling
 Sample size determination for mean and
proportion

2
 Introduction
 The population is too large for us to consider
collecting information from all its members

 If the whole population is taken there is no need

of statistical inference

 A representative sample has all the important

characteristic of the population from which is it
drawn.

3
 Common terms used in
sampling
 Population:-any specified groups of persons, things
and measurements
 Target/source/references population
 A collection of items that have something in common
for which we wish to draw conclusion at a particular
time
 The whole groups of interest that we want to
generalize the results of the study
 Study or sample population
 The subset of target population that has at least
some chance of being sampled
 The specific population from which data are collected
 Sample
 The subset of the population, about which information
is actually obtained or
 A subset of population whose properties are to be
generalized to the larger population
4
 Common terms used in
sampling
 Sampling unit
 The unit of selection in sampling process

 Study unit:-the unit on which information is

collected, persons, housing unit etc.
 Note: the sampling unit is not necessarily the same as the
study unit

 Sampling frame:-the lists of all the units in the

references population from which a sample to be
picked.
 Sampling fraction (interval):-is the ratio of the
number of units in the sample to the number of
units in the references population (n/N).
5
 SAMPLING….
 Sampling is a process by which we study a small
part of a population to make inference about the
entire population.
 A main concern in sampling
 Ensure that the sample represents the
population, and
 The findings can be generalized

6
 The hierarchy of sampling

Study subjects
The actual participants in the study

Sample
Subjects who are selected

Sampling Frame
The list of potential subjects from which the sample is drawn

Study population
The Population from whom the study subjects would be obtained

Source population
The population to whom the results would be applied
7
 Advantages of sampling
 Feasibility: Sampling may be the only feasible
method of collecting information
 Reduced cost: Sampling reduces demands on
resource such as finance, personnel, and
material.
 Greater accuracy: Sampling may lead to better
accuracy of collecting data
 Sampling error: Precise allowance can be made
for sampling error
 Greater speed: Data can be collected and
summarized more quickly
 Time efficient

8
 Disadvantages of
sampling
 There is always a sampling error

 Sampling may create a feeling of

discrimination within the population

 May be inadvisable where every unit in the

population is legally required to have a record

9
 Key Points During Sampling

 Selecting the right people

 Selecting the right number of the right
people.
 Generally, the larger the sample size the
smaller the statistical error but the
greater the cost

10
 Steps in sampling
1. State the objectives
2. Define the target population
3. Define the data to be collected
4. Define the required precision and accuracy
5. Define the measurement instrument
6. Define sample frame, size and method then
select the sample

11
 Sampling methods
 There are two types of sampling

methods:

A. Probability Sampling methods

B. Non-probability methods

12
 Probability sampling

methods
A sample obtained in a way that every number of
the population has a known &non-zero

 Probability of being include in the sample i.e.

involves random selection of sample
 Involves the selection of a sample from a
population, based on chance

 Probability sampling is
 more complex,
 more time-consuming
 Usually more costly than non-probability
13
sampling.
 Probability sampling…
 There are several different ways in which a probability
sample can be selected.

 The method chosen depends on a number of factors,

such as
 The available sampling frame

 How spread out the population is

 How costly it is to survey members of the

population
 Involve random selection procedures

 Homogeneity of the characteristics to be studied

 Generalization is possible

14
 Advantages and disadvantages

Advantages
 Selection is based on the principle of randomization
or chance
 Reliable estimates can be produced
 Generalization can be made about the population

Disadvantages
 More complex, more time-consuming and usually
more costly than non probability sampling

15
 Classification
Probability Sampling

Simple Random
Sampling Stratified Multistage
Sampling sampling

Systematic Cluster
Homogenous
Frame Sampling Sampling Stratified
Cluster
Homogenous Homogenous Sampling
No Frame Wide area
Not
Not Homogenous
Homogenous Wide area

16
 Simple random
sampling
17  This is the most common and the simplest of the
sampling methods
 In this method, the subjects are chosen from the
population with equal probability of selection
 Assumption of the study population:
 Homogeneous population
 Availability of frame
 One may use a random number table;
 Lottery method
 Table of random number
 Computer generated random number
 SRS….
 Lottery methods
 used for small population
 each unit in the population is represented by a slip
of paper
 put in a box and mixed
 A sample of required size is drawn from the box.

 Table of random numbers

 It is a table of random numbers constructed by a
process that
 In any position in the table, each of the numbers 0
through 9 has a probability 1/10 of occurring.
 The occurrence of any number in one part of the
table is independent of the occurrence of any
number in any other part of the table. E.g. scientific
calculator
18
 SRS….
 Using table of random numbers
 begin at any randomly chosen column or row
 Proceed in any direction until the required numbers of
subjects are selected.
 Numbers not appearing in the sampling frame are ignored.
 Numbers reappearing are also ignored

 Advantage
 It is simple, easy to apply when the population is

small.
 Known and equal chance of selection

 Each unit in the sampling frame has an equal chance of

being selected.
 It does not require additional information on the frame (such

as geographic areas)
 The formulas are easy to use.

 Easy to apply to small populations.

19
 SRS…
 SRS has certain limitations:
 Requires a sampling frame.
 Difficult if the reference population is
dispersed.
 Minority subgroups of interest may not be
selected.
 Time consuming

20
 Example
 Age at first sex and associated factors for early
sexual initiation among female students at BHU,
Ethiopia
 There are a total of 6, 000 students
 We want to select 800 sample students
 In this case, we assumed homogeneity with respect to age at
first sex
 Their ID can be taken as frame
 Hence we can use computer generated random number to
select 800 students randomly

21
 Systematic sampling
 In systematic sampling individuals are chosen at
regular intervals from the sampling frame

 Ideally we randomly select a number to tell us where

to start selecting individuals from the list

 The skip interval can be determined by calculating k

 Individuals are taken at regular intervals down the list
 The starting point is chosen at random

22
 Systematic sampling
 Important if the reference population is arranged in
some order:
 Order of registration of patients
 Numerical number of house numbers
 Student’s registration books

Population size/sample size=N/n=k

 By randomly selecting a starting value between 1 and k

(a), we can pick every kth individual as the sampling

unite
 Thus, a, a+k, a+2k, a+3k+…… will be the final sample to be
studied.

23
 Steps in systematic random
sampling
1. Number the units on your frame from 1 to N (where N is the
total population size).
2. Determine the sampling interval (K) by dividing the number
of units in the population by the desired sample size.
K=N/n k=sampling interval
N=population size n=sample size
3. Draw a random number between one and K. This number is
called the random start and would be the first number
included in your sample.
 Let the selected number be j
4. Select every Kth unit after that first number j, j+k,
j+2k, j+3k---------------------j+nk

24
 Example
 A systematic sample is to be selected from 1200
students of a school. The sample size selected is
100. The sampling fraction is (skip interval)
k=1200/100=12
 The number of the first student to be included in
the sample is chosen randomly, for example by
blindly picking one out of twelve pieces of paper,
numbered 1- 12.
 If number 6 is picked, then every twelfth student
will be included in the sample, starting with
student number 6, until 100 students are
selected: then numbers selected would be 6, 18,
25
30, 42, etc.
 Merits/
advantages
 Less time consuming
 Easier to perform than simple random sampling.

 It provides a good approximation to SRS.

 Unlike SRS, systematic sampling can be

conducted without a sampling frame (useful in

some situations where a sampling frame is not
readily available).i.e. contraction of sampling
frame is not necessarily required
E.g. In patients attending a health centre, where it
is not possible to predict in advance who will be
attending
26
 Demerits
Demerits/disadvantages
 However, there is a risk of bias if there is cyclic

trend of the study population

 Should not be used when a cyclic pattern is

inherent in the sampling frame.

Example:
 For instance, if we want to select a random sample of
days on which to count clinic attendance, systematic
sampling with a sampling interval of 7 days would be
inappropriate, as all study days would fall on the same
day
 List of married couples arranged with men's names
alternatively with the women's names will result in a
sample of all men or women
Note: Systematic sampling is essentially considered to be
equivalent to SRS if there is no any cyclic pattern.
27
 Stratified random
sampling
 If it is important that the sample includes
representative groups of study units with specific
characteristics (for example, residents from urban
and rural areas, or different age groups), then the
sampling frame must be divided into groups or
strata, according to these characteristics.

 Stratified sampling is only possible when we know

what proportion of the study population belongs to
each group we are interested in

28
 Stratified random
 sampling
It is used when the population is known to have
heterogeneity with regard to some factors and those
factors are used for stratification
 The population is 1st divided into homogeneous,
mutually exclusive groups called strata, and
 A population can be stratified by any variable that is
available for all units prior to sampling (e.g., age,
sex, province of residence, income, etc.).
 A separate sample is taken independently from each
stratum, by simple sampling or systematic
sampling

 Stratifying random sampling basically assumes

 homogeneity within strata
 heterogeneity b/n strata

29
 Stratified sampling
…
 The procedures are:
 Divided the total population in to different
homogeneous subgroups (strata)
 Allocate sample for each strata (n i)
 Proportional (ni =Ni(n/N))
 Where ni =sample for each strata
Ni=total population of each strata
n=required sample size
N=total population of the
 Disproportional (equal allocation)

30
 Example
 A survey is conducted on household water supply in
a district comprising 20,000 households, of which
20% are urban and 80% rural

 It is suspected that in urban areas the access to

safe water sources is much more satisfactory. The
total population of the district is 10, 000
(urban=4000 and rural=6000). The sample size
required has been decided to be 300

 Allocate the sample proportionally for both strata?

nurban= 4000*300/10,000=120
nrural= 6000*300/10,000=180
31
 Stratified sampling…
 Advantages
 Less sampling variable than other sampling
techniques.
 Representativeness of the sampling is
improved.
 Disadvantages
 Sampling frame is needed for each strata.

32
 Cluster sampling
 In many administrative surveys, studies are done on
large populations which may be geographically quite
dispersed
 Using simple random sample method will require large
costs and will be inconvenient
 Sometimes it is too expensive to carry out SRS
 Population may be large and scattered
 Complete list of the study population unavailable
 Travel costs can become expensive if interviewers
have to survey people from one end of the
country to the other
 Cluster sampling is the most widely used to reduce the
cost
33
 Cluster sampling…
 Clusters appropriate in such conditions and
random samples of clusters will be included in the
study

 Then every member of the cluster will also be part

of the study

 The clusters should be homogeneous, unlike

stratified sampling where the strata are
heterogeneous

34
 Steps in cluster
sampling
 The reference population (homogeneous) is divided
into clusters.
 These clusters are often geographic units (e.g. districts,
villages, etc.).
 A number of clusters are selected randomly to
represent the total population, and then all units
within selected clusters are included in the sample.
 No units from non-selected clusters are included in
the sample—they are represented by those from
selected clusters
 This differs from stratified sampling, where some units are
selected from each group
 All the units in the selected clusters are studied

35
 Example
 In a study of knowledge, attitudes, and practices
related to family planning in rural communities of
a region, a list is made of all the villages.

 Using this list, a random sample of villages is

chosen and all the adults in the selected villages
are interviewed

36
 Merit and demerit
 Merit/advantages-
 A list of all the individual study units in the
reference population is not required.
 It is sufficient to have a list of clusters.
 Demerit/disadvantage -
 Sampling error is usually higher than for a
simple random sample of the same size.
 modified cluster sampling
 divide into sub cluster
 and then take equal number from each
 WHO-EPI coverage evaluation technique 3-7 subject
from each

37
 Multi-stage sampling
 Many studies, especially large nationwide surveys,
will incorporate different sampling methods for
different groups, and may be done in several
stages

 In experiments, or common epidemiological studies

such as case-control or cohort studies, this is not a
common practice

38
 Multi-stage

sampling…
Is appropriate when the reference population is large
and widely scattered.
 Similar to the cluster sampling, except that it involves

picking a sample from within each chosen cluster, rather

than including all units in the cluster.
 This type of sampling requires at least two stages.

The primary sampling unit (PSU)

 Is the sampling unit in the first sampling stage (usually

large size).
The secondary sampling unit (SSU)
 Is the sampling unit in the second sampling stage, etc.

 E.g. the PSU could be kebele and the SSUs could be

house hold.
39
 Merit and demerit
 Merit/advantage
 Cuts the cost of preparing sampling frame
 save a great amount of time and effort
 Demerit/disadvantage
 Sampling error is increased compared with a
simple random sample
 Gives less precise estimates than simple
random sampling for the same sample size

40
 Example
 In a study of utilization of pit latrines in a district,
150 homesteads are to be visited for interviews
with family members as well as for observations on
types and cleanliness of latrines.
 The district is composed of six wards and each
ward has between six and nine villages.
 The following four stage sampling procedure could
be performed:
 Select three wards out of the six by simple
random sampling
 For each ward, select five villages by simple
random sampling (15 villages in total)
41
 Example
 For each village select ten households. Because
simply choosing households in the center of the
village would produce a biased sample, the
following systematic sampling procedure is
proposed:
 Go to the center of the village
 Choose a direction in random way
 Walk in the chosen direction and select every third or every
fifth household (depending on the size of the village) until
you have the ten you need.
 Decide beforehand whom to interview (for example the head
of the household, if present, or the oldest adult who lives there
and who is available.)

42
 Non Probability sampling
techniques
 A type of sampling where each study unit in the
population has an unknown probability of
inclusion in the sample

 There is an assumption that there is an even

distribution of a characteristic of interest within
the population.

 The selection of subjects is arbitrary or subjective

43
 When to use Non probability
Sampling
 Group that represents the target population
already exists.

 Difficult or impossible to obtain the list of names

for sampling (Homeless, IV Drug user).

 For rare population.

44
 Non probability
sampling…
Advantages
 Used when a sampling frame does not exist.

 They are quick, inexpensive and convenient.

 Good for pretests, pilot studies, In-depth interviews.

 Used when precise representativeness is not

necessary.
Disadvantage
 No random selection (unrepresentative).

 Reliability cannot be measured.

 No way to measure the precision of the resulting

sample.
 Inappropriate for generalizing findings.

45
 Classification
Non-probability
Sampling

Convenience Judgment Snowball

Quota

Sampling based Investigator Used in studies

For convenience on some interviews as involving respondents
sake the study judgment, gut- many people in who are rare to find.
units that feelings or each category of Start by identifying
happen to be experience of the study unit as he someone who met the
available at the researcher. If can find until he criteria. Then let
time of data inference drawing has filled his him/her bring the
collection are is not necessary, quota other he/she knows
selected in the these samples
46
sample are quite useful.
 Judgmental

/Purposive
The researchers choose the sample based on who
they think would be appropriate for the study.
 Selecting sample on the basis of knowledge of the
research problem to allow selection of "typical"
persons for inclusion in the sample.
 Primarily used when there is a limited number of
people that have expertise in the area being
researched.

 Appropriate when the study subjects are difficult to

locate
 More efficient and economic where the sample
sizes are small.
 Used where randomization is not expected to
provide representative samples
47
 Quota
 The population is first segmented in to mutually
exclusive sub-groups as in stratified sampling.

 Select subjects until a specific number of

units/quota/ for various sub-groups has been
filled.
 It is really a means for satisfying sample size
objectives for certain sub-populations.

 No rules for selecting the subjects.

 This is one of the most common forms of non-

probability sampling.

48
 Quota sampling…
 Advantages
 Is generally less expensive than random sampling.
 Easy to administer, especially considering the tasks
of listing the whole population, randomly selecting
the sample and following-up on non-respondents can
be omitted from the procedure.
 Is an effective sampling method when information is
urgently required and can be conducted without
sampling frames.
 Disadvantages
 It does not meet the basic requirement of
randomness.
 Some units may have no chance of selection or the
chance of selection may be unknown
 Therefore, the sample may be biased

49
 Convenience/Haphazard
 Sometimes referred to as haphazard or accidental
sampling
 Selection of subjects based on easy availability &
accessibility
 Examples :People who just happen walking
 Is a method in which for convenience sake the
study units that happen to be available at the time
of data collection are selected
 Often used in face to face interviews

50
 Convenience or haphazard
sampling…
Advantage
 Easy to use
 Can delivery accurate results when the population
is homogeneous.
Disadvantage
 Not normally representative of the target
population
 E.g. scientists could use this method to determine
whether a lake is populated or not.
 Assuming that the lake water is well-mixed, any
sample would yield similar information
 A scientist could safely draw water anywhere on
the lake without bothering about whether or not
the sample is representative
51
 Snowball
 Involves a process of “chain referrals”
 Thus the sample group appears to grow like a
rolling snowball
 Suitable for locating key informants
 You start with one or two key informants
and ask them if they know persons who
know a lot about your topic of interest
 Used when trying to interview hard to
reach groups

52
 Snowball sampling…
 Often used in hidden populations which are difficult
for researchers to access;
 example populations would be drug users or commercial
sex workers.

 Because sample members are not selected from a

sampling frame, snowball samples are subject to
numerous biases.

 For example, people who have many friends are

more likely to be recruited into the sample.

53
 Exercise 1: Group
 Choose appropriate sampling technique for the
following proposed research topics
1. Prevalence of Depression and Associated Factors Among
Urban Civil Servants, in AA city, Ethiopia

2. Prevalence of Aggressive Behavior And Associated Factors

Among Patients With Schizophrenia Attending At Amanuel
mental specialized hospital, Addis Ababa, Ethiopia

3. Place of delivery among women attained antenatal care

service in Bule hora rural district, West Guji, Ethiopia
4. Challenges of homosexuality in Ethiopia

54
 ERROR IN SAMPLING

Sampling error(random error)

 Error that arising from the sample process itself.

 The difference b/n value (a statistic) computed

from a sample and corresponding value

(parameter) computed from a population.
Sampling error=x -̅μ
 Sampling error is minimized by increasing the size

of the sample: when n=N → sampling error =0.

55
 ERROR IN SAMPLING….
 Non-sampling error(systematic error)
 A bias to which the investigator may not aware
that it distorts the results.
 It is a type of systematic error in the design or
conduct of a sampling procedure which results
in distortion of the sample size
 We can eliminate or reduce the non sampling
error by careful design of the sampling
procedures and by increasing the sample size.
 There are several possible sources of bias in
sampling.

56
 Sample Size Determination…
Introduction
 Among the questions that a researcher should
ask when planning a survey is that "How large a
sample do I need?“
 The answer will depend on the aims, nature and
scope of the study and on the expected result.
 In planning a sample survey, the decision on the
size of sample is very important.

57
 Introduction…
 In general, sample size depends on:
 The objective of the study
 The study design
 Degree of confidence with which to conclude.
 The type of data analysis to be performed
 The desired precision of the estimates one wishes
to achieve
 The kind and number of comparisons that will be
made
 The number of variables that have to be examined
simultaneously
 How heterogeneous the population is

58
 Sample size-Qualitative
studies
 There are no fixed rules for sample size in
qualitative research.
 The size of the sample depends on
 What you try to find out
 From what different informants or

perspectives you try to find that out.

59
 Sample size-Quantitative
studies
 Calculations made
 The bigger the sample ,the better the study
becomes to a certain point
 The desirable sample size depends on the
expected variation in the data (of the most
important variables):
 The more varied the data are, the larger the
sample size we would need to attain the desired
level of accuracy

60
 We can use either of the following methods
 Using a census for small populations
 calculate the sample size
1. Using a computer package
2. Using formulas to calculate a Sample Size

61
 Using A Census For Small Populations

 Use the entire population as the sample

 Attractive for small populations ( 200 or less)
 Eliminates sampling error and provides data on all the
individuals in the population.
 Financial considerations make this impossible for large
populations

62
 Using formulas to calculate a Sample Size

 Three criteria usually will need to be specified

to determine the appropriate sample size:

1. The level of precision:

2. The level of confidence or risk, and

3. The degree of variability in the

attributes being measured

63
 1. Rules of thumb
1. For smaller samples (N ‹ 100), there is little point
in sampling. Survey the entire population.
2. If the population size is around 500 (give or take
100), 50% should be sampled.
3. If the population size is around 1500, 20% should
be sampled.
4. Beyond a certain point (N = 5000), the population
size is almost irrelevant and a sample size of 400
may be adequate.
5. Statistician – maximalist – at least 500

64
2. Confidence interval
approach
 Given confidence interval

mean ( proportion ) z  s.e

2
 Hence the absolute precision denoted by
d is given as
d  z  s.e
2

 Where s.e is the standard error of the

estimator of the parameter of interest.
65
 Sample size calculation…

 Calculate the sample size

1. Using a computer package (recommended)
2. Using formulas to calculate a Sample Size

 Using dependents variable(s)

 Using associated factors
 Select the optimum sample of those calculated

66
For prevalence study

Estimating a single population mean and

proportion

67
 Example:-
 A population of cancer patients has survival
standard deviation of 43.4 months. If one wants to
conduct a study on these populations how large
sample size is needed, so that 95% of the sample
mean of this size will be within ±6 months of the
population mean. Population size is 480 patients.
 Solution: - δ=43.4 month, CI=95% d= ±6 months, α/2=0.025
(z=1.96), N=480 patients.
n= (Z α/2)2 δ2/d2 = (1.96)2x (43.4)2/62
=200

68
 Example:-
 In a survey of school children to determine the
population of immunized children against polio, an
investigator determined the maximum discrepancy b/n
sample and population proportion of immunized to be
0.04, at level of confidence of 99%.further the
investigator had a previous knowledge on the
prevalence among children in a similar community to be
90% and the total population of school children is 800.
 Solution:-d=0.04, α=0.01, α/2 =0.005(2.58), p=0.9, q=0.1,
N=800, n=?
 n= (Z α/2)2 p.q/d2 = (2.58)2x0.9x0.1/ (0.04)2 =374.4 = 374
 ni/N=374/800=4.7% which is <5% there fore the initial
sample size is sufficient.

69
 3. Hypothesis testing
approach
Comparison of two PROPORTIONS
 The number of samples required in each group to

compare the proportion is given by:

 n = (p1q1 + p2q2) (f(,)) / ((p1 - p2)²
  = type I error (level of significance)
  = type II error ( 1- = power of the study)
 power = the probability of getting a significant
result
 f (,) =10.5, when the power = 90% and the level
of significance = 5%

70
 Hypothesis…
 Sample sizes to estimate the difference
between two populations mean (assuming
that the two populations have common
standard deviations).

72
 Cont,…
Where n1 = the size of sample one,

r = the ratio of sample 2 to sample 1

** when the two population are of equal, the

size of each sample is twice given by a
single population

73
Sample-case control, cohort
and cross sectional studies
 Sample size - for test of significant difference
between two proportions, the following formula can
be used:

n
Z  Z    p 1  p  p 1  p 
2
2
1 1 2 2

p1  p2 
2

Parameters:
n - size of sample in each group
P1 ,P2–estimated population prevalence in the comparison
groups
β = 1- Power (the probability that if the two proportions differ
the test will produce a significant difference)
 Usually a power of 80% is used
74
Sample-case control, cohort and
cross sectional studies…
To test a hypothesis about the difference between
two population proportion
For Cohort, case control and cross sectional studies.

75
 Sample size using statistical
software
 As an alternative method, we can use EPI INFO
statistical software to calculate the sample size
required for the study.

76
 Start page
 Step 1: Go to the main menu at the top left
corner and you will see the following window

77
 ‘statCalc’
 Step 2: Go to ‘statCalc’

78
 Sample size
Population survey
 For single population, the information needed are
 Size of the study population (N)
 Proportion of the variable of interest from previous study
(p)
 Worst acceptable result (d)
 Design effect

79
 Sample size
 Let us assume the population that we want to
conduct the study has target population of size
N=120,000
 The proportion of the variable of interest is not
known which means there is no previous study
done and hence we decided to use 50 percent as
an estimate of the prevalence for that variable
 Finally different level of confidence with the
corresponding sample size will be calculated and
displayed.
 The choice depends on
 The level of confidence you fixed

80
Sample size

81
 Sample size
 Sample size for two populations: unmatched case
control study

82
Sample size
 Sample size for two populations: cohort or cross
sectional

83
 n k
h a
T

ou
y 84