Sampling Design: DR Rasidah Arshad Fep, Ukm

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 55

SAMPLING DESIGN

Dr Rasidah Arshad
FEP, UKM
SAMPLING TERMINOLOGIES
2

 Population
 The entire group of people, events, or things
of interest that the researcher wishes to
investigate.
 An element
A single member of the population.
 A sample
A subgroup of the population.
 Some members selected from the population.
SAMPLING TERMINOLOGIES
3

 A subject
 A single member of the sample, just as an
element is a single member of the
population.
 Sampling unit
 The elements that is available for selection
in some stage of the sampling process.
 Sampling Frame
 Complete list of all the elements in the
population, from which a sample is drawn.
SAMPLING (Persampelan)
4

Definition:
The process of selecting a sufficient
number of elements from the
population; so that results from
analyzing the sample are
generalizable to the population.
PARAMETERS OF POPULATION
5

 Parameters are the characteristics


of the population
(Parameter = ciri2 populasi)
 Examples:
µ Population mean (min populasi)
σ Population standard deviation
σ2 Population variance
SAMPLE STATISTICS
6

 The sample statistics is used as estimates of


the population parameters.
(Sampel statistik: Anggaran kepada
parameter populasi)
X Sample mean
S Standard deviation in the sample
S2 Variation in the sample
RESEARCH DESIGN
7
ADVANTAGES OF SAMPLING
8

 Less costs
 Less errors due to less fatigue
 Less time
 Destruction of elements avoided
 Gives results with known accuracy that can
be calculated mathematically
REASONS FOR SAMPLING
9

(1) Representative
If the sample is chosen in a scientific way, then the
sample statistic is reasonably close to the population
parameters
(2) Normality of distribution (taburan normal)
Characteristics of populations are generally normally
distributed.
(3) Central Limit Theorem
If the samples are sufficiently large and the sampling
design is right, then the sample mean X will be close to
the population mean µ
SAMPLING PROCESS
10

(1) Define the population (tentukan populasi)


 Elements, geographical boundaries, time:
 Example: UKM students, factory workers etc.
(2) Determine the sample frame (tentukan rangka
persampelan)
 Listing of elements in the population
 Example: list of students/ employees

(3) Determine the sampling design / technique


(Tentukan teknik persampelan)
 Probability sampling
 Non-probability sampling
SAMPLING PROCESS (Cont.)
11

4) Determine the appropriate sample size


(Tentukan saiz sampel yang sesuai)

5) Execute the sampling process


(Laksanakan proses persampelan)
ERRORS
 “Sampling error is the error that arises in a
data collection process as a result of
taking a sample from a population rather
than using the whole population.

 “Non-sampling error is the error that arises


in a data collection process as a result of
factors other than taking a sample.
“Sampling error is the error that arises in a data
collection process as a result of taking a sample from
a population rather than using the whole population
“Non-sampling error is the error that arises in a data
collection process as a result of factors other than
taking a sample
Respondent errors

 Respondents differ greatly in motivation to answer


correctly and in ability to do so.

 Obtaining an honest response to sensitive


questions is difficult.

 Basic errors
 Recall bias: simply does not remember
 Prestige bias: exaggerates to ‘look’ better
 Intentional deception: lying
 Incorrect measurement: does not understand the units or
definition
CENSUS vs. SAMPLE
SAMPLING PROCESS
12

Determine the
Define the Determine the
sampling
population sampling frame
design

Determine the
sample size

Execute
sampling
process
SAMPLING TECHNIQUES
13

1) Probability sampling
(Persampelan kebarangkalian)

2) Non-probability sampling
(Persampelan bukan kebarangkalian)
SAMPLING TECHNIQUES
14

Non-Probability
Probability Sampling
Sampling
1) Simple Random 1) Convenience Sampling
Sampling (Persampelan (Persampelan mudah)
rawak mudah) 2) Judgment Sampling
2) Systematic Sampling (Persampelan
(Persampelan pertimbangan/bertuju
sistematik) an)
3) Stratified Random 3) Quota Sampling
Sampling (Persampelan (Persampelan kuota)
rawak berstrata) 4) Snowball sampling
4) Cluster Sampling
(Persampelan kluster)
Simple random sampling

• Applicable when population is small, homogeneous &


readily available
• All subsets of the frame are given an equal probability.
Each element of the frame thus has an equal probability
of selection.
• It provides for greatest number of possible samples. This
is done by assigning a number to each unit in the
sampling frame.
• A table of random number or lottery system is used to
determine which units are to be selected.
SIMPLE RANDOM SAMPLING
(Persampelan Rawak Mudah)
17

 Procedure:
 Drawing elements from the population
(Membuat cabutan)
 Each element has a known and equal chance of
being selected.
(Setiap elemen mempunyai kebarangkalian yang
sama dan diketahui untuk dipilih)
Example: 1000 UKM students. We need 100
students. Draw out 100 names.

 Characteristics: highly generalisable, easily


understood, reliable population frame necessary.
 Also known as unrestricted probability sampling.
SIMPLE RANDOM SAMPLING
SYSTEMATIC SAMPLING
19
(Persampelan Sistematik)
 Procedure:
 Choosing every nth element in the
population for the sample.
 Must have sampling frame

 Example:
 Pilih setiap pelajar ke-10 daripada senarai
pelajar dalam kursus tertentu
 Iaitu pilih pelajar ke-10, ke-20, ke-30 etc..
SYSTEMATIC RANDOM SAMPLING
SYSTEMATIC SAMPLING:
20
EXAMPLE
 Example:
 Want to select a sample of 30 household
from 500 in an area.
 Choose every 7th house from random
numbers 1 to 7.
 If the random number picked is 7, then
choose houses number 7, 14, 21, 28, …..
until 30 houses are selected.
STRATIFIED SAMPLING
(Persampelan Berstrata)
21

 Involves the division of population into non-


overlapping subgroups.
 Stratum: Each of the subgroups
 Strata: Two or more subgroups.
 The strata are formed based on members'
shared attributes or characteristics.
 Each stratum is the same within

 Each stratum is different between


STRATIFIED RANDOM SAMPLING
STRATIFIED SAMPLING
22

 Procedure
 Divide population in strata
 Include all strata
 Pick samples using simple random sampling (SRS)
from each stratum.
 Example of strata: Job level in organisation
 Top management

Middle-level management

Lower-level management

Supervisors

Clerks

Faculties in UKM

FEP, FSKK, FST, FSTM, FPI etc
Stratified random sampling

Draw a sample from each stratum


TYPES OF STRATIFIED
23
SAMPLING
(1) Proportionate Sampling
 The number of sample subjects drawn from each
stratum is proportionate to the total number of
elements in the respective strata
(2) Disproportionate Sampling
 The number of sample subjects chosen from
various strata is not directly proportionate to the
total number of elements in the respective strata.
Decision due to some stratum are too large
Variability suspected within particular stratum
Easier, easier, less expensive to collect from one or more
strata than from others
STRATIFIED SAMPLING: EXAMPLE
24

 Example: Proportionate sampling


 10 top managers, 30 middle managers, 50 lower
level managers, 100 supervisors, 500 clerks, 20
secretaries. Need 140 sample (based on stats),
decides 20% each group
 2 top managers, 6 middle managers, 10 lower
level managers, 20 supervisors, 100 clerks, 4
secretaries = 142 samples
 Example: Disproportionate sampling
 7 top managers, 15 middle managers, q20 lower
level managers, 30 supervisors, 60 clerks, 10
secretaries = 142 samples
CLUSTER SAMPLING
(Persampelan Kluster)
25

 Involves the division of population into


already existing grouping
 Each subgroup is called a Cluster
 Natural grouping of the population
 Kluster: kumpulan elemen yang
mempunyai kepelbagaian dalam
kumpulan (intragroup heterogeneity) dan
keseragaman antara kumpulan
(intergroup homogeneity).

CLUSTER SAMPLING: EXAMPLES
26

 Examples of clusters:
 Blok pangsapuri
 City blocks/ buildings
 School districts
 Schools
 Classes in schools
 Departments in organizations
 Bandar dalam Selangor
 Geographical areas
CLUSTER SAMPLING
CLUSTER SAMPLING
27

 Procedure
(1) Divide population into clusters
(2) A sample of the cluster is selected
randomly from the population
(SRS)
(3) Collect data from all people within
the clusters that are sampled.
CLUSTER SAMPLING
28

Example (1)
Want to survey academic performance of

secondary school students in Selangor.


Divide the entire population (population of Selangor)

into different clusters (cities).


Selects a number of clusters through simple or

systematic random sampling.


Example (2)
Want to survey residents of a city

Take a sample of city block (clusters)

Select respondents (samples) within each city block.


STRATIFIED VS. CLUSTER
SAMPLING
29

STRATIFIED SAMPLING (SS) CLUSTER SAMPLING (CS)


Group same within: Group has no defining
homogeneous characteristics – are
representatives of the
population - heterogeneous

Groups are different from Groups are roughly the same


each other of each other
SRS taken from each group Census taken from each
group
Knows examples of strata Knows examples of cluster
Knows reason for stratified Knows reason for cluster
sampling sampling
CONVINIENCE SAMPLING
CONVENIENCE SAMPLING
43

 Collection of information from members of


the population who are conveniently
accessible to the researcher.
 Example: Pepsi challenge
 To determine whether people prefer Pepsi
over Coke
 Held at a shopping mall
 Sample: shoppers
CONVENIENCE SAMPLING…….
JUDGEMENT SAMPLING
45

 Choose the subjects who are in the best


position to provide the information required
 A type of purposive sampling
 Example:
A researcher wants to find out what is
needed to drive women to the top position of
organization
 Select sample among women who holds top
position in organizations
JUDGEMENTAL SAMPLING
QUOTA SAMPLING
47

 A pre-determined proportion of people


from different subgroup is sampled.
 A type of purposive sampling
 Example:
 To determine the difference in work attitude
between white-collar and blue-collar workers.
 60% blue-collar, 40% white-collar workers
 Number of subjects: 30
Blue-collar: 18
White-collar: 12
QUOTA SAMPLING
Snowball sampling
 Snowball sampling is a special
nonprobability method used when the
desired sample characteristic is rare.
 in studies which need to be carried out to
understand subjects which are difficult to trace

 It may be extremely difficult or cost


prohibitive to locate respondents in these
situations.

 This technique relies on referrals from initial


subjects to generate additional subjects.
Snowball sampling (cont’d)
 This sampling method is implemented in situations
where the topic is highly sensitive and not openly
discussed such as conducting surveys to gather
information about HIV Aids.

◼ Not many victims will readily respond to the questions


but researchers can contact people they might know
or volunteers associated with the cause to get in touch
with the victims and collect information

 It lowers search costs; however, it introduces bias


because the technique itself reduces the likelihood
that the sample will represent a good cross section
from the population.
Difference between Probability Sampling and Non-Probability Sampling
Methods
Probability Sampling Non-probability

Definition Probability Sampling is a sampling Non-probability sampling is a sampling


technique in which sample from a larger technique in which the researcher selects
population are chosen using a method samples based on the subjective
based on the theory of probability. judgment of the researcher rather than
random selection.

Alternatively Known as Random sampling method. Non-random sampling method.

Population selection The population is selected randomly. The population is selected arbitrarily.

Nature The research is conclusive in nature. The research is exploratory in nature.

Sample Since there is method to deciding the Since the sampling method is arbitrary, the
sample, the population demographics is population demographics representation
conclusively represented. is almost always skewed.

Time taken Take a longer time to conduct since the This type of sampling method is quick since
research design defines the selection neither the sample or selection criteria of
parameters before the market research the sample is undefined.
study begins.

result This type of sampling is entirely unbiased This type of sampling is entirely biased and
and hence the results are unbiased too hence the results are biased too rendering
and conclusive. the research speculative.
Response rate
 About 20 – 30% usually return a questionnaire

 Follow up techniques could bring it up to about


50%

 Still, response rates under 60 – 70% challenge the


integrity of the random sample

 How the survey is distributed can affect the


quality of sampling
SAMPLING : SUMMARY
SAMPLING
TECHNIQUE

PROBABILITY NON-PROBABILITY
SAMPLING SAMPLING

SIMPLE RANDOM CONVENIENCE


SAMPLING SAMPLING

SYSTEMATIC JUDGEMENT
SAMPLING SAMPLING
PROPORTIONATE Purposive
SAMPLING
sampling
STRATIFIED RANDOM
SAMPLING QUOTA SAMPLING
DISPROPORTIONATE
SAMPLING

SNOWBALL
CLUSTER SAMPLING SAMPLING
RESEARCH DESIGN
54
L4-Literature Review
L5-Theoretical Framework

L3-
Problem
Statement
THE RESEARCH PROCESS
35 Research design & analysis

Planning & Defining RESEARCH ✔ Reporting


DESIGN
L6
PROBLEM ✔ REPORT
STATEMENT WRITING
✔ L3 ✔ L13
INTRO SAMPLING
L1
LITERATURE ✔ L10

REVIEW
✔ L4
✔ Measurement 
SCI. MEASUREMENTS Scaling 
RESEARCH Reliability & validity 
L7
HYPOTHESIS ✔
L2
DEVELOPMENT
DATA ✔ Questionnaire 
L5 Interviews 
COLLECTION Observation 
L8-

RESEARCH ✔
9

FRAMEWORK
L5
DATA ANALYSIS L11-
12

You might also like