Académique Documents
Professionnel Documents
Culture Documents
sanizah@tmsk.uitm.edu.my 1 sanizah@tmsk.uitm.edu.my 2
sanizah@tmsk.uitm.edu.my 3 sanizah@tmsk.uitm.edu.my 4
1
Types of Sampling Techniques
Why sampling? NON-
PROBABILITY
PROBABILITY
SAMPLING
Get information about large populations when it’s SAMPLING
impossible to study the whole population
Convenience Simple
Less costs random
sampling sampling
Less field time
Judgemental Systematic
Eliminate any BIAS sampling sampling
More accuracy i.e. Can Do a Better Job of Data
Collection Snowball Cluster
sampling sampling
Once a sampling frame has been established, you can
choose a SAMPLING TECHNIQUE Quota
Stratified
random
sampling
sampling
sanizah@tmsk.uitm.edu.my 5 6
2
Strengths and Weaknesses of
Non-Probability sampling Probability sampling
Technique Strength Weakness
Convenience Less expensive, • Selection bias, The items/individuals are selected randomly,
sampling less time, convenient • Not representative of the based on known probabilities
No need list of pop pop
Random means the item has an equal chance of being
Judgemental •Less expensive, less time, • Bias due to experts’ selected (unbias)
sampling convenient belief may make sample
unrepresentative Used when a researcher plans to make
inferences about the population
Quota sampling Sample can be controlled for • High bias because
certain characteristics sample units not Advantage
independent
• Time consuming The sample represent the population
Snowball Useful in reaching/locating • Selection bias maybe in
Disadvantage
sampling rare populations/characteristic researcher’s clasification
of subjects Sample selected not representative of the population
• Time consuming
sanizah@tmsk.uitm.edu.my 9 sanizah@tmsk.uitm.edu.my 10
sanizah@tmsk.uitm.edu.my 11 sanizah@tmsk.uitm.edu.my 12
3
Suppose you want Let say we get 2, 5, 8, and Table of random numbers
to select a sample of 4 10. Our sample would then
people look this:
from a group of 12.
204029273215632140
582032154785962024
362333254789120325
985263017424503686
sanizah@tmsk.uitm.edu.my 13 sanizah@tmsk.uitm.edu.my 14
4
SYSTEMATIC SAMPLING Systematic Random 1 26 51 76
2 27 52 77
3 28 53 78
Sampling N = 100
4
5
29
30
54
55
79
80
6 31 56 81
7 32 57 82
Suppose you want to select a sample
Want n = 20 8 33 58 83
9 34 59 84
of 4 people from a group of 12 10 35 60 85
5
Stratified Sampling
Cluster Sampling
The target population is first divided into
List of clients N
subpopulations or clusters.
Malays Chinese Others
Then a random sample of clusters is selected
based on a probability sampling technique
Strata N1 N2 N3 such as SRS.
For each selected cluster, all elements are
included in the sample.
n1 n2 n3
sanizah@tmsk.uitm.edu.my 21 sanizah@tmsk.uitm.edu.my 22
sanizah@tmsk.uitm.edu.my 23 sanizah@tmsk.uitm.edu.my 24
6
Multi-Stage Sampling
Designed to reduce time and cost when
working with samples from very large
populations.
Example:
Suppose we need a random sample of 2000
residents from the Malaysian population.
How to choose the sample using multi-stage
sampling?
sanizah@tmsk.uitm.edu.my 25 sanizah@tmsk.uitm.edu.my 26
Telephone the questioned less expensive than personal only short question can
Interview interview asked based on interview be asked
prepared speed of data collection restricted to respondent
questionnaire who have telephone
Telephone limited duration
Direct
Observation Mail questionnaire
cheapest
easiest
low response rate
simple question can be
Data collection Direct no interviewer influence asked
cover wide area
respondent has more time to
Questionnaire answer
Indirect Direct observation not influenced by others’ need a high skilled and
perception unbiased
Others (e-mail; not effected by the
video record) respondent itself
7
Designing a questionnaire
Before you begin drafting your questionnaire, it is important to consider: Questionnaire checklist
Who is the questionnaire for?
What is it intending to find out or measure?
i. Objectives of the study
Guidelines in Designing a questionnaire ii. Answers sought from the study
Design questions to meet the objective of the research.
Questionnaires should be short , simple and easy to understand. iii. Variables used in the study
Begin with simple and less controversial questions.
iv. Methods of data analysis
Avoid:
doubt, confusion, and vagueness.
bias questions. Once the above procedures are understood
sensitive questions.
double barrel question. by the researchers, a proper questionnaire
asking questions that are beyond the respondents' capabilities. can be designed.
questions that involve calculation.
REFERENCES
sanizah@tmsk.uitm.edu.my 31