RESEARCH METHODS

Prepared by:
Associate Prof. Dr. Roziah Mohd Rasdi
Dept. of Professional Development & Continuing Education
Faculty of Educational Studies
Universiti Putra Malaysia
[email protected]
 Topic 11

Sample Selection
 Introduction
Strength of quantitative research method – its ability to use
smaller group of people to make inferences about larger
group (Bartlett, Kotrik & Higgins, 2001).
To making generalization findings from sample
back to population

To do so – you need to pick the most accurate smaller group

to represent the larger group. This is called the Sampling.

Smaller group = SAMPLE (n)

Larger group = POPULATION (N)
 Definition
Population (or target population)
entire group of people, events or things of interest that
the researcher wishes to investigate
Element
a single member of the population

Sampling Frame
a listing of all the elements in the population from
which the sample is drawn
Sample
a subset of the population

Subject / Respondent
a single member of the sample
 Factors affecting the inferences drawn
from a sample
• The size of the sample – the larger the sample size, the
more accurate the findings.

• The extend of variation in the sampling population – the

greater the variation in the study population with respect to
the characteristics under study, for a given sample size, the
greater the uncertainty.
 Aims in selecting a sample

To achieve maximum precision in your

estimates within a given sample size, and
avoid bias in the selection of your
sample.

Population
Sample

Elements
 Types of Sampling

Probability Non-probability Mixed

Sampling sampling sampling

Simple Stratified Cluster Quota Systematic

random random sampling sampling
sampling sampling Purposive

Accidental
Proportionate
Single
stratified
stage Snowball
sampling
Double Expert
Disproportionate stage sampling
stratified
sampling Multi
stage Figure: Types of sampling in
quantitative research
 Probability Sampling
 The elements in the population have some known
chance or probability of being selected as sample
subjects.
 Allows to make inference from sample about
population (generalization).
 Example:

Sample = Consumer. Rate the prize of Novel Z.

75% said expensive.

Inference: 75% of all consumer feels the same.

 Use of inferential statistics – the significant values (p-
value) and confidence interval.
 Types of Probability Sampling Techniques
Simple Random Sampling

 Every element in the population has a known and equal chance

of being selected as a subject.
 Is the most representative of the population for most purposes.
 Easy to implement.

 Disadvantages are:
Most cumbersome and tedious
The entire listing of elements in population frequently
unavailable
Very expensive
Not the most efficient design
Time-consuming
Random Sampling using SPSS
Stratified Random Sampling

 Comprises sampling from populations segregated into a number of

mutually exclusive sub-populations or strata.
 E.g. University students divided into juniors, seniors, etc
Employees stratified into clerks, supervisors, managers, etc
 Homogeneity within stratum and heterogeneity between strata.
 Statistical efficiency greater in stratified samples.
 Sub-groups can be analysed.
 Different methods of analysis can be used for different sub-groups.
 Stratified Sampling
o Proportionate sampling
o Disproportionate sampling

Substrata/subsets University

Substrata/subsets Faculty

Substrata/subsets Students

Seniority
 PROPORTIONATE DISPROPORTIONATE

The number of each elements from

The number of each elements from
each stratum is selected without
each stratum is selected according to
consideration to the size of the
its proportion in the population.
stratum.

Example:
• Total population N = 500
Example:
(Male = 300, Female = 200)
• Total population N = 500
comprise of Male and Female.
• Proportion of male & female:
Male = 60%
• Sample size (n) = 350
Female = 40%
Select 50% male & 50% female
• Sample size (n) = 350. proportion of
Male (n) = 350 x 50% = 175
male & female in the sample:
Female (n) = 350 x 50% = 175
Male = 350 x 60% = 210
Female = 350 x 40% = 140
Advantages: Disadvantages:

 Control of sample size in  Increased error if

strata subgroups are selected
at different rates
 Increased statistical
efficiency  Expensive if strata on
population must be
 Provides data to created
represent and analyze
subgroups  High cost

 Enable use of different

methods in strata
Cluster Sampling

 Take clusters or chunks of elements for study.

E.g., sample all students in DCE5900 and DCE5131 to study
the characteristics of Management Science majors.

 Divide the population into discrete groups.

 The complete lists of the clusters will serve as the sampling

frame.

 Select a few cluster using Simple Random Sampling.

 Statistically it is less efficient than other probability sampling

procedures discussed so far.

 Area Sampling:
Cluster sampling confined to a particular area.
E.g., sampling residents of a particular locality, county, etc
Provide an unbiased estimate
of population parameters of
properly done. ADVANTAGES
Economically more efficient
than simple random
Lower cost per sample
Easy to do without list

Often lower statistical

efficiency due to subgroups
DISADVANTAGES being homogeneous rather
than heterogeneous.
Moderate cost.
 Stage Cluster Sampling

1. Choose the cluster grouping in your sampling frame.

2. Number each of the cluster with a unique number.

1, 2, 3, 4, 5, ………………….

3. Select the cluster using Simple Random Sampling.

4. Then randomly select certain proportion from the

cluster as your sample.
 THINGS TO CONSIDER IN SAMPLE
SELECTION USING PROBABILITY SAMPLING

 Identify correct population and sampling frame

( complete list of all cases in the population)

 Ensure representativeness of the sample

(avoid sampling error)

 Avoid sampling bias

 Adequate sample size

 High response rate

Total response rate = total number of responses – (ineligible + unreachable)

total number in sample
 Sample
representativeness
– not well represented

Sample
representativeness
– well represented
Estimating Sample Size
 Formula for calculating sample size depending on
type of statistical analysis

 Regression/Correlation analysis
(Tabachnik & Fidell, 2001)

• n ≥ 104 + m (independent variable) – Simple

linear regression

• n ≥ 50 + m (independent variable) – Multiple

linear regression
 Non-Probability Sampling

Used when the number of elements in a

population is either unknown/cannot be
identified.

Cannot make inference from the sample

about the population.

Most often used in qualitative studies.

 in some quantitative studies it may not be

possible to use probability sampling.
 Types of Non-Probability Sampling
Techniques

Quota Sampling

 The researcher non-randomly select subjects from

indentified strata until the planned number of subjects is
reached

 Quotas for numbers or proportion of people to be sampled

Examples:
1) survey for research on dual career families: 50%
working men and 50% working women surveyed.
2) Women in management survey: 70% women
surveyed and 30% men surveyed.
Accidental Sampling

 Similar as Quota sampling, but will stop collecting data when you reach
the required number of respondents you decide to have in your sample.
 Common among market research and newspaper reporters.
 If you are not guided by any obvious characteristics, some people
contacted may not have the required information.

Judgemental / purposive Sampling

 Involves the choice of subjects who are in the best position to provide the
information required.
 Researcher deliberately selects the subjects against one/more trait to be a
representative sample.
 Experts’ opinions could be sought.
E.g. Doctors surveyed for cancer causes.
Expert Sampling

 Your respondent must be known experts in the field

of interest to you.

 When use it in qualitative research:

- the number of people is dependent upon
the data saturation point.

 When use it in quantitative research:

- you decide on the number of experts to be
contacted without considering the
saturation point.
Snowball Sampling

 Used when elements in population have specific

characteristics or knowledge, but are very difficult to locate
and contact.

 Researcher identifies a small number of subjects who in turn

identifies others in the population.

 Initial sample group can be selected by probability or non-

probability methods, but new subjects are selected based on
information provided by initial subjects.
E.g. Used to locate members of different stakeholder
groups regarding their opinions of a new public works
project.
 Mixed Sampling
Systematic Sampling

 Every nth element in the population starting with a randomly

chosen element.
 Select sample at regular intervals from sampling frame.

Example:
To sample 35 households from a total of 260 houses - sample
every 7th house starting from a randomly chosen number from
1 to 10. If that random number is 7, sample 35 houses starting
with 7th house (14th house, 21st house, etc)

Possible problem is that there could be systematic bias. e.g.

every 7th house could be a corner house, with different
characteristics of both house and dwellers.
 Systematic
Disadvantages Sampling

Periodicity within
population may skew
sample and results.

Trends in list may bias

results.
Advantages
Moderate cost.
Simple to design.

Easier than simple

random.
Choice Points in Sampling Design
 Procedure
Step 1: Define the population
for Drawing
Step 2: Identify the sampling frame
– listing of all units in the
a Sample
population from which the sample
will be selected

Step 3: Select a sampling

procedure

Step 4: Determine the sample

units

Step 5: Select the sample units

Step 6: Collect data from the

sampled units