MA in TESOL

Course: TESOL 2002

Lecturer: SAM Chanphirun

Chapter Review
Week 3: SAMPLING
Jack R. Fraenkel (2012) Chapter 6

Submitted by: Din Koldevinkea

1. What is a Sample?
Samples are a small part of individual members, all of whom/which represent a bigger body/ group
in a research study; or is the group on which information is obtained.
It is a group of people you selected from the population.
Sampling refers to the process of selecting the individuals or samples for a research study.

2. Samples and Population

Population is the larger group to which researchers hope to apply the results based on the study on
the samples; or is the group of interest to the researcher, the group to whom the researcher would
like to generalize the results of the study.
When it is possible, researchers would prefer to study the entire population of interest, but this is
very difficult.

Population census is one example of a research with the entire population being studied.

3. Targeted and Accessible Population

The Targeted Population is the actual population to which a researcher would really like to
generalize is rarely available, but is rarely available; it is the researchers’ ideal choice.
The Accessible Population is the population to which a researcher is able to generalize. Tt is the
researchers’ realistic choice.
The more narrowly researchers define the population, the more they save on time, effort, and
money, but the more they limit generalizability.
It is essential that researchers describe the population and the sample in sufficient detail so that
interested individuals can determine the applicability of the findings to their own situations.
The results of the finding will be based on the targeted population not just the accessible population.
What researchers want is to make sure that the samples can represent the entire population. How?
With the right method, the findings can be used to generalize on the entire population.
Errors in selecting the samples will result in problems with the quality of the research.

4. Random Sampling vs. Nonrandom Sampling

When using Random Sampling or Probability Sampling, the researchers choose the samples
randomly in the given or chosen population and all the sample presumably had an equal chance of
being selected; it is to select a Representative Sample and it can represent the entire population; if
not, you cannot make a generalization.
while in all the samples are chosen Nonrandomly or Non-probability Sampling; each member of
the population (the entire faculty of the university) does not have an equal chance of being selected;
some, in fact, have no chance. There are some factors or reasons that affect the selection.

Quantitative Research uses Random and Qualitative uses Non-random.

There are 3 types of Random Sampling

A Simple Random Sampling is when each and every member of the population has an equal and
independent chance of being selected. There are a few methods under this method.
Fishbowl Method: There are a lot of fish in the bowl. The fish are all the same. You want to
select some fish. When you take one out, you cannot choose or decide which fish to take because all
the fish are alike. Aka, Lucky Draw; the selection is equal and independent. In order to do this, you
need Sampling Frame which is the list of all the samples/ individuals. This is to avoid bias and to
prove that the individual selected can represent the entire population and the findings can be
generalized into the whole population.
Using Table of Random Number: the numbers are generated and placed randomly by
software or books. You need a sampling frame to do this.

Your Sampling Frame is 200.

First you need to decide to use the Row or the Column.
Say, we are going for the Column
Because our Sampling Frame is 200, it is 3 digits. So you can decide to use the first 3 digits 011, the
last 3 723 or the middle 3 172. It’s up to you. I will go with the first 3.
Then you run the list down consistently will get 011, 912, 086, 059,… you will stop when you get
the number of sample you want to study. If we want 20, then stop at the 20 th number. But, because
our total population is only 200, we need to cross out the 912 because it exceeds the total number.
You can also go with row.

011723
912334
086401
059397
666278
051965
063045
560132
727009
000037
667899
042397
987650
091126

Or we can use Excel. Say our Sampling Frame is 15. We need 5 random samples.
Formular: =randbetween(1,15). Then, a number appears. The example shows 8. Then we need to
drag downward until we have 5 cells. Then the numbers will determine the selected for the study.

We cannot increase our sample size after we begin our research or when we go to the research field.
In the excel, if we drag it down more cells, all the shown number will be changed. Reasons? First,
we select 5 from 15, and only 3 show up and you want to select 2 more to make it 5, it will mean
that the 2 will be selected from the 12 not the entire 15.

Stratified Random Sampling is a process in which certain subgroups, or strata, are selected for
the sample in the same proportion as they exist in the population; percentage. We want the people
from both or all the groups/strata equally; when you want your sample to have some sorts of certain
characteristic, you use Stratified Random Sampling.

For example; the total population of MA Students is 120 with 100 males (83.33%) and 20 females
(16.67%). We have 2 strata here, male and female.
Therefore, when we select samples, the 100% of sample size must come from the 83.33% from the
male and the 16.67% from the female exactly. Say we want to select 20 samples, then we need
16.66 or 17 males and 3.33 or 3 females.

Formula (Sample Size * Stratum / Total Population) 20 * 100 / 120 = 3.33 and 20 * 100 / 120 =
16.66

Proportionate Stratified Random Sampling

In-proportionate Stratified Random Sampling

The selection of groups, or clusters, of subjects rather than individuals is known as Cluster
Random Sampling. It is used when we cannot pick individual from the population.
It is often useful to combine cluster random sampling with individual random sampling. This is
accomplished by Two-stage Random Sampling.

5. Non-random Sampling Method

Systematic Sampling: In systematic sampling, every nth individual in the population list is selected
for inclusion in the sample.; sometimes it is a hybrid-both random and non-random.
You need to have sampling frame. Then you decide how many do you need. Say, you have 200
people and you need 20. You need to select the Interval= Total Population/Sample Size. 200/20=10
(Interval)
Therefore, you are going to select every 10th item in the sampling frame. There could be a problem
if the numbers of the items are arranged intentionally. Also, there is an argument about where to
start. If we start counting from number 1 to the n number, then number 1 will never be selected. In
order to avoid that, you put all the number within the interval, in our case 10, in a random manner,
so 1 may not be always the starting point. Say, we don’t have 1 2 3 4 5 6.. but we may have 6 5 7 1
8 9 … To guard against bias, researchers can use random start when the starting number is
random.

The sampling interval is the distance in the list between each of the individuals selected for the
sample. (Population size divided by desired sample size)
The sampling ratio is the proportion of individuals in the population that is selected for the sample.
(Sample size divided by Population size)
Periodicity is a danger in systematic sampling. f the population has been ordered systematically—
that is, if the arrangement of individuals on the list is in some sort of pattern that accidentally
coincides with the sampling interval—a markedly biased sample can result; therefore, researchers
should carefully examine the list to make sure there is no cyclical pattern present

6. Convenience Sampling
Many times, it is extremely difficult (sometimes even impossible) to select either a random or a
systematic nonrandom sample. At such times, a researcher may use convenience sampling which is
the selection of a group of individuals who conveniently are available for study; this is quite likely
be biased; cannot be considered representative of any population and should be avoided if possible.
You select samples that you can easily reach out to. It may not be reliable but it is convenience for
us, the researchers. Some researchers use the Convenience Sampling but they find other reasons to
justify their selection.

7. Purposive Sampling
On occasion, based on previous knowledge of a population and the specific purpose of the research,
investigators use personal judgment to select a sample.
Purposive sampling is different from convenience sampling in that researchers do not simply study
whoever is available but rather use their judgment to select a sample that they believe, based on
prior information, will provide the data they need.
It is a very powerful sampling method for Qualitative but is not used in Quantitative because it is
biased. The purpose is to enable researchers to deepen their understanding of the samples before
selecting, and it also enable them to get the samples who can provide rich data for the study.

8. Sample Size
Drawing conclusions about a population after studying a sample is never totally satisfactory, since
researchers can never be sure that their sample is perfectly representative of the population.
The question remains, therefore, as to what constitutes an adequate, or sufficient, size for a sample.
Unfortunately, there is no clear-cut answer to this question.
If it is 2-3% of the population, it is too small, but if it is 25%, it is too large.
The best answer is that a sample should be as large as the researcher can obtain with a reasonable
expenditure of time and energy.

9. External Validity: Generalizing from a Sample

The whole notion of science is built on the idea of generalizing. Most researchers wish to generalize
their findings to appropriate populations. Both the nature of the sample and the environmental
conditions—the setting—within which a study takes place must be considered in thinking about
generalizability. The extent to which the results of a study can be generalized determines the
external validity of the study. We are talking about the Generalization of the finding to the
population.

10. Population Generalizability

Population generalizability refers to the degree to which a sample represents the population of
interest.
Representativeness refers to the essential, or relevant, characteristics of a population; the fact that
samples can represent the population in Random Sampling.
The more generalizable the findings are, the better your study is.
The finding in Study A can be reflected in Study B. But researchers must take a look at the
resources/ characteristics/ conditions at both studies first before making the generalization.

11. When Random Sampling is not Feasible

As we have shown, sometimes it is not feasible or even possible to obtain a random sample.
Replication is when researcher repeats the study using different groups of subjects in different
situations.
If a study is repeated several times, using different groups of subjects and under different conditions
and if the results obtained are essentially the same in each case, the researcher may have additional
confidence about generalizing the findings.

12. Ecological Generalizability

Ecological generalizability refers to the degree to which the findings/results of a study can be
extended to other settings or conditions, or from one place to another place.
Researchers must make clear the nature of the environmental conditions—the setting— under
which a study takes place. These conditions must be the same in all important respects in any new
situation in which researchers wish to assert that their findings apply.