ALDERSGATE COLLEGE Instructional Module in Educational Statistics

Module 2: SAMPLING and DATA GATHERING TECHNIQUES

Prerequisite Skills:  Skills in performing basic operation

in mathematics
 Skills in understanding definitions
Instructors: Emerson Y. Castañeto

Overview

This module serves a guideline into the exploration of the basics of statistics particularly sampling and data gathering
techniques.

Objectives

At the end of this lesson, students are expected to:

1. determine the sample size ‘n’ from a given population “N”
2. explain the meaning of margin of error used in research
3. perform sampling from any target population
4. differentiate probability sampling from non-probability sampling
5. identify the appropriate data-gathering technique in a research

Learning Focus

What is Sampling?
Sampling is a statistical procedure that is concerned with the selection of the individual observation from a larger
population; it helps us to make statistical inferences about the population.

Why use samples?

Here are some reasons why researchers do not often use the entire population research.
a) We hardly ever know who makes up the entire population.
b) It is too costly in terms of human resources and other expenses.
c) It is time consuming and costly.
d) There is a lot of error to control and monitor.
e) Lists are rarely up to date.

Determining Sample Size

In doing a research, if the population is too big to handle, a substantial number of samples is acceptable.
Determining sample size is a very important issue because samples that are too large may waste time, resources and
money, while samples that are too small may lead to inaccurate results.
There are many approaches to determine the sample size. This includes using a census for small populations,
imitating the sample size of similar studies, using published tables and applying formulas to calculate a sample size.
 Using A Census For Small Population: One approach is to use the whole population as the sample. Although this
is not possible for large populations, a census is attractive for small populations, say, N≤200. A census eliminates sampling
error and provides data on all the members or elements in the population.

Prepared by Emerson Y. Castañeto Page 1

 ALDERSGATE COLLEGE Instructional Module in Educational Statistics

 Using Sample Size Of A Similar Study: Another approach is to use the sample size as those of studies similar to
the one you plan to work on. However, by using this method, you may run the risk of repeating errors that were made in
determining sample size for that study.
 Using Published Tables: A third way to determine sample size is to rely on published tables which provide the
sample size for a given set of criteria. You can visit this site: https://www.research-advisors.com/tools/SampleSize.htm
Gay (1976) offers some minimum acceptable sizes depending on the type of research as follows:
1. Descriptive research – 10% of the population. For smaller population, a minimum of 20% may be required.
2. Correlational research – 30 subjects
3. Ex-post facto research or causal comparative research – 15 subjects or groups
4. Experimental research – 15 subjects per group some authorities believe that 30 per group should be
considered minimum.
 Using Formulas To Determine Sample Size: The fourth approach to determine sample size is the use of formulas.
The simplest formula to use is the Slovin's formula.
𝑵
Slovin’s Formula: 𝒏=
𝟏+𝑵𝒆𝟐
Where: n is the sample size
N is the population size
e is the margin of error
To use the formula, first figure out what you want your error of tolerance to be or the margin of error. For
example, you may be happy with a confidence level of 95 % (giving a margin error of 5%), or you may require a
tighter accuracy of a 98% confidence level (a margin of error of 2%). The margin of error is the percentage which
defines the lower and upper bounds of the confidence interval. Hence we indicate how much error we live with and
try to ensure that our sample estimate doesn’t differ from the true population by more than this percentage. Plug
your population size and required margin of error [Do not forget to convert "e" to decimal, ex. 5% = 0.05] into the
formula. The result will be the number of samples you need to take.

Example 1: Find n if N = 10,000 and e = 5% Example 2: Find n if N = 10,000 and e = 1%

𝑁 10,000 𝑁 10,000
𝑛= = 𝑛= =
1+𝑁𝑒 2 1+(10,000)(0.05)2 1+𝑁𝑒 2 1+(10,000)(0.01)2
10,000 10,000
= =
1+25 1+1
n = 384.615 ≈ 385 n = 5000

In research, the more samples you have the better result you will get. Therefore, Slovin’s formula is just a guide
for obtaining the number of samples. You can get more than what is suggested by the formula but not below it.

Sampling Techniques
Another important is survey research is the type of sampling done. Since we already know how to compute for the
appropriate sample size, your next concern is how to select the sample from the population. This activity is referred to as
sampling.
Figure 1. Schematics of The Two Types Of Sampling Techniques
Types of Sampling Techniques

Non-Probability Sampling Probability Sampling

Convenience Quota Purposive Simple Systematic Stratified Cluster

Prepared by Emerson Y. Castañeto Page 2

 ALDERSGATE COLLEGE Instructional Module in Educational Statistics

Types of Sampling Techniques

A. Non-Probability Sampling: Each member of the population does not have a known chance of being included in the
sample. Instead, personal judgment plays a very important role in the selection.

Types of Non-Probability Sampling

1. Convenience Sampling – This type is used because of the convenience it offers to the researcher. It is used when
a researcher needs to conduct a study quickly or he/she is on a shoestring budget. It is also one of the only
methods you can use when you can’t get a list of all the members of a population.
For example, let’s say a researcher is conducting a survey for a company who wanted to know what
Shoemart employees think of their wages. It’s unlikely he/she will be able to get a list of employees, so
he/she may have to resort to standing outside of Shoemart and grabbing whichever employees come
out of the door (hence the name “grab sampling”).
2. Quota Sampling – This is very similar to the stratified random sampling. The only difference is that the selection
of the members of the samples in stratified is done randomly and has no formal rules or proportions. It is useful
when the time frame to conduct a survey is limited, the research budget is very tight, or survey accuracy is not the
priority.
For example, job interviewers with a limited time frame to hire specific types of individuals can use
quota sampling. For example, an interviewer who wants to hire people from particular schools can
isolate applicants from those schools into particular subgroups. Similarly, an interviewer who wants
racial or ethnic diversity in his hires can separate a huge group of applicants into groups based on a
person’s ethnicity or race.
3. Purposive sampling- Choosing the respondents on the basis of predetermined criteria set by the researcher. The
participants are selected based on the purpose of the sample, hence the name. Participants are selected
according to the needs of the study (hence the alternate name, deliberate sampling); applicants who do not meet
the profile are rejected.
For example, you may be conducting a study on why high school students choose community college
over university. You might canvas high school students and your first question would be “Are you
planning to attend college?” People who answer “No,” would be excluded from the study.
4. Snowball Sampling – It is also called as chain sampling. One element of the population is approached at a time
and then is asked to refer the investigator to the other elements of the population.
For example, a research purpose is to investigate what kind of personality billionaires possess.

B. Probability Sampling: Sampling are chosen in such a way that each member of the population has known though
not necessarily equal chance of being included in the sample.

Types of Probability Sampling

1. Simple Random Sampling – Samples are chosen at random with members of the population having a known or
sometimes equal probability or chance of being included in the samples.
a. Lottery: This needs a complete listing of the members of the population. You write the names or codes on
a piece of paper or cards, place them in a large container, and then randomly draw the desired number
of samples. The process is relatively easy for small population but relatively complicated and time-
consuming for large population.
b. Generation of random numbers/digits: This is a better and perhaps more efficient method for selecting a
simple random sample. Computers and •even your calculators can be used to generate random digits.
The randomly produced digits can be used to pick your samples. However, a complete listing of the
members of the population is needed in this type of random selection.
2. Systematic Sampling – Samples are randomly chosen following certain rules set by the researcher. This involves
𝑁
choosing the kth member of the population with 𝑘 = 𝑛 , but should be a random set.

Prepared by Emerson Y. Castañeto Page 3

 ALDERSGATE COLLEGE Instructional Module in Educational Statistics

3. Stratified Random Sampling –This method is used when the population is too big to handle, thus dividing N into
subgroup, called strata, is necessary. Sample per strata are then randomly selected, but considerations must be
given to the sizes of the random samples to be selected from the subgroups.
A process that can be used is proportional allocation. This procedure chooses sample sizes proportional
to the sizes of the different subgroups or strata.
Another process that could be uses is equal allocation. This procedure chooses sample sizes equally
from the different subgroups or strata.
Example:
Below is the number of students taking up Statistics from each college
department. A study will be conducted to determine the effectiveness of the modules
being used in the course.
Table 3. Number of Students Enrolled in Statistics
Course Number of
Students
Accountancy 110
HRM 89
Agriculture 57
IT 44
Total 300
1) At 5% margin of error, how many participants should be taken as samples? Use
Slovin’s Formula.
2) How many participants will be taken from each department using:
a. proportional allocation
b. equal allocation

Solution:
1) Use the Slovin’ s Formula to determine the number of samples.
𝑁 300
𝑛= 2
=
1+𝑁𝑒 1+(300)(0.05)2
300
=
1+0.75
= 171.43 ≈ 172
Therefore, a total of 172 students (or more) will be taken as samples to partake in the study.

2a) To determine the number of student-participants from each department considering a

proportional allocation, perform the following:

i. Compute the percentage of student from each department.

Course Number of Students Percentage (%)

Accountancy 110 (110 / 300 x 100) = 36.7
HRM 89 89 / 300 x 100) = 29.7
Agriculture 57 (57 / 300 x 100) = 19.0
IT 44 (44 / 300 x 100) = 14.7
Total 300

Prepared by Emerson Y. Castañeto Page 4

 ALDERSGATE COLLEGE Instructional Module in Educational Statistics

ii. Multiply the percentage with the number of samples (n).

Course Number of Students Percentage (%) Proportional Allocation
Accountancy 110 36.7% or 0.367 0.367 x 172 = 63
HRM 89 29.7% or 0.297 0.297 x 172 = 51
Agriculture 57 19.0 % or 0.19 0.19 x 172 = 33
IT 44 14.7% or 0.147 0.147 x 172 = 25
Total 300

Therefore, There shall be 63 students from the accountancy department, 51 students

from HRM department, 33 students from the Agriculture department and 25 students from IT
department to join the research. In this way, the selection of samples is proportional to the
number of students from each department.

2b ) To determine the number of student-participants from each department considering an equal

allocation, divide the number of sample (n) by the number of groups or strata, Hence, 172(samples) ÷
4 (departments) = 43.
Therefore, considering an equal allocation, 43 students in each department should
be tapped to participate in the study. In this way, there is equal distribution of samples from
each department.

4. Cluster Sampling. Cluster sampling is sometimes called area sampling because it is usually applied when the
population is large. In this technique, groups or clusters instead of individuals are randomly chosen. Recall that
in the simple random sampling, we select members of the samples individually. In cluster sampling, we will draw
the members of the sample by group or cluster and then we select a sample from each group or cluster
individually.
To illustrate its use, for instance, we want to determine the average daily expenses of families living
in Quezon City. Let us assume that there are 300 barangays in Quezon City. We can draw a random
sample of 30 barangays using simple random sampling and then a certain number of families from
each of the 30 barangays are chosen.

Or, Education department wants to inspect quality of education in schools of ABC City. Instead of
moving around the whole city locating the selected schools that are dispersed, through cluster
sampling, the researcher divides the city’s population into 21 towns, thus having 21 clusters, then
selecting 7 clusters using simple random sampling. Once the selection is finalized, the researcher
goes to the selected clusters and examines each and every element (school) of them.

Prepared by Emerson Y. Castañeto Page 5

 ALDERSGATE COLLEGE Instructional Module in Educational Statistics

Data Gathering Methods

Now that you have already known how to get the acceptable number of samples from the target population, your
next step is to focus on how to gather the information or data which you need from your samples or from your subjects or
research.
Following is a diagram of the four more popular data-gathering techniques and their advantages and
disadvantages
Figure 2. Data Gathering Techniques

DIRECT IN-DIRECT REGISTRATION EXPERIMENTAL

or INTERVIEW or QUESTIONAIRE METHOD METHOD
EXPERIMENTAL
VS
CONTROL

 Clarification can  Saves time, effort  Most reliable  Can go beyond

be done easily and money because it is plain description
 Easy to tabulate governed by law
 Time consuming
 Costly and time-  Large number of  Data are limited to  Lots of threats to
consuming samples can be what are listed in internal and
 Need lots of reached the documents external validity
effort  Problem on retrieval
 Limited to what are
included in the
questionnaire

Data Gathering Techniques

a) Direct or Interview: In this method, th e researcher has direct contact with the respondents. If the direct method
is used, the researcher can get more accurate answers since clarification can be made is the interviewee does
not understand the question.
Example: a researcher interviews respondents regarding their stand or view on a particular issue.

b) Indirect or Questionnaire/Survey: The researcher gives or distributes the questionnaire to the respondents
either by personal delivery or by mail.
These are some of the characteristics/features of a good questionnaire:
1. It should contain a short letter to the respondents which includes:
a. The purpose of the survey
b. An assurance of confidentiality
c. The name of the researcher or writer of the questionnaire
2. There is a descriptive title/name of the questionnaire
3. It is designed to achieve objectives.
4. The directions are clear.
5. It is designed for easy tabulation
6. It avoids use of double negatives
7. It also avoids double-barreled questions.
8. It phrases questions well for all respondents.

Prepared by Emerson Y. Castañeto Page 6

 ALDERSGATE COLLEGE Instructional Module in Educational Statistics

c) The Registration Method: This method of data gathering is governed by law.

Example: If a research wants to know the number of registered cars, s/he has to go to the Land Transportation
Office; the list of registered voters in the Philippines is found at the COMELEC.

d) The Experimental Method. This method of gathering data is used to find out cause and effect relationships.
Example: The researcher wants to know if Online Statistics will increase the academic performance of the
students. He/she has to do the following: Get two Statistics classes of equal intelligence. Give ordinary
classroom lecture to one group while enroll the other group online. At the end of the term, give the same
test to both groups, compare their scores and by the use of some statistical tools, find out if their
academic performances are significantly different.

Prepared by Emerson Y. Castañeto Page 7