Module 2

Collection of Data

Introduction

Any statistical investigation must necessarily be based on accurate

data and in order to ensure the accuracy of data, one must know the right
sources and methods of collecting them.

Objectives:

At the end of this module, you should be able to:

1. Differentiate sample from population.

2. Compare the different methods of data collection.
3. Enumerate and differentiate the different types of probability
sampling methods.

Types of Data

There are 2 types of data – primary and secondary data. Primary

data refer to information which are gathered directly from an original
source or which are based on direct or first hand experience. Secondary
data on the other hand refer to information which are taken from
published or unpublished data which were previously gathered by other
individuals or groups.

Methods of Collecting Data

 11

There are several methods of collecting data and these are:

1. The Interview method. This is a method where there is a person-

to-person contact or exchange of information between the
interviewer and the interviewee. This method provides consistent
and more precise information since clarification may be given by the
interviewee. The questions may be repeated or modified such that
questions asked are understandable to the level of understanding of
the interviewee. However, this method is time-consuming, expensive
and has a limited field coverage.

2. The Questionnaire method. The researcher uses a prepared set

of questions and information are solicited from the respondents by
means of written responses based on the list of questions which are
relevant to the problems of the study. Questionnaires may be mailed
or hand-carried. This method is inexpensive and can cover a wide
area in a shorter period of time. However, there is a high possibility
that the respondents may not answer some of the questions or they
may not return the questionnaire, especially if it is mailed.

3. The Registration method. This method of gathering information

is enforced by certain laws such as registration of births, deaths,
motor vehicles, marriages and licenses. The advantage of this
method is that information is kept systematized and made available
to all because of the requirement of the law.

4. The Observation method. In this method the researcher observes

the behavior of persons or organizations and their outcomes. The
researcher can employ either the participant observation method
where the researcher is included in the group. Or he can employ the
non-participant observation method where the researcher is outside
of the group and not noticed by the subject.

5. The Experiment method. This method is used when the objective

is to determine the cause and effect relationship of certain
phenomena under controlled conditions. It is usually used in
laboratory experiments or field experiments.
 12

In the conduct of any research investigation, the researcher is not

concerned with collecting a great mass of data because sometimes it is
impossible or impractical to observe the entire set of observations (called
population), but rather to observe only a small fraction (called samples) of
the population.

Population and Sample

In Statistics, population refers to the set of all observations made on

all objects under study for a given characteristics of interest or variable.
The number of observations in a given population is referred to as the size
of the population and is designated as N. a population may be so large
that it may be impossible or impractical for the researcher to study all its
elements. In such a case, the study of a sample from the given population
would be more appropriate.

A sample is a subset of the population. It is any subgroup of

observations drawn from the population by some appropriate method so
that the characteristics of the population can be estimated. The number
of observations in a given sample is termed as sample size, and is
designated by n.

To compute for the sample size, different statistical formulas can be

used. One formula popularly used in computing for the sample size, n, is

N
n
1  Ne 2

Where:

N = the population size,

n = the sample size, and
e = the marginal error (1%, 5% or 10%)

Example: Given a population size, N, of 5,000, the sample size, n, is

3333 at 1% marginal error,

370 at 5% marginal error, and
 13

98 at 10% marginal error.

If our inferences from the sample to the population are to be valid,

then we must obtain samples that are representative of the population.
Samples which are not representative of the population will often result to
bias and to eliminate any possibility of a bias, then it is desirable that our
samples must be a random sample. Likewise, there other reasons why we
random samples instead of a population such as:

1. Reduced cost. 2. Greater speed.

3. Greater efficiency and accuracy.
4. Greater scope. 5. Convenience.
6. Necessity. 7. Ethics.

Sampling Methods

Sampling is the process of choosing a sample. As a rule, a

representative or sample should be obtained in such a manner that the
characteristics and variations of the population being sampled are
reflected. Basically, there are two broad classifications of sampling,
namely, probability sampling and non-probability sampling.

Two broad classifications of sampling:

1. Non-probability sampling
In non-probability sampling, the element of the population is taken
depending to a large extent on the personal feelings of the researcher or
purpose and without regard for some chance mechanism for choosing an
element. Some common types of non-probability sampling are:

a. The samples are selected haphazardly.

b. The samples are restricted to a part of the population that is readily
accessible.
c. The samples essentially consist of volunteers, in studies wherein the
measuring process is unpleasant to the person being measured.
d. The samples are arbitrarily selected because there is a “good”
evidence that it is very representative of the total population.
 14

2. Probability sampling
In probability sampling, every element belonging to the population
has a known and non-zero probability of being included in the sample.
Only probability sampling allows us to calculate sampling errors, and
therefore permits us to judge the goodness of our sample statistics.

Types of Probability Sampling:

1. Simple random sampling – is a sampling method in which all

members of the population have an equal chance of being included in the
sample.

2. Systematic sampling - Suppose the elements of a population of size

N are numbered from 1 to N in some order: consider taking a sample of
size n such that N/n = k. A systematic sample consists of an element
selected randomly from the first k elements and every k th subsequent
element.

3. Stratified random sampling - If the population of size N is

subdivided into non-overlapping subpopulations of sizes N1, N 2, …, NL
respectively, such that N1 + N2 + … + NL = N1, then samples of sizes n1,
n2, …, nL drawn independently from one subpopulation to another is a
stratified sample of size n = n1 + n2 + … + nL.

The subpopulations are called strata, and the procedure of grouping

is known as stratification.

Depending on the size of the sample taken from the strata, stratified
sampling can be categorized into:
a. Equal allocation - In stratified sampling, if equal-sized
samples are drawn from each stratum, we have equal
allocation, i.e. n1 = n2 = … = nL.
 15

b. Proportional allocation - In stratified sampling, when the

stratum sample sizes n1, n2, …, nL are proportional to the
respective stratum subpopulation sizes N1, N2, …, NL, then we
have proportional allocation, i.e.

n1 = N1 , n2 = N2 , …, nL = NL
n N n N n N

Stratified samples may be drawn in different manners.

a. If samples are drawn using simple random sampling in each stratum,
the procedure is called stratified random sampling.

b. If samples are drawn using systematic sampling in each stratum, the

procedure is called stratified systematic sampling.

4. Cluster sampling - A cluster sample is a sample of groups from our

sampling frame wherein information is obtained from all individual units
contained in the groups selected. Such groups are called clusters.

5. Multi-stage sampling - Sampling may also be done at several stages

or hierarchical classifications and this is referred to as multi-stage
sampling.
 16

Activity
1. Given the following population data on the number of schools
located in 30 municipalities:

Mun. No. 1 2 3 4 5 6 7 8 9 10
No. of Schools 5 4 6 7 8 5 6 7 10 8

Mun. No. 11 12 13 14 15 16 17 18 19 20
No. of schools 12 7 4 5 6 8 9 9 15 11

Mun. No. 21 22 23 24 25 26 27 28 29 30
No. of schools 3 5 8 5 2 4 9 16 9 6

From the above data, draw a random sample of size 10, using

a. simple random sampling

b. systematic sampling

2. A big University has a total of 375 faculty members. Of these,

110 are doctoral degree holders, 130 are masters degree
holders, and the rest are bachelor degree holders.

a. Determine the appropriate sample size. Use a 5%

marginal error.
b. Apply stratified sampling using proportional allocation to
determine number of individuals to be included in each
stratum.