See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/374061078

Sampling in Educational Research

Presentation · September 2023

DOI: 10.13140/RG.2.2.14372.65927

CITATION READS

1 658

1 author:

Satish Pathak
Maharaja Sayajirao University of Baroda
5 PUBLICATIONS 1 CITATION

SEE PROFILE

All content following this page was uploaded by Satish Pathak on 21 September 2023.

The user has requested enhancement of the downloaded file.

Sampling in Educational Research

By:
Prof. Satish P. Pathak
Department of Education (CASE, IASE, IUCTE)
Faculty of Education and Psychology,
The Maharaja Sayajirao University of Baroda.
 Sampling : The basic concept
• Collection of appropriate evidences : an important stage
• Hypotheses are tested
• Conclusions are drawn
• Conclusions are expected to be applicable on population
• Sampling becomes a fundamental aspect in this context
 Need of Sampling
• Collection of all evidences / data is impossible
• Similarity among the elements reduces the necessity of
collecting all evidences (e.g. A study of primary school
teachers……)
• Sample is enough, if it is fully representative. (e.g. To study
the opinion of the M.S. Uni. teachers about online teaching.)
• Sample produces more accurate results compared to total
census (e.g. To study the family background of school going
children in a district.)
 Need of Sampling
• Intervention studies can be conducted effectively. (e.g.
Effectiveness of Blended learning strategy on higher
education students.)
• It provides better scope for in-depth investigation.
• Limited resources suggest for sampling (e.g. To study the
impact of feedback being provided by video on the
development of teaching skills among teacher trainees.)
• Sampling economizes time
• Sampling yields reasonable estimates of population values.
 The Process of Sampling
The process of selecting a fractional part representing the whole population.

❑ Delineating the universe [ Researcher portrays all possible events /

objects similar to the one under investigation.]
*Delineate: To mark the boundary of something
❑ Defining the population [ Researcher will define the portion of the
universe to which he / she has access and from which he/she will
draw the sample.]
❑ Defining the sample and deciding its size
❑ Deciding the sample units
❑ Preparing the sampling frame (listing the units of the population)
❑ Selecting the sample from the frame
Defining the sample and deciding its size
The fractional part of the population which contains all its characteristics and
thereby represents the whole population is called the sample.

• What is a good sample?

▪ Representative,
▪ Accuracy and adequacy (appropriate size)
• What shall be the size? It depends on;
▪ Homogeneity and Heterogeneity of population
▪ Size of population
▪ Nature of the study
▪ Level of precision / level of significance / acceptable level of sampling
error
 Deciding the sampling Units
It means identifying the constituent elements and deciding the sampling units

Example Sample Unit [ Sources of information ]

To assess the nature of Science instruction at One instructional session OR One or more
higher secondary level in Baroda district. instructional session/s for the content of one unit
To study the classroom climate of secondary School OR individual class [ A Unit is composed of
schools of Baroda city. elements, which are sub-units. e.g., Teachers,
Students, Physical conditions etc. …. will provide
data for the sample unit. ]
Studying the opinion of teachers and parents Parents and teachers
regarding the relevance of online education.

Major considerations: Objectives of the study, Nature of the Unit, Balance

between precision and cost, and Overlapping of Units
 Preparing the Sampling Frame
An accurate, complete and up-to-date list of all the units of
the population is called Sampling frame.

• Identify the source of units

• Tailor-Made list may be used
• Tailor-Made list must be examined
• Inaccurate and incomplete Tailor-Made list must be modified
 Selecting the Sample
Researcher must know the different Sampling techniques through
which sample can be selected scientifically.

Two broad categories of Sampling techniques:

➢ Probability Sampling

➢ Non-Probability Sampling
 Probability Sampling (PS)
If the sampling procedure is such that it can specify the probability of each unit
of the population to be included in the sample, then, it is called PS

▪ The chance of the unit to be selected is known.

▪ Random selection, as the fundamental element of control
▪ Main techniques under PS are:
1. (Simple) random sampling
2. Stratified random sampling
3. Systematic random sampling
4. Multi-stage sampling
5. Cluster sampling
 Simple Random Sampling (SRS)
It is the technique of selecting sample where each unit of
population has an equal chance of being selected.

• Based on the principle of randomness in selection of given units

• “Random” means “without purpose or by accident”
• More suitable for homogenous population
• Different methods used for the selection of sample:
➢ Use of cards as mechanical device
➢ Use of paper slips and container as mechanical device
➢ Use of Random number table
➢ Use of Rotating Machine
Specimen of Portion of Table of Random Digits
 Stratified Random Sampling (StRS)
The process of selecting a sample where total population is divided into certain strata on the
basis of the characteristics related to the phenomena under study and selecting a sample
randomly and independently from each stratum is called StRS.

• Identify the strata of interest and then draw a specified number of subjects
from each stratum.
• The basis for stratification may be geographical or it may involve characteristics of
population such as income, occupation, sex, age, year in college or teaching level
etc.
• Stratified sampling with equal allocation (for equal size of all strata)
• Stratified sampling with proportional allocation (for different sizes of strata)
• Stratified sampling with optimum allocation ( for the strata with different sizes and
variances)
• Stratified sampling guarantees representation of defined groups in the population.
 Systematic Random Sampling (SyRS)
When sample is drawn from fixed intervals, it is called SyRS.

• For example, if you want to study the reactions of the people who come to see
the Science fair regarding the educational value of the fair.
• Rough estimate: 1000 people will visit the fair (Size of the population : N = 1000)
• Sample to be selected is 10 % i.e. 100 (Size of the sample : n = 100)
• Interval K = N/n (1000/100 = 10th Unit will be selected by adding the size of the
interval)
• Let us assume the randomly selected no. is 7, then units would be 7, 17, 27, 37,….
• Sy. Random sampling is convenient, economical and more precise.
 Multi-stage Sampling (MS)
Particularly in conducting large-scale surveys, one often finds that population consists of more
than one set of units like; States, Districts, Blocks, Schools, Students etc..

Study Units and Sub-units Stage wise Sampling

To study the vocational ▪ List of schools ▪ Stage – 1 : Selection of schools
interests of IX grade students ▪ List of IX grade ▪ Stage – 2 : Selection of students
in Baroda. students
Selecting the sample at two different stages independently is called Two stage Sampling.
To study the vocational ▪ List of Blocks ▪ Stage – 1 : Selection of Blocks
interests of IX grade students ▪ List of schools ▪ Stage – 2 : Selection of schools
of district Baroda in Gujarat. ▪ List of IX grade ▪ Stage – 3 : Selection of students
students
Selecting the sample at three different stages independently is called Three stage Sampling.

▪ Stratification increases representativeness

▪ Multi-stage sampling is scientific and economical
 Cluster Sampling (CS)
• For example, to study the cluster of tribal in the forest areas in a stage where no
educational facilities exist so far.
➢ to list out all the clusters
[ externally homogeneous but internally heterogeneous groups ]
➢ to select the required no. of clusters randomly
• The division of population is found in groups / clusters
• The groups are naturally together and have common characteristics
• For example, to study the attitude of school teachers towards co-education in
Baroda city.
➢ Sample unit and sub-unit : Schools and Teachers (Male and Female)
➢ Schools with only Male / Female teachers will be clusters as well as strata
 Non-Probability Sampling (NPS)
The sampling procedures which do not have a provision to estimate the probability
of an element with which it has to be included in the sample are known as NPS.

▪ NPS techniques do not use random method.

▪ It is thought to be inferior to probability sampling.
▪ The guiding principles in non-probability sampling are :
➢ The availability of subjects
➢ The personal judgment / choice of the researcher
➢ Convenience in selection of sample
▪ Where goal is not to achieve generalization but to understand the
complexities of variables
 Non-Probability Sampling (NPS)
▪ Main techniques under NPS are:
1. Convenient sampling (Accidental sampling)
2. Quota sampling
3. Purposive sampling (Judgmental sampling)
4. Snowball sampling (Network sampling)
Accidental Purposive
 Convenient Sampling (CS)
The procedure of sampling where convenience is the major
deciding factor for drawing the sample is known as CS.

• Where researcher is unable to apply the principles of randomness

• For example:
• To study the impact of some psychological inputs on students’
achievement.
• To study the reactions of the people towards an educational film to which
they have been exposed.
• To study the effectiveness of non-formal education programme launched
in the rural areas of Gujarat.
• Researcher’s bias may influence sample
• Representativeness, Generalizability
 Quota Sampling (QS)
When the researcher decides to draw sample based on;
▪ Sample units from all strata
▪ S/he decides the size of the total sample and number of units from
each stratum.
▪ Ensure representation from all divisions/strata (according to
Convenient sampling technique)
▪ Steps of quota sampling;
➢ Choose the relevant classification of population
➢ Calculating the quota (equal / proportional) for each stratum
➢ Continuing invite cases till the quota is filled.
 Purposive Sampling (PS)
▪ It is also known as judgmental sample. The sample is
selected based on the judgement of researcher.
▪ Sample size is small.
▪ It is not to represent the population but gather pertinent
data from sample units that other units may not have.
 Types of Purposive Sampling
1) Maximum variation sampling; Wide variety is covered
2) Homogeneous sampling; (units are similar and data may
not vary)
3) Typical case sampling; Extreme cases, Few sample units
can give data
4) Extreme case sampling
5) Total population sampling
6) Expert choice sampling
Snowball Sampling (SS)

To conduct the study, the

researcher first selects few
sample that he thinks are a
part of the phenomenon.
These few give the
name/reference of the others
from whom data can be
collected.
 References:
• Kumar, S. (2002). Sampling in Educational Research. A CASE
Publication. Vadodara. Centre of Advanced Study in Education,
The M.S.University of Baroda.
• Krejcie, R.V. and Morgan, D.W. (1970). Determining Sample Size for
Research Activities. Educational and Psychological Measurements.
Retrived from
https://journals.sagepub.com/doi/10.1177/001316447003000308
 Thank You

View publication stats