MHM-2

RESEARCH METHODOLOGY

CHAPTER-3
Sample Design:
● Censure and sample Survey

● Implication of Sample design

● Steps in sampling design

● Criteria for selecting a sampling procedure

● Characteristics of a good sample design.

● Different types of Sample design.

● Measurement Scales.

● Important scaling Techniques

Sample Design:
Sample design refers to the plan or strategy used to select a subset of individuals or items from
a larger population for inclusion in a research study. It is a crucial aspect of research
methodology, as the quality and validity of study results often depend on the appropriateness of
the sample design.
Key components of sample design include:
1. Population: The entire group of individuals or items that the researcher is interested in
studying. It's important to clearly define the population to ensure that the sample
represents it accurately.
2. Sampling Frame: A list or source from which the sample will be drawn. This could be a
list of households, students, companies, etc. It should ideally cover the entire population
of interest and be up-to-date and accessible.
3. Sampling Method: The technique used to select individuals or items from the sampling
frame to be included in the study. Common sampling methods include simple random
sampling, stratified sampling, cluster sampling, systematic sampling, and convenience
sampling.
4. Sample Size: The number of individuals or items included in the sample. Determining an
appropriate sample size is essential for achieving statistical power and precision in the
study results.
5. Sampling Procedure: The specific steps followed to implement the sampling method
and select the sample. This includes any protocols or guidelines for randomization,
stratification, clustering, etc.
6. Sampling Bias: Any systematic error introduced into the sample selection process that
results in a non-representative sample. It's important to minimize bias as much as
possible to ensure the generalizability of study findings.
Sample design plays a critical role in the validity and reliability of research results, as it
determines the extent to which findings can be generalized to the larger population. Researchers
must carefully consider various factors and choose appropriate sample designs based on their

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 1 of 9

 research objectives, resources, and constraints

Census and sample Survey

A census is a method of data collection that aims to gather information from every individual or
item in a population.
In a census, researchers attempt to collect data from every member of the population rather than
just a subset.
Because a census aims to include every individual or item in the population, it provides a
comprehensive and accurate picture of the entire population.
Census data can be used for various purposes, including government planning, demographic
analysis, and market research.
Sample Survey:

A sample survey, on the other hand, involves collecting data from only a subset or sample of
the population, rather than from every individual or item.
In a sample survey, researchers select a representative sample from the population and collect
data from this sample.
The goal of a sample survey is to obtain information about the population by studying a smaller,
manageable subset of individuals or items.
Statistical techniques are often used to ensure that the sample is representative of the population
and that inferences drawn from the sample can be generalized to the population as a whole.
Sample surveys are commonly used in various fields, including social science research, market
research, and public opinion polling.

Implication of Sample design

The implications of sample design in research are significant and can impact the validity,
reliability, and generalizability of study findings. Here are some key implications:

Representativeness: The sample design directly influences how well the selected sample
represents the population of interest. If the sample is not representative, the study results may not
accurately reflect the characteristics of the population, leading to biased conclusions.

Generalizability: The extent to which findings from a study can be generalized to the larger
population depends on the quality of the sample design. A well-designed sample that accurately
reflects the population increases the generalizability of the study results.

Precision: Sample size and sampling method choice affect the precision of estimates derived
from the sample. Larger sample sizes generally lead to more precise estimates, while certain
sampling methods may introduce variability or error into the estimates.

Bias: Sample design can introduce bias into the study results if certain groups or characteristics
are overrepresented or underrepresented in the sample. It's essential to minimize bias through
appropriate sampling techniques and procedures.

Statistical Power: Sample design impacts the statistical power of a study, which is the ability to
detect true effects or relationships. Inadequate sample sizes or biased samples can reduce
statistical power, limiting the study's ability to detect significant findings.

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 2 of 9

Resource Utilization: The choice of sample design should consider available resources such as
time, budget, and personnel. More complex sampling methods or larger sample sizes may require
greater resources but can lead to more robust study findings.

Ethical Considerations: Sample design should also consider ethical principles, such as fairness,
confidentiality, and informed consent. Researchers must ensure that sample selection procedures
protect the rights and well-being of study participants.

Analysis and Interpretation: The sample design influences the appropriate statistical methods
for analyzing the data and interpreting the results. Researchers must account for the sampling
design in their analytical approach to accurately estimate population parameters and draw valid
conclusions.

Steps in sampling design

Designing a sampling plan involves several key steps to ensure that the selected sample is representative
of the population of interest and meets the objectives of the research study. Here are the typical steps in
sampling design:

Define the Population: Clearly define the target population that the study aims to generalize findings to.
The population should be well-defined and specific, allowing for the identification of sampling units.

Specify the Sampling Frame: Identify a list or source from which the sample will be drawn. This is
known as the sampling frame and should ideally cover the entire population of interest. Ensure that the
sampling frame is up-to-date, comprehensive, and accessible.

Choose a Sampling Method: Select an appropriate sampling method based on the research objectives,
resources, and constraints. Common sampling methods include:

Simple Random Sampling: Every member of the population has an equal chance of being selected.
Stratified Sampling: The population is divided into homogeneous groups (strata), and random samples
are drawn from each stratum.
Cluster Sampling: The population is divided into clusters, and a random sample of clusters is selected
for inclusion in the study.
Systematic Sampling: Every nth member of the population is selected from a randomly chosen starting
point.
Convenience Sampling: Sampling units are selected based on convenience or accessibility.
Determine Sample Size: Calculate the appropriate sample size needed to achieve the desired level of
precision and confidence for the study objectives. Consider factors such as the population size,
variability, expected effect size, and available resources.

Implement Sampling Procedure: Execute the sampling plan according to the chosen sampling method
and sample size determination. Follow established protocols and guidelines for randomization,
stratification, clustering, or other sampling procedures.

Select Sampling Units: Randomly select sampling units from the sampling frame according to the
sampling method and sample size requirements. Ensure that the selected sample is representative of the
population and free from bias.

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 3 of 9

 Monitor Sampling Process: Monitor the sampling process to ensure that it is carried out correctly and
that any deviations or issues are addressed promptly. Document any challenges or limitations
encountered during sample selection.

Validate Sample Representativeness: Assess the representativeness of the selected sample by

comparing its characteristics to those of the population. Conduct preliminary analyses to check for any
systematic biases or discrepancies.

Adjust Sampling Design (if necessary): If the initial sample does not adequately represent the
population or meet the study objectives, consider revising the sampling design or sample size.
Additional sampling rounds or adjustments may be necessary to achieve the desired sample quality.

Document Sampling Design: Document the sampling design, including details of the sampling method,
sample size determination, sampling frame, and sampling procedure. Transparency and documentation
are essential for ensuring the reproducibility and reliability of study findings.

By following these steps in sampling design, researchers can ensure that the selected sample is
representative, unbiased, and suitable for making valid inferences about the population of interest.

Criteria for selecting a sampling procedure

Selecting the appropriate sampling procedure is crucial for ensuring that the sample is
representative of the population and that study findings are valid and reliable. Several criteria
should be considered when choosing a sampling procedure. These include:
1. Population Characteristics: Understanding the characteristics of the population is
essential for selecting an appropriate sampling procedure. Consider factors such as size,
heterogeneity, geographical dispersion, and accessibility of the population.
2. Research Objectives: The goals and research questions of the study play a significant
role in determining the most suitable sampling procedure. Different sampling methods
may be better suited for different research objectives, such as estimating population
parameters, testing hypotheses, or exploring relationships.
3. Resource Constraints: Considerations such as time, budget, and available personnel can
influence the choice of sampling procedure. Some sampling methods may require more
resources or expertise to implement effectively.
4. Sampling Frame Availability: The availability and quality of the sampling frame (i.e.,
the list or source from which the sample will be drawn) can influence the choice of
sampling procedure. Ensure that the sampling frame covers the entire population and is
up-to-date and accessible.
5. Level of Precision: Consider the desired level of precision or accuracy in the study
findings. Larger sample sizes generally result in more precise estimates, but they may not
always be feasible due to resource constraints.
6. Population Homogeneity: If the population is relatively homogeneous, simple random
sampling or systematic sampling may be appropriate. For heterogeneous populations,
more complex sampling methods such as stratified sampling or cluster sampling may be
needed to ensure representation of different subgroups.
7. Sampling Bias Considerations: Evaluate potential sources of bias and determine how
the chosen sampling procedure will minimize or mitigate bias. Ensure that the sampling
procedure avoids systematic underrepresentation or overrepresentation of certain groups
or characteristics.

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 4 of 9

 8. Feasibility and Practicality: Assess the practicality of implementing the sampling
procedure given the research context, logistical considerations, and available resources.
Choose a sampling procedure that is feasible to implement within the constraints of the
study.
9. Statistical Considerations: Consider the statistical properties of the sampling procedure,
such as its efficiency, robustness to violations of assumptions, and compatibility with the
planned data analysis techniques.
10. Ethical Considerations: Ensure that the chosen sampling procedure adheres to ethical
principles and guidelines, such as fairness, transparency, confidentiality, and respect for
the rights and well-being of study participants.
By considering these criteria in selecting a sampling procedure, researchers can choose the
method that best aligns with the objectives of the study, the characteristics of the population, and
the practical constraints of the research context. This ensures that the selected sample is
representative and suitable for making valid inferences about the population of interest.

Characteristics of a good sample design.

A good sample design is essential for ensuring that the selected sample accurately represents the
population of interest and produces valid and reliable research findings. Several characteristics
define a good sample design:
1. Representativeness: Perhaps the most critical characteristic, a good sample design
ensures that the sample is representative of the population in all relevant characteristics.
This means that the sample accurately reflects the diversity and variability present in the
population, allowing for valid generalizations from the sample to the population.
2. Unbiased Selection: The sample design should minimize or eliminate bias in the
selection process. Bias occurs when certain groups or characteristics are systematically
overrepresented or underrepresented in the sample, leading to inaccurate or misleading
conclusions. Randomization and careful sampling techniques can help reduce bias.
3. Precision: A good sample design results in estimates that are precise and reliable.
Precision refers to the degree of variability or uncertainty in the sample estimates. Larger
sample sizes generally lead to greater precision, but the sample design should balance the
trade-off between precision and resource constraints.
4. Efficiency: An efficient sample design maximizes the information gained from the
sample while minimizing the resources (time, cost, personnel) required for data
collection. Efficient sampling methods and techniques can help achieve the desired level
of precision with minimal resource expenditure.
5. Feasibility: The sample design should be practical and feasible to implement within the
constraints of the research context. This includes considerations such as the availability
of a sampling frame, logistical feasibility of data collection, and compatibility with the
research timeline and budget.
6. Flexibility: A good sample design is flexible enough to accommodate changes or
adjustments as needed during the course of the study. It should allow for modifications
based on unforeseen challenges, changes in research objectives, or new information that
arises during the study.
7. Transparency: The sample design should be transparent and well-documented, allowing
others to understand and evaluate the sampling procedures used in the study. Transparent
reporting of the sample design facilitates replication, validation, and comparison with
other studies.
8. Ethical Considerations: The sample design should adhere to ethical principles and
guidelines, ensuring that the rights and well-being of study participants are protected.
This includes obtaining informed consent, maintaining confidentiality, and minimizing

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 5 of 9

 any potential harm or discomfort to participants.
9. Generalizability: A good sample design enables valid inferences to be made from the
sample to the larger population. The sample should be sufficiently representative and
unbiased to support generalizations and conclusions that apply beyond the sample itself.
10. Compatibility with Analytical Techniques: The sample design should be compatible
with the planned data analysis techniques and statistical methods. This ensures that the
data collected from the sample can be appropriately analyzed and interpreted to address
the research objectives.
Overall, a good sample design is characterized by its representativeness, lack of bias, precision,
efficiency, feasibility, flexibility, transparency, adherence to ethical principles, generalizability,
and compatibility with analytical techniques. By incorporating these characteristics into the
sample design, researchers can enhance the validity and reliability of their research findings.

Different types of Sample design.

There are several types of sample designs commonly used in research, each with its own
advantages, limitations, and applicability. Here are some of the main types of sample designs:
1. Simple Random Sampling:
● In simple random sampling, each member of the population has an equal chance
of being selected for the sample.
● This method is straightforward and easy to implement, making it ideal when the
population is relatively homogeneous and the sampling frame is well-defined.
● However, it may not be practical for large populations, and it does not guarantee
representation of specific subgroups within the population.
2. Stratified Sampling:
● Stratified sampling involves dividing the population into homogeneous subgroups
or strata based on certain characteristics (e.g., age, gender, income), and then
selecting samples independently from each stratum.
● This method ensures representation of different subgroups within the population
and can improve the precision of estimates by reducing variability within strata.
● However, it requires prior knowledge of the population characteristics and may be
more complex and costly to implement compared to simple random sampling.
3. Cluster Sampling:
● Cluster sampling involves dividing the population into clusters or groups, and
then randomly selecting entire clusters as the sampling units.
● This method is useful when it is impractical or costly to obtain a complete list of
individuals in the population, as it can reduce logistical challenges and costs
associated with sampling.
● However, it may introduce greater variability within clusters and require larger
sample sizes to achieve the same level of precision as simple random sampling.
4. Systematic Sampling:
● Systematic sampling involves selecting every nth member from a list or sampling
frame after randomly choosing a starting point.
● This method is easy to implement and less time-consuming than simple random
sampling, while still providing a degree of randomness in the selection process.

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 6 of 9

 ● However, it may introduce bias if there is a periodic pattern or structure in the
sampling frame, and it may not be suitable for populations with inherent order or
periodicity.
5. Convenience Sampling:
● Convenience sampling involves selecting sampling units based on their
accessibility or availability to the researcher.
● This method is quick, easy, and inexpensive, making it suitable for exploratory or
pilot studies or when access to the entire population is not possible.
● However, it may introduce bias if the sample does not accurately represent the
population, and the results may not be generalizable beyond the sample itself.
6. Snowball Sampling:
● Snowball sampling involves selecting initial participants who then refer additional
participants, who in turn refer more participants, and so on.
● This method is commonly used in studies where the population of interest is hard
to reach or hidden, such as marginalized or stigmatized groups.
● However, it may lead to biased samples if certain subgroups are overrepresented
or if there are limitations to the network of referrals.
These are some of the main types of sample designs used in research. The choice of sample
design depends on various factors, including the research objectives, characteristics of the
population, available resources, and practical considerations. Researchers should carefully
consider these factors when selecting the most appropriate sample design for their study.

Measurement Scales.
Measurement scales, also known as scales of measurement or levels of measurement, refer to the
different ways in which variables can be categorized, measured, and expressed in statistical
analysis. There are four main types of measurement scales:
1. Nominal Scale:
● The nominal scale is the simplest level of measurement and involves categorizing
variables into distinct categories or groups.
● In this scale, variables are named, labeled, or assigned to categories without any
inherent order or ranking.
● Examples include gender (male, female), ethnicity (Asian, African American,
Hispanic), and marital status (single, married, divorced).
● Nominal data can be summarized using frequencies and percentages, but
arithmetic operations such as addition and subtraction are not meaningful.
2. Ordinal Scale:
● The ordinal scale involves categorizing variables into ordered or ranked
categories, where the categories have a meaningful sequence but the intervals
between them are not necessarily equal.
● In this scale, variables can be ranked or ordered, but the differences between the
categories are not consistent or measurable.
● Examples include rankings (1st, 2nd, 3rd), Likert scales (strongly agree, agree,
neutral, disagree, strongly disagree), and levels of satisfaction (very satisfied,

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 7 of 9

 satisfied, neutral, dissatisfied, very dissatisfied).
● Ordinal data can be summarized using median, mode, and percentile, but
arithmetic operations such as addition and subtraction are not meaningful.
3. Interval Scale:
● The interval scale involves categorizing variables into ordered categories with
equal intervals between them, but without a true zero point.
● In this scale, variables can be measured on a continuous scale, and the intervals
between the categories are consistent and measurable.
● Examples include temperature measured in Celsius or Fahrenheit, calendar dates,
and IQ scores.
● Interval data can be summarized using mean, standard deviation, and range, and
arithmetic operations such as addition and subtraction are meaningful. However,
multiplication and division are not meaningful because there is no true zero point.
4. Ratio Scale:
● The ratio scale is the highest level of measurement and includes all the properties
of the interval scale, with the addition of a true zero point.
● In this scale, variables can be measured on a continuous scale with equal intervals
between categories and a true zero point, which represents the absence of the
attribute being measured.
● Examples include height, weight, age, income, and time measured in seconds.

● Ratio data can be summarized using mean, standard deviation, range, and
arithmetic operations such as addition, subtraction, multiplication, and division
are all meaningful.
These measurement scales provide a framework for understanding the characteristics and
properties of different types of variables, which in turn guide the selection of appropriate
statistical techniques for analysis. It's important for researchers to consider the level of
measurement when designing studies, collecting data, and interpreting results.

Important scaling Techniques.

Scaling techniques are used to transform raw data into a standardized format that allows for
meaningful comparisons and analysis. Several important scaling techniques are commonly used
in various fields of research. Here are some of the most important ones:
1. Likert Scale: Likert scales are widely used in survey research to measure attitudes,
opinions, and perceptions. Respondents are asked to indicate their level of agreement or
disagreement with a series of statements using a predefined scale, typically ranging from
"strongly disagree" to "strongly agree." Likert scales produce ordinal data.
2. Semantic Differential Scale: Similar to Likert scales, semantic differential scales are
used to measure attitudes and perceptions. Respondents are asked to rate an object,
concept, or experience on a set of bipolar adjective pairs (e.g., good-bad, happy-sad, fast-
slow). This scale provides interval-level data.
3. Thurstone Scale: Thurstone scales are designed to measure attitudes by asking
respondents to evaluate a series of statements according to their degree of agreement or
disagreement. However, unlike Likert scales, Thurstone scales use a technique called
"equal-appearing intervals" to assign scores to each statement based on its perceived level
of agreement. Thurstone scales produce interval-level data.

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 8 of 9

 4. Guttman Scale: Guttman scales, also known as cumulative scales or scalogram analysis,
are used to measure attitudes or traits that are assumed to be hierarchical in nature.
Respondents are presented with a series of statements or items ordered by difficulty or
severity, and their responses are used to infer their position along the latent trait being
measured. Guttman scales produce ordinal-level data.
5. Rasch Model: The Rasch model is a psychometric approach used to develop
measurement scales and assess the validity and reliability of responses to categorical
items. It is based on the principles of item response theory and provides a framework for
analyzing responses to test items in terms of both item difficulty and respondent ability.
The Rasch model produces interval-level data.
6. Likert-Type Scaling: Likert-type scaling is a broad category of scaling techniques that
includes Likert scales, semantic differential scales, and other similar approaches. These
scales involve presenting respondents with a series of statements or items and asking
them to rate their level of agreement, satisfaction, frequency, or other attributes. Likert-
type scaling produces ordinal-level data.
7. Visual Analog Scale (VAS): VAS is a scaling technique commonly used in healthcare
and psychology to measure subjective experiences such as pain, mood, or quality of life.
Respondents are asked to indicate their level of agreement or intensity by placing a mark
along a continuous line anchored by two extreme endpoints. VAS produces interval-level
data.
8. Magnitude Estimation: Magnitude estimation is a scaling technique used to measure the
perceived intensity or magnitude of stimuli such as brightness, loudness, or taste.
Respondents are asked to assign numerical values to stimuli based on their perceived
magnitude, with the values typically being logarithmically related to the stimulus
intensity. Magnitude estimation produces ratio-level data.
These are some of the most important scaling techniques used in research across various
disciplines. The choice of scaling technique depends on the nature of the variables being
measured, the research objectives, and the preferences of the researcher.

Prof. Kuldeep Choudhary MHM-2 MHM 206-18 Research Methodology Page 9 of 9