UNIT II

Chapter 1: Tools for Data Collection

⮚ Collection of primary data

In research methodology, primary data refers to data collected directly

from original sources, tailored specifically to address the research
question or objective. Unlike secondary data, which has been collected for
purposes other than the current research, primary data is fresh and
original. The collection of primary data is fundamental in research to
ensure the validity and reliability of findings.

Advantages of Primary Data Collection:

● Accuracy: Direct collection ensures the data is relevant and specific

to the research question.

● Control: Researchers have control over how the data is collected,

allowing for consistency and clarity.

● Timeliness: The data is collected in real-time, so it reflects current

conditions and is more reliable for studying ongoing issues.
Disadvantages of Primary Data Collection:

● Cost: Collecting primary data can be expensive in terms of time,

resources, and labor.

● Time-Consuming: The process of designing instruments, collecting

data, and analyzing it is typically more time-consuming than
secondary data collection.

● Sampling Challenges: For methods like surveys or interviews,

ensuring a representative sample can be difficult.
⮚ Collection of data through questionnaires and schedules
Collection of data through questionnaires and schedules is a
common method of gathering primary data in research. Both
methods involve the use of pre-designed sets of questions that
respondents are asked to answer.

1. Questionnaires
● A questionnaire is a research tool consisting of a series of written
questions designed to gather data from respondents. Questionnaires
are typically self-administered, meaning the respondent fills them
out independently, without direct involvement from the researcher.
● Formats:
o Closed-Ended Questions: Respondents choose from
predefined options (e.g., multiple choice, yes/no).
o Open-Ended Questions: Respondents answer in their own
words, providing more detailed insights.

Advantages:
● Cost-effective: Particularly when distributed to large numbers of
people, either online or via postal mail.
● Efficiency: Can be distributed widely and responses can be
collected quickly.
● Anonymity: Respondents may feel more comfortable answering
honestly, especially on sensitive topics.
Disadvantages:
● Response Rate: Can suffer from low response rates, especially if
sent via mail or email.
● Limited Interaction: Lack of opportunity for clarification or probing
further into answers.
● Misunderstanding Questions: Respondents may interpret
questions differently, leading to inaccurate responses.
Example: A researcher studying consumer preferences may send
out a questionnaire asking people to choose their preferred brand of
cereal, rate customer service experiences, or select age group
ranges.

2. Schedules (Interview Schedules)

● A schedule is similar to a questionnaire but differs in that it is
typically administered by a researcher (interviewer) who asks the
questions and records the responses. This process is more direct
and personal.
● It’s often used when the researcher needs more control over how
questions are asked or when complex or detailed responses are
required.
● Formats:
o Structured Schedule: Questions are fixed, and the interviewer
asks them in a specific order. There is little to no flexibility for
deviating from the set questions.
o Semi-Structured Schedule: Some questions are pre-set, but
the interviewer can adjust the order or ask follow-up questions
depending on the responses.
o Unstructured Schedule: The interviewer is free to explore
responses and ask any relevant questions as the conversation
evolves.
Advantages:
● Clarification: The interviewer can clarify questions, ensuring the
respondent understands them correctly.
● Higher Response Rate: Face-to-face or telephone interactions tend
to yield better response rates compared to questionnaires.
● Flexibility: The interviewer can probe for deeper insights or follow up
on unexpected answers.
Disadvantages:
● Interviewer Bias: The way questions are asked or responses are
recorded could influence the answers.
● Time-Consuming: Administering schedules requires more time and
resources because of the interviewer’s involvement.
● Cost: More expensive than questionnaires due to the labor involved
in conducting interviews.
Example: A researcher conducting a study on employee satisfaction
may visit employees in person or conduct phone interviews, asking
predetermined questions but allowing room for deeper exploration
based on responses.
Key Differences Between Questionnaires and Schedules
Schedule (Interview
Feature Questionnaire
Schedule)
Self-
Administered by the
Administere administered
researcher/interviewe
d By by the
r
respondent
Minimal, only
Direct interaction with
Interaction the written
the researcher
questions
Often more Can be structured,
Question
structured semi-structured, or
Format
(closed-ended) unstructured
Limited control
More control over
Response over how
clarifying and probing
Control questions are
responses
interpreted
Low cost (no Higher cost
Cost interviewer (interviewers are
needed) needed)
Flexibility Less flexibility High flexibility for
for follow-up or follow-up or clarifying
 Schedule (Interview
Feature Questionnaire
Schedule)
clarification questions
Fast to
distribute, Time-consuming to
Time
slower to collect and transcribe
Efficiency
collect responses
responses
Generally more
superficial Allows for deeper
Data Depth (especially with insights due to the
closed-ended ability to probe
questions)

⮚ Collection of Secondary Data

Collection of Secondary Data refers to the process of gathering data that has
already been collected, processed, and made available by other researchers or
organizations for purposes other than the current research. Unlike primary data,
which is collected firsthand, secondary data is sourced from existing materials.

Advantages of Secondary Data Collection

1. Cost-Effective
o Secondary data is often more affordable to access compared to
collecting primary data, which can require substantial resources and
effort.
2. Time-Saving
o Data is already collected, processed, and analyzed, saving researchers
time in the research process.
3. Large Data Sets
o Secondary data often provides access to extensive datasets, such as
national surveys or longitudinal studies, which might be challenging or
impossible for a researcher to gather independently.

Disadvantages of Secondary Data Collection

1. Data Relevance
 o The existing data may not perfectly align with the researcher’s specific
needs or may be outdated, incomplete, or lacking in detail.
2. Quality Concerns
o The researcher has little control over how the data was collected or its
accuracy, which may lead to concerns about validity or reliability.
3. Limited Control Over Data Collection
o Researchers cannot modify the way the data was originally collected or
the variables that were considered, which may limit the ability to address
specific research questions.
4. Access Restrictions
o Some secondary data may not be publicly available or may require
special permissions to access, such as proprietary market reports or
private research datasets.

selection of appropriate method for data collection

Selecting the appropriate method for data collection depends on various

factors, including the research objectives, the type of data needed, the
resources available, and the target population. Below are some common
data collection methods and guidelines on when to use each:

1. Surveys and Questionnaires

● When to use: When you need to collect standardized information

from a large group of people.
● Advantages:
○ Can reach a large audience.
○ Can be administered online, in person, or by phone.
○ Easy to analyze if structured well.
● Disadvantages:
○ Responses may not always be truthful or accurate.
○ Low response rates if not well-designed or incentivized.

2. Interviews (Structured, Semi-structured, and Unstructured)

● When to use: When you need in-depth information, understanding
of individual experiences, or complex responses.
● Advantages:
○ Allows for rich, qualitative data.
○ Can explore topics in detail.
○ Allows for clarification and probing.
● Disadvantages:
○ Time-consuming and resource-intensive.
○ May be biased by the interviewer.
○ Difficult to analyze, especially for large samples.

3. Focus Groups

● When to use: When you want to explore ideas or opinions from a

small group and facilitate discussion.
● Advantages:
○ Group dynamics can generate ideas that individuals might not
think of.
○ Provides a deeper understanding of social processes and
attitudes.
● Disadvantages:
○ Limited generalizability.
○ Group dynamics may affect individual responses (e.g., peer
pressure).
○ Requires skilled facilitation.

4. Observation

● When to use: When you want to understand behaviors, processes,

or events in their natural context.
● Advantages:
○ Provides direct insights into how people behave in real
settings.
○ Can capture non-verbal data that surveys cannot.
● Disadvantages:
○ Observer bias can affect findings.
○ May not always capture internal thoughts or motivations.
 ○ Time-intensive and may require a significant amount of
resources.

5. Experiments (Controlled or Field)

● When to use: When you want to establish causal relationships

between variables.
● Advantages:
○ Can provide strong evidence of cause-and-effect
relationships.
○ Allows for control over variables.
● Disadvantages:
○ May be expensive and require specialized equipment or
settings.
○ Ethical concerns, especially in human subjects research.
○ May lack external validity if not generalizable to real-world
situations.

6. Document/Content Analysis

● When to use: When you need to analyze existing data such as

reports, publications, online content, etc.
● Advantages:
○ No need to interact with people.
○ Can be useful for historical or longitudinal studies.
● Disadvantages:
○ May be biased based on the availability and quality of the
documents.
○ Can be time-consuming to sift through large amounts of data.

7. Case Studies

● When to use: When you need an in-depth understanding of a single

unit or case over time (e.g., individual, group, or organization).
● Advantages:
○ Provides a deep understanding of the subject.
○ Can offer insight into unique or rare cases.
 ● Disadvantages:
○ Not easily generalizable.
○ Time-consuming and resource-intensive.

8. Secondary Data (Existing Data Analysis)

● When to use: When you want to analyze data that has already been
collected by others (e.g., public datasets, government reports, etc.).
● Advantages:
○ Cost-effective and less time-consuming.
○ Can cover a large scope and sample size.
● Disadvantages:
○ Data may not exactly match your research question.
○ Limited control over data quality and accuracy.

Chapter 2: Sampling Design

Sampling design is a critical aspect of research that refers to the process

of selecting a subset (sample) from a larger population to gather data. The
purpose of sampling is to make inferences about the population without
needing to collect data from every member. The choice of sampling design
influences the accuracy, reliability, and generalizability of the research
findings.

Steps in Sampling Design

The process of sampling design involves several steps to ensure that the
sample chosen for the study is representative, accurate, and aligned with the
research objectives. Below are the key steps involved in creating a sampling
design:
 1. Define the Population

● Objective: Clearly identify the group of individuals or items that your

research will focus on.
● What to Do: Determine the boundaries of your population in terms of
location, time, and characteristics. For example, if you're studying
customer satisfaction, your population may be "all customers who
purchased a product in the last six months."

2. Determine the Sampling Frame

● Objective: Develop a list or a database from which you will draw the
sample.
● What to Do: Identify and obtain a sampling frame, which is the actual list
or representation of the population. For example, this could be a list of
registered voters, a company’s employee directory, or an academic
database of enrolled students.

3. Choose the Sampling Method

● Objective: Select the appropriate sampling method based on the

research objectives and available resources.
● What to Do: Decide between probability sampling (where each member
of the population has a known chance of being selected) or non-
probability sampling (where selection is not based on random chance).

4. Determine the Sample Size

● Objective: Decide how many individuals or units to include in the sample.

● What to Do: Calculate the optimal sample size, which depends on
factors.

5. Define the Sampling Units

● Objective: Identify the individual units or elements from the population

that will be included in the sample.
● What to Do: Decide whether your sampling unit is an individual, a
household, an organization, or any other unit of interest. For example, if
you’re studying consumer behavior, the sampling unit may be an
individual customer; if you’re studying workplace culture, the sampling
unit may be an entire department.

6. Choose the Sampling Technique

● Objective: Select the specific technique or procedure for selecting units

from the sampling frame.
● What to Do: Apply the chosen sampling method (e.g., random selection,
stratification) to ensure that the units are selected appropriately.

7. Collect Data

● Objective: Gather data from the selected sample.

● What to Do: Implement the data collection process using the research
tools (surveys, interviews, observations) you’ve prepared. Ensure that the
data collection follows the ethical guidelines, including obtaining
informed consent if applicable.

8. Analyze and Assess the Sample

● Objective: Evaluate the quality and representativeness of your sample

and the data collected.
● What to Do: After data collection, assess whether the sample accurately
represents the population and if there are any noticeable biases. You
might also want to check for sample size adequacy and possible non-
response or dropout rates.

9. Address Sampling Bias

● Objective: Minimize any errors or bias introduced by the sampling

process.
● What to Do: Identify any possible sources of sampling bias, such as non-
random selection, underrepresented groups, or issues with the sampling
frame. Take corrective actions if necessary, such as adjusting for stratified
sampling or increasing the sample size.

10. Report Sampling Methodology

● Objective: Clearly document and report how the sample was selected in
the research methodology section.
● What to Do: Provide a detailed description of your sampling process,
including how you defined the population, constructed the sampling
frame, selected the sampling method, and determined the sample size.
Mention any limitations that could affect generalizability or bias.

 Characteristics of a good sample design

A good sample design is essential to ensure that the results of a
study or survey are reliable and representative of the larger
population. Here are some key characteristics of a good sample
design:
1. Representativeness
 The sample should accurately reflect the characteristics of the population
from which it is drawn. This ensures that conclusions drawn from the
sample can be generalized to the entire population.
2. Randomness
 Random sampling methods reduce bias by giving each member of the
population an equal chance of being selected. This helps in making the
sample more representative and reduces the influence of confounding
factors.
3. Adequate Sample Size
 The sample size should be large enough to produce reliable estimates
with a specified level of confidence and precision. A small sample may
lead to inaccurate or misleading results, while an overly large sample may
be inefficient.
4. Clear Definition of the Population
 A good sample design begins with a clear definition of the population
being studied. It ensures that the sample is drawn from the correct group
and minimizes the risk of including irrelevant individuals.
5. Minimized Bias
 A good sample design aims to reduce various types of bias (selection bias,
nonresponse bias, etc.) through careful planning and implementation.
This includes addressing how the sample is selected and how
nonrespondents are handled.
6. Stratification (if applicable)
 Stratified sampling involves dividing the population into distinct
subgroups (strata) and then sampling from each stratum. This ensures
that important subgroups are adequately represented, particularly if they
differ significantly on key variables.
7. Cost-effectiveness
 The sample design should consider budget and time constraints. It should
be efficient in terms of both financial and logistical resources, without
sacrificing the quality of the data.

 Different types of sample designs

The following are the main types of sample designs:

1. Probability Sampling Designs
These designs rely on random selection, meaning each member of the
population has a known (non-zero) chance of being selected. They are
generally considered more accurate and reliable.
 Simple Random Sampling (SRS)
o Description: Every member of the population has an equal chance
of being selected.
o Advantages: Simple to understand and implement, unbiased.
o Disadvantages: May require a complete list of the population and
can be inefficient for large populations.
 Systematic Sampling
o Description: The first member is selected randomly, and then every
kth member is chosen after that.
o Advantages: Easy to implement and can be more efficient than
simple random sampling.
 o Disadvantages: Can introduce bias if there is a hidden pattern in the
population structure.
 Stratified Sampling
o Description: The population is divided into distinct subgroups
(strata) based on specific characteristics, and random samples are
taken from each subgroup.
o Advantages: Ensures that all key subgroups are represented and
can increase precision.
o Disadvantages: Requires knowledge of the strata and can be more
complex to implement.
 Cluster Sampling
o Description: The population is divided into clusters (often
geographically), and a random sample of clusters is selected. All
members of the chosen clusters may be surveyed.
o Advantages: Cost-effective for large, geographically dispersed
populations.
o Disadvantages: May be less precise if the clusters are
heterogeneous (different within the clusters).
 Multistage Sampling
o Description: A combination of different sampling techniques is
used in stages. For example, you might first select clusters and then
sample within those clusters using stratified or simple random
sampling.
o Advantages: Flexible and can handle large, complex populations.
o Disadvantages: More complex and can lead to higher sampling
errors.
2. Non-Probability Sampling Designs
These designs do not rely on random selection, which may lead to bias.
They are often used when probability sampling is not feasible due to
constraints such as time or budget.
 Convenience Sampling
o Description: The sample is chosen based on convenience or
accessibility (e.g., interviewing people who are nearby or willing to
participate).
o Advantages: Quick and inexpensive.
o Disadvantages: Likely to be biased, as the sample may not be
representative of the population.
 Judgmental (Purposive) Sampling
 o Description: The researcher selects the sample based on their
judgment or specific characteristics of interest, typically aiming to
select individuals who are particularly knowledgeable or
experienced.
o Advantages: Useful for specific case studies or expert opinion
research.
o Disadvantages: Highly subjective, can lead to bias, and may not be
representative of the broader population.
 Quota Sampling
o Description: The population is segmented into groups (like in
stratified sampling), and participants are selected non-randomly to
meet a predetermined quota for each subgroup.
o Advantages: Ensures representation from specific subgroups,
faster and cheaper than probability sampling.
o Disadvantages: Still prone to bias because participants are not
randomly selected.
 Snowball Sampling
o Description: Often used for hard-to-reach or hidden populations,
where existing participants refer new participants to the study.
o Advantages: Useful for studying rare or difficult-to-access
populations.
o Disadvantages: The sample may be biased, as participants are
selected based on referrals.
3. Other Hybrid or Advanced Sampling Designs
These approaches combine aspects of the techniques above or use
additional methodologies to enhance accuracy or reduce costs.
 Double Sampling
o Description: Involves collecting data from a subset of the
population in two stages: first with a simpler method, then a more
detailed method.
o Advantages: More detailed and accurate data can be obtained for a
subset of the population.
o Disadvantages: Time-consuming and complex to execute.
 Adaptive Sampling
o Description: Used when studying rare or elusive events. The sample
design can adapt during data collection to focus on areas or groups
that are more likely to exhibit the characteristics of interest.
o Advantages: Efficient for studying rare phenomena.
 o Disadvantages: Can be more complex and difficult to implement.

 How to select a random sample

Selecting a random sample is an essential step in ensuring that your

sample is representative of the population you're studying. Here's a step-
by-step guide on how to select a random sample:
1. Define the Population
 Description: Clearly identify and define the population from which the
sample will be drawn. The population should include all individuals or
units that you want to make inferences about.
 Example: If you're studying the average age of college students in a
university, your population is all students enrolled at that university.
2. Decide on the Sample Size
 Description: Determine how many individuals or units you want in your
sample. This decision depends on factors like the desired level of
precision, available resources, and how large the population is.
 Factors influencing sample size:
o Confidence level (e.g., 95%)
o Margin of error (e.g., ±5%)
o Population size
 Tools: You can use online sample size calculators or statistical formulas
to calculate the ideal sample size.
3. Create a List of the Population (Sampling Frame)
 Description: Prepare a complete list of all the individuals or units in your
population. This is often called a "sampling frame."
 Example: If you're sampling from students at a university, the sampling
frame might be a list of all enrolled students, like a student registry or a
database.
4. Choose a Random Sampling Method
 There are different methods to randomly select individuals. Here are some
common methods:
a. Simple Random Sampling (SRS)
 Description: Every member of the population has an equal chance of
being selected. This is the most basic form of random sampling.
 How to do it:
1. Assign a unique number to each member in the sampling frame.
 2. Use a random number generator (either manually or using software)
to select numbers that correspond to the individuals you’ll include
in your sample.
3. The selected individuals become your sample.
o Example: If your population is 500 students, and you need a sample
of 50, generate 50 random numbers between 1 and 500.
b. Systematic Sampling
 Description: Instead of selecting randomly, you select every kth individual
from the list, starting at a randomly chosen point.
 How to do it:
1. Determine the sampling interval (k) by dividing the population size
by the desired sample size.
2. Randomly select a starting point within the first k individuals.
3. Select every kth individual after that.
o Example: If your population has 1,000 individuals and you need a
sample of 100, select every 10th person (1000 / 100 = 10). If you
randomly select person 7 as your starting point, you will then select
persons 7, 17, 27, etc.
c. Stratified Sampling
 Description: The population is divided into distinct subgroups (strata), and
a random sample is taken from each stratum. This ensures each subgroup
is represented.
 How to do it:
1. Identify the strata (e.g., males and females, different age groups,
etc.).
2. Randomly select a sample from each stratum, either proportionally
or equally depending on your research goals.
o Example: In a university setting, you could stratify by major (e.g.,
business, engineering, arts) and then randomly select students
from each major.
d. Cluster Sampling
 Description: The population is divided into clusters, and entire clusters are
selected randomly. This method is often used when a population is
geographically dispersed.
 How to do it:
1. Divide the population into clusters (e.g., schools, neighborhoods, or
regions).
2. Randomly select some of the clusters.
 3. Include all individuals within the selected clusters in your sample.
o Example: If your population consists of students from multiple
schools, randomly select a few schools, then survey all students
within those selected schools.
5. Random Number Generation or Drawing
 Tools for Random Selection:
o Manual methods: If the sample size is small, you could use
methods like drawing names out of a hat.
o Software methods: For larger samples, you can use random
number generators or software tools like Excel, R, or Python to
automate the random selection process.
o Excel: You can use the =RAND() function to assign a random
number to each individual in your sampling frame, then sort them
by the random number to select your sample.
o Online Random Number Generators: Websites like random.org can
generate random numbers or draw random samples for you.
6. Check for Bias or Errors
 After selecting your sample, ensure that no biases have been introduced
during the random selection process. Double-check the method and
confirm that the sampling process was properly followed.
 Potential issues:
o Sampling frame errors (e.g., missing individuals from the list)
o Nonresponse bias (e.g., individuals refusing to participate)
7. Conduct the Study with the Sample
 Once you've selected the random sample, proceed with data collection or
experimentation as planned.
Example: Simple Random Sampling
1. Population: 1000 students.
2. Desired sample size: 100 students.
3. Sampling frame: List of all 1000 students.
4. Selection process: Assign each student a unique number from 1 to 1000.
Use a random number generator (e.g., in Excel or random.org) to select
100 random numbers between 1 and 1000. The students corresponding to
those numbers are your sample.