Module 2: Data Collection & Sampling Methods

Introduction
Data collection is a crucial step in the research Random Sampling:
process, providing the foundation for analysis, Random sampling is a method where every
interpretation, and drawing meaningful individual or element in the population has an
conclusions. The methods employed in collecting equal chance of being selected. This minimizes
data play a pivotal role in ensuring the accuracy, bias and ensures that the sample is representative
reliability, and relevance of the information of the entire population, making it a commonly
gathered. Additionally, sampling methods are used and statistically robust technique.
essential in selecting a subset of the population to
represent the whole, making the research process Stratified Sampling:
more feasible and manageable. Stratified sampling involves dividing the
population into subgroups or strata based on
Purpose of Data Collection: specific characteristics and then randomly
selecting samples from each stratum. This method
Understanding the purpose behind data collection ensures representation from different demographic
is essential. Whether it is to explore relationships, or characteristic groups within the population.
describe phenomena, make predictions, or test
hypotheses, the objectives influence the choice of
Cluster Sampling:
methods and tools used in the process.
Cluster sampling involves dividing the population
into clusters and then randomly selecting entire
Types of Data: clusters for analysis. This method is particularly
useful when the population is naturally grouped,
Data can be classified into two main types: making it more feasible to sample entire clusters
qualitative and quantitative. Qualitative data is instead of individual elements.
non-numeric and often involves words,
descriptions, or visual representations, while
quantitative data consists of numerical values, Convenience Sampling:
facilitating statistical analysis. Convenience sampling involves selecting
individuals who are readily available or easily
accessible. While this method is convenient, it may
Data Collection Methods: introduce bias, as the sample may not be
Various methods can be employed to collect data, representative of the entire population.
depending on the nature of the research. Common
methods include surveys, interviews, observations,
experiments, and archival research. Each method Primary & Secondary data
has its strengths and limitations, and researchers
must choose the most suitable approach based on The most important factor in any statistical data is
their research questions and objectives. original collected data must be correct and proper.
If there are shortcomings at the very source of data,
Sampling Techniques: no useful and valid conclusions can be drawn even
In many cases, it is impractical or impossible to after applying the best techniques of data analysis.
collect data from an entire population. Sampling In this context, we quote the remarks made by a
involves selecting a subset of individuals or judge an Indian statistics: "Cox, when you are bit
elements from the larger population to represent it older you will not quote Indian statistics with that
accurately. Sampling techniques include random accuracy". The governments are very keen on
sampling, stratified sampling, cluster sampling, amassing statistics - they collect them, add them,
and convenience sampling, each with its own raise them, to the nth power take the cube root and
advantages and drawbacks. prepare wonderful diagrams.
But what you must never forget is that every one of It is collected by the researcher or research team
those figures comes in the first instance from the directly from individuals, groups, experiments, or
"Chowkidar" (i.e. the village watchman) who just observations.
puts down what he down pleases."
Purpose:
Data may be Collected from the Following Two Gathered with a specific research question or
Sources objective in mind.
(i) The investigator or the organising agency may
conduct the enquiry originally or Customization:
(ii) Necessary data may be obtained for enquiry Researchers have control over the design and
from some other sources or agencies who had collection process, allowing them to tailor the data
already collected the data on that subject. collection methods to suit the research needs.

Definition of Primary Data and Secondary Data Freshness and Relevance:

Primary data is fresh, relevant, and specific to the
Primary data: The data collected by an agency for current research project.
the first time for any statistical investigation and
used for statistical analysis is called primary data. Reliability:
Since it is collected firsthand, there is a higher
Secondary data: level of accuracy and reliability compared to
(i) Published or unpublished data collected by secondary data.
some agency and
(ii) Processed by some agency and Examples:
(iii) Collected from there and used for the Surveys, interviews, observations, experiments,
statistical work, is called secondary data, focus groups, and any data collected directly for
The second agency, which publishes this data the research study.
becomes the secondary source.
Thus secondary source is the agency who Secondary Data:
publishes the data for use by others but which was
not originally collected and processed by it. Definition:
Secondary data refers to information that has
The clear distinction between primary and already been collected by someone else for a
secondary data is as follows: purpose other than the current research.
(i) The distinction is a matter of degree only.
(ii) The same data may be secondary, in the hands Source:
of one and primary in the hands of others. It is obtained from existing sources, such as
(iii) In general, the data are primary to the source government publications, academic journals,
who collects and processes them for the first time books, reports, or online databases.
and are secondary who later use this data
Purpose:
Difference between Primary and Used to support or supplement the current research
Secondary Data rather than being collected specifically for it.

Primary Data: Customization:

Researchers have limited control over the design
Definition: and collection process, as the data was collected
Primary data refers to original and firsthand for a different purpose.
information collected directly from the source for a
specific research purpose. Freshness and Relevance:
May not be as current or directly applicable to the
Source: current research question.
Reliability: and local persons, the investigators can cover a
There may be issues of relevance, accuracy, or bias wide area.
in secondary data, as it was not originally collected
for the researcher's specific purpose. Information through Questionnaires and
Schedules:
Examples: In this method of collecting primary data, the
Published reports, government statistics, academic investigator, while keeping in mind the motive
journals, historical data, and any information of the study, prepares a questionnaire. The
collected for purposes other than the current investigator can collect data through the
research project. questionnaire in two ways:

Sources of data Mailing Method:

This method involves mailing the questionnaires
to the informants for the collection of data. The
investigator attaches a letter with the
questionnaire in the mail to define the purpose
of the study or research. The investigator also
assures the informants that their information
would be kept secret, and then the informants
note the answers to the questionnaire and return
the completed file.

Enumerator’s Method:
This method involves the preparation of a
Direct Personal Investigation:
questionnaire according to the purpose of the
As the name suggests, the method of direct
study or research. However, in this case, the
personal investigation involves collecting data
enumerator reaches out to the informants
personally from the source of origin. In simple
himself with the prepared questionnaire.
words, the investigator makes direct contact with
Enumerators are not the investigators
the person from whom he/she wants to obtain
themselves; they are the people who help the
information. This method can attain success only
investigator in the collection of data.
when the investigator collecting data is efficient,
diligent, tolerant and impartial. For example,
Sources of Collecting Secondary Data
direct contact with the household women to
obtain information about their daily routine and
1. Published Sources
schedule.
Government Publications:
Indirect Oral Investigation:
Government publishes different documents
In this method of collecting primary data, the
which consists of different varieties of
investigator does not make direct contact with
information or data published by the Ministries,
the person from whom he/she needs information,
Central and State Governments in India as their
instead, they collect the data orally from some
routine activity. As the government publishes
other person who has the necessary required
these Statistics, they are fairly reliable to the
information. For example, collecting data of
investigator. Examples of Government
employees from their superiors or managers.
publications on Statistics are the Annual Survey
of Industries, Statistical Abstract of India, etc.
Information from Local Sources or
Correspondents: In this method, for the
Semi-Government Publications:
collection of data, the investigator appoints
Different Semi-Government bodies also publish
correspondents or local persons at various
data related to health, education, deaths and
places, which are then furnished by them to the
births. These kinds of data are also reliable and
investigator. With the help of correspondents
used by different informants. Some examples of
semi-government bodies are Metropolitan Surveys and Questionnaires:
Councils, Municipalities, etc.
Surveys involve asking individuals a set of
Publications of Trade Associations: structured questions to gather information about
Various big trade associations collect and their opinions, attitudes, behaviors, or
publish data from their research and statistical demographics. Questionnaires are often self-
divisions of different trading activities and their administered forms.
aspects. For example, data published by Sugar
Mills Association regarding different sugar mills Interviews:
in India. Interviews involve direct interaction between the
researcher and the participant, where questions are
Journals and Papers: asked, and responses are recorded. Interviews can
Different newspapers and magazines provide a be structured, semi-structured, or unstructured.
variety of statistical data in their writings, which
are used by different investigators for their Observations:
studies. Observational research involves systematically
watching and recording behaviors, events, or
International Publications: phenomena. This method is particularly useful
Different international organizations like IMF, when studying natural behavior in a specific
UNO, ILO, World Bank, etc., publish a variety context.
of statistical information which are used as
secondary data. Experiments:
Experimental research involves manipulating
Publications of Research Institutions: variables to observe their effects on outcomes.
Research institutions and universities also Experiments allow researchers to establish causal
publish their research activities and their relationships between variables.
findings, which are used by different
investigators as secondary data. For example Focus Groups:
National Council of Applied Economics, the Focus groups involve a small group of participants
Indian Statistical Institute, etc. discussing a topic facilitated by a researcher. This
method is useful for exploring diverse perspectives
2. Unpublished Sources and opinions.
Another source of collecting
secondary data is unpublished sources. The data Case Studies:
in unpublished sources is collected by different Case studies involve in-depth examination of a
government organizations and other particular individual, group, or situation. Data for
organizations. These organizations usually case studies can be gathered through interviews,
collect data for their self-use and are not observations, and analysis of existing documents.
published anywhere. For example, research
work done by professors, professionals, teachers Archival Research:
and records maintained by business and private Archival research involves the examination of
enterprises. existing records, documents, and historical data to
answer research questions. This method is often
used in historical and qualitative research.
Methods of collecting data
Biological Sampling:
There are various methods of collecting data, and In medical and biological research, data may be
the choice of method depends on the research collected through biological samples such as
question, the nature of the study, and the available blood, tissues, or DNA.
resources. Here are some common methods of
collecting data: Internet and Online Sources:
Researchers can collect data from online platforms,
websites, and social media. This method is Types of Sampling Methods:
particularly useful for studying online behaviors Different sampling methods, such as stratified
and trends. sampling, cluster sampling, systematic sampling,
convenience sampling, and others, may be
Sensor and Technology Data: employed based on the characteristics of the
Data can be collected through sensors, devices, or population and the
technology, such as wearable devices, satellite research objectives.
imagery, and monitoring systems.
Stratified Sampling:
Personal Documents and Diaries: In stratified sampling, the population is divided
Personal documents, diaries, and autobiographies into subgroups (strata) based on certain
can provide insights into individuals' experiences, characteristics, and then samples are randomly
thoughts, and emotions. selected from each stratum. This ensures
representation from various demographic or
Telephone and Mail Surveys: characteristic groups.
Surveys can be conducted over the phone or
through mailed questionnaires. These methods are Cluster Sampling:
useful when face-to-face interaction is not feasible. Cluster sampling involves dividing the population
into clusters and randomly selecting entire clusters
Sampling Methods: for the study. This method is efficient when it is
Techniques like random sampling, stratified impractical to sample individual elements.
sampling, and cluster sampling are used to select a
representative subset of the population for study. Sampling Frame:
A sampling frame is a list or representation of the
Sampling entire population from which the sample will be
drawn. It is essential that the sampling frame is
accurate and up-to-date to ensure the validity of the
Sampling is a critical aspect of data collection, sampling process.
particularly when it is not feasible or practical to
study an entire population. The goal of sampling is Sample Size:
to select a subset of individuals or elements from Determining an appropriate sample size is crucial.
the population that accurately represents its A sample that is too small may not accurately
characteristics. Here's an overview of the role of represent the population, while a sample that is too
sampling in data collection: large may be unnecessarily resource-intensive.
Statistical techniques can be employed to calculate
Representativeness: an optimal sample size.
The primary objective of sampling is to ensure that
the selected sample is representative of the larger Sampling Bias:
population. A representative sample increases the Researchers must be aware of potential sources of
likelihood that the findings from the sample can be bias in the sampling process. Bias occurs when
generalized to the entire population. certain members of the population are more or less
likely to be included in the sample, leading to
Random Sampling: skewed results. Efforts should be made to
Random sampling involves each member of the minimize bias and increase the generalizability of
population having an equal chance of being the findings.
selected. This method minimizes bias and ensures
that each subgroup of the population is Randomization:
proportionally represented in the sample. Random Randomization is a key principle in sampling. It
sampling is often considered the gold standard in ensures that each member of the population has an
sampling methods. equal chance of being included in the sample,
reducing the risk of systematic errors and
enhancing the external validity of the study.
 Population censuses are often conducted by
government agencies. Governments use census
Ethical Considerations: data for policy planning, resource allocation, and
Researchers must adhere to ethical standards in the
selection and treatment of study participants. various socio-economic initiatives.
Informed consent and confidentiality are essential
considerations in the sampling and data collection Regular Intervals:
process. Population censuses are typically conducted at
regular intervals, such as every ten years. This
Census and Sampling Method periodicity allows for tracking changes in the
population over time.
A census is a method of data collection that
involves gathering information from every Mandatory Participation:
individual or element in a defined population. It Participation in a census is usually mandatory for
aims to include every member of the population, individuals residing in the surveyed area.
leaving no one out. The term "census" is often Governments may enforce compliance through
associated with population censuses conducted by legal mechanisms to ensure a high response rate.
governments, where data is collected from every
resident in a country or a specified geographic Examples:
area. However, the concept of a census can extend National censuses conducted by governments, such
to various fields, including economics, agriculture, as the decennial census in the United States, the
and social sciences. national census in India, or other similar efforts
globally.
Census:
Sampling Method:
Definition:
A census involves collecting data from every Definition:
individual or element in a defined population. It Sampling involves selecting a subset of individuals
aims to include every member, leaving no one out. or elements from a larger population for the
purpose of making inferences about the entire
Scope: population. It is a more practical and cost-effective
The scope of a census is comprehensive, covering alternative to conducting a census.
a wide range of information about the entire
population. It provides a detailed snapshot of Scope:
various characteristics of individuals or elements Sampling focuses on a representative subset of the
within the population. population rather than the entire population. It is
based on the assumption that the characteristics of
Complete Enumeration: the sample can be generalized to the larger
A census seeks to enumerate the entire population, population.
leaving no gaps or exclusions. It is a method of
complete enumeration, aiming to gather data from Selection Process:
everyone. Sampling involves a systematic selection process,
whether it be random, stratified, cluster-based, or
Accuracy: another method. The goal is to ensure that the
Census data is considered highly accurate as it selected sample is representative of the larger
covers the entire population. It eliminates sampling population.
errors, providing a precise representation of the
population's characteristics. Resource Efficiency:
Sampling is generally more resource-efficient than
Government Involvement: conducting a census. It requires fewer resources in
terms of time, cost, and manpower, making it a
practical approach for large populations.

Accuracy:
While sampling introduces the potential for
sampling errors, statistical methods are used to
estimate and control these errors. With proper
sampling techniques, accurate and reliable results
can be obtained.

Feasibility:
Sampling is often more feasible, especially when
dealing with large populations. It allows
researchers to gather valuable information without
Probability Sampling
the challenges associated with complete
Probability sampling involves the use of random
enumeration.
selection, ensuring that every element in the
population has a known and nonzero chance of
Examples:
being included in the sample. This approach
Random sampling, stratified sampling, cluster
enhances the representativeness of the sample,
sampling, convenience sampling, and other
making it more likely to mirror the characteristics
sampling techniques used in various research
of the entire population.
studies, surveys, or experiments.
Simple Random Sampling:
Comparison of Census and Sampling Method
Explanation: Each individual in the population has
an equal chance of being selected.
Scope:
Example: If you have a list of 500 students, you
Census covers the entire population, while
could use a random number generator to pick 50
sampling focuses on a representative subset.
students for your sample.
Accuracy:
Stratified Sampling:
Census is considered highly accurate, while
Explanation: The population is divided into strata
sampling introduces the potential for sampling
(subgroups) based on certain characteristics, and
errors.
random samples are then drawn from each stratum.
Example: In a survey on academic performance,
Resource Intensity: Census can be resource-
you might stratify the population by grade levels
intensive, while sampling is generally more
and then randomly select students from each grade.
resource-efficient.
Systematic Sampling:
Feasibility:
Explanation: Every kth individual is selected from
Census may be impractical for large populations,
a list after randomly choosing a starting point.
while sampling is often more feasible.
Example: If you have a list of employees in
alphabetical order, you might select every 25th
Methods of sampling employee after randomly choosing a starting point.

Cluster Sampling:
Sampling methods are broadly categorized into
two main types: probability sampling and non-
Explanation: The population is divided into
probability sampling. Each type has its own set of
clusters, and a random sample of clusters is
techniques with distinct characteristics.
selected. All individuals within the chosen clusters
are included in the sample.
Example: If you're studying educational outcomes
in schools, you might randomly select a few
schools and include all students from those schools
in your sample.

Non-Probability Sampling
Non-probability sampling methods do not rely on
random selection, and individuals in the population
do not have a known chance of being included in
the sample. While non-probability sampling is less
rigorous in terms of representativeness, it is often
more practical and cost-effective.

Convenience Sampling:
Explanation: Participants are chosen based on their
easy accessibility.
Example: Surveying people in a shopping mall or
on the street.

Purposive Sampling:
Explanation: Participants are selected based on
specific characteristics relevant to the study.
Example: Recruiting individuals who have
experienced a specific event for an in-depth
interview.

Snowball Sampling:
Explanation: Initial participants refer others to
participate in the study.
Example: Studying a hard-to-reach population by
starting with a few individuals who then refer
others with similar characteristics.

Quota Sampling:
Explanation: Quotas are set for different subgroups
based on certain characteristics, and individuals are
selected to meet those quotas.
Example: Ensuring a certain number of
participants from different age groups in a survey
on consumer preferences.

Judgmental Sampling:
Explanation: Researchers use their judgment to
select participants who they believe are
representative of the population.
Example: A researcher selects individuals for an
interview based on their knowledge of the
community.