Sampling- Sampling consists of obtaining information from a larger group or a universe.

Populations vs sample-

• A population or universe can be defined as any collection of persons or objects or event

in which one is interested.

• If we are studying the voting behaviour or political participation of the citizens of India
then all the citizens of India will come under population.

• By sample we mean the aggregate of objects, persons or elements, selected from the
universe. It is a portion or sub part of the total population. Example: In the above
example for assessing the voting behavior of the citizens, instead of getting information
from the whole population we can select a part of the population of the country and
then can ask their opinion.

Advantages of sampling

• Relatively low cost: Where conducting a census requires immense resources, a survey is
much more cost-effective.

• Fast and Convenient: It’s much easier to examine a selected sample of the population
than to take a census.

• The results are more accurate: If the sample is selected keeping in mind the right quality
checks, then its possible to achieve highly accurate results

Application- New product development, Customer satisfaction

Sampling unit- An element or a group of elements on which observations can be taken is called
a sampling unit.

• The objective of the survey helps in determining the definition of sampling unit. Like if
the objective is to determine the average income of the persons in the household and
suppose the income per person is not known but the income of the household is known,
then the sampling unit is household.

• Similarly if the average income of any person in the household is known then the
sampling units are the persons in the household.

Representative sample- A sample which contains all the salient (important )features of the
population.

• Example: if a population is having 30% males and 70% females, then a representative
sample should have nearly 30% males and 70% females.
 Sampling frame- List of all the units of the population to be surveyed constitutes the
sampling frame. All the sampling units in the sampling frame have identification particulars.
Example: All the students in a particular university listed along with their roll numbers
constitute the sampling frame.

Advantages of Sampling over Complete Enumeration-

 Reduced cost and wider scope: Sampling costs less than full enumeration.

 Extra data, like health practices during a health survey, can be gathered at minimal
additional cost.

 Greater Accuracy, skilled interveiewer

 Urgent information required- The data from a sample can be quickly summarised.
 Feasibility- Conducting the experiment on smaller number of units, particularly when
the units are destroyed is more feasible.


Organisation of work

It's easier to manage and organize data from smaller groups than from an entire population. For
example, drawing small samples from each city is more efficient and accurate than sampling the
whole state at once. Better organization leads to better data and more accurate statistical
inferences.

Ways to ensure representativeness

 Random sample or probability sample: The selection of units in the sample from a
population is governed by the laws of chance or probability. The probability of selection
of a unit can be equal as well as unequal

• Non random sample or purposive sample: The selection of units in the sample from
population is not governed by the probability laws.
For example the units are selected on the basis of personal judgement of the
surveyor.
The process volunteering to take some medical test or to drink a new type of
coffee also constitute the sample on non-random laws.

Population

• Finite- If the number of objects or units in the population is countable, it is said to be a

finite population.
 • Infinite Population- If the number of objects or units in the population is infinite, it is
said to be an infinite population

Target Population- A finite or infinite population about which we require information is called
target population. For example, all 10 year old children in the United States of America.

Study population- This is the basic finite set of individuals we intend to study.

Properties-

The population must be clearly defined to avoid ambiguity about unit inclusion, covering all
relevant details. For example, in a math achievement survey, the researcher should specify
student age or grade, school type, location, and academic year.

The units in the population must be clearly defined; without this, a researcher can't accurately
select a sample or draw valid inferences.

Characteristic of a Sample

In inferential research, the sample must be unbiased and representative, reflecting the
population's characteristics in the same proportion or intensity. (Free from bias)

A clear sampling plan is essential. If properly followed, it ensures that results from repeated
samples closely match those from studying the entire population.

A good sampling plan ensures that repeated samples yield results close to those from the full
population. For example, with 100 samples, 95 may give correct estimates. A representative
sampling plan increases the likelihood of selecting a sample that reflects the population by
including and balancing diverse elements.

Principal Steps in a Sample Survey

1. Objective of Survey
Clearly define the purpose of the survey. Objectives must align with available resources—time,
money, and manpower.

2. Define the Population

Specify the target population in unambiguous terms. This ensures clarity about who or what is
being studied.
3. Frame and Sampling Unit
Divide the population into distinct, non-overlapping sampling units that cover the entire
population.

4. Data to be Collected
Focus only on data relevant to the survey objectives. Avoid collecting unnecessary or irrelevant
information.

5. Questionnaire or Schedule
Design a brief, clear, and unambiguous questionnaire or schedule that avoids confusion and is
easy for respondents to answer.

6. Method of Collecting Information

 Interview Method: The interviewer personally asks questions.

 Mailed Questionnaire: Sent to respondents with a deadline for return.

7. Non-Respondents
Account for non-responses due to absence or refusal. Apply proper procedures to handle these
cases.

8. Selection of Proper Sampling Design

Define the sample size, selection method, and estimation approach to ensure
representativeness and reliability.

9. Summary and Analysis of Data

Scrutinize, edit, and tabulate the data. Use proper coding and perform statistical analysis for
accurate results.

Errors

Sampling Errors: Originate from using a sample instead of the entire population to
estimate parameters and make inferences. Absent in complete enumeration.

Reasons for Sampling Errors:

 Faulty sample selection: Bias from non-random sampling techniques (e.g., purposive
sampling). Avoid by using suitable random sampling.

 Substitution: Bias introduced by replacing a difficult-to-enumerate unit with a

convenient one.
  Faculty demarcation of sampling units: Bias in area surveys due to subjective decisions
on borderline cases.

 Improper choice of statistics: Using a biased statistic (e.g., sample variance as a direct
estimate of population variance).

Key Observation: Increasing the sample size generally reduces sampling error.

Non-Sampling Errors: Arise during data observation, ascertainment, and processing. Present
in both complete censuses and sample surveys.

Major Reasons for Non-Sampling Errors:

1. Faulty Planning or Definitions:

o Inadequate or inconsistent data specifications relative to survey objectives.

o Errors in unit location, characteristic measurement, recording, and ill-designed

questionnaires.

o Lack of trained investigators and adequate supervision.

2. Response Errors: Errors due to respondent answers.

o Accidental: Misunderstanding questions.

o Prestige Bias: Overstating positive attributes or understating negative ones due to

pride.

o Self-Interest: Providing incorrect information to protect personal interests.

o Bias due to Interviewer: Influencing responses through questioning or recording

methods.

3. Non-Response Biases: Occur when full information isn't obtained from all selected units
due to unavailability, inability, or refusal to answer. Leads to bias by excluding a
potentially distinct population segment.

4. Errors in Coverage: Result from imprecise survey objectives, leading to:

o Inclusion of ineligible units.

o Exclusion of eligible units.

5. Compilation Errors: Mistakes during data processing (editing, coding, tabulation,

summarization). Controllable through verification and consistency checks.
 6. Publication Errors: Errors in presenting and printing results, including proofing errors and
failure to highlight data limitations.

Additional Points:

 Sample surveys can also have non-sampling errors due to defective frames and faulty
unit selection.

 Non-sampling errors are often more serious in complete censuses due to the larger
scale, making quality control harder.

 Sampling error typically decreases with larger sample size, while non-sampling error may
increase.

 In many cases, the non-sampling error in a census can outweigh the combined sampling
and non-sampling errors of a well-conducted sample survey, making the latter
preferable.

Types of sampling

• Probability sampling is defined as a sampling technique in which the researcher chooses

samples from a larger population using a method based on the theory of probability.

• Here every unit of the population has some probability of being included in the sample.

• If the probability is equal for every unit, then it is called equal probability of selection
scheme(EPS).

Simple Random Sampling, Stratified Random Sampling, Systematic Sampling, Cluster Sampling
Simple Random Sampling:

Merits:

 Eliminates Bias: Random selection gives each unit equal chance, removing subjectivity
and personal bias. More representative of the population.

 Simple Procedure: Easy to select a sample and less prone to bias in the selection
process.

 Cost-Effective (Sometimes): Desirable when data collection costs are not excessively
high.

Demerits:

 Requires Updated Frame: Needs a complete and current list (catalog) of the entire
population, which is often unavailable. This limits its applicability.

 Geographic Dispersion: Can result in a geographically spread-out sample, significantly

increasing the time and cost of data collection.

Stratified Sampling:

Stratification- Population is divided into homogeneous groups (strata) based on auxiliary

information relevant to the study.

Merits:
  Ensures Representation: Guarantees desired representation of all strata in the sample,
preventing over/under-representation or exclusion seen in unstratified random
sampling. Provides a more representative cross-section.

 Increased Precision: Yields estimates with higher precision compared to simple random
sampling.

 Most Efficient (Often): Frequently considered the most efficient sampling method due to
controlled representation.

Problems:

 Principle of Stratification: Determining the appropriate criteria for dividing the

population into meaningful and homogeneous strata.

 Allocation of Sample Size: Deciding how to distribute the total sample size effectively
across the different strata.

 Number of Strata (k): Determining the optimal number of strata to use.

Systematic Sampling:

 Definition: Selects the first unit randomly from an ordered list, then selects subsequent
units at a fixed interval (k). Requires a complete and up-to-date list.

 Sampling Interval (k): Calculated as N/n (Population size / Sample size).

 Random Start (i): A random number between 1 and k, determining the first selected
unit. The sample consists of units i, i+k, i+2k, ..., i+(n-1)k.

Merits:

 Operational Convenience: Easier to implement than simple or stratified random

sampling.

 Efficiency: Less time and work involved in sample selection.

 Even Spread: Tends to yield a sample evenly distributed across the population.

Demerits:

 Not Truly Random: Systematic samples are generally not considered truly random
because the frame is rarely perfectly random.

 Potential for Bias (Periodicity): Can produce highly biased estimates if the sampling
interval (k) aligns with a periodic pattern in the population frame.
Cluster sampling-

Cluster Sampling:

 Definition: The population is divided into groups (clusters), and then a random sample
of these clusters is selected. All or a sample of elements within the selected clusters are
included in the final sample.

 Sampling Unit: The cluster (a group of elements), not individual elements.

  Rationale: Used when a complete list of individual population elements is unavailable or
impractical to obtain.

 Example: Sampling households (clusters) in a city when a list of all individuals is not
available.

Merits:

 Resource Efficient: Requires fewer resources (time, cost, administrative, travel) for the
sampling process compared to simple random or stratified sampling because only
certain groups are selected.

Demerits:

 (Repeated Merit - Likely a Typo in the Original Text): Since cluster sampling selects only
certain groups from the entire population, the method requires fewer resources for the
sampling process which makes it generally cheaper than simple random or stratified
sampling as it requires fewer administrative and travel expenses. (This appears to be a
repetition of the merit and should likely be a demerit related to potential lower precision
or higher sampling error if clusters are not homogeneous).

Likely Demerit (Inferred):

 Lower Precision (if clusters are heterogeneous): If the elements within a cluster are not
very similar (heterogeneous), the sample may not be as representative of the overall
population as a simple random or stratified sample of the same size, potentially leading
to lower precision and higher sampling error.

Non-Probability Sampling:

 Definition: Sampling techniques where the researcher selects samples based on their
subjective judgment, rather than random selection. Less stringent and relies heavily on
researcher expertise.

 Types:

o Convenience Sampling: Selecting participants who are easily accessible to the

researcher.

 Example: Choosing the first five names of students from an attendance

register.

o Quota Sampling: Creating a sample that reflects the proportions of different

subgroups within the population based on known characteristics.
  Example: Ensuring a sample of 100 individuals has 40 women and 60
men if the population has those proportions. Sampling continues until
these quotas are met.

o Snowball Sampling: Used when the target population is difficult to locate.

Initial participants are asked to refer other individuals who fit the study
criteria. Works like a referral program to build the sample.

 Example: Surveying homeless persons. Identifying a few individuals and

asking them to help locate others in the same situation.

Auxiliary Information:

 Definition: Additional data collected alongside the primary information of interest. The
variable providing this extra data is the auxiliary variable.

 Purpose: Enhances the efficiency of estimators for population parameters.

  Collection: Can be gathered from past data, pilot surveys, or concurrently with the main
survey (either without additional cost or with dedicated resources).

Examples:

 Estimating Income: If direct income reporting is unavailable, income can be estimated

using expenditure and savings (Income = Expenditure + Savings).

 Estimating Crop Yield: Previous year's average wheat yield for the same plot can serve as
auxiliary information.

Stages of Use:

1. Designing/Pre-selection Stage: Used to structure the sample, such as in stratifying the

population or forming clusters.

2. Selection Stage: Used to select units with varying probabilities based on the auxiliary
variable, with or without replacement.

3. Estimation Stage: Used to formulate more efficient estimators like ratio, product, and
regression estimators.