Business Research Methods (KMBN 203)

Unit 4

Sampling
Basic Concepts
1. Defining the Universe
The universe, also known as the population in statistical terms, refers to the entire group of
individuals, items, or elements that share a common set of characteristics and are of interest to
the researcher.
Characteristics:
• Well-defined: The universe should be clearly defined based on the research objectives
and scope.
• Homogeneous: Elements within the universe should share common characteristics
relevant to the research.
• Accessible: It should be feasible to access and study elements within the universe.
Example: In market research, the universe might be all potential customers of a particular
product or service within a specific geographical area.
2. Concepts of Statistical Population
A statistical population is a subset of the universe that represents the specific group from which
the sample will be drawn. It shares characteristics with the universe and serves as the basis for
making generalizations.
Characteristics:
• Representativeness: The population should accurately reflect the characteristics of the
universe.
• Measurable: It should be possible to define and measure the attributes of the
population.
• Clear Boundaries: There should be clear criteria for inclusion or exclusion from the
population.
Example: If the universe is all adults in a country, the statistical population might be adults
aged 18-65 residing in urban areas.
3. Sample
A sample is a subset of the statistical population selected for study and analysis. It is chosen to
represent the larger population accurately and efficiently.
Characteristics of a Good Sample:
• Representativeness: The sample should accurately reflect the characteristics of the
population from which it is drawn.

1
 • Randomness: Selection should be based on random or probability-based methods to
minimize bias.
• Adequacy: Sufficient sample size to ensure reliable statistical analysis and valid
conclusions.
• Feasibility: Practical considerations such as time, resources, and accessibility of the
sample.
• Ethical Considerations: Ensuring voluntary participation and confidentiality of
respondents.
Characteristics of a Good Sample
Representativeness:
The sample should accurately mirror the characteristics of the population in terms of
demographics, behaviors, and other relevant factors. This ensures that findings from the sample
can be generalized to the population.
Randomness:
Random sampling methods reduce bias and ensure that every element in the population has an
equal chance of being included in the sample. This increases the likelihood that the sample is
unbiased and representative.
Adequacy:
The sample size should be large enough to provide sufficient statistical power for analysis.
Adequate sample sizes vary depending on the research design, population variability, and
desired confidence level.
Feasibility:
Practical considerations such as time, budget, and resources influence the sampling method
chosen. Researchers must balance the ideal sampling technique with practical constraints to
achieve the best possible representation of the population.
Ethical Considerations:
Participants should be selected and treated ethically, ensuring informed consent, privacy, and
confidentiality. Protecting participants' rights and well-being is crucial in all stages of sampling
and research.
Sampling Frame
A sampling frame is a list or source from which a sample is drawn. It serves as a practical
approach to identify and select potential participants or elements for inclusion in the sample.
Practical Approach for Determining the Sample Frame
• Census Lists: Utilizing official lists or databases that comprehensively cover the target
population (e.g., voter registration lists, customer databases).
• Sampling from Organizations: Accessing organizational records or membership lists
relevant to the study (e.g., employee lists, professional associations).

2
 • Geographical Areas: Sampling based on defined geographical units (e.g., census
tracts, postal codes) relevant to the study population.
• Virtual Sampling Frames: Using online platforms, databases, or social media profiles
to identify and recruit participants for online surveys or studies.
Sampling Errors
Sampling errors occur when the sample selected from the population does not accurately
represent the entire population due to random variation in the selection process.
Types:
• Random Sampling Error: Variability that occurs by chance when drawing a sample.
• Systematic Sampling Error: Bias introduced by the sampling method or approach
used.
Methods to Reduce Sampling Errors:
• Random Sampling: Employing random sampling techniques minimizes systematic
biases and ensures representativeness.
• Increasing Sample Size: Larger sample sizes reduce random variability and increase
the accuracy of estimates.
• Stratified Sampling: Dividing the population into subgroups (strata) and sampling
proportionately from each stratum reduces variability and improves representation.
Non-Sampling Errors
Non-sampling errors occur due to factors other than the sampling process itself, affecting the
accuracy and reliability of research findings.
Types:
• Coverage Error: When the sampling frame does not accurately represent the target
population.
• Measurement Error: Errors in data collection, recording, or analysis.
• Non-Response Bias: Differences between respondents and non-respondents that affect
the validity of results.
Methods to Reduce Non-Sampling Errors:
• Improving Data Collection Methods: Training interviewers, using standardized
protocols, and validating measurement instruments.
• Pilot Testing: Conducting a pilot study to identify and address potential issues before
full-scale data collection.
• Minimizing Non-Response: Implementing effective strategies to encourage
participation and reduce non-response bias.
Sample Size Constraints
Sample size constraints refer to limitations imposed by practical considerations such as time,
budget, and resources when determining the size of the sample.
Considerations:

3
 • Statistical Power: Ensuring the sample size is adequate to detect meaningful effects or
differences.
• Precision: Balancing the trade-off between sample size and precision of estimates.
• Logistics: Practical considerations such as accessibility and feasibility of data
collection.
Non-Response
Non-response occurs when selected individuals or units in the sample do not participate or
provide incomplete responses, potentially biasing the results.
Causes:
• Lack of interest or motivation to participate.
• Inability to contact or reach selected respondents.
• Privacy concerns or distrust in research.
Strategies to Address Non-Response:
• Pre-survey Communication: Clearly communicating the purpose and importance of
the study.
• Incentives: Offering incentives to encourage participation.
• Follow-up: Implementing follow-up strategies to reach non-respondents.
• Weighting: Adjusting weights for respondents to correct for non-response bias.
Probability Sampling Methods
Probability sampling methods are techniques that involve random selection of elements from
a population, ensuring that every element has a known and non-zero chance of being selected.
These methods allow researchers to make statistical inferences about the population based on
the sample. Here are the key types of probability sampling methods:
1. Simple Random Sample
A simple random sample involves selecting elements from a population entirely by chance,
where each element has an equal probability of being chosen.
Process:
• Assign a unique identifier (e.g., numbers) to each element in the population.
• Use a random number generator or table to select elements without any systematic
pattern.
Advantages:
1. Simple to implement.
2. Ensures every element has an equal chance of selection.
Disadvantages:
• Challenging to ensure complete randomness without bias.
2. Systematic Sample

4
A systematic sample involves selecting elements from a population at regular intervals after an
initial random starting point is determined.
Process:
• Calculate the sampling interval (total population size divided by desired sample size).
• Randomly select a starting point between 1 and the sampling interval.
• Select every kth element from the population using the determined interval.
Advantages:
• Easier and more practical than simple random sampling.
• Ensures every element has an equal probability of selection.
Disadvantages:
• Vulnerable to periodicity if there is a pattern in the population list.
3. Stratified Random Sample
A stratified random sample divides the population into homogeneous subgroups (strata) based
on certain characteristics, and then samples are randomly selected from each subgroup.
Process:
• Identify relevant strata based on demographic, geographic, or other characteristics.
• Randomly sample from each stratum in proportion to its size or importance.
Advantages:
• Ensures representation from all important subgroups.
• Increases precision and reduces sampling error compared to simple random sampling.
Disadvantages:
• Requires prior knowledge of population characteristics to define strata effectively.
4. Area Sampling (or Cluster Sampling)
Area sampling involves dividing the population into clusters or geographical areas, and then
randomly selecting clusters to sample from. All elements within selected clusters are included
in the sample.
Process:
• Identify clusters based on geographical boundaries (e.g., neighborhoods, city blocks).
• Randomly select clusters and sample all elements within the selected clusters.
Advantages:
• Cost-effective and efficient for large geographical areas.
• Practical when a sampling frame for individual elements is not available.
Disadvantages:
• Potential for increased sampling variability compared to other methods.

5
 • Requires accurate and up-to-date geographical information.
Non-Probability Sampling Methods
Non-probability sampling methods involve selecting elements from a population in a non-
random manner, meaning that not every element has a known or equal chance of being selected.
These methods are often used when it's difficult to obtain a complete list of the population or
when practical considerations limit the use of probability sampling methods. Here are the key
types of non-probability sampling methods:
1. Judgment Sampling
Judgment sampling involves the researcher's subjective judgment or expertise in selecting
elements that are believed to be representative of the population.
Process:
• The researcher selects individuals or cases based on their knowledge and judgment of
the population characteristics.
• Typically used in exploratory research or when specific expertise is required.
Advantages:
• Useful when time and resources are limited.
• Allows for selection of knowledgeable or typical cases.
Disadvantages:
• Susceptible to bias and researcher subjectivity.
• Limited generalizability to the broader population.
2. Convenience Sampling
Convenience sampling involves selecting elements that are readily available and accessible to
the researcher.
Process:
• The researcher selects individuals or cases based on convenience or accessibility.
• Commonly used in pilot studies, quick surveys, or when access to the population is
challenging.
Advantages:
• Quick and easy to implement.
• Cost-effective and practical in certain situations.
Disadvantages:
• Prone to selection bias (e.g., over-representing certain groups).
• Limited generalizability to the entire population.
3. Purposive Sampling

6
Purposive sampling involves selecting elements based on specific criteria or characteristics
relevant to the research objectives.
Process:
• The researcher deliberately selects individuals or cases that meet predefined criteria or
characteristics.
• Used when specific expertise, characteristics, or conditions are of interest.
Advantages:
• Allows for targeted selection of cases relevant to the research.
• Useful in qualitative research or when specific traits are important.
Disadvantages:
• Potential for bias if criteria are not well-defined or applied consistently.
• Limited generalizability beyond the selected criteria.
4. Quota Sampling
Quota sampling involves selecting individuals or cases based on pre-defined quotas or
proportions to ensure representation of certain characteristics.
Process:
• The researcher sets quotas for different categories (e.g., age, gender, income) based on
their prevalence in the population.
• Selects individuals or cases to fill each quota until the desired sample size is reached.
Advantages:
• Ensures representation of specific subgroups or characteristics.
• More flexible than stratified sampling in terms of implementation.
Disadvantages:
• Vulnerable to bias if quotas are not properly defined or adhered to.
• Difficult to generalize findings beyond the selected quotas.
5. Snowball Sampling
Snowball sampling involves selecting initial participants who then refer or nominate other
potential participants, creating a chain or "snowball" effect.
Process:
• The researcher identifies and recruits initial participants who meet the criteria.
• Participants then refer others who meet the criteria, expanding the sample size
iteratively.
Advantages:
• Useful when the population is difficult to identify or access.
• Effective for studying hidden or hard-to-reach populations.

7
Disadvantages:
• Potential for bias if initial participants are not representative.
• Difficult to control the size and composition of the sample.
Determining the Size of the Sample
Practical Considerations in Sampling and Sample Size Determination:
• Research Objectives: Clear understanding of what the study aims to achieve and the
level of precision required.
• Population Characteristics: Knowledge of the population size, variability, and
diversity.
• Resource Constraints: Consideration of time, budget, and availability of resources for
data collection.
• Statistical Power: Ensuring the sample size is sufficient to detect meaningful effects
or differences.
• Ethical Considerations: Ensuring the rights and well-being of participants are
protected throughout the research process.
Sample Size Determination:
• Statistical Methods: Using formulas or software to calculate sample sizes based on
population size, desired confidence level, and margin of error.
• Pilot Studies: Conducting pilot studies to estimate variability and refine sample size
calculations before full-scale data collection.
• Consultation: Seeking advice from experts or statisticians to ensure appropriate
sample size determination based on study design and objectives.