Simple Random Sampling: A method where every member of the population has an equal chance of being selected for the sample
Population: The entire group being studied (1,200 students)
Sample: The subset of the population being surveyed (60 students)
- Assign a unique number to each student (1 to 1,200)
- Use a random number generator to select 60 unique numbers
- Contact the students corresponding to those numbers
- Survey only those selected students
Population: All 1,200 students in the school
Sample: 60 students to be surveyed
Give each student a unique identification number from 1 to 1,200
Generate 60 random numbers between 1 and 1,200 without replacement
Every student has an equal probability (60/1,200 = 1/20 = 0.05) of being selected
Sample: 60 students
Selection probability: 1/20 = 5%
The population is all 1,200 students. The sample is 60 students selected using simple random sampling, where each student has a 5% chance of being selected.
• Equal Probability: Every member has the same chance of selection
• No Bias: Selection is completely random
• Representative: Sample reflects population characteristics
Stratified Sampling: A method that divides the population into subgroups (strata) and samples proportionally from each subgroup
Stratum (plural: strata): Subgroup of the population with shared characteristics
Proportional Allocation: Sample size from each stratum proportional to stratum size in population
Sample size / Population size = 80 / 800 = 1/10 = 0.1 (10%)
Grade 6: (200/800) × 80 = 0.25 × 80 = 20 students
Grade 7: (300/800) × 80 = 0.375 × 80 = 30 students
Grade 8: (300/800) × 80 = 0.375 × 80 = 30 students
20 + 30 + 30 = 80 students ✓
Each grade contributes the same percentage to the sample as it does to the population
Grade 7: 30 students
Grade 8: 30 students
Total: 80 students
From Grade 6: 20 students, Grade 7: 30 students, Grade 8: 30 students. This maintains the same proportions as the population.
• Proportional Allocation: Sample from each stratum proportional to its size
• Representativeness: Ensures all subgroups are represented
• Accuracy: Provides more precise estimates than simple random sampling
Systematic Sampling: A method where every kth member of the population is selected after a random start
Sampling Interval (k): The fixed interval between selected members (k = N/n)
Random Start: A randomly chosen starting point between 1 and k
k = Population size ÷ Sample size = 2,000 ÷ 100 = 20
Choose a random number between 1 and 20 (inclusive)
Suppose we select 7 as our starting point
Select employees: 7, 27, 47, 67, 87, 107, 127, ..., up to the 100th employee
Every employee has an equal chance (100/2,000 = 1/20 = 5%) of being selected
Starting point: Random number 1-20
Select: Every 20th employee after start
The sampling interval is 20. Select a random starting point between 1 and 20, then select every 20th employee after that until 100 employees are selected.
• Fixed Interval: k = N/n determines selection pattern
• Random Start: Prevents bias in selection
• Systematic Pattern: Easy to implement and understand
Population: The entire group of individuals or items being studied.
Sample: A subset of the population selected for study.
Sampling Frame: A list of all members of the population from which the sample is drawn.
Sampling Bias: When some members of the population are more likely to be included in the sample than others.
Representative Sample: A sample that accurately reflects the characteristics of the population.
- Define the population: Clearly identify who or what is being studied
- Choose the sampling method: Based on population characteristics and research goals
- Determine sample size: Balance accuracy needs with available resources
- Implement the method: Follow the specific steps for the chosen method
- Collect data: Survey or measure the selected sample members
Cluster Sampling: A method where the population is divided into clusters, and entire clusters are randomly selected for the sample
Cluster: A naturally occurring group within the population (e.g., schools, neighborhoods)
Within-Cluster Homogeneity: Members within a cluster are diverse, but clusters themselves are similar
The 100 schools serve as the clusters in this sampling method
Randomly select 5 schools from the 100 schools using simple random sampling
Survey all students in the 5 selected schools
Average students per school: 50,000 ÷ 100 = 500 students per school
Expected sample size: 5 × 500 = 2,500 students
This method is cost-effective as researchers only need to visit 5 schools instead of 2,500 individual students
Survey all students in selected schools
Expected sample: 2,500 students
Randomly select 5 schools from the 100 schools using simple random sampling, then survey all students in those 5 schools. This will yield approximately 2,500 students.
• Cluster Selection: Randomly select entire clusters
• Cost Efficiency: Reduces travel and administrative costs
• Practicality: Useful when population is geographically dispersed
Method Appropriateness: Choosing the best sampling method based on research goals and population characteristics
Grade Distribution: Equal distribution across grades suggests homogeneity between groups
Research Goals: Whether to compare grades or get overall picture influences method choice
Population: 2,000 students (500 per grade 9-12)
Research goal: Study habits across all grades
Stratified sampling is most appropriate because:
- Grades are distinct subgroups with potentially different study habits
- Want to ensure all grades are represented
- Equal grade sizes make proportional allocation straightforward
Sampling ratio: 200/2,000 = 1/10 = 0.1 (10%)
Grade 9: (500/2,000) × 200 = 0.25 × 200 = 50 students
Grade 10: (500/2,000) × 200 = 0.25 × 200 = 50 students
Grade 11: (500/2,000) × 200 = 0.25 × 200 = 50 students
Grade 12: (500/2,000) × 200 = 0.25 × 200 = 50 students
Simple random: Might not represent all grades equally
Systematic: Could work but doesn't guarantee grade representation
Cluster: Would require selecting entire grades, which isn't practical here
Grade 9: 50 students
Grade 10: 50 students
Grade 11: 50 students
Grade 12: 50 students
Stratified sampling is most appropriate. Each grade should contribute 50 students to the sample (25% each), ensuring equal representation across all grade levels.
• Method Selection: Match method to research goals
• Stratified Appropriateness: Use when subgroups matter
• Proportional Allocation: Maintain population ratios in sample
Definitions of Key Concepts
Population: The entire group of individuals or items that researchers want to study and draw conclusions about. For example, all students in a school district.
Sample: A subset of the population that is actually studied. The sample should be representative of the population to allow valid inferences.
Sampling Frame: A list or database of all members of the population from which the sample is drawn. For example, a student roster.
Sampling Bias: Occurs when some members of the population have a higher probability of being selected than others, leading to unrepresentative results.
Representative Sample: A sample that accurately reflects the characteristics of the population, allowing for valid generalizations.
Core Rules and Principles
Simple Random Sampling:
- Every member of the population has an equal chance of being selected
- Selection is completely random with no pattern
- Most basic probability sampling method
Stratified Sampling:
- Population is divided into subgroups (strata) based on shared characteristics
- Sample is taken from each stratum proportional to its size in the population
- Ensures all subgroups are represented in the sample
Systematic Sampling:
- Every kth member of the population is selected after a random start
- k = N/n, where N is population size and n is sample size
- Easier to implement than simple random sampling
Cluster Sampling:
- Population is divided into clusters (natural groups)
- Entire clusters are randomly selected for the sample
- More cost-effective when population is geographically dispersed
Step-by-Step Methods
Method 1: Simple Random Sampling
- Create a sampling frame of all population members
- Assign a unique number to each member
- Use a random number generator to select the required number of members
- Survey only the selected members
Method 2: Stratified Sampling
- Divide the population into strata based on relevant characteristics
- Determine the sample size for each stratum using proportional allocation
- Within each stratum, perform simple random sampling
- Combine samples from all strata
Method 3: Systematic Sampling
- Calculate the sampling interval: k = N/n
- Select a random starting point between 1 and k
- Select every kth member after the starting point
- Continue until the desired sample size is reached
Method 4: Cluster Sampling
- Identify natural clusters in the population
- Randomly select a number of clusters
- Survey all members within the selected clusters
- Combine data from all selected clusters
Multiple Concrete Examples
Simple Random Example: Surveying 100 out of 1,000 employees by assigning each a number and using a random number generator.
Stratified Example: From a school with 600 students (200 in each grade), surveying 60 students with 20 from each grade.
Systematic Example: From a list of 10,000 customers, selecting every 50th customer after a random start to get 200 customers.
Cluster Example: From 100 classrooms, randomly selecting 10 classrooms and surveying all students in those rooms.
Tips, Tricks, and Common Pitfalls
Tips:
- Choose the sampling method based on research goals and population characteristics
- Use stratified sampling when subgroups are important to represent
- Use cluster sampling when the population is geographically dispersed
- Always ensure the sampling frame is complete and up-to-date
Common Mistakes:
- Using convenience sampling instead of proper probability methods
- Not maintaining equal probability in random methods
- Failing to maintain proportional allocation in stratified sampling
- Creating biased samples by using incomplete sampling frames
Key Notes for Memorization
- SIMPLE RANDOM: Equal probability for all members
- STRATIFIED: Proportional allocation from subgroups
- SYSTEMATIC: Every kth member after random start
- CLUSTER: Entire groups selected randomly
- SAMPLING INTERVAL: k = N/n
- PROPORTIONAL ALLOCATION: ni = (Ni/N) × n
When to Use Each Method
1. Simple Random: When population is homogeneous and small enough to manage
2. Stratified: When subgroups are important and you want to ensure representation
3. Systematic: When population is listed in a random order and large
4. Cluster: When population is geographically dispersed and cost is a concern
Questions & Answers
Question: In Exercise 2, why do we need to maintain the same proportions in the sample as in the population? Can't we just select 80 students randomly?
Answer: That's an excellent question! You can certainly select 80 students randomly using simple random sampling, but there are important trade-offs:
Advantages of stratified sampling:
- Guarantees representation from each grade level
- Provides more precise estimates for each subgroup
- Prevents the possibility of having too few students from a particular grade
Simple random sampling risks:
- Could accidentally select too many grade 7 students and too few grade 6 students
- Results might not reflect the actual distribution in the population
- Less reliable for subgroup analysis
Stratified sampling ensures your sample is truly representative of the population structure.
Question: What's the difference between stratified sampling and cluster sampling? They both divide the population into groups.
Answer: Great question! While both methods involve dividing the population into groups, they work differently:
Stratified Sampling:
- Population divided into strata based on shared characteristics
- Sample is taken from EACH stratum
- Goal: Ensure representation from all subgroups
- Example: Take 20% from grade 6, 20% from grade 7, etc.
Cluster Sampling:
- Population divided into clusters (natural groups)
- ENTIRE clusters are selected randomly
- Goal: Reduce costs and improve logistics
- Example: Randomly select 5 classrooms and survey ALL students in those classrooms
In stratified sampling, we sample FROM each group. In cluster sampling, we sample ENTIRE groups.
Question: Why can't we just ask our friends to participate in a survey? It would be much easier than using these complex methods.
Answer: What you're describing is called "convenience sampling," and while it's definitely easier, it leads to serious problems:
Problems with convenience sampling:
- Bias: Your friends may share similar characteristics and opinions
- Non-representative: May not reflect the broader population
- Invalid conclusions: Results cannot be generalized to the whole population
- Systematic error: Consistently misses certain perspectives
Why proper methods matter:
Probability sampling methods (like those we studied) ensure that every member of the population has a known chance of being selected. This allows us to make valid statistical inferences about the entire population based on our sample.
Think of it this way: If you only asked people at a basketball game about sports preferences, you'd get very different results than if you surveyed the general public!