Cal Poly Pomona
Random Samples
Selection of students from a class
Scenario: 33 students in the classroom.
Objective: Choose a random sample of 4 students.
Method: Use of random number generator from 1 to 33.
Random number generator precautions
Issue: Avoiding duplicates (e.g., student 11 appearing twice).
Solution: Use "random integer no repeats" function to ensure unique selections.
Spreadsheet approach
Process: Names assigned random numbers, ranked by size.
Example: Identifying top 4 (like Kevin, Rachel, Leonardo, Russell) based on random numbers.
Considerations in Sampling
Potential sampling bias
Concern: Small samples may not represent larger populations, e.g., all freshmen selected.
Implication: Could skew results based on the demographic representation.
Stratified Sampling
Definition
Concept: Population divided into strata (groups) based on a variable (e.g., grade level).
Application: Select members from each stratum randomly.
Example scenario
Setup: 14 freshmen and 19 non-freshmen in the class.
Selection method: Randomly choose 2 from each group to ensure balanced representation.
Further stratification potential
Greather granularity: Separate by class year (freshmen, sophomores, juniors, seniors) and select from each.
Cluster Sampling
Definition
Concept: Population is grouped into clusters; random clusters selected.
Application: All members from selected clusters comprise the sample.
Example scenario
Setup: Classes as clusters; select 4 classrooms and sample all students within them.
Advantages and disadvantages
Efficiency: Easier data collection from fewer clusters.
Risk: Potential for bias if clusters share similar demographics.
Systematic Sampling
Definition
Process: Choose the first individual randomly, then select every kth individual afterward.
Example: If k=8, after selecting an individual, every 8th student in line is included.
Convenience Sampling
Definition
Concept: Sample selected based on ease of access.
Examples:
First 100 students in the cafeteria.
Students in a specific teacher's classes.
Disadvantages
Bias: Likely unrepresentative of the population.
Implication: Results may be skewed due to demographic or situational factors.
Terminology
Parameter: A value that describes the entire population.
Example: Average enjoyment rating for the entire population.
Statistic: A value describing a sample.
Example: Average rating for the specific sample chosen.
Conclusion
Importance of selecting appropriate sampling methods to reduce bias and ensure representative samples for accurate results.
Homework and quizzes related to sampling methods and statistics upcoming.