Unit 3 Lesson 2: Sampling Methods and Bias
Sampling Methods and Bias
Unit 3, Topics 3.3–3.4: Sampling Methods and Bias
Overview
This lesson explores different ways to select a sample from a population to gather reliable data, along with common types of bias that can make results inaccurate or unrepresentative. Sampling methods ensure the sample reflects the whole group, such as simple random sample (SRS, where every individual has an equal chance of being chosen), stratified sampling (dividing the population into subgroups by key traits like age or grade and sampling from each), cluster sampling (randomly selecting entire groups, like classes or neighborhoods), and systematic sampling (picking every kth person from a list, such as every 10th student).
Example: For a school survey on lunch preferences, SRS randomly picks 100 from 1,000 students using numbers; stratified by grade samples 25 from each year to balance freshmen and seniors.
Bias occurs when the sample doesn't represent the population fairly, including undercoverage (missing entire groups, like only surveying city residents in a rural area), nonresponse bias (people who don't reply differ from those who do, such as busy parents skipping surveys), voluntary response bias (people self-select to participate, like only fans responding to a poll), and response bias (answers are influenced by how questions are worded or by social desirability, like underreporting bad habits).
Example: In a health poll, undercoverage skips low-income areas without phones, missing diverse views; voluntary response from fitness enthusiasts overstates exercise rates.
Context, such as the overall population and how the sample is taken, is crucial because it determines if the results can be generalized (e.g., polling only online users in a diverse town biases against those without internet access, leading to skewed views on technology preferences). Effective sampling boosts the reliability of your findings, while bias limits how well they apply to everyone.
Example: A town election poll using stratified by neighborhood generalizes better than voluntary online responses, which bias toward urban tech users and fail to represent rural voters' priorities.
Assignment:
Part 1: Guided Practice Activity
Consider the example of polling a school population of 1,000 students to learn their favorite subjects (e.g., math vs. art). Your goal is to survey 100 students fairly. Follow the tasks below to describe methods and spot potential biases.
Example Scenario: You plan to ask students about their subject preferences during lunch breaks.
Tasks:
- Describing Sampling Methods:
- Provide a clear description of each method: simple random sample (SRS), stratified sampling, cluster sampling, and systematic sampling (e.g., choosing every 10th student from an alphabetical list).
Example: SRS gives every student an equal chance using random numbers, avoiding favoritism; stratified divides by grades and samples proportionally from each for balance; cluster picks random classes and surveys all inside for efficiency; systematic selects every 10th from a list for simplicity. - Write 1-2 full sentences explaining which method would best fit this school poll and why.
Example: Stratified sampling by grade would ensure fair representation across all years, preventing overemphasis on seniors who might prefer advanced subjects like calculus over art, making results more reliable for the whole school. - Extra Practice: For a city-wide health survey on exercise habits, describe two suitable sampling methods and briefly explain why each works well in that context.
- Provide a clear description of each method: simple random sample (SRS), stratified sampling, cluster sampling, and systematic sampling (e.g., choosing every 10th student from an alphabetical list).
- Identifying Bias Types:
- Identify and define each bias type with a specific example related to the school poll: undercoverage, nonresponse (e.g., shy students avoiding the survey), voluntary response, and response bias.
Example: Undercoverage misses commuters who skip lunch surveys, underrepresenting their preferences; nonresponse from quiet students skews toward vocal groups; voluntary response favors club members who opt in; response bias leads to polite answers on sensitive topics like disliked subjects. - Write 1-2 full sentences assessing the impact of one bias on the results.
Example: Voluntary response bias from math fans could overstate math's popularity by 20-30%, leading to misleading conclusions about school curriculum needs and wasting resources on unnecessary math expansions. - Extra Practice: For your city health survey, identify two potential bias types and explain their specific impacts in that setting.
- Identify and define each bias type with a specific example related to the school poll: undercoverage, nonresponse (e.g., shy students avoiding the survey), voluntary response, and response bias.
- Reflection:
- Evaluate one sampling method in the school poll context, including how to implement it and its strengths/weaknesses.
Example: SRS for the school poll reduces bias by giving every one of the 1,000 students an equal chance through a random number generator, but it requires a complete list and time to implement; its strength is broad coverage, though weakness is potential for uneven grade representation. - Write 2-3 full sentences explaining how the sampling design overall affects the reliability and generalizability of the results.
Example: A well-designed method like stratified sampling boosts reliability by mirroring the school's diversity in grades and backgrounds, making findings more applicable to the entire population for balanced curriculum planning. In contrast, cluster sampling of only a few hallways is cheap and quick but risks similar opinions from close friends, limiting generalizability and potentially overlooking preferences from remote parts of the school.
- Evaluate one sampling method in the school poll context, including how to implement it and its strengths/weaknesses.
Part 2: Independent Practice
Imagine polling 200 voters out of a town of 5,000 to understand their election preferences (e.g., policy priorities like education vs. economy). Follow the tasks to design and critique your approach.
Tasks:
- Describe two sampling methods that would work well for this poll, including a step-by-step explanation of how to implement each.
- Identify two bias types that could occur in this poll and explain their specific impacts on the results.
- Write 2-3 full sentences evaluating one method in the town context, discussing how it affects the reliability and generalizability of the findings.
- Extra Activity: Invent your own poll scenario. Describe one sampling method, identify one bias, and evaluate its impact with context.
Homework Assignment
- Think of a real-world poll you could conduct (e.g., favorite foods in your family or neighborhood). Describe one sampling method in detail, identify at least one potential bias, evaluate its impact with specific context, and explain how the design influences the reliability of your results to share next class.