Data Analysis and Statistics · Term 4

Sampling Strategies

Distinguishing between biased and representative samples to ensure valid conclusions.

About This Topic

Sampling strategies in Grade 7 introduce students to the ethics and logic of data collection. In the Ontario curriculum, students learn to distinguish between biased and representative samples, ensuring that the conclusions they draw are valid. This topic is vital as it teaches students to be critical consumers of information. They explore how the size and selection process of a sample can completely change the results of a survey.

Students investigate various methods, such as random sampling, convenience sampling, and systematic sampling. They also look at how the phrasing of a question can introduce bias. By analyzing real-world Canadian examples, like census data or election polling, students see the importance of getting a 'fair' snapshot of a population. This topic is particularly effective when students can conduct their own mini-surveys and analyze the results. Students grasp this concept faster through structured discussion and peer explanation.

Learning Objectives

• Compare the reliability of conclusions drawn from random samples versus convenience samples for a given population.
• Analyze how specific wording in survey questions can introduce bias and affect data interpretation.
• Evaluate the potential risks of making generalizations about a large population based on a small sample size.
• Design a simple survey that uses a random sampling method to collect data on a specific topic.
• Identify examples of biased sampling in real-world scenarios and explain the source of the bias.

Before You Start

Data Collection and Organization

Why: Students need to be familiar with collecting and organizing data in tables before they can analyze different sampling methods.

Introduction to Probability

Why: Understanding basic probability concepts helps students grasp the idea of equal chances in random sampling.

Key Vocabulary

Sample	A small group of individuals or items selected from a larger group, used to represent the whole population.
Population	The entire group of individuals or items that a study is interested in understanding.
Random Sample	A sample where every member of the population has an equal and independent chance of being selected, minimizing bias.
Convenience Sample	A sample selected based on ease of access or availability, often leading to biased results.
Biased Sample	A sample that does not accurately represent the population due to a systematic error in the selection process or question design.
Representative Sample	A sample whose characteristics closely match those of the population it is drawn from, allowing for valid generalizations.

Active Learning Ideas

Simulation Game: The Jelly Bean Census

Each group gets a large jar of mixed-colour beans. They must test different sampling methods (e.g., taking the top 10 vs. shaking and picking 10) to see which method best predicts the actual percentages in the whole jar.

⏱ 40 min·Small Groups

Generate mission

Formal Debate: Survey Bias

Provide students with biased survey questions (e.g., 'Don't you agree that school should start later?'). Students work in teams to identify the bias, rewrite the question to be neutral, and debate why the original would produce 'bad' data.

⏱ 30 min·Small Groups

Generate mission

Gallery Walk: Sampling in the News

Post various news headlines that cite statistics. Students walk around and use a checklist to evaluate the likely sampling method and identify potential sources of bias (e.g., 'Only 10 people were asked' or 'The survey was only on Twitter').

⏱ 35 min·Pairs

Generate mission

Real-World Connections

Market researchers use sampling to gauge consumer preferences for new products. For example, polling a small group of shoppers at a specific mall about a new snack food might be a convenience sample, while a randomly selected group from a national customer database would be more representative.

Political pollsters use sampling to predict election outcomes. A pollster might conduct a random digit dial survey to contact a wide range of voters across different demographics, aiming for a representative sample of the electorate.

Scientists collecting environmental data, such as water quality in a lake, will use specific sampling strategies. Taking water samples only from the shore would be a convenience sample, whereas collecting samples from various depths and locations across the lake would provide a more representative picture of the water quality.

Watch Out for These Misconceptions

Common Misconception

A larger sample is always better, regardless of how it was picked.

What to Teach Instead

Students often think quantity beats quality. A peer discussion comparing a large biased sample (e.g., 1000 people at a hockey game) to a small random sample (e.g., 50 people from across Canada) helps them see that representation is more important than size.

Common Misconception

Convenience sampling is 'good enough' for most things.

What to Teach Instead

Students often want to just ask their friends. Having them compare their 'friend group' data to a random sample of the whole school helps them realize how much data can change when the sample isn't representative.

Assessment Ideas

Exit Ticket

Provide students with two scenarios: Scenario A describes a survey of students in the school cafeteria about their favorite lunch options. Scenario B describes a survey where every 10th student on the school's enrollment list is asked about their favorite lunch options. Ask students to identify which scenario is more likely to yield a representative sample and justify their answer.

Quick Check

Present students with three survey questions about school policies. For example: 'Do you agree that the school should have mandatory uniforms?' vs. 'Should the school consider implementing uniforms to improve safety and reduce distractions?' Ask students to identify which question is more neutral and explain how the phrasing might influence responses.

Discussion Prompt

Pose the following question: 'Imagine you want to know the average number of hours Grade 7 students in your city spend on homework each week. You can only survey 50 students. What are the risks of surveying only students in your own class? How could you try to get a more reliable sample, even with only 50 students?'

Frequently Asked Questions

What is a representative sample?

A representative sample is a small group that accurately reflects the characteristics of the larger population. For example, if a school is 50% boys and 50% girls, a representative sample should also be roughly 50/50.

How does bias affect data?

Bias makes data 'lean' in one direction, leading to incorrect conclusions. This can happen if the sample isn't random or if the questions are leading, which is why scientists and pollsters work hard to eliminate it.

How can active learning help students understand sampling?

Active learning, like the 'Jelly Bean Census,' allows students to see the immediate impact of different sampling methods. When they see that their 'convenience' sample was way off the actual total, the concept of bias becomes a practical problem to solve rather than just a definition to learn.

Why is the Canadian Census important for sampling?

The Census is one of the few times a government tries to reach the *entire* population rather than just a sample. It provides the 'gold standard' data that researchers use to see if their smaller samples are actually representative of the country.