Welcome to the World of Sampling!
In this chapter, we are exploring one of the most practical parts of Statistics: Sampling. Have you ever wondered how TV ratings are calculated or how scientists know what the most popular snack in the UK is? They don't ask every single person! Instead, they use a sample.
By the end of these notes, you’ll understand how to pick a group of people to represent a whole population, why it’s important to be fair, and how to spot mistakes that lead to wrong conclusions.
1. Population vs. Sample
To understand sampling, we first need to define the two main groups we are dealing with.
What is a Population?
The Population is the entire group that you want to find out about. It doesn’t have to be people—it could be all the lightbulbs made in a factory or all the trees in a forest.
What is a Sample?
A Sample is a smaller group picked from the population. We study the sample to learn about the whole population.
The Soup Analogy: Imagine you are cooking a massive pot of vegetable soup.
- The Population is the whole pot of soup.
- The Sample is the one spoonful you taste to see if it needs more salt.
- If that one spoonful (the sample) tastes good, you assume the whole pot (the population) is good!
Quick Review:
• Population: The whole group.
• Sample: The part of the group you actually test.
2. Why do we use Samples?
You might think, "Wouldn't it be more accurate to just ask everyone?" While that’s true, it’s usually impossible for three main reasons:
1. Time: It would take years to interview every person in a country.
2. Cost: Paying researchers to talk to everyone is too expensive.
3. Destructive Testing: If a factory wants to see how much pressure a glass bottle can take before it breaks, they can't test every bottle, or they’d have nothing left to sell!
Key Takeaway: Sampling is a faster and cheaper way to get a good idea of what a whole group is like.
3. Simple Random Sampling
To make sure our sample is fair, every member of the population should have an equal chance of being chosen. This is called Simple Random Sampling.
How to pick a Random Sample:
1. Assign a number to every member of the population.
2. Use a random number generator (like on a calculator or computer) to pick the numbers.
3. The people/items with those numbers become your sample.
Memory Aid: Think of a Lottery. Every ball in the machine has the same chance of being picked. That is a perfect random sample!
Don't worry if this seems a bit technical! The main thing to remember is that it isn't "random" if you just pick the people standing closest to you. That would be "convenience sampling," which isn't as fair.
4. Understanding Bias
A sample is Biased if it doesn't represent the whole population fairly. Bias leads to misleading results.
Common Causes of Bias:
• Location: If you want to know how people in the UK feel about football, but you only ask people outside a stadium on match day, your results will be biased!
• Sample Size: If the sample is too small (e.g., asking only 2 people), it won't represent the group well.
• Time of Day: If you survey people in a supermarket at 10:00 AM on a Tuesday, you probably won't get many views from people who work 9-to-5 jobs.
Common Mistake to Avoid: A lot of students think "random" means "haphazard." Picking the first 10 people you see is not random sampling; it is biased because people who arrive first might have different characteristics than those who arrive later.
Did you know? In 1936, a famous magazine predicted the wrong US President because they only polled people who had telephones and cars. At the time, only wealthy people had those, so the sample didn't represent the whole population!
5. Inferring Properties from a Sample
Once we have our sample data, we can "scale it up" to estimate the properties of the whole population. This is called Inference.
The Scaling Formula:
To estimate a total in the population, you can use this simple calculation:
\( \text{Estimated Total} = \frac{\text{Number in Sample with Characteristic}}{\text{Total Sample Size}} \times \text{Population Size} \)
Step-by-Step Example:
A school has 1000 students. A random sample of 50 students is asked if they like school dinners. 10 students say "Yes." Estimate how many students in the whole school like school dinners.
Step 1: Find the fraction of the sample that said yes: \( \frac{10}{50} = 0.2 \) (or 20%).
Step 2: Multiply this by the total population: \( 0.2 \times 1000 = 200 \).
Answer: We estimate that 200 students like school dinners.
Key Takeaway: The larger the sample, the more reliable your estimate for the population will be.
Quick Summary Checklist
✓ Population: The whole group you are interested in.
✓ Sample: A small section used to represent the group.
✓ Simple Random Sampling: Everyone has an equal chance of being picked.
✓ Bias: When a sample isn't representative (unfair).
✓ Inference: Using sample results to guess the population results.