Welcome to the World of Sampling!
Ever wondered how pollsters predict election results or how scientists decide if a new medicine works without testing every single person on Earth? That is the power of Statistical Sampling. In this chapter, you will learn how to pick a small group of people (or things) to represent a much larger group. It is a bit like tasting a single spoonful of soup to see if the whole pot needs more salt—you don't need to eat the whole pot to find out!
By the end of these notes, you will understand the different ways we collect data and why choosing the right method is the secret to getting accurate results.
1. Population vs. Sample
Before we start picking groups, we need to know the "who" and the "what."
The Population: This is the entire group of people or objects you are interested in. If you want to know the average height of students in your school, the population is every single student in that school.
The Sample: This is a subset (a smaller part) of the population that you actually collect data from. If you ask 30 students in the canteen their height, those 30 students are your sample.
The Census: This is when you observe or measure every single member of a population. In the UK, the government conducts a census every 10 years.
Pros and Cons: The Big Trade-off
Why wouldn't we just do a census every time? It seems more accurate, right? Well, it is, but it's not always practical.
Census Advantages:
- It gives a completely accurate result.
- No one is left out.
- It is expensive and time-consuming.
- It can be difficult to reach everyone.
- If the testing involves destroying the items (like testing how long a lightbulb lasts), a census would leave you with nothing to sell!
Sample Advantages:
- Much quicker and cheaper.
- Fewer people are needed to collect the data.
- The data may not be perfectly accurate.
- The sample might not represent the whole population (this is called bias).
Quick Takeaway
A census is great for accuracy but bad for your wallet and schedule. A sample is fast and affordable but carries the risk of not being a perfect "mini-me" of the population.
2. Random Sampling Techniques
To make a sample fair, every member of the population should have a chance of being picked. This is where Random Sampling comes in. To do this, you usually need a Sampling Frame—a list of everyone in the population (like a register or a phone book).
Simple Random Sampling
This is the purest form of random sampling. Imagine putting everyone’s name in a giant hat and pulling them out. In the modern world, we use random number generators.
How to do it:
- Assign a unique number to every person in your sampling frame.
- Use a calculator or computer to generate random numbers.
- Pick the people whose numbers match the ones generated.
Systematic Sampling
Think of this as sampling with a "system." You pick a starting point at random and then pick every \(k^{th}\) person.
How to do it:
- Calculate the interval \(k = \frac{\text{population size}}{\text{sample size}}\).
- Pick a random number between 1 and \(k\) as your starting point.
- Pick every \(k^{th}\) person after that.
Stratified Sampling
Sometimes a population has clear groups (called strata), like year groups in a school or different ages. To be fair, you want your sample to have the same proportions as the population.
The Formula: \( \text{Number in sample from stratum} = \frac{\text{Number in stratum}}{\text{Number in population}} \times \text{Sample size} \)
Don't worry if this seems tricky! Just remember: it’s all about percentages. If 60% of your school are girls, 60% of your sample should be girls.
Memory Aid
Stratified = Strata = Segments. Think of the layers (strata) of a cake. You want a slice that includes a bit of every layer!
3. Non-Random Sampling Techniques
Sometimes we don't have a list of everyone (no sampling frame), so we have to use non-random methods.
Quota Sampling
This is like stratified sampling but without the randomness. An interviewer is told to find 20 men and 20 women on the street. Once they have 20 men, they stop asking men and only look for women.
Pro: Very easy and inexpensive. Con: It can be biased because the interviewer might choose people who look "friendly."
Opportunity (Convenience) Sampling
This is simply picking whoever is available at the time. If you stand outside a gym and ask the first 10 people you see about their diet, that is opportunity sampling.
Pro: The easiest way to get data. Con: Highly unlikely to represent the whole population. (People at a gym probably eat healthier than the average person!)
Did you know?
Opportunity sampling is the most common method used by students for their own projects, but it's often the one most criticized for being biased!
4. Critiquing and Bias
In your exam, you might be asked to "critique" a sampling method. This just means "spot the mistakes."
Common Pitfalls to Watch For:
- Small Sample Size: If you only ask 2 people, you can't possibly know what 1,000 people think.
- Bias: Does the sample only include one type of person? (e.g., only interviewing people at 10 AM on a Tuesday ignores everyone who works a 9-to-5 job).
- Sampling Frame Errors: Is the list out of date? Does it exclude certain people (like people without a landline)?
The "Different Samples" Rule
Important Point: Remember that different samples from the same population will lead to different conclusions. This is natural! The goal of a good sampling technique is to make sure those differences are as small as possible.
Quick Review Box
Simple Random: Every person has an equal chance. Needs a list.Systematic: Every \(k^{th}\) person. Simple and quick.
Stratified: Proportional to groups. Most representative.
Quota: Non-random groups. Fast but potentially biased.
Opportunity: Whoever is there. Very biased but very easy.
Summary: How to Choose?
When you are solving a problem in Statistics and Mechanics (Paper 3), always ask yourself these three questions:
- Do I have a list of the population? (If yes, use random sampling. If no, use quota or opportunity).
- Is the population divided into obvious groups? (If yes, stratified or quota is best).
- Do I have a lot of time and money? (If no, avoid a census!).
Final Tip: When the exam asks for a disadvantage of a random sample, "it's hard to get a sampling frame" is almost always a winning answer!