Sampling and estimation

Imagine you have a giant jar full of colourful sweets, and you want to know what percentage are red without counting every single one. That's a bit like what sampling and estimation (making an educated guess) is all about!

• Population: This is the entire group you're interested in. In our sweet jar example, the population is all the sweets in the jar. If you're studying teenagers, the population is all teenagers.
• Sample: This is a smaller, manageable group that you pick from the population. It's like taking just a handful of sweets from the jar. You study this small group to learn about the big group.
• Sampling: This is the process of choosing that small group (your sample). You want to pick your sample carefully so it's a good mini-version of the whole population.
• Estimation: Once you've studied your sample, you use what you found out to make an educated guess (an estimate) about the whole population. If 30% of your handful of sweets are red, you might estimate that 30% of all the sweets in the jar are red.

Think of it like a chef tasting a spoonful of soup to know if the whole pot needs more salt. The spoonful is the sample, and the whole pot is the population.

Let's say a big mobile phone company wants to know how many of their customers are happy with their new phone model. They have millions of customers – it's impossible to call every single one!

1. Identify the Population: The population is all their customers who bought the new phone model.
2. Choose a Sampling Method: They decide to use random sampling (like drawing names out of a hat, but with computers). This means every customer has an equal chance of being picked, which helps make the sample fair.
3. Collect the Sample: They randomly select 1,000 customers from their huge list. This group of 1,000 is their sample.
4. Gather Data: They call these 1,000 customers and ask them if they are happy with their new phone.
5. Calculate an Estimate: Let's say 800 out of the 1,000 customers (80%) say they are happy. The company would then estimate that around 80% of all their customers are happy with the new phone.

This way, they get a good idea about their millions of customers without having to contact every single one, saving lots of time and money!

Here's a general breakdown of how you go from a big question to a smart guess:

1. Define your Population: Clearly state who or what you are interested in studying. This is your 'whole cake'.
2. Decide on your Sample Size: Figure out how many items or people you will include in your sample. This is how big your 'slice' will be.
3. Choose a Sampling Technique: Select a fair way to pick your sample from the population. This ensures your slice isn't just the icing.
4. Collect your Data: Gather information from each member of your chosen sample. This is like tasting your slice of cake.
5. Calculate Sample Statistics: Find numbers like the average or proportion from your sample data. For example, the average height of students in your sample.
6. Estimate Population Parameters: Use your sample statistics to make an educated guess about the entire population. This is where you say, 'Based on my slice, the whole cake probably tastes like this!'
7. Consider the Confidence: Think about how sure you are that your estimate is close to the real value. This is like saying, 'I'm 95% sure the cake is chocolate.'

Picking your sample correctly is super important. If your sample isn't like the population, your estimate will be way off! Imagine trying to guess the flavour of a whole cake by only tasting the cherry on top.

• Random Sampling (or Simple Random Sampling): This is like putting everyone's name in a hat and drawing them out. Every single member of the population has an equal chance of being chosen. This is generally the best way to get a fair sample.
• Systematic Sampling: Imagine you have a list of people. You pick a random starting point, say the 5th person, and then pick every 10th person after that (5th, 15th, 25th, etc.). It's systematic because there's a pattern.
• Stratified Sampling: If your population has distinct groups (like different age groups or genders), you make sure your sample has the same proportion of these groups. For example, if 60% of students are girls, you make sure 60% of your sample are also girls. This is like making sure your cake slice has some sponge, some cream, and some jam if the cake has layers.
• Quota Sampling: Similar to stratified, but instead of random selection within the groups, you just pick people until you meet a 'quota' (a target number) for each group. It's less random and can introduce bias.
• Opportunity Sampling (or Convenience Sampling): This is the easiest but often the least reliable. You just pick people who are readily available, like asking your friends. It's like only tasting the part of the cake closest to you – it might not be representative of the whole thing.

Even smart people make mistakes with sampling. Here are some common traps:

• ❌ Mistake 1: Biased Samples: This happens when your sample isn't truly representative of the population. For example, asking only people at a gym about their favourite sport will likely over-represent fitness enthusiasts. ✅ How to Avoid: Use random sampling methods whenever possible. Ensure every part of your population has a chance to be included. Think about who might be missing from your sample.
• ❌ Mistake 2: Too Small a Sample Size: Using a tiny sample, like asking only 3 people about a city's opinion. Your estimate will be very unreliable, like trying to guess the whole book from reading just one sentence. ✅ How to Avoid: While there are formulas, a general rule is that larger samples tend to give more reliable estimates. For A-Level, you'll often be given a suitable sample size or asked to justify one.
• ❌ Mistake 3: Misinterpreting Confidence Intervals: Thinking a 95% confidence interval means there's a 95% chance the sample mean is within the interval. It's about the population mean. ✅ How to Avoid: Remember, a confidence interval (a range of values) tells you that if you repeated your sampling many times, 95% of those intervals would contain the true population mean (the real average of the whole group). It's about the method, not a single sample's certainty.
• ❌ Mistake 4: Confusing Population and Sample Statistics: Using the symbol for a sample mean (e.g., x̄) when you're talking about the population mean (e.g., μ). ✅ How to Avoid: Always be clear about whether you're referring to the whole population or just your sample. Use the correct notation (e.g., μ for population mean, x̄ for sample mean; σ for population standard deviation, s for sample standard deviation).

When you make an estimate, you're rarely 100% sure it's exactly right. A confidence interval is like saying, 'I'm pretty sure the true answer lies somewhere between this number and that number.'

• What it is: It's a range of values, calculated from your sample data, that is likely to contain the true value of the population parameter (like the true average height of all oak trees).
• Confidence Level: This tells you how 'confident' you are that your interval contains the true value. Common levels are 90%, 95%, or 99%. A 95% confidence level means that if you repeated your sampling many, many times, 95% of the intervals you create would contain the true population value.
• Margin of Error: This is how wide your interval is. A smaller margin of error means a more precise estimate. Think of it as how much 'wiggle room' there is around your estimate.

Imagine you throw a hoop (your confidence interval) at a target (the true population mean). A 95% confidence level means that if you throw the hoop 100 times, you'd expect to hit the target about 95 times. The bigger your hoop (wider interval), the easier it is to hit the target, but your aim isn't as precise!