Introduction: Why Do We Estimate?
Imagine you want to know the average height of every 15-year-old in the UK. Measuring all of them would be nearly impossible—it would take years and cost a fortune! In Statistics, we solve this by looking at a smaller group, called a sample, and using what we find to make a "best guess" about the whole group, known as the population.
This process is called estimating population characteristics. It is a vital tool for researchers, doctors, and even business owners who need to make big decisions without having every single piece of data.
Note: For more specific techniques on how to estimate wildlife populations, see the chapter on Petersen capture-recapture. For details on how we check if a process is staying consistent, see Control charts.
1. Estimating the Population Mean
The most common characteristic we estimate is the mean (the average). If we have a representative sample, the simplest and best estimate for the population mean is the sample mean.
The Rule: We assume that the mean of the sample (\(\bar{x}\)) is a "point estimate" for the mean of the population (\(\mu\)).
How to do it:
- Calculate the mean of your sample by adding all values and dividing by the number of items (\(n\)).
- State that this value is your estimate for the population mean.
Example: A scientist weighs 50 squirrels in a forest. The mean weight of these squirrels is \(420g\). What is the estimated mean weight of all squirrels in that forest?
Answer: The estimated population mean is \(420g\).
Common Mistake: Don't worry if the estimate isn't "perfect." In an exam, if you are asked to "estimate the population mean," you are simply being asked to use the sample mean you have calculated or been given.
2. Estimating Population Proportions
Sometimes we don't want an average; we want to know how many or what percentage of a population has a certain characteristic. We call this a proportion.
The "Scaling Up" Method
If you know the proportion in a sample, you can estimate the total number in the population using this simple logic:
\(\text{Estimated Total in Population} = \frac{\text{Number in Sample}}{\text{Sample Size}} \times \text{Population Size}\)
Example: A factory produces \(10,000\) lightbulbs a day. A quality control manager tests a sample of \(200\) bulbs and finds that \(6\) are faulty.
1. The sample proportion is \(\frac{6}{200} = 0.03\) (or \(3\%\)).
2. To estimate the total faulty bulbs in the whole population: \(0.03 \times 10,000 = 300\).
Estimate: There are likely \(300\) faulty bulbs in the total batch.
Quick Review: To estimate a population total, find the fraction in your sample and multiply it by the total population number.
3. Why Does Sample Size Matter?
You might wonder: "Can I just sample 2 people and guess the height of the whole country?" Technically you could, but it wouldn't be very reliable!
Key Factors for Reliability:
- Sample Size (\(n\)): Generally, the larger the sample, the more reliable the estimate. This is because large samples are less likely to be affected by one or two unusual "outliers."
- Replication: If you take several different samples and they all give similar estimates, your confidence in that result increases.
- Bias: If your sample isn't random (e.g., only measuring basketball players for a height study), your estimate will be biased and won't represent the population correctly.
Did you know? As a sample size gets larger, the sample mean usually gets closer and closer to the true population mean. This is a foundational idea in Statistics!
4. Comparing Samples to Populations
In your exam, you might be asked to compare a sample result to a known population characteristic. We often use the median and Interquartile Range (IQR) for this.
Example Context:
A national report says the median house price in the UK is \(\$250,000\). A student samples 20 houses in their local town and finds a median of \(\$310,000\).
Conclusion: The student can infer that house prices in their specific town are likely higher than the national average.
Key Takeaway: When comparing, always use context. Don't just say "the number is bigger"; say "the sample suggests that the population in this area has a higher average than the national population."
Summary: Key Points to Remember
- Point Estimate: Use the sample mean (\(\bar{x}\)) as your best guess for the population mean (\(\mu\)).
- Proportions: Calculate the percentage or fraction in your sample and "scale it up" to the size of the whole population.
- Reliability: Larger samples (\(n\)) give more reliable estimates and reduce the impact of anomalies.
- Bias: Always ensure the sample is representative, or the estimate will be misleading.
Top Tip for the Exam: If a question asks why an estimate might be unreliable, look at the sample size. If the sample is small (e.g., less than 30), that is almost always a valid point to raise!