Sample (statistics) facts for kids
In statistics, a sample is a small part of a larger group. This larger group is called a population. A sample is chosen carefully. It should represent the whole population fairly. This means it should be chosen without bias, or unfairness.
We use samples because populations can be very large. It might be impossible to count or study every single person or item. So, to solve a problem in statistics, we often start with sampling. Sampling means choosing which data to collect. This data is then used for analysis.
For example, imagine you want to study the pollution in a lake. If you take water samples from only one spot, your study might not be accurate. The results could be different depending on where you take the samples.
A good rule is that samples need to be random. This means every person or item in the population has an equal chance of being chosen.
In real life, random samples are usually taken using a clear procedure. A procedure is a set of rules or steps that are followed exactly. Even with a good procedure, some bias can still happen.
Think about trying to guess the result of an election poll. All ways of collecting opinions have challenges. The final election results are often different from what samples predicted. If you call people on the phone, you won't reach those who don't answer. If you ask people on the street, you miss those who stay home. So, a completely neutral sample is not always possible. In these cases, a statistician tries to figure out how much bias there is. There are ways to estimate this.
Scientists also face this when measuring things. For example, weighing a piece of metal. Even with sensitive equipment, you might get slightly different results each time. No measurement system is perfect. These measurements are like samples, and they have some error. Statistics helps us understand this error and analyze this kind of data.
There are different kinds of samples:
- A complete sample includes all items that have a certain feature.
- An unbiased or representative sample is made by picking items from a complete sample. This picking process does not depend on the items' features.
The way a sample is collected, and its size, affects how the data is understood.
What is Stratified Sampling?
If a population clearly has smaller groups inside it, then each of these smaller groups needs to be sampled. This method is called stratified sampling. It's also known as a stratified random sample. Stratified sampling often uses proportions, like percentages (%).
Let's say you want to study the incomes of adults. You might notice that college graduates often earn different amounts than non-graduates. So, you would divide your population into these groups.
Imagine that 30% of all adult males are college graduates. Then, in your sample, you would make sure 30% of the males you pick are graduates. The other 70% would be non-graduates. You would do the same for females, as their percentage of graduates might be different. This creates a sample that is divided by gender and education.
Next, you might divide each of these groups by age. For example, graduates might earn more compared to non-graduates when they are middle-aged. This helps make the sample even more accurate.
Another type of stratified sample deals with variation. This is when some groups have a wider range of data. In these cases, larger samples are taken from the more varied groups. This helps make the summary numbers, like the mean (average) and standard deviation (how spread out the data is), more reliable.
Related pages
See also
In Spanish: Muestreo (estadística) para niños
