An excellent example of every-day statistics is the use of a meat thermometer on your supper.

In our household, meat must reach a safe cooking temperature before we eat it. When we pull the chicken out of the oven or off the stove, we always test it with a meat thermometer. When you use a meat thermometer, you choose some random places in the meat and measure the temperature. If you see that the temperature for all of those places is considered safe, you assume that all of the meat is safe.

In statistics, we don’t measure every single point in a population. In the same way, we don’t measure the temperature in every single square millimetre of the meat. We choose a sample (the points in the meat), and we make an inference about the rest of the population from what we have learned.

The key to making a reasonable inference is your random sample. If we simply test the thin pieces of meat or the edges, we might miss the thickest middle section that is still a little bit frozen. But if we randomly choose ‘samples’ from all different parts of the meat, we can feel fairly confident that the average meat temperature is high enough to be safe. You also need to have a large enough number of tries to be sure that you haven’t missed any possibilities.

## Statistics And Sampling Of People

Sampling people in a population can work the same way. If you take a large enough random sample of people, you can find out what the ‘temperature’ or makeup of the group, is really like.

There is much more to statistics than sampling and observing, but the next time you take the temperature of meat, think about the random sample you are taking and the way that you assume a result based on the few sample points that you observe.