A scientist might, for example, collect data from a pair of dice by rolling them over and over again in the lab. The results would form a frequency distribution; given enough data, it's possible to draw conclusions about which numbers are most likely to come up.