Within these two broad categories, there are two types of sampling designs that are specially designed for observation: event sampling and time sampling.2 In event sampling, the researcher records that a specific behavior or act occurs; the duration of the observation is determined by when the act begins and ends. In some instances, the researcher may also capture its antecedents—events and conditions immediately before the behavior of interest—and the consequences—what happens immediately after the event. For example, assume a machine malfunctions or stops; the event of interest is the stoppage itself. When did it occur, what was the machine producing, what or who stopped the machine (an internal device, a person operating the machine, etc.), who was operating the machine at the time it stopped, etc. While it is unlikely we could hire an observer to wait and watch for machines to malfunction, we can use records of such events. Either event or time sampling can be used when observing human behavior.