ABSTRACT
In most scientific investigations, data are drawn from samples across the time dimension as well as across subjects. There are a number of reasons why the time dimension is important for observational data. As a practical alternative to continuous observation, time-sampling techniques are frequently employed. For intrasession observation, there are three basic approaches to data collection: continuous observation, semicontinuous sampling, and discrete sampling. These are partial-interval sampling and whole-interval sampling. Because of the retention of some form of continuous observation in these two methods, people refer to them as semicontinuous sampling. Partial-interval sampling has been reported as widely used or as the most popular intrasession sampling method. Two discrete sampling methods have been commonly used. These are the ad libitum method and the momentary sampling method. The ideal process will enable meaningful assessments of reliability and elimination of systematic session sampling error.
