The maintenance of behaviour
The cumulative records and other data given in the discussion of reinforcement schedules depicted not transitional effects, but rather the patterns of behaviour which develop after prolonged exposure to the schedules. A schedule such as fixed interval does not develop a consistent overall rate of responding throughout a session, but nevertheless a steady state of behaviour also eventually becomes predictable within fairly precise limits. In practice, a steady behavioural state is normally claimed if a specified criterion is reached. M. Sidman has distinguished two major types of experimental interest in these steady states of behaviour. The first he calls descriptive and it is this which has been used in comparing the effects of the various reinforcement schedules. The second interest Sidman has termed manipulative. Sidman has argued that the research methods made possible by steady states of behaviour are preferable to the alternative statistical methods.