Association for Behavior Analysis International

Time is ubiquitous; an inherent component of every experimental procedure and natural behavioral phenomenon. Investigating the behavioral processes governing the acquisition, adaptation, and discrimination of temporal relations between responses and stimuli is important for understanding how these processes function outside of strictly "timing" experiments. This symposium features analysis of timing in3 distinct experimental paradigms. Fox and Kyonka will present results of an experiment comparing response-initiated versus stimulus-initiated fixed-interval schedules. Rice and Kyonka will describe a model for acquisition of temporal control in a highly variable environment using both signaled and nonsignaled changes. Marshall and Kirkpatrick will discuss the role of timing in a choice in a delay discounting paradigm. Understanding how the arrangement of experimental contingencies affects timing, even when timing is not the primary interest of study is important for a more thorough understanding of behavior. The3 papers that will be presented will approach timing using3 distinct experimental arrangements.

Response-initiated fixed-interval (RIFI) schedules are tandem fixed-ratio (FR) 1 and fixed-interval (FI) schedules that may produce ratio- and interval-like responding. Very little research has systematically assessed responding on RIFI schedules compared to FI schedules. To compare patterns of responding in the 2 types of interval schedules, 4 pigeons were exposed to RIFI- and FI-15 s, 30 s, 60 s, 120 s, 240 s schedules of reinforcement in counterbalanced order across 10 conditions. For all conditions, post-reinforcement pauses were a function of the interval duration. Post-reinforcement pauses were more variable and included more long pauses in RIFI conditions than in FI conditions. Response rates in RIFI conditions were higher than in corresponding FI conditions. The shape of response rate distributions differed for RIFI and FI conditions in a way that suggests lower temporal control in RIFI conditions. RIFI schedules are not analogous to FI schedules, and may have emergent properties that are different from those of FI and FR schedules. For these pigeons, the onset of visual stimuli (red and green key lights) were more effective time markers than the pigeons own key pecks.

When intervals vary pseudorandomly across sessions, temporal control of behavior is acquired rapidly with minimal carryover from previous sessions. To determine whether temporal control can carry over from previous sessions in a highly variable environment, 4 pigeons were exposed to multiple peak-interval schedules in which the programmed intervals on food trials were changed for both schedules midsession. In a signaled condition, a 5 minute blackout preceded the midsession change. In an unsignaled condition, the midsession change occurred with no accompanying signal. Each pair of intervals was in effect for the second half of 1 session and the first half of the following session; a total of 72 trials separated by approximately 23 hours of nonexperimental time. To determine how temporal acquisition varied as a function of this unsignaled mid-session change, start and stop times were obtained from occasional no-food trials. Temporal control in both conditions was weaker than reported in previous research where intervals were changed at the start of a session. Further analyses will identify whether the addition of a signal to the mid-session change to intervals affects acquisition of temporal control and sensitivity of start and stop times to current and previous intervals.

The choice between alternatives often requires a delay until receipt of the chosen alternative; how an individual behaves during such a delay may be critical to the choice itself. In 2 experiments, rats made choices between 2 alternatives. In a delay discounting experiment, rats reared in either an enriched or an impoverished environment made a choice between a small reward that would be available after a short period of time and a larger reward that would be available after a longer period of time. In a probability discounting experiment, different rats housed in pairs made a choice between a small reward that was certainly delivered and a larger reward that was probabilistically delivered. Measures of timing behavior during the choice-outcome interval are generally lacking, even though the ability to time such intervals may affect choice behavior. An analysis of interval timing behavior was conducted in the two discounting tasks to explore the mechanisms that factor into the choices individuals make, particularly with regard to timing processes.