Pigeons and Humans Time Miscued Intervals

Cues signaling time to reinforcer availability can be highly informative, somewhat informative, or uninformative of the specific time to reinforcement. The purpose of this series of studies was to characterize the functional relation between miscues and pigeon and human response patterns in a peak-interval (PI) procedure. In Experiment 1, pigeons pecked keys to earn food in a cued and miscued PI procedure. Each session consisted of fixed-interval (FI) 2-s and 4-s schedules of food and occasional, 12-s PI trials during which pecks had no programmed consequences. The correlation between visual cues (red or green key light colors) and the schedule of reinforcement (FI 2 s or FI 4 s) was manipulated across cue-informativeness conditions and represented by the phi coefficient (ϕ). Different key light colors were either highly (ϕ = 1), moderately (ϕ = 0.8, 0.6, 0.4, or 0.2), or not correlated (ϕ = 0) with specific FI schedules. Temporal discrimination in PI trials was accurate; peaks in response distributions occurred at around 2 s and 4 s. Red and green FI and PI response distributions converged when the correlation between key color and FI was less than ϕ = 0.8. In Experiments 2 and 3, humans participated in a modified PI procedure. Participants pressed buttons on a computer keyboard to shoot at the center of the screen while a target moved from left to right across the screen. Participants earned points for shots hitting the target (around 2 s and 4 s) and lost points for misses. The target was masked during 12-s PI trials, but participants could still earn and lose points. The correlation between the background color (light or dark green) and time to point availability (2 s or 4 s) was manipulated. Different background colors were either highly (ϕ = 1), moderately (ϕ = 0.8, 0.4), or not correlated (ϕ = 0) with specific times to point availability. In Experiment 2, each participant experienced all 4 cue-interval correlations and in Experiment 3, participants were randomly assigned to one of the four cue-interval correlations. Consistent with Experiment 1, temporal discrimination in PI trials was accurate, peaks in response distributions occurred at around 2 s and 4 s. Response distributions in light and dark green PI trials were influenced by cue-informativeness order in Experiment 2 and were converged in groups with cue-interval correlations ϕ \u3c 0.8 in Experiment 3. Across all experiments, the convergence in PI distributions was not a sign that temporal discrimination accuracy was eliminated (peak times were always stable), but rather an adaptive change in response patterning involving adjustment of the time of abrupt shifts in response rate