Renata Cambraia

Tese de doutoramento em Psicologia BásicaInterval timing pertains to the ability of organisms to adapt their behavior to the temporal regularity of stimuli such as lights, sounds, or food. Although a set of properties have been established to characterize timing, researchers will encounter phenomena such as temporal distortions, or biases. That is, temporally controlled behaviors may be sensitive to nontemporal variables, such as payoffs (reinforcement probabilities) and base-rates (frequency of stimulus intervals). How timing is affected by these variables in different tasks and species is debatable. In the present dissertation, we discuss these issues throughout four studies investigating how payoff and base-rate affect pigeons and humans in a temporal bisection task. This temporal discrimination task is tested in its standard version – subjects choose one arbitrary key after a short interval sample and another after a long sample (e.g., short—green, long—red) – and a novel versio...

We investigated how base rates affect temporal discrimination. In a temporal bisection task, pigeons learned to choose one key after a short sample and another key after a long sample. When presented with a range of intermediate samples they produced a psychometric function characterized by a bias and a scale parameter. When one of the trained samples was more frequent than the other, only the location parameter changed, with the pigeons biasing their choices toward the key associated with the most frequent sample. We then reproduced the bisection task in a long operant chamber, with choice keys far apart, and tracked the pigeons' motion patterns during the sample. Pigeons learned to approach the short key following sample onset, wait on the "short side" for a few seconds, and then, when the sample continued, depart toward the long key. This time-place curve was affected by sample base rate: The probability of pigeons going directly to the long side after sample onset ...

We investigated how differential payoffs affect temporal discrimination. In a temporal bisection task, pigeons learned to choose one key after a short sample and another key after a long sample. When presented with a range of intermediate samples they produced a Gaussian psychometric function characterized by a location (bias) parameter and a scale (sensitivity) parameter. When one key yielded more reinforcers than the other, the location parameter changed, with the pigeons biasing their choices toward the richer key. We then reproduced the bisection task in a long operant chamber, with choice keys far apart, and tracked the pigeons' motion patterns during the sample. These patterns were highly stereotypical-on the long sample trials, the pigeons approached the short key at sample onset, stayed there for a while, and then departed to the long key. The distribution of departure times also was biased when the payoff probabilities differed. Moreover, it is likely that temporal control decreased while control by location increased. No evidence was found of changes in temporal sensitivity. The results are consistent with models of timing that take into account bias effects and competition of stimulus control.