Master of Science (MSc)
Faculty of Science
According to the instructional failure hypothesis, the contextual similarity of the intertrial interval (ITI) and the delay interval (DI) is responsible for the choose-short bias that occurs when memory for filled intervals is tested. This hypothesis may also explain the choose-long bias for empty intervals, if birds confuse an extended DI with a long empty interval. In the present study, pigeons were trained in a within-subjects design to discriminate durations of a filled interval (2-s and 8-s of light), and durations of an empty interval (2-s and 8-s bound by two 1-s light markers). In order to disambiguate the ITI, sample presentation phase, and the DI, the colour displayed on the video monitor was different during the various phases of a trial. In Experiment 1, the ITI was magenta, the sample presentation phase was dark, and the DI was grey. Although accuracy was greater for short than for long samples, the retention functions were parallel. Thus, there was no evidence of a biased-forgetting effect. In Experiment 2, additional tests indicated that the interval from the offset of the ITI to the onset of the DI, as well as the filled and empty intervals themselves affected choice responding in Experiment 1. In Experiment 3, the colour displayed on the video monitor was manipulated such that the illumination was magenta during the ITI, and dark during the sample presentation and comparison phases in order to prevent pigeons timing from the ITI offset to the DI onset, and to create confusion for one interval-type, but not the other. A choose-long bias was obtained for empty intervals, and no-bias was obtained for filled intervals. In Experiment 4, the colour displayed on the video monitor was manipulated such that the illumination was dark during the ITI and sample presentation phase, and grey during the comparison phase in order to prevent pigeons from timing from ITI offset to DI onset, and to eliminate the possibility of confusion for both filled and empty intervals. The results of this experiment revealed no biases for either filled or empty intervals. In Experiment 5, the colour displayed on the video monitor was dark for all of the phases of the trial in order to create confusion, and thus produce the choose-short and long biases. The results of this experiment revealed no bias for filled intervals, and a choose-long bias for empty intervals. The results are discussed in the context of the instructional confusion hypothesis.
Gagne, Stephen, "Biased forgetting effects in the assessment of memory for filled and empty intervals: Evidence for the instructional failure/confusion hypothesis" (2005). Theses and Dissertations (Comprehensive). 762.