Manatee Psychophysical Testing Are Results Biased by Sequence Learning?

Athena Rycyk

Date of Award

2004

Document Type

Thesis

Degree Name

Bachelors

Department

Natural Sciences

Second Department

Social Sciences

First Advisor

Bauer, Gordon

Keywords

Manatee, Sequence Learning, Gellermann, Fellows

Area of Concentration

Biological Psychology

Abstract

Researchers investigating sensory abilities of animals employ counterbalanced schedules in detection and discrimination tasks. These schedules determine the order for stimulus presentation, for example, signal present vs. signal absent in a detection task, or left vs. right orientation for the correct choice in a discrimination task. If property applied, such schedules can prevent subjects from using sequential cues to influence performance. Two of the most commonly used schedules, Gellermann Schedules and Fellows Schedules, are constrained by short sequences with no more than three of the same trial in a row. Unfortunately, evidence indicates that some species, including harbor seals, are sensitive to constrained sequences, introducing a bias to the results of psychophysical testing, potentially causing overestimates of sensory abilities. One way to test for sensitivity to the schedules is to compare the percentage of correct trials after short and long sequences. The probability of a change in schedule sequence increases with the number of same trials. For example, the probability of a change of the correct trial from the left to the right side is greater after three left-correct trials in a row, when there will always be a change, than after only one left trial, when both change and no-change are possible. Recent experiments with two manatees (Trichechus manatus latirostris) indicated that they were not sensitive to trial sequences in two-choice discrimination tests of visual acuity using Gellermann and Fellows Schedules. After more than 5,000 trials for one manatee and over 2,000 trials for the other manatee, the subjects' performances did not differ after long trial runs compared to short trial runs. The absence of sequence bias was found even on those trials where discriminations were difficult and alternative strategies would be more likely to influence results.

Rights

This bibliographic record is available under the Creative Commons CC0 public domain dedication. The New College of Florida, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.

This document is currently not available here.

Share

COinS