How task format affects cognitive performance: a memory test with two species of New World monkeys

Michèle N. Schubiger,Alexandra Kissling,Judith M. Burkart

DOI: https://doi.org/10.1016/j.anbehav.2016.08.005

IF: 3.041

2016-11-01

Animal Behaviour

Abstract:In cognitive tests, animals are often given a choice between two options and obtain a reward if they choose correctly. We investigated whether task format affects subjects' performance in a physical cognition test. In experiment 1, a two-choice memory test, 15 marmosets, Callithrix jacchus, had to remember the location of a food reward over time delays of increasing duration. We predicted that their performance would decline with increasing delay, but this was not found. One possible explanation was that the subjects were not sufficiently motivated to choose correctly when presented with only two options because in each trial they had a 50% chance of being rewarded. In experiment 2, we explored this possibility by testing eight naïve marmosets and seven squirrel monkeys, Saimiri sciureus, with both the traditional two-choice and a new nine-choice version of the memory test that increased the cost of a wrong choice. We found that task format affected the monkeys' performance. When choosing between nine options, both species performed better and their performance declined as delays became longer. Our results suggest that the two-choice format compromises the assessment of physical cognition, at least in memory tests with these New World monkeys, whereas providing more options, which decreases the probability of obtaining a reward when making a random guess, improves both performance and measurement validity of memory. Our findings suggest that two-choice tasks should be used with caution in comparisons within and across species because they are prone to motivational biases.

zoology,behavioral sciences