Long-tailed macaques extract statistical information from repeated types of events to make rational decisions under uncertainty

Abstract

Human children and apes seem to be intuitive statisticians when making predictions from populations of objects to randomly drawn samples, whereas monkeys seem not to be. Statistical reasoning can also be investigated in tasks in which the probabilities of different possibilities must be inferred from relative frequencies of events, but little is known about the performance of nonhuman primates in such tasks. In the current study, we investigated whether long-tailed macaques extract statistical information from repeated types of events to make predictions under uncertainty. In each experiment, monkeys first experienced the probability of rewards associated with different factors separately. In a subsequent test trial, monkeys could then choose between the different factors presented simultaneously. In Experiment 1, we tested whether long-tailed macaques relied on probabilities and not on a comparison of absolute quantities to make predictions. In Experiment 2 and 3 we varied the nature of the predictive factors and the complexity of the covariation structure between rewards and factors. Results indicate that long-tailed macaques extract statistical information from repeated types of events to make predictions and rational decisions under uncertainty, in more or less complex scenarios. These findings suggest that the presentation format affects the monkeys' statistical reasoning abilities.

Figure 1

Experimental setup. Subjects stood in an extension of the testing cage. A Plexiglas panel with two holes allowed them to touch the objects or the hands that were presented to them. A wheeled table stood in front of the cage, on which the testing material was placed, and behind which stood the experimenter (Exp. 1) or the different social agents (Exp. 2 and 3).

Figure 2

Example of the procedure of one session for each experiment. In Experiment 1, the experimenter would approach a first object to a hole in the Plexiglas and wait for the monkey to touch it, and then reward the monkey or not. This would be repeated a certain number of times with the first object, and then with the second (sampling phase). The numbers (1 or 0) represent an example of rewarding pattern of Exp. 1c and 1d. In the test trial, both objects would be presented simultaneously, and the monkey would have to touch one in order to indicate her choice. In Experiment 2 and 3, first one human, then the other would repeatedly present her fist to the monkey and then try to open the box. If she succeeded, she would grab a grape in the box and hand it to the monkey. The numbers (1 or 0) are an example of the rewarding pattern in Exp. 2b and Exp. 3b, and indicate the success of the human and the resulting reward.

Figure 3

Group performance in Experiment 1. This figure shows the group mean percentage of choosing the option leading to more probable rewards in each condition. Filled circles represent the group means, and the black bars represent the standard errors. The empty circles represent individual performances.

Figure 4

Group performance across trials in Experiment 1. This figure shows the correlation between group performance and trial number for each condition. The dark lines are regression lines and the empty circles represent the group percentage of correct choices in a specific trial.

Figure 5

Group performance in Experiment 2. (a) Shows the group mean percentage of choosing the favourable option, in both conditions. Filled circles represent the group means, and the black bars represent the standard errors. The empty circles represent individual performances. (b) Shows the correlation between group performance and trial number for each condition. The dark lines are regression lines and the empty circles represent the group percentage of correct choices in a specific trial.

Figure 6

Group performance in Experiment 3. (a) Shows the group mean percentage of choosing the favourable option, in both conditions. “Humans” stands for the condition in which the rewarding patterns covaried with the social agents (Exp. 3a) and “Boxes” stands for the condition in which the rewarding patterns covaried with the boxes. Filled circles represent the group means, and the black bars represent the standard errors. The empty circles represent individual performances. (b) Shows the correlation between group performance and trial number for each condition. The dark lines are regression lines and the empty circles represent the group percentage of correct choices in a specific trial.