A Comparative Perspective on Three Primate Species' Responses to a Pictorial Emotional Stroop Task

Abstract

The Stroop effect describes interference in cognitive processing due to competing cognitive demands. Presenting emotionally laden stimuli creates similar Stroop-like effects that result from participants' attention being drawn to distractor stimuli. Here, we adapted the methods of a pictorial Stroop study for use with chimpanzees (N = 6), gorillas (N = 7), and Japanese macaques (N = 6). We tested all subjects via touchscreens following the same protocol. Ten of the 19 subjects passed pre-test training. Subjects who reached criterion were then tested on a standard color-interference Stroop test, which revealed differential accuracy in the primates' responses across conditions. Next, to test for an emotional Stroop effect, we presented subjects with photographs that were either positively valenced (a preferred food) or negatively valenced (snakes). In the emotional Stroop task, as predicted, the primates were less accurate in trials which presented emotionally laden stimuli as compared to control trials, but there were differences in the apes' and monkeys' response patterns. Furthermore, for both Stroop tests, while we found that subjects' accuracy rates were reduced by test stimuli, in contrast to previous research, we found no difference across trial types in the subjects' response latencies across conditions.

Keywords: Stroop effect; affect; attentional bias; cognitive bias; emotions; snake detection theory; touchscreen; welfare; zoo.

Figure 1

A photograph of one of the Japanese macaque subjects participating in a voluntary touchscreen test session at Lincoln Park Zoo. Here, the subject’s target stimulus-border color is yellow and she is selecting the correct stimulus in a congruent test trial in Test Phase 1.

Figure 2

Example stimuli for Training Phase 1 and Phase 2. Each stimulus was a human-made object unfamiliar to the subjects presented on a white background. The stimuli were created as pairs of two identical photographs, one with a yellow border and one with a blue border.

Figure 3

Examples of stimuli used in Test Phase 1 showing potential stimuli pairings for the three conditions (congruent, incongruent, and control). In congruent trials the color of the shape matched the subject’s target stimulus-border color for both target and distractor, while in incongruent trials the color of the shape is the opposite to that of the subject’s target stimulus-border color. In control trials, the images were neutral black and white photographs of human-made objects.

Figure 4

Stimuli categories across trial types and conditions in Test Phase 2. Within each trial, stimuli only differed by their border colors, but within each test condition (negative Stroop, positive Stroop, and Control) two trial types were presented to subjects (test trials and neutral trials) (Table 2). This figure provides examples of the stimuli presented across trial types within conditions, but note that different images were used across trials and conditions as described in the text. In this figure, images in the Stimulus A columns represent examples of correct choices for subjects whose target stimulus-border color was yellow, while images in the Stimulus B columns represent examples of correct options for subjects whose target stimulus-border color was blue.

Figure 5

The subjects’ average accuracy in selecting the stimuli with their target stimulus-border color across the three experimental trial types in Test Phase 1 when the ITI was short (3 s or 4 s).

Figure 6

The macaques’ average accuracy across the three experimental trial types in Test Phase 1 as tested with short (3 s) and long (8 s) ITIs.

Figure 7

The subject’s average accuracy in neutral and test trials across the three experimental conditions (control, negative Stroop, and positive Stroop) in Test Phase 2. Here, test trials refer to the control, snake, and food trials in each of the three experimental conditions respectively (see Figure 4 and Table 2).