Page 195 - Emotions through the eyes of our closest living relatives- Exploring attentional and behavioral mechanisms

Appendix A
Supplementary Materials for Chapter 2
Experiment 1: Bonobos’ attentional bias towards emotions of familiar and unfamiliar conspecifics.
Table S1. Definition of emotion categories used for bonobos in Experiment 1, and number of pictures per emotion category and per familiarity category.
Appendices
Picture category
Distress
Sex Playing
Grooming Yawning
Self- scratching
Neutral
Description
Aggressive displays (e.g., charges and direct displays); submissive behaviors (e.g., grin faces, fleeing and crouching)
Mating, genito-genital rubbing, prominent full swelling, penile erection
Together or alone, with a relaxed open mouth, without an object
Two or more individuals grooming
Wide-open mouth with or without teeth exposure
Roughly rubbing the body, face, or one of the limbs
Walking, lying down, or sitting
No. of individuals per picture
M = 1.55, SE = 0.16 M = 2.25, SE = 0.25
M = 2.25, SE = 0.25 M = 2.38, SE = 0.26
M = 1.17, SE = 0.17 M = 1.00, SE = 0.00 M = 1.67, SE = 0.14
No. of unfamiliar pictures
15
27 22
53 20
18 155
No. of familiar pictures
25
16 36
50 16
30 173
Note. All bonobo participants saw the same number of unique pictures, but for the familiar models the composition of the stimulus set differed per participant because we replaced pictures of the participants themselves with pictures of other familiar individuals. Furthermore, the number of pictures differs per Familiarity and also per Picture category because some behaviours were easier to photograph or occurred more frequently than other behaviours.
Appendix to stimuli and validation
Six primate experts scored the pictures of bonobos based on valence and intensity.
Three experts worked with the bonobos on a daily basis, and the others worked with
bonobos or chimpanzees in the past. Experts were presented with one picture at a
time and were asked to 1) rate how negative or positive they thought bonobos would A experience each picture (1= very negative, 7 = very positive) and 2) how intense the
picture was (1= not intense, 7 = very intense). Pictures were shown until a response was given. We calculated intraclass correlations (ICCs) for valence and intensity ratings using a two-way mixed model and a consistency definition and found a high reliability for both ratings (ICCvalence = .82, 95% CI: .79 - .84, F(653, 3265) = 5.45, p < .001; ICCintensity = .87, 95% CI: .86-.89, F(653, 3265) = 7.96, p < .001). A generalized linear
193