Page 138 - Emotions through the eyes of our closest living relatives- Exploring attentional and behavioral mechanisms

Chapter 6
Procedure
The experiment was carried out between 21-01-2019 and 13-03-2019. In this period, the park was closed for visitors. A test session involved the presentation of two different trials, each consisting of a specific video sequence, and each trial followed by an observation period. The video sequence consisted of a primer, followed by either a yawn or control video (lasting 14 s), which was repeated 4 times and with a colored screen (again to grab attention) for 1 s in between each video. The length of one video sequence was thus 90 s (cf. Massen et al., 2013): primer (30 s) – colored screen (1 s) – yawn/control video (14 s) – colored screen (1 s) – yawn/control video (14 s) – colored screen (1 s) – yawn/control video (14 s) – colored screen (1 s) – yawn/control video (14 s). The presentation of one video sequence (representing one trial) was then followed by a 3.5-min observation period, after which the second trial started. If the first trial involved yawn videos, the second trial involved control videos and vice versa. The second trial was also followed by a 3.5-min observation period, completing one test session. Within one test session we always showed the same stimulus individual. See supplements for more information on the procedure.
We cycled through the entire video database four times (i.e., 4 blocks) over the course of the experiment to ensure sufficient data points. The order of control and yawn trials were counterbalanced per subject, and was further counterbalanced over the subjects per block. Within each block, trigger (i.e., familiar/unfamiliar/ avatar) was also randomized per subject. We designed a testing schedule based on eight test subjects, but two of those subjects involved a mother-infant pair and a mother-juvenile pair in which the infant/juvenile never left the mother. As such, we created a test schedule for six individuals rather than eight. With these six test subjects, three types of triggers, two conditions (yawn and control), two orders of condition presentation (yawn-control, or control-yawn), and finally four repetitions, we had a total of 288 test sessions and 576 trials planned (see Tables S2.1 to S2.3 and S3 for an overview). However, one video sequence was accidentally presented an extra time, resulting in 289 rather than the planned 288 sessions after data collection finished. On any given testing day, individuals participated in one or two sessions with 30 min breaks between video presentations to the tested subject. Furthermore, subjects never saw a video sequence more than once on any given day.
APDL and OEJM recorded all occurrences of yawning and self-scratching, and self-scratching was recorded as a measure for arousal and tension (Troisi, 2002). It was not possible to reliably quantify the amount of time spent looking at the screen due to the lack of continuous visibility of the gaze of the orangutans. To nonetheless ensure
136