Procedure

This international project collected data from the AT centers in Sweden and special needs schools in Dubai and the USA [26].

To ensure consistency of data collection, the research team in Sweden provided a checklist for the research coordinators responsible for each participating organization to collect film clips in participants' natural contexts by the following instructions: (1) choose an activity that is meaningful and motivating for the children/youths to do, especially related to communication in natural routines (e.g., playing a game, performing an educational task, or the child communicating with the adult about something that had happened). The types of chosen activities are similar between the EGAT and NEGAT conditions; (2) choose a communication partner who know this child well; (3) film the activity about 5–10 min/day during three consecutive days in the contexts with a cell phone or a tablet on a stand; (4) choose a time of day when the children/youths would perform at their best; (5) try the best to film the faces, gaze direction and body movements of the child and his/her communication partner, and the computer screen with EGAT or low-tech communication devices.

After completion of data collection, these videos were sent to the research team in Sweden via the Cloud service, encrypted using Secure Sockets Layer.

#### *2.2. Participants*

### 2.2.1. Recruitment

The original project recruited 17 children/youths with severe physical disabilities and complex communication needs and 17 of their communication partners from the organizations in three countries (nine in Sweden, five in Dubai, and three in the USA). These children/youths were candidates for and in need of EGAT after testing other aided AAC devices that had limited functional use or were unsuccessful, and they were new to EGAT before the research started.

#### 2.2.2. Selection of Participants and Film Clips

Fourteen of 17 dyads had completed video filming in their natural contexts. The first and second authors (Y-H.H., M.B.) conducted participant screening referring to the film instruction checklist. The inclusion criteria were: (1) the children/youths age was between 1 and 21 years old; (2) the language was English or Swedish; (3) at least two available videos of each participant and his/her communication partner showed they performed similar communicative interaction activities in the EGAT and NEGAT conditions (e.g., play or school learning); (4) the videos captured the dyadic facial expressions, gaze, body movements and gestures in addition to the computer screen or low-tech communication tools in order to observe their communicative interaction clearly; (5) the videos were

informative, including activities generating dyadic communicative interaction instead of the child/youth only using EGAT to play games on a computer screen without interaction with the partner; (6) the videos showed the best ability of the child/youth in both the EGAT and NEGAT conditions, for instance, having more mouse clicks on screen using EGAT, showing longer attention/gaze towards the computer screen or longer interaction with partners, and/or demonstrating a high engagement level in that activity without a bored face. The exclusion criteria were (1) videos were too short (less than five minutes) and not informative; (2) videos failed to record the interactions between the dyads.

Eight dyads were not included due to language issues (*n* = 1), older age (*n* = 1), dissimilar tasks in the EGAT and NEGAT conditions (*n* = 2), or lack of communicative interaction (*n* = 4, where videos only showed the children conducting cause-effect games instead of communication activities). Based on the selection criteria, six children/youths and their communication partners were included. Dyads included three from Sweden, two from Dubai, and one from the USA. Each included dyad had four to 10 film clips of different lengths collected during the research period.

Following screening and after several meetings with two authors (M.B. and H.H.), the first author selected the best quality of two videos for each dyad, which showed similar activities, informative communicative interactions, and the best ability of children/youths in the EGAT and NEGAT conditions. A total of 12 videos from 47 film clips in six dyads were included, and each pair of videos in each dyad were concurrent or within two to five months. The videos ranged in length from five to 12 min, with a median length of eight minutes. The included videos with EGAT were recorded when the children/youths had accumulated three to six-month experiences of using EGAT, with four videos at three months of intervention and two at six months.
