"I felt successful!" Assessing autistic adolescent game usability from randomized control trial to improve sensitivity to eye-gaze cues

We demonstrated that Social Games for Adolescents with Autism (SAGA) effectively improves perception and understanding of eye gaze cues among autistic adolescents. SAGA is a serious game in which participants learn to interpret non-verbal social communicative behaviors (e.g., pointing, head turns, eye gaze cues) from characters to solve narrative-related quests. Importantly, SAGA was initially evaluated using a waitlist control. In our current study, SAGA2, we used feedback from autistic adolescents to refine the intervention game and develop an active control game. The intervention and control games were identical in narrative structure and game logistics, including increasing task difficulty and personalized progression. However, instead of focusing on the characters’ non-verbal social cues, the control game required participants to complete an object matching task on a screen adjacent to the characters’ face. In this randomized controlled trial, 46 autistic adolescents (10-18 years) were randomly assigned to the intervention or control condition. Participants received computers to play the game at home for 45 minutes, thrice a week, over 12 weeks. Post-training, we administered a game usability survey which inquired about participants’ impressions regarding the game logistics (e.g., The instructions were clear, My character was easy to control), and their affect experiences during and after the training (e.g., I thought the game was fun, I felt frustrated). Participants in both conditions reported similar levels of ease in game navigation and a general willingness to recommend the game to their friends. Both games elicited similar levels of positive and negative affect during, and after the training. However, the control game elicited slightly more positive affect after the training, likely due to participants getting farther along in the narrative. Together, these usability findings indicate that the control game is well matched to the intervention game in terms of motivating factors, engagement, and difficulty from participants’ perspective.