Teaching Reading Skills to Minimally Verbal School-Aged Children with Autism Spectrum Disorders with a Serious Game. a Controlled Study

Background: School-aged children with Autism Spectrum Disorders (ASD) and minimally verbal skills are usually judged “low-functioning” in accordance with poor performance on conventional assessment of global cognitive skills. However, recent studies suggest that some of these children have spared non-verbal cognitive abilities despite failing to acquire spoken functional language. As reading instruction for these children are limited by the traditional phonics approach, we developed the serious game SEMA-TIC based on non-verbal cognitive skills. SEMA-TIC is a computer-based intervention with 10 series of 10 games, progressively teaching reading skills based on specific learning strategies adapted to the autistic profile and without verbal instructions.  

Objectives: The study investigates the efficiency of SEMA-TIC for the improvement of reading skills of minimally verbal school-aged children with ASD. It also compares the results with basic reading skills of typically developing children after formal academic reading training.

Methods:  Twenty five ASD children (6 -11 years) with no functional language were recruited. Two groups were constituted: (1) ASD intervention group (n=12) who received four thirty-minute SEMA-TIC sessions per week over an average of 22 weeks; (2) ASD non-intervention group (n=13). The two groups completed at baseline and after the follow-up period five experimental reading skill tasks (alphabet knowledge, words reading, words / non-words discrimination, sentences reading and words segmentation) and two standardized reading tasks (ALOUETTE and ODEDYS). Furthermore, a group of 16 typically developing children (TD group) from 6 to7 years schooled at the end of the first year of primary school and already readers were assessed by means of the same tasks.  

Results: The two ASD groups showed no significant difference concerning clinical characteristics and results on experimental tasks at baseline. Results revealed significant main effects of group (ASD intervention versus ASD non- intervention group, ANOVA; F(1,23)=256; p<.001) and session (ANOVA; F(1,23) = 301; p<.001) on experimental tasks. A significant group x session x experimental tasks interaction was also found (ANOVA; F(4,92) = 4.6; p<.01). Post-hoc analysis revealed that ASD intervention group significantly improved in 4 out of 5 experimental tasks after training compared to ASD non-intervention group. Results also showed a main effect of group (ASD intervention versus TD group) on experimental tasks (ANOVA; F(1,26) = 11; p<.01).  However and most importantly for the purpose of this study, results of post-hoc analysis (Tukey tests) show no significant difference between ASD intervention and TD group on each experimental task. Moreover, SEMA-TIC training enabled 25% of ASD intervention group to complete standardized tasks (ALOUETTE and ODEDYS) and to become decoding readers.

Conclusions: The present study shows that minimally verbal school-aged children with ASD are able to be instructed reading skills by means of a specific intervention using non-verbal cognitive skills. We therefore suggest that teaching reading skills to ASD children is not ineluctably linked to spoken language. The development of specific serious games as SEMA-TIC will be a major advance for these children.