Purpose: The purpose of this study was to compare social interaction and communication in autistic children with the approach of architectural spaces and ordinary centers. Methodology: The data were causal-comparative. The statistical population was all children with autism in centers with architectural spaces approach and children with autism in normal centers of Tehran. The statistical sample consisted of 3 centers with architectural approach and 3 normal centers selected by available sampling method. Data collection was done based on the Childhood Autism Rating Scale (GARS) and Edited Autism Checklist in Toddlers (M-CHAT) Robins, Feln & Barton (1999). Data analysis was performed by SPSS software in descriptive and inferential statistics at the significant level of 0.05 using independent t-test. Findings: According to the data analysis, the significance level in both variables of the study is greater than 0.05, which means that the scores of social interaction variables and the relationship of both groups follow the normal distribution. Independent t-test was then used to compare the mean social interaction of children with autism with the centers of architecture spaces and ordinary centers and showed that the obtained t was significant at the level of 0.01. Therefore, the null hypothesis can be rejected at the level of 0.01 in favor of the research hypothesis and with 99% confidence that there is a significant difference in social interaction between the two groups. Comparison of means indicates that social interaction of children with autism in architectural centers is higher than that of children with autism in normal centers. Conclusion: Social interaction and communication rate in children with autism in centers with architecture approach were more than children with autism in normal centers. As the number of these children is increasing, especially in Iran, it seems necessary to design centers and spaces appropriate to the needs of these children.
