Face-Expression Expert System: a New Teaching Program Using Equivalent Relations for Children with Autism Spectrum Disorder

Background:  Individuals with autism have various kinds of difficulties on the cognition of facial expressions.  Therefore, we need to consider comprehensive analysis for perception, conceptualization, comprehension, verbal-naming, imitation, appreciation of the situation, prosodic inference, self-other mapping for examining the cognition of faces, and facial expressions.  We have developed the comprehensive face and facial expression learning support program called Face-Expression Expert Program (FEEP), which is based on the framework of the equivalent relation among four types of stimuli.  The program consisted of four classes (“facial expression,” “emotion-word,” “descriptive sentence,” and “prosody of emotion”) covering wider developmental age.  In the current studies, we evaluated the effect of FEEP on the development of facial expression in individuals with autism.  

Objectives:  Study 1 examined the condition for establishing concept of facial expressions by sorting facially expressed pictures with younger individuals with autism.  In Study 2, we assessed the emergence of equivalent relations among facial expression, emotion-word, sentence, and prosody.

Methods:  In Study 1, young children with autism were given pictures of “happy” and “sad” facial expressions of two adult males and two females (stimulus set A and B), and illustrations (stimulus set C).  As generalization stimuli, pictures of another adult female (stimulus set D) were used.  Tasks were identical matching-to-sample and categorical matching-to-sample.  In Study 2, we used four facial expressions; “happy,” “sad,” “surprised,” and “angry.” We examined the emergence of equivalent relations among three stimulus sets; facial expression, emotion-word, and descriptive sentence.  We evaluated the following six relations; 1) Comprehension task: “word - selecting picture” (i.e., the child was required to select the corresponding picture of facial expression when the emotion-word was presented as sample stimulus), 2) Naming task: “picture - word naming,” 3) Imitation task:  “facial expression - imitating facial expressions,” 4) Expression task: “word name - making facial expression,” 5) Context Comprehension task: “descriptive sentence - selecting picture,” and 6) Context Production task: “facial expression - describing with sentence.”

Results:  Study 1 showed the generalization to the stimulus set D, the equivalent relations, even though they were not trained.  Study 2 showed that 1) the intervention of Comprehension task facilitated the correct naming in Naming task, 2) the intervention of Imitation task facilitated the correct facial expression in Expression task, and 3) both of the Context Comprehension and Production tasks showed high level of correct responding.

Conclusions:  Results from Study 1 and 2 indicated that the concept formation and equivalent relations (symmetry and transitivity) of facial expression did emerge by newly developed teaching program, FEEP.  Study 1 indicated that sorting task of facial expressions established the concept formation of facial expressions. Study 2 showed several equivalent relations: 1) symmetrical relation emerged betweencomprehension and naming, 2) equivalent relation emerged among word naming, other’s facial expressions, and self-facial expressions, and 3) symmetrical relation emerged between facial expressions and descriptive sentences.  We will make FEEP as a computer-based teaching (CBT) program.  It will enhance more comprehensive research, such as comparison study and intervention study on the same platform.