RCT of Mind Reading and in Vivo Rehearsal on the Emotion Encoding and Decoding of Children with High-Functioning ASDs

Background: Children with high-functioning ASD (HFASD) often exhibit deficits in encoding and decoding of emotions (e.g., Baron-Cohen et al., 1997; Volker et al., 2009) which may contribute to their clinical difficulties.  Computer programs, such as Mind Reading (MR; Baron-Cohen et al., 2004) have been proposed to address the emotion-recognition deficits.  Although MR studies have yielded preliminary support of efficacy for children with HFASD (e.g., LaCava et al., 2007), randomized trials of the software are needed that include larger samples, control groups, comprehensive assessment of treatment impact (proximal and distal measures), and evaluation of maintenance.

Objectives: This RCT evaluated the efficacy of MR and in vivo rehearsal on targeted emotion encoding and decoding skills (proximal measures), and broader decoding skills, autism symptoms, and social skills (distal measures) of children with HFASD. 

Methods: Participants: Forty-three children, ages 8-12 with HFASD met inclusion criteria (short-form IQ >70; language score >80; and ADI-R score meeting ASD criteria) and were randomly assigned to either the treatment or wait-list control.  Proximal outcome measures: Cambridge Mindreading Face-Voice Battery for Children (CAM-C Faces and Voices subscales) and Emotion Recognition and Display Survey (ERDS Receptive and Expressive subscales).  Distal outcome measures: Diagnostic Assessment of Nonverbal Accuracy2 (DANVA2 Adult Faces and Child Faces subscales), Social Responsiveness Scale (SRS), and Behavior Assessment System for Children-Second Edition-Parent Rating Scale (BASC-2 PRS; Social Skills subscale).  Participants were assessed at pretest, posttest, and 5 weeks following treatment. Procedures: The manualized MR and in-vivo rehearsal protocol was administered during 24 staff-supervised 90-minute sessions.  Children completed MR exercises (Emotions Library, Learning Center, etc.) and in vivo rehearsal according to a manualized protocol.  Children earned MR generated computer-based rewards, and earned points for identifying and displaying emotions and following program and social rules.  

Results: Proximal measures:  Significant between-condition omnibus F results were examined using Bonferroni-corrected post hoc comparisons of posttest and follow-up scores (controlling for pretest).  Results indicated significantly better emotion decoding of faces and voices taught in MR (CAM-C) and encoding of targeted emotions (ERDS Expressive) for the treatment group (vs. controls) at posttest and follow-up.  Parent ratings of decoding (ERDS Receptive) were non-significant at posttest but significant at follow-up (favoring the treatment group). Distal measures:  Significant between-condition omnibus F results were examined using uncorrected post hoc comparisons of posttest and follow-up scores (controlling for pretest).  Results indicated significantly better decoding of emotions in children’s faces not directly taught (DANVA2 Child Faces) at posttest for the treatment group; this was not maintained at follow-up.  The treatment group also received significantly lower ratings of ASD symptoms at posttest and follow-up versus controls.  Results were non-significant for the DANVA2 Adult Faces and BASC-2 Social Skills subscales. 

Conclusions: Children who used MR learned to identify emotions in faces and voices directly taught in the program, and exhibited some degree of skill generalization to stimuli not directly taught.  They also exhibited significant improvements in parent-rated expressions of emotions and significant decreases in ASD symptoms.  Broader social skills were not significantly impacted, suggesting the need for enhancement of the protocol.