Understanding Theory of Mind Improvements As a Result of Face Processing Instruction

Background:  FaceSay is a computer-based intervention that offers students simulated practice with eye gaze, joint attention, and facial recognition (Symbionica, LLC). Previous research has demonstrated that FaceSay improves emotion recognition and theory of mind (ToM) in school-aged children with ASD (Rice et al., 2014). ToM skills are not directly targeted with FaceSay, and the mechanism behind these improvements is not yet understood. In order to determine the specific components of ToM that are affected by this intervention, between-group differences in categories of items on the NEPSY-II ToM subscale were examined (Korkman, Kirk, & Kemp, 2007).

Objectives:  To explore the components of ToM that are influenced by intervention with FaceSay. Potential differences were assessed using a: 1) theoretical approach to item grouping based on the NEPSY-II manual, and 2) statistical approach to item grouping.

Methods: Thirty-one school-aged children (M_age=7.76 years) meeting the educational criteria for ASD participated in a randomized-control trial studying the effects of intervention with FaceSay (Experimental group) versus an educational program (Control group) for 10 weeks. Students’ ToM skills were assessed pre- and post-intervention using the ToM subtest of the NEPSY-II. Items were grouped into constructs using the NEPSY-II manual and exploratory factor analyses (EFA) with Varimax rotation. The Verbal Task and Contextual Task were treated separately for all analyses. Theoretical constructs were excluded if they were indexed by only one item. Factors were generated using pre-intervention data, and regression scores were calculated for each student at both time points using pre-intervention models. Repeated-measures ANOVAs were conducted to assess the effects of the intervention on each set of items.

Results: Results of the theoretical analyses indicated that groups differed on items assessing the understanding of false belief (F(29,1)=5.59, p<.05, partial η²=.19), imitation (F(29,1)=8.55, p<.01, partial η²=.28), and figurative language (F(29,1)=10.33, p<.01, partial η²=.26), with those undergoing the intervention experiencing greater improvement. Results of the EFA recommended extracting five components, accounting for 63.7% of the variance. Repeated-measures ANOVAs resulted in a significant effect for the fourth factor (F(29,1)=7.86, p< .01, partial η²=.21). This factor represented higher performance on items 4 (imitation/pretending) and 15 (understanding figurative language) of the NEPSY-II, with individuals in the the Experimental group seeing greater improvement. These items were similar, and unique, in that they both mentioned fingers and required students to think abstractly about this body part.

Conclusions:  Results highlight two methods for assessing change in ToM as a result of intervention with FaceSay. There was some overlap in the items that were noted by both sets of analyses. Intervention with FaceSay led to improvements in imitation and figurative language items. The construct most closely targeted by FaceSay is imitation; the game “Follow the Leader” asks students to manipulate one avatar’s face to match the facial expression of another. In addition, the game often offers instruction or reinforcement using figurative language, such as “Let’s roll!” that offers students additional exposure to this manner of speaking. Future work should continue to explore the beneficial effects of FaceSay intervention.