Results From An RCT of FaceSay Software Games

Background: In vitro generalization of social skills - i.e. to another in vitro setting - have been reported in a number of RCTs of promising technology interventions. Generalization of social skills to everday life, however, has been limited to anecdotal reports. Even the often cited Bernard-Opitz study asks at the conclusion "whether behavior learned in the computer setting generalizes to the real setting...". In this RCT, the FaceSay HFA and LFA groups showed improvements in social interactions with peers in blinded playground observations (Hauck, 1995).

Objectives:
Aim 1: The primary aim of the study was to examine the effect of FaceSay on emotion and facial recognition skill development in children with an ASD. It was predicted that the intervention groups would improve both emotion recognition (Hypothesis 1) and facial recognition (Hypothesis 2) skills after the intervention.
Aim 2: The secondary aim of the study was to examine the impact of FaceSay on social behaviors in the natural environment. It was predicted that the intervention groups would improve in observed and reported social skills after the intervention.

Methods: Forty-nine children met the inclusion criteria of a prior diagnosis of ASD (DSM IV) and confirmation of the diagnosis using the CARS scale. Participants were grouped into High Functioning Autism (HFA, N=24; KBIT > 70) and Low Functioning Autism (LFA, N=25; KBIT < 70). Participants were randomized to either treatment (FaceSay) or control (Tuxe Paint). Participants attended 12 sessions over approximately 6 weeks at their school.

Outcome Measures:
Emotion Recognition: Ekman and Friesen static photos and schematic drawings of emotions
Facial Recognition: Benton
Social Validity: Parent SSRS
Playground Observation: Blinded research assistants observed each child interacting on their normal playground during recess for five minutes twice on two separate days at baseline and twice again after the intervention using the Social Skills Observation (Hauck, 1995).

Results: For all of the Hypotheses, separate ANCOVAs were run for the LFA group and the HFA group. For these analyses the independent variable was the group (training or control). The dependent variable was the post-test score, and the covariates were the pre-test score and KBIT score.
Hypothesis 1 - Improved Emotion Recognition LFA Results: F(1, 21) = 4.52, p < 0.05. HFA Results: F(1, 20) = 29.31, p < 0.001
Hypothesis 2 - Improved Facial Recognition. LFA Results: Not significant. HFA Results: Benton-Short form F(1, 20) = 10.86, p < 0.01, Benton-Long form F(1, 20) = 4.67, p < 0.05
Hypothesis 3 - Improved social interaction in natural environment: LFA Results for SSRS F(1, 21) = 14.42, p < 0.01 and Social Skills Observation, F(1, 21) = 5.05, p < 0.05. HFA results for SSRS F(1, 20) = 4.36, p = 0.05 and Social Skills Observation, F(1, 20) = 13.61, p < 0.001.

Conclusions: The HFA group improved more than the LFA group, but both FaceSay groups showed significant improvment in playground social interactions with their peers, where it counts, an important breakthrough for the students and the field.