Congruence Among Parent and Teacher Ratings and Observational Assessments of Social-Communication and Play In Preschoolers with ASD

Background:  An important question in intervention research pertains to the impact on individuals’ daily functioning. Subjective methods of measurement generally ask post-hoc whether an intervention made a difference in a person’s life.  A more objective approach entails collecting pre- and post-intervention data on everyday functioning. This study reports on pilot testing of a parent-teacher rating scale to assess whether preschoolers with autism spectrum disorders (ASD) make noticeable gains on social-communication and play over a school year. 

Objectives:  (a) Examine correlations between parent and teacher ratings of social-communication and play at the beginning (pretest) and end (posttest) of the school year, and between teacher and parent rating change scores from pretest to posttest; (b) examine the correlations of teacher and parent ratings of social-communication and play skills with direct observational assessments of social-communication and play; and (c) determine whether teachers and/or parents rated children as significantly higher in social-communication and play at posttest than pretest. 

Methods:  Parents and teachers of 32 preschoolers with ASD were given a 5-minute video of a typically-developing preschooler engaging in toy play with an adult, along with a 5-item Likert scale. Each item was rated on a 10-point scale (“strongly disagree” to “strongly agree”) at pretest and posttest.  Items asked the adult to rate their student/child in comparison to the child in the video, e.g., “My student/child engages with people in similar ways as the child in the video.” Social-communication and play skills were directly assessed at pretest and posttest using a researcher-developed scoring system (Dykstra et al., 2011) of social-communication behaviors observed during the ADOS, and the Structured Play Assessment (Ungerer & Sigman, 1981).

Results:  At pretest and posttest, teacher and parent ratings were significantly correlated for 3 of 5 rating items, and for the average rating across all items, rhos from .42 to .57, p-values < .05; however, the correlation between parent and teacher rating change scores (i.e., difference between pretest and posttest ratings) approached 0. At pretest, average teacher and parent ratings correlated significantly with observational measures of social-communication and play, rhos from .37 to .63, p-values < .05. At posttest, average teacher ratings also correlated significantly with observational measures of social-communication and play, rho(31) = .53 and .47, p = .002 and .008, respectively. Average posttest ratings by parents were not significantly correlated with the observational measure of play, rho(23) = .32, p = .10, but were significantly correlated with the observational measure of social-communication, rho(23) = .42, p = .04. Both teachers and parents gave children significantly higher mean ratings at posttest than pretest, t(30) = 6.2, p < .001, and t(21) = 2.9, p = .008, respectively. 

Conclusions:  The moderate congruence between parent and teacher ratings, and between social validity ratings and observational measures, suggests that the rating scale is a valid measure of children’s social-communication and play as manifested in everyday activities at school and home. Such social validity measures can add an important and unique dimension to the assessment of change in intervention studies.