17211
A Novel Teacher Implemented Protocol to Assess Early Social Communication Skills and Play in Preschool Children with Autism
Initiations of early nonverbal social communication skills including gestures to request (behaviour regulation) and share (joint attention) as well as play skills are fundamental to young children’s development. Yet, educators and community practitioners have limited access to validated tools to assess these foundational skills and select developmentally appropriate skill targets.
Objectives:
The assess the feasibility and validity of a novel teacher implemented brief assessment designed to capture the presence of preschool children’s nonverbal social communication and play skills in classroom settings and facilitate teachers’ selection of developmentally appropriate target skills for students.
Methods:
Participants. Six teachers and two paraprofessionals from six public preschool classrooms administered the assessment with a total of 68 preschool students with autism who were enrolled in a larger intervention study.
Measures. Three assessments were administered at entry into the larger study including two well-established protocols used in research: the Early Social Communication Scale (ESCS: Mundy et al., 2003) and Structured Play Assessment (SPA: Ungerer & Sigman, 1981). In addition, a novel assessment was designed to function as a brief protocol of the ESCS and SPA for use by teachers to identify skill targets in core areas of early development.
Outcomes. Children’s spontaneous initiations of joint attention (including coordinated joint looks, points, shows, and gives), behaviour regulation (give and point to request) as well as four levels of play skills (simple, combination, pre-symbolic, and symbolic) were examined. Teachers recorded the presence or absence of the behaviours on a three-point scale (never, once, two or more).
Results:
On average, teachers delivered the novel assessment with 86.57% fidelity (SD= 8.15%). Logistic regression was applied to examine the probability of teachers’ agreement of their in vivo selection of children’s skill targets with the researcher’s target selected after review of the videotaped administration of the brief assessment. The probability of skill target agreement between each of the eight teachers and the researcher varied by skill domain including: JA skills (0.40-1.00), BR skills (0.30-1.00), and play skills (0.40-0.60). Agreement on JA skills and BR skills was greater than chance (JA: chi2 (1)= 4.76, p<0.05; BR: chi2 (1)= 11.53, p<0.01) while play was not.
Further, agreement between researcher selected JA target skills from the brief assessment and JA targets selected from the ESCS was above chance (chi2 (1)= 5.06, p<0.05) while agreement on BR targets was not (p=0.32). Finally, agreement between the researcher selected play target and the target obtained from the SPA was significantly above chance (chi2 (1)= 5.06, p<0.01).
Conclusions:
Although lower agreement occurred on BR target selection, findings from this preliminary study indicate that similar JA and play targets were obtained from the brief assessment and the established research protocols. Several factors related to variability in BR will be discussed including ESCS administration fidelity and prompting protocols.
Further, findings indicate teachers could learn to administer the assessment and often select accurate JA and BR target skills yet agreement not above chance was found for play. Further training regarding developmental play levels may enhance teachers’ accurate identification of play targets.