Search
Browse By Day
Browse By Time
Browse By Panel
Browse By Session Type
Browse By Topic Area
Search Tips
Register for SRCD21
Personal Schedule
Change Preferences / Time Zone
Sign In
X (Twitter)
Introduction: Children with autism spectrum disorder (ASD) demonstrate highly variable writing skills (Finnegan & Accardo, 2018; Zajic & Asaro-Saddler, 2019; Zajic & Wilson, 2020). Existing studies have emphasized analyzing written products, leaving less known about engagement during commonly implemented psychoeducational assessments. Further, available studies often adopt variable-centered rather than person-centered approaches. This presentation describes both approaches to examining task engagement (total time engaged, visual engagement with task components, and time spent writing) during a writing assessment. The variable-centered approach compares task-level and within-task engagement in children with ASD to typically developing (TD) peers and peers with attention-deficit/hyperactivity disorder (ADHD). The person-centered approach examines profiles of task engagement and writing performance within the ASD group using latent profile analysis.
Hypotheses: Using a variable-centered approach, the ASD group will show heterogeneous engagement behaviors that appear lower in quality than TD peers and similar to peers with ADHD. Using a person-centered approach, latent profiles will highlight heterogeneous profiles within the ASD group that differ in time spent engaged with different task components and in their resulting writing scores.
Study Population: Data come from a large study of school-age children that included children with ASD (n = 60), children with ADHD (n = 32), and TD children (n = 29). Children with ASD and ADHD came into the study with clinical diagnoses that were confirmed using reliable, valid assessments. All children were between the ages of 9 and 17 (M = 12.80, SD = 2.16) and had FIQ estimates above 73 (ASD: M[SD] = 98.73[15.13], ADHD: M[SD] = 99.31[15.00], TD: M[SD] = 115.62[12.82]. Groups did not differ on age or gender but did on IQ.
Methods: All children completed the spontaneous writing task of the Test of Written Language, 4th Edition, a narrative assessment scored for writing conventions and quality. All samples were scored by multiple trained raters (αs > .84). A video camera recorded writing task activity that was human coded using Noldus Observer XT 14.0. All samples were coded (30% subsample interrater: αs > .90) for visual task behaviors (looking at picture, at outline, at draft, and away) and writing behaviors (writing on outline and draft). Variable-centered analyses (SPSS 25) examined group differences in overall task time and task time engaged in visual and writing behaviors using mixed analysis of covariance (ANCOVA; controlling for FIQ). Person-centered analyses (Mplus 8) examined latent profiles for the ASD group using indicators of task time, visual task behaviors, writing behaviors, and writing scores.
Results: Variable-centered analyses (see Figure 1) found the ASD group wrote lower quality narratives and spent the least amount of time completing the task. The ASD group was lowest in within-task engagement with looking at and writing on the draft. Person-centered analyses (see Figure 2) found a 4-profile model fit best: 1) Low engagement, below-average writing scores (27.1%), 2) High engagement, above-average writing scores (39.1%), 3) Average engagement, above-average writing scores (18.5%), and 4) High engagement, below-average writing scores (15.3%). Findings highlight group-level challenges using a variable-centered approach but profile-specific strengths and difficulties using a person-centered approach.
Matthew Zajic, Teachers College, Columbia University
Presenting Author
Emily J. Solari, University of Virginia - Charlottesville
Non-Presenting Author
Nancy McIntyre, University of Central Florida
Non-Presenting Author
Lindsay Lerro, The Listening Center
Non-Presenting Author
Peter Clive Mundy, University of California - Davis
Non-Presenting Author