Measurement Invariance of Foundational Learning Skills in Kindergarten Boys With and Without Autism: A Secondary Data Analysis

• In Event: 2-009 - Using Secondary Data to Identify Educational Outcomes among Special Populations

Abstract

Background
Autism Spectrum Disorder (ASD) is characterized by pervasive impairment in social communication and the presence of fixated interests and/or repetitive behaviors (APA, 2013). Current figures estimate that 1 in 68 U.S. children has ASD (Christensen et al., 2016); yet, few studies have examined constructs related to classroom participation and learning skills in children with ASD. There is a critical need for increased understanding of assessments that will guide achievement and skill development for these children (Fleury et al., 2014). This study evaluated the structure of foundational learning skills based on items from two commonly used teacher-report measures and examined whether measurement differences exist between boys with and without ASD.
Method
Participants
Drawing from a nationally representative sample of children participating in the Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011; NCES, n.d.), the sample for this secondary analysis included ≈5,660 boys with teacher-report data available during spring of kindergarten who met subpopulation criteria. The ASD group (≈1.6%) included boys with a primary disability of autism as identified by their IEP, as indicated during any of the first seven waves. The non-ASD group included boys who did not have an IEP or parent-reported autism diagnosis.

Measures
Fifteen items from the CBQ-SF (Putnam & Rothbart, 2006) and ECLS-K Approaches to Learning subscale (NCES, 2010) were combined to measure three hypothesized constructs: 1) strong concentration; 2) high distractibility; and 3) learning readiness.

Procedures
Confirmatory factor analysis was performed to evaluate evidence of configural (pattern), metric (loadings), and scalar (intercepts/thresholds) invariance. Sampling weights and linearization variance estimation were used to account for the complex sampling design.
Results
Global fit information is presented in Table 1, and final parameter estimates in Figure 1. Preliminary analyses supported the hypothesized structure for both groups and indicated partial invariance. Four items—adapts (AL21), becomes absorbed (CBQ6), works independently (AL15), and follows instructions (CBQ10)—exhibited noninvariance, indicating that teachers interpreted these skills differently for boys with and without ASD. There were also significant factor mean differences, with the ASD group exhibiting less foundational learning skills, on average.
Discussion
This study contributes to a limited body of literature evaluating children with ASD in educational settings and highlights the advantage of using large, secondary datasets to study small subpopulations. Findings documented measurement-related differences in teachers’ ratings of learning skills between boys with and without ASD, suggesting the importance of considering ASD symptomology when evaluating children’s participation and learning.

Authors

• Nicole Sparapani, University of California, Davis School of Education and the MIND Institute
  Presenting Author

• Natalie A Koziol, Nebraska Center for Research on Children, Youth, Families & Schools

• Bonnie Mackintosh, Nebraska Center for Research on Children, Youth, Families & Schools