Poster #4 - Subtyping Learning Disabilities Using Data from the Early Childhood Longitudinal Study (ECLS-K: 2011)

• In Event: 2-001 - Poster Session 1 during Breakfast

Abstract

Approximately 5% of students are identified with learning disabilities (LD) in the United States. Different methods are used to identify LD in schools (e.g., discrepancy model or simple low achievement), and students with a variety of emotional, behavioral, and academic challenges end up being identified with LD. Additionally, there are valid concerns about disproportionality and LD identification by race/ethnicity, socio-economic status (SES), home language, and gender. Examining associations between risk factors and LD identification and creating subtypes of LD might address some of these issues.

The present study seeks to answer the following research questions: (1) To what extent does kindergarten performance on academic, cognitive, and behavioral measures predict identification with LD in 4th/5th grade? (2) At school entry (fall, kindergarten), are there relevant profiles/subtypes of students who go on to be identified with LD? For RQ1, I am using structural equation modeling. For RQ2, I am using weighted latent class analysis. For both questions, I include relevant covariates (e.g., gender, race/ethnicity, home language, SES, average school achievement level).

Data for this study comes from secondary data from the Early Childhood Longitudinal Study, Kindergarten Class of 2010-2011 (ECLS-K 2011). Assessments were given annually in kindergarten through fifth grade to 18,174 children who attended one of the 1,319 sampled schools across the United States. Data includes information from teachers on individual student’s “approaches to learning” and social skills. Data used also includes direct measures of executive functioning (i.e., the Numbers Reversed Task and the Dimensional Change Card Sort). The ECLS-K: 2011 also includes data from a Special Education Teacher Questionnaire, which includes child’s receipt of special education services, primary disability, and any other disabilities. Finally, data used includes information from school administrators on school resources, overall student population characteristics, and methods used to determine LD (e.g., IQ-achievement discrepancy and/or RTI).

There are several limitations to this study. First, data collection ends at fifth grade, and some students are identified with LD after fifth grade. Additionally, some more nuanced data is missing from the data set (e.g., IQ scores, why students were referred to special education). Despite the above limitations, this study has several important implications. Results from this study could help inform decisions around which kindergarteners are most likely to benefit from early interventions. Additionally, this study could introduce meaningful subtypes of students at-risk of, or with, LD. Future research could examine effects of specific interventions for these subtypes of students.

Author

• Laura Rhinehart, UCLA
  Presenting Author