Poster #123 - Approaches to Learning in the ECLS-K: Defining and Measuring the Construct

• In Event: 1-169 - Poster Session 04
  In Poster Session: PS 04 Section - Methods, History, Theory

Integrative Statement

Children’s Approaches to Learning (AtL) has been identified in research and policy as a key domain of children’s development and school readiness. Nevertheless, there remains a lack of consensus around the exact definition and specific dimensions of AtL. Measures of the construct are few and tend to vary in both content and psychometric rigor. One of the most widely used measures of AtL is the brief scale developed for the Early Childhood Longitudinal Study – Kindergarten Cohort (ECLS-K). Scales scores have consistently been found to predict important child developmental and educational outcomes in both cross-sectional and longitudinal studies, suggesting that this measure taps a critical component of children’s early development. Yet, despite its broad application in national and other research over the past 20 years, very little has been published regarding the scale’s psychometric properties and validity. No studies of convergent or divergent validity have been conducted to define the construct being measured, and though it has been employed longitudinally in both rounds of the ECLS-K, no evidence has been put forth regarding its ability to measure the AtL construct consistently over time. The current study aimed to provide much-needed information regarding the measurement and validity of the ECLS-K AtL scale, as well as to clarify the construct of AtL generally. Research questions included: (a) What is the degree and distribution of shared variance between AtL scores and various measures of related constructs? and (b) Does the scale demonstrate longitudinal invariance over three years from kindergarten through second grade? Canonical correlation analyses were conducted using spring semester data in each of three academic years – kindergarten, first grade, and second grade – to measure the association of AtL scale scores with concurrent measures of children’s executive functioning, self-regulation, social skills, academic performance, and internalizing and externalizing behavior problems. Measurement invariance was assessed via structural equations modeling across six time points (fall and spring of the same three years). Large, statistically significant correlations were observed between children’s AtL and measures of self-regulation and social skills. Further, results indicate that the scale is longitudinally invariant, making it appropriate for growth analysis and other longitudinal applications. Implications of these findings for further clarification of AtL as a construct and validation of the ECLS-K AtL scale, in particular, are discussed.

Author

• Katharine Buek, University of Texas Systems
  Presenting Author