Construct Validity of the Revised Assessing Classroom Sociocultural Equity Scale (ASCES): Confirmatory Factor Analysis

• In Event: Poster Session 11
  In Poster Session: PS 11 Topic -Diversity, Equity & Social Justice

Abstract

Purpose
Racially marginalized (Black, Latinx, Asian/Pacific Islanders, and Native American) learners (RMLs) consistently face challenges in academic settings that have far-reaching impacts across early childhood and into secondary education. Racial inequalities and teacher-child relationships in early childhood classrooms can impact achievement, social competence, and executive function (Curenton et al., 2020). The Assessing Classroom Sociocultural Equity Scale (ASCES) is a measurement tool established in response to the need to assess equitable sociocultural interactions for racially marginalized children in early childhood classrooms (pre-K to Grade 3). Internal consistency with a five-factor solution has previously been demonstrated (Curenton et al., 2019), with Cronbach alphas ranging 0.74-0.90 and factor loadings above 0.40. Following further development, the ASCES tool is now 40 items across six proposed dimensions: 1) Challenging Status Quo Knowledge (CSQK), 2) Equitable Learning Opportunities for RMLs (ELO), 3) Inequitable Discipline (ID), 4) Connection to Home Life (CHL), 5) Personalized Learning Opportunities (PLO), and 6) RMLs’ Participation and Engagement. Items are scored on a five-point scale based on frequency of observed behaviors, from 1 (never) to 5 (consistently nearly all the time). As shown in Figure 1, we anticipate models to depict all ASCES items as related to their associated dimensions. The purpose of this project is to confirm a proposed factor structure of the ASCES tool using the revised 40-item scale in order to determine its internal consistency and validity.

Method
Video data for this project is from the My Teaching Partner-Mathematics/Science (MTP) project (Whittaker et al. 2019). Presently, randomly selected video recordings (10-20 minutes in length) of preschool classroom instruction (N=70) were rated using the ASCES measure by the developer and four research assistants. The measure developer trained research assistants via a multi-day training session in which trainees reviewed coding manuals and coded practice videos alongside certified raters. Research assistants then took a reliability test , which included coding four videos independently and achieving at least 80% adjacent agreement (scores within one point) with master-coded scores. For the proposed study, coding will continue until a sufficient sample of 150 is met to fit analytical models. With two research assistants coding approximately 5 videos a week, all videos will be coded and analyzed by February 2021.

Future Analyses
To establish inter-rater reliability, 10% of all videos (n=15) will be dual coded. A two-way, single-measures mixed effects interclass correlation coefficient (ICC) model will be calculated, and reliability values will be interpreted based on guidelines from Cicchetti (1994). Based on the previous factor structure (Curenton et al. 2019), confirmatory factor analysis using robust Maximum Likelihood (ML) estimators will be executed using the lavaan package in R (Rosseel, 2012). Factor loadings for items will be a minimum of 0.40 to be considered for each factor. These data will provide further internal consistency and validity of the revised ASCES measurement tool.

Implications
Observational tools like ASCES can be used to assess equitable interactions in the classroom as well as provide an avenue for professional development to support teachers in creating racially equitable classroom environments.

Authors

• Gullnar Syed, SRI International - District of Columbia
  Presenting Author

• Stephanie Michelle Curenton, Boston University
  Non-Presenting Author

• Shana Ernestine Rochester, University of Maryland - Baltimore County
  Non-Presenting Author