The Psychometric Properties of the Berkeley Assessment of Social Emotional Learning - Youth Form.

• In Event: Co-Developing Approaches for Critical Data-Based Decision Making to Improve Social and Emotional Learning Practices for Equity

Abstract

Background: Quality assessments of social emotional learning (SEL) are required to provide formative data to educators seeking to improve the equitable practice of SEL using critical data-driven decision making (cDDDM; Dodman et al., 2023). Yet, there are a number of gaps in the literature on SEL assessments: 1) they overly focus on individual social emotional competencies (SECs), ignoring critical school contexts, 2) are rarely designed to be used for cDDDM, and 3) few consider their role in promoting equity. Given inequitable school contexts, considering the school climate students are embedded in is critical to successful SEL measurement and implementation and, in turn, student well-being. The inclusion of climate (i.e. safety and connection, opportunities for student voice and leadership, perceptions of culturally responsive practices) captures critical information about (in)equitable school experiences in line with cDDDM, and its pairing with SEC and well-being represents a critical step forward from the literature.

Objectives: In an effort to fill these gaps, we co-developed the "Berkeley Assessment of Social Emotional Learning - Youth Voice (BASEL-YV)" with practitioners to measure students’ perceptions of school climate, SEC, and well-being. Race and gender inequities may impact student engagement with SEL and how measures are interpreted by students from various identities, rendering the establishment of measurement invariance a critical step prior to making use of data for decision-making (DiPerna et al., 2022). The present study reports on the psychometric measurement invariance of the [SURVEY].

Methods: Data are from 1627 students (927 middle and 745 high school students) in California. 53.3% were self-identified as female, 46.4% as male, and 0.3% as gender X. 30.1% self-identified as Hispanic, 28.7%, White, 25.3%, Asian, 11.7% multiracial, 3.3% Black/African American, 0.7% Native Hawaiian or Pacific Islander, and 0.2% AIAN. The "BASEL-YV" consists of 47 items: 11 items measure well-being factors, 13 items measure school climate factors, and 23 items measure SEC factors. The response options are 1 = “NO!,” 2 = “no,” 3 = “yes,” 4 = “YES!”. Descriptive statistics, confirmatory factor analysis (CFA), reliability, and measurement invariance testing by grade level (middle/high), gender (male/female), and race (Asian, Hispanic, Multiracial, White, Black and Other) are reported.

Results: Mean scores ranged from 2.12 to 3.44 and standard deviations ranging from .62 to .95. CFA confirmed that the proposed measurement model fit the data (CFI= .91, TLI= .90, RMSEA=.06), informing the removal of one cross-loaded item. Latent factors of each construct were found to be invariant across gender, race, and school level. Resulting scales were found reliable (Cronbach’s alpha .71-.84, except for a single 2-item scale = .61).

Authors

• Tiffany M. Jones, Colorado State University

• Addison Duane, Sacramento State University

• Juyeon Lee, The University of Hong Kong

• Ashley Metzger, University of California - Berkeley

• Kamryn Morris, University of California - Berkeley

• Amia Nash, University of California - Berkeley

• Hui Hu, University of Hong Kong

• Chenxiao Wang, University of Wisconsin - Madison

• Valerie Shapiro, University of California - Berkeley