When districts bring in SEL assessment: A cautionary tale for school psychologists

With an increase in social-emotional learning (SEL) interventions in schools, comes the need for social-emotional learning assessments. With this rise in SEL assessment, there has been an increase in third-party assessment companies offering their services to school districts. School psychologists, as experts trained in intervention and assessment, are often excluded from the initial decision-making process and are often brought in at a later stage, where the opportunity to leverage their knowledge and expertise is missed. This dissertation aimed to explore if an SEL assessment that was conducted, effectively measured the intended constructs. Through a confirmatory factor analysis (CFA), this dissertation aimed to examine whether the selected measurements effectively captured the intended constructs that a school district sought to assess. The study examined the test-retest reliability and Standard Error of Measurement (SEM) to ensure the tool’s efficacy in detecting meaningful changes across the different grade levels. Lastly, a test of proportions was conducted to assess the variance between the number of students flagged by the assessment as needing intervention and the students identified by the school counselors as needing such support. Through a CFA, this study found a moderate to poor fit between the observed and predicted covariance matrices. This means that there was not a good fit between the scales chosen to assess self-awareness and self-management. The internal consistency of variables for the social awareness construct ranged from questionable to excellent (α = 0.642-0.917) across grade levels while the internal consistency for the self-management construct ranged from acceptable to poor (α = 0.505-0.788) across grade levels. Test-retest reliability revealed a moderate negative relationship (Pearson’s r = -0.374, p = 0.042) for grades K-2, a weak positive relationship for grades 3-6 (Pearson’s r = 0.295, p = 0.002) and a strong positive relationship for grades 7-8 (Pearson’s r = 0.807, p < .001). The test of proportions found no significant difference between the students identified by the assessment as needing intervention and the students identified by the counselors as needing intervention. It is recommended that further research be completed to assess how this tool can be useful for educators and the assessment of intended constructs.