Assessing Social-Emotional Skills Using SENNA: Development, Psychometrics, and Validity

• In Event: Framework and Assessment Considerations for Social and Emotional Learning

Abstract

Supporting social-emotional skill development has become a key objective of contemporary educational policies to prepare students for the challenges of the volatile, uncertain, complex and ambiguous world. Social-emotional skills (SEMS) have been defined as individual characteristics that originate from biological predispositions and environmental factors, that manifest themselves as consistent patterns of thoughts, feelings, and behaviors, develop through formal and informal learning experiences, and influence different outcomes throughout the individual’s life. In the past years, several initiatives intended to integrate the multitude of different SEMS’ lists into coherent comprehensive frameworks.
Relying on a thorough empirical analysis of the underlying structure of several SEMS inventories, the authors proposed an integrative model distinguishing five broad SEMS domains, i.e. Self-management, Engaging with others, Amity, Negative emotion regulation and Open-mindedness, further grouping more specific skills. The five broad SEMS domains are conceptually close to the domains of the Big Five of personality, with Self-management affiliated with Conscientiousness, Engaging with others and Amity conceptually akin to the interpersonal dimensions of Extraversion and Agreeableness respectively, Negative emotion regulation related to Neuroticism, and finally Open-Mindedness associated with Openness to experience.
In the present contribution we will describe the development of SENNA, an instrument developed by the Institute Ayrton Senna to assess this comprehensive SEMS model, and will document its usefulness, reliability and validity for large-scale assessment in Brazilian education. SENNA is a self-report inventory, aligning 162 items in 5 broad SEMS domains, further grouped in 18 facets, with 6 items per facet phrased as typicality/identity items and 3 items phrased as self-efficacy items.
In Study 1, using a sample of over 30,000 students aged 11 to 18 years (grades 6 to 12), exploratory structural equation modeling of the 18 facets showed that these had their highest loading on the expected domain with congruency coefficients greater than .80. Internal consistencies were in the range of .70 to .80 for most scales. In Study 2, relying on a sample of 1041 students aged 10 to 18 years (from grade 5 elementary school to grade 3 in high school), SENNA was administered with alternative SEMS’ measures including Student’s Approach to Learning (SAL), Mindsets, Essential Skills, and Habits (MESH) and Character Growth Card (CGC) to examine its construct and concurrent validity. Study 3, explored the criterion validity of SENNA skills to predict standardized tests scores on math and language, using a sample of 12,987 students from 425 schools of the State of São Paulo aged 11 to 18 (grades 6 to 12). Facet’s unique contribution on multiple regressions predicting language and math respectively were around magnitude of  = .17. Overall facets increased the explained variance from 12% to 18% compared to SEMS domains only.

Authors

• Ricardo Primi, Universidade São Francisco
  Presenting Author

• Daniel Santos, Universidade de São Paulo
  Non-Presenting Author

• Oliver P John, University of California - Berkeley
  Non-Presenting Author

• Filip De Fruyt, Ghent University
  Non-Presenting Author