Many Pathways to Well-Being: The Development and Validation of the Sources of Well-Being Scale for Teachers

• In Event: What We Learned Measuring Educator Well-Being During the 2021–2022 School Year: Social and Emotional Learning Assessment Science and Practice

Abstract

Objective. The objective of this study was to test the structural, convergent, and predictive validity of a new measure designed to capture core factors underlying teacher well-being, called the Sources of Well-Being (SOW) scale. Theoretical Foundation. Based on the Job Demands-Resources (JD-R) model of occupational stress, teacher well-being is fundamentally a function of the ratio of job demands to school supports (Granziera, Collie, & Martin, 2021). Rooted in JD-R, and the empirical literature on teacher burnout and well-being (e.g., McIntyre, McIntyre, & Francis, 2017), 55 items were co-developed by a team of researchers and experienced educators based on teachers' open-ended reporting about their leading sources of stress and well-being at school (see Table 3.1 for sample items). Method. Our sample (N = 678) was composed primarily of white (87.5%) female (82.6%) pre-K-12 teachers (96.9%). All data were self-reported and collected online in 2021. We employed confirmatory factor analysis to test the structural validity, and structural equation modeling to test the convergent and predictive validity of the SOW. For all analyses, we used maximum likelihood estimation adjusting the standard errors for school-level nesting in Mplus (version 8.8). Results. We found that a nine-factor model provided an acceptable to fit the data (x2 = 3965.96, p < .001; RMSEA =.05; CFI =.88; SRMR =.06), and consisted of the following subscales: Teacher-Student Relationships (standardized factor loading range = .40 to .83), Student Academic and Social-Emotional Behavior (standardized factor loading range = .53 to .83), Teacher-Principal Relationships (standardized factor loading range = .61 to .88), Teacher-Colleague Relationships (standardized factor loading range =.60 to .86), Teacher-Family Relationships (standardized factor loading range =.79 to .82), Educational Policy Demands (standardized factor loading range =.51 to .71), Time Pressure (standardized factor loading range=.58 to .81), School Resources (standardized factor loading range = .60 to .76), and Physical and Emotional Safety (standardized factor loading range = .54 to .82). Regarding convergent validity, the nine SOW factors were associated with previously validated measures of related constructs (bs = .24 to .92, ps <.01), including the Teacher Well-Being Scale (Collie et al., 2015), Effective School Functioning Scale (Caprara et al. 2003), and Delaware School Climate Survey for Teachers (Bear et al., 2014). Further, most SOW factors predicted teacher emotional exhaustion (Maslach et al., 1996; bs = -.11 to -.52, ps <.05), job satisfaction (New Teacher Center, 2017; bs =.24 to .26, ps <.01), and/or turnover intentions (Moeller et al., 2018; bs = -.14 to -.42, ps<.05). The SOW factors that were most predictive of teacher well-being were Physical and Emotional Safety, Time Pressure, and Student Academic and Social-Emotional Behavior. Significance. Together, these results provide preliminary support for the structural, convergent, and predictive validity of the SOW. Also, mean completion time of the SOW was 2.50 minutes, suggesting it measures multiple factors central to teacher well-being economically, making it useful for brief studies and applications in time-pressured schools. The utility of the SOW to guide efforts to foster teacher well-being, as well as for advancing the science of school well-being, is discussed.

Authors

• James Floman, Yale University
  Presenting Author

• Annette R. Ponnock, Yale University
  Non-Presenting Author

• Marc A. Brackett, Yale University
  Non-Presenting Author