Centering the Emotional Experiences of Educators of Color With Matched-Pairs Design During the COVID-19 Pandemic

• In Event: Managing Emotions in Times of Uncertainty and Stress: Findings From a Free National Course for Educators 2020–2022

Abstract

In 2020, a free online course was launched to help educators manage their emotions in times of uncertainty and stress, which included an opt-in survey of workplace well-being (De-identified, 2020;2021). The survey, designed to collect relevant information for policy decisions, provided a dataset of over 20,000 participants. However, it is well known that the U.S. teacher population is predominantly white, female identifying individuals (Hussar et al., 2020). Theory. Utilizing traditional random sampling to conduct qualitative analyses, such as deriving coding schemas, has the potential to dilute or eliminate experiences of minority voices in the group (Marshall, 1996). Objectives. To ensure findings and subsequent policy decisions do not perpetuate systemic inequities, the current study compares methodologies to capture experiences of diverse participants (N = 3,157). We hypothesized that a random sampling methodology would reveal different results than a methodology intentionally over-sampling educators of color (EOC). Method. Two samples were generated using these contrasting methodologies: Sample A was randomly generated from a representative sample of educators (N = 500); and Sample B paired EOC and white educators, matching on characteristics including grade levels taught and years of experience (N = 500). Next, six educators were recruited to identify themes within each sub-sample to create two unique codebooks. Themes across raters were synthesized to create one final, codebook for each sub-sample. Finally, four third-party post-baccalaureate researchers established reliability on the codebooks (A: Cohen’s Kappa = 0.718; B: Cohen’s Kappa = 0.746). Two of these researchers coded the remaining responses for each sample. Results. Chi square analyses revealed significant differences in both samples. In the random sample (A), white educators reported significantly more stressors in the “Limited Resources, related to Time and Workload” category than EOC (χ2 (1) = 4.04, p =.044; 17.4% EOC; 28.9% white). Additionally, within the matched sample (B) significantly more white educators reported stressors related to “Socio-emotional Needs or Physical Needs,” while significantly more EOCs endorsed stressors related to “Communication with Administration” (socio-emotional needs or physical needs: χ2 (1) = 9.22, p = .002; 29.8% EOC; 44.6% white; communication with admin: χ2 (1) = 3.18, p = .051; 36.1% EOC; 28.6% White). Significance. The sampling methodologies produced contrasting results. While the random sample indicated white educators experiencing increased stressors related to time, workload, and limited resources, centering the experiences of EOC in the matched-pairs sub-sample revealed that EOCs experience significantly more stressors related to communication with administration. In addition, this matched-pairs design no longer finds a significantly higher rate of stress related to time and workload for white educators, and rather indicates significantly higher rates of stressors related to social-emotional and physical needs for white educators, highlighting how the dilution of an identity can alter findings. Importantly, these results confirm current literature regarding school-level factors that impact the retention of EOCs, such as communication with administration (Farinde et al., 2016). It is integral that researchers are intentional in considering repercussions of methodological choices, which voices are centered in research, and the resulting impact on policy decisions.

Authors

• Morgan Mannweiler, Yale University
  Presenting Author

• Jennifer Seibyl, Yale University
  Non-Presenting Author

• Tse Yen Tan, Yale University
  Non-Presenting Author