4.1.2. Measures

The online survey collected information about teacher characteristics and 4 normed scales on health and well-being. Where relevant, timescales on the surveys were changed (e.g., from “the last 4 weeks” to “since the onset of COVID-19”).

Demographic characteristics included race, age, gender, highest degree attained, marital status, and parenting status. Teaching characteristics included years of experience, distance between home and school, and impetus for becoming a teacher. All characteristics were collected on the survey as categorical variables. Table 1 summarizes descriptive statistics for all teacher characteristics. In addition to racial identity variables, a subset of these teacher characteristics was used as covariates in regression models. For data analysis purposes, age was split into binary variables for 35 or under and 36 or older. Teaching experience was split into 0–2 years, 3–5 years, and 6 or more years because we hypothesized that very new teachers and moderately new teachers (0–2 and 3–5 years of experience, respectively) might have different experiences teaching during the pandemic compared to more experienced teachers, and because teachers of Color are more likely to leave the field within 5 years (Achinstein, Ogawa, Sexton & Freitas, 2010). Distance between home and school was split into a binary variable for whether teachers lived in the same city/county where they taught or in a different city/county; no teachers in the sample lived in the community where they taught.

The PSS is a commonly used psychological instrument measuring perceived stress by determining the degree to which situations are appraised as stressful (Cohen & Williamson, 1988). Ten Likert-style items were asked on a scale from never (0) to very often (4). Higher PSS scores indicate higher levels of perceived stress. Norm/reference group information for all scales is provided in Table 2 .

Health and well-being scales.

Notes:

The MOS-36 is a generic patient-reported outcome measure that quantifies health status and measures health-related quality of life (Ware & Sherbourne, 1992). The scale includes 9 subscales comprised of a mix of question types (e.g., Likert style, true/false), each transformed to a 0–100-point scale. The subscales measure: physical functioning, role limitations due to physical health, role limitations due to emotional problems, energy/fatigue, emotional well-being, social functioning, general health, and health change over time (RAND Corporation, n.d.). In the current sample, the subscales had high internal consistency (Cronbach's α=0.80); we thus took the mean of the 9 subscales to create a single overall scale for use as a regression model outcome variable. Across subscales, lower scores indicate poorer health/well-being outcomes.

PROMIS measures self-reported alertness, sleepiness, tiredness, and functional impairments associated with sleep problems (National Institutes of Health, 2021). This Likert-style survey includes eight items ranging from not at all/never/very poor (1) to very much/always/very good (5). Raw scores are generated by taking the sum of all items, ranging from 8–40. Adjusted standardized t-scores range from 0–100 and are normed around a mean of 50 and standard deviation of 10 (HealthMeasures, 2021). Higher PROMIS scores indicate greater sleep disturbance.

WHOQOL-BREF is an abbreviated generic Quality of Life Scale developed by the World Health Organization (1996). It includes 26 Likert-style items ranging from not at all/very poor/very dissatisfied/never (1) to completely/very good/very satisfied/always (5). These questions are sorted into 4 quality of life domains ranging from 0–100: physical health, psychological well-being, social relationships, and environment. In addition, 2 items ask about overall perceived quality of life and health (also transformed to a 0–100 scale). In the current sample, the subscales had high internal consistency (Cronbach's α=0.86) so we took the mean of the 6 subscales to create a single overall scale for use as a regression model outcome variable.