Abstract
Educator mental health sits at the intersection of multiple pressing educational issues. We are among the first to provide estimates of school system employee (SSE) stress, anxiety, and depression during the COVID-19 pandemic. Most participants reported clinically meaningful anxiety and depressive symptoms (77.96% and 53.65%, respectively). Being in the lowest strata of family income was associated with higher stress, a greater likelihood of clinically significant depressive symptoms, and reduced intentions to continue in the same job, portending the current staffing shortages affecting schools. Supporting SSE mental health should become a policy priority.
Keywords: COVID-19, educational policy, mental health, psychology, regression analyses, retention, school systems, stress/coping, teachers, teacher cognition
The COVID-19 pandemic has placed unprecedented strain on school systems. Teachers are experiencing high levels of burnout (Pressley, 2021), and many school system employees (SSEs) are considering leaving their jobs (Kaufman et al., 2020). Although SSE mental health sits at the intersection of multiple pressing educational issues, including staff attrition and retention, school climate, and occupational effectiveness, research on the topic is scarce (Harding et al., 2019). The purpose of this study is to provide some of the first estimates of SSE stress, anxiety, and depression during the pandemic, using standardized instruments.
Method
Through public web postings and a midsize urban district’s newsletter, we recruited 662 SSEs (65.11% [431] instructional staff, 48.34% [320] lead classroom teachers) in Wisconsin to participate in a fully remote randomized controlled trial of a meditation-based well-being smartphone application (Hirshberg et al., 2022). Participants came from 40 districts and were slightly more racially diverse and female than SSEs across the state (Supplemental Materials [SM] Table S1). Any preK–12 SSE (e.g., coach, food service, teacher, administrator) was eligible (see Online SM for details). In secondary analyses of pretest data (linked in SM) collected between mid-June and August 2020, a relatively stable period of virus transmission in Wisconsin but a time of considerable social unrest nationwide, we examined four research questions:
What are the average levels of stress, anxiety, and depressive symptoms?
What is the prevalence of clinically meaningful and severe anxiety and depressive symptoms?
Are there associations between symptoms and SSE characteristics?
Are there associations between symptoms and SSE characteristics and intentions to continue in the same professional role in the coming year (2020–2021)?
Psychological stress over the prior 30 days was assessed with the Perceived stress scale (10 items; Cohen et al., 1983). Scores less than or equal to 23, 24–36, and 37 or higher represent low, moderate, and high stress, respectively. The Patient Reported Outcomes Measurement Information Systems (PROMIS; 4–28 items, adaptive; Pilkonis et al., 2011) were used to assess anxiety and depressive symptoms over the prior 7 days. PROMIS scale T-scores have a population average of 50 and a standard deviation (SD) of 10. Scores less than 55, 55–59.9, 60–69.9, and 70 or higher represent normal, mild, moderate, and severe symptom categories, respectively. Mild or greater symptoms are considered clinically meaningful. Details on measures are provided in SM Table S2.
We calculated sample averages and the proportion of the sample in each symptom severity category for stress, anxiety, and depressive symptoms. Using multivariable and logistic regression analyses, we regressed participant age, race, gender, education level, family income level, and employment category (i.e., classroom instructional staff/other) onto stress and, separately, the likelihood of clinically meaningful anxiety and depressive symptoms. Using logistic regression, we regressed all covariates and clinically meaningful anxiety and depressive symptoms on the intention to continue in the same job next year. Listwise deletion was used because no more than five participants (less than 1%) were missing data on any variable except for race (n = 21, 3.17% missingness).
Results
Average levels of stress (28.78, SD = 5.79), anxiety (59.92, SD = 7.03), and depressive (55.43, SD = 6.31) symptoms were in the mild/moderate range. Anxiety and depressive symptoms were respectively 1 and 0.5 SD above the population average. A large majority of SSEs reported clinically meaningful anxiety symptoms (77.96%). Of these, 61.60% (48.02% of sample) reported moderately severe or greater symptoms. A majority of SSEs also reported clinically meaningful depressive symptoms (53.65%). Of these, 50.34% (27.05% of sample) reported moderately severe or severe symptoms (Table 1).
Table 1.
Stress, Anxiety, and Depressive Symptoms Means (SD) and Prevalence Rates Overall and by Demographic Subgroup (n / %)
| Outcome | Stress | Anxiety | Depression | |
|---|---|---|---|---|
| Mean (SD) | 28.78 (5.79) | 59.93 (7.03) | 55.43 (6.31) | |
| Low/None to mild | 126 (19.0%) | 145 (21.9%) | 305 (46.1%) | |
| Moderate | 478 (72.21%) | 197 (29.76%) | 175 (26.4%) | |
| Moderately severe | NA | 285 (43.1%) | 174 (26.3%) | |
| High/Severe | 53 (8.01%) | 31 (4.7%) | 4 (0.6%) | |
| Missing | 5 (0.76%) | 4 (0.6%) | 4 (0.6%) | |
| Covariate | Stress | Anxiety | Depression | |
| Age (years) | 42.58 (10.67) | |||
| Race/Ethnicity* | ||||
| American Indian/Alaskan Native | 4 (0.60%) | 33.75 (2.99) | 66.73 (4.47) | 63.58 (4.43) |
| Asian/Pacific Islander | 13 (1.96%) | 31.25 (6.58) | 62.94 (6.98) | 58.45 (6.54) |
| Black/African American | 26 (3.93%) | 30.17 (4.80) | 60.00 (9.77) | 56.47 (7.16) |
| Hispanic/Latino | 27 (4.08%) | 29.29 (7.30) | 58.79 (7.21) | 55.23 (7.63) |
| White/Caucasian | 571 (86.25%) | 28.59 (5.72) | 59.88 (6.88) | 55.25 (6.17) |
| Two or more races | 21 (3.17%) | 29.30 (6.19) | 59.22 (6.80) | 55.75 (6.42) |
| Gender* | ||||
| Female | 578 (87.3%) | 28.97 (5.78) | 60.20 (6.85) | 55.63 (6.33) |
| Male | 79 (11.9%) | 27.59 (5.62) | 57.90 (7.99) | 54.09 (5.96) |
| Nonbinary or other | 5 (0.76%) | 25.25 (7.63) | 60.78 (7.35) | 53.30 (7.43) |
| Highest education level | ||||
| Less than college degree | 70 (10.6%) | 28.80 (6.64) | 59.50 (9.51) | 54.77 (7.76) |
| College degree | 203 (30.7%) | 28.99 (5.60) | 60.44 (7.04) | 55.97 (6.11) |
| Advanced degree | 386 (58.3%) | 28.69 (5.72) | 59.79 (6.45) | 55.30 (6.11) |
| Family income (U.S. dollars) | ||||
| < $40,000 | 58 (8.76%) | 31.21 (5.44) | 61.70 (9.11) | 58.42 (7.34) |
| $40,000–$80,000 | 216 (32.6%) | 29.59 (5.81) | 60.69 (6.71) | 56.13 (6.05) |
| $80,000–$100,000 | 107 (16.2%) | 28.33 (5.54) | 59.98 (5.77) | 54.97 (5.99) |
| > $100,000 | 277 (41.8%) | 27.79 (5.72) | 58.94 (7.12) | 54.47 (6.17) |
| Employment category | ||||
| Non-instructional staff | 183 (27.64%) | 28.28 (6.03) | 59.69 (7.20) | 54.84 (6.87) |
| District personnel | n = 25 | 28.27 (6.88) | 59.24 (6.51) | 55.11 (7.70) |
| School support | n = 109 | 28.57 (5.60) | 60.40 (6.78) | 55.04 (6.58) |
| School staff | n = 39 | 26.71 (6.70) | 57.99 (8.79) | 54.00 (7.61) |
| Instructional staff | 431 (65.11%) | 28.89 (5.76) | 60.01 (6.93) | 55.60 (6.11) |
| Lead teachers | n = 320 | 29.06 (5.67) | 60.22 (6.61) | 55.50 (5.93) |
| Assistants | n = 88 | 28.32 (5.81) | 60.02 (6.83) | 56.03 (6.83) |
| Specialists | n = 23 | 28.48 (6.27) | 59.88 (7.33) | 55.23 (6.63) |
Note. Non-instructional staff: District personnel include district administrators (e.g., curriculum directors) and district staff (e.g., district secretaries). School support include occupational therapists, librarians, teacher instructional coaches, and afterschool program employees. School staff include building custodial workers, bus drivers, food service, and secretaries. Instructional staff: Lead teachers include regular education, special education, and specials (e.g., art, physical education) lead classroom teachers. Assistants include classroom aides, English language learner support staff, substitute teachers, and special education assistants. Specialists include literacy specialists and speech and language pathologists. NA = the category does not exist; SD = standard deviation.
Participants were able to endorse more than one category.
In regression analyses (see Figure 1; SM Table S3), SSE age (β = –.07, p = .005) and family income were significantly associated with stress. SSEs with family income $80,000–$100,000 and above $100,000/year had significantly lower levels of stress than did participants with family income below $40,000/year (β = −2.34, p = 0.028 and β = −2.86, p = 0.003, respectively). Age (β = −0.03, p = 0.003, odds ratio [OR] = 0.98, 95% CI [0.96, 0.99]) and family income ( > $100,000 – < $40,000/year, β = −0.78, OR = −2.18, 95% CI [0.22, 0.91]) were significantly associated with lower odds of clinically meaningful anxiety symptoms. The odds of clinically meaningful depressive symptoms were significantly lower in all income levels greater than $40,000/year ($40,000–$80,000, β = −0.94, p = 0.007, OR = 0.40, CI [0.20, 0.77]; $80,000–$100,000, β = −1.31, p = 0.001, OR = 0.29, CI [0.12, 0.59]; > $100,000, β = −1.34, p < 0.001, OR = 0.27, CI [0.13, 0.54]) and in male SSEs (β = 1.18, p = 0.004, OR = 3.24, CI [1.54, 7.79]).
Figure 1.
OR from logistic regressions of clinically meaningful depressive (A) and anxiety (B) symptoms, and intention to continue in the same professional role (C).
Note. Reference levels for educational attainment and family income are “less than a college degree” and “less than $40,000/year,” respectively. In A and B, OR less than 1 and greater than 1 represent lower and greater odds of clinically meaningful depressive and anxiety symptoms, respectively. In C, OR less than 1 and greater than 1 represent lower and greater odds of intending to continue in the same professional role next year. Upper bound for moderate stress confidence interval (C) is 219.43. All variables with confidence intervals that do not cross the dotted vertical line at 1 and in C greater than $100,000/year and $40,000–$80,000/year are statistically significant (p < .05). OR = odds ratio.
Of the 566 SSEs who reported future employment intentions, 30 (5.30%) did not intend to continue next year. Controlling for covariates and clinically meaningful anxiety, SSEs with clinically meaningful depressive symptoms had 0.33 times the odds of intending to continue (CI [.12, .88], p = .027). Controlling for covariates and clinically meaningful anxiety and depressive symptoms, compared to SSEs with family incomes less than $40,000/year, odds of intending to continue were 3.93 CI [1.00, 15.28], p = 0.045, and 3.69 CI [0.94, 13.27], p = 0.050 times higher among those with incomes of $40,000–80,000 and more than $100,000/year, respectively.
Discussion
This study is among the first to provide estimates of SSE mental health during the COVID-19 pandemic, using standardized instruments. The picture is concerning. Meaningful levels of anxiety and depressive symptoms were normative (i.e., > 50% of the sample), with 48.02% and 27.05% of SSEs reporting moderate or severe anxiety or depressive symptoms, respectively. These prevalence rates are higher than those observed in general population samples collected during the same period (Czeisler, 2020). Consistent with population trends in depression, it is notable that SSEs identifying as female were significantly more likely to report clinically meaningful depressive symptoms.
Complementing research characterizing teacher burnout and COVID-related anxiety during the school year (e.g., Pressley, 2021), we observed concerning levels of SSE general anxiety and depression during the summer of 2020. It was impossible to know at the time, but the years that followed included the two of the most severe COVID-19 transmission waves yet, escalating social unrest, and staffing shortages that forced schools nationwide to reduce critical services. Perhaps portending these staffing shortages, SSE families earning less than $40,000/year reported significantly more stress, were more likely to report clinically meaningful depressive symptoms, and were less likely to intend to return in their job relative to higher earning peers. This group was predominately core instructional staff (> 70% classroom teachers or assistants; SM Table S4). Most teachers in this income range were likely early career, a group known to have high rates of career attrition. Across the sample, clinically meaningful depressive symptoms were associated with lower odds of intending to return. These data suggest that the lowest income SSEs, which may include early career teachers, are at elevated risk for poorer mental health outcomes and occupational intentions and should be a focal point of future policy.
Generalizing these results to SSE nationwide should be done cautiously. This relatively small convenience sample of Wisconsin SSE was predominately instructional staff (65%), female (87%), and White (86%). In addition, these data were collected at a particular point in time under a unique set of circumstances. However, the COVID-19 pandemic is only one among several factors that might affect SSE mental health. For example, the murder of George Floyd in May 2020 and continuing struggle for racial justice may also affect SSE mental health, perhaps particularly for SSEs of color. It is possible that we overestimated symptoms among Wisconsin SSEs and yet underestimated symptoms of SSEs nationwide, who are racially more diverse. In addition, working in educational systems may be more difficult today than it was in the summer of 2020. These results therefore underscore the importance of continued research on SSE mental health so that decision-makers are informed about the needs of their workforce and how to best support them. They also highlight the importance of sampling diverse SSEs to enable a more complete picture of the relationships between SSE characteristics (e.g., race, gender, employment category, income) and mental health.
Schools are tasked with equitably recovering lost student learning resulting from the pandemic in students suffering from greater mental health concerns. Psychologically healthy teachers and SSEs are an essential part of any conceivable solution. These data suggest that additional SSE supports are necessary and may be a prerequisite to student and educational system pandemic recovery efforts.
Supplemental Material
Supplemental material, sj-pdf-1-edr-10.3102_0013189X221142595 for Educators Are Not Alright: Mental Health During COVID-19 by Matthew J. Hirshberg, Richard J. Davidson and Simon B. Goldberg in Educational Researcher
Authors
MATTHEW J. HIRSHBERG, PhD, is a scientist at the Center for Healthy Minds at the University of Wisconsin–Madison, 625 West Washington Avenue, Madison, WI 53703; hirshberg@wisc.edu. His research focuses on interventions that promote social-emotional competencies, mental health, and well-being in educational contexts.
RICHARD DAVIDSON, PhD, is the founder and director of the Center for Healthy Minds at the University of Wisconsin–Madison and William James and Vilas Professor of psychology and psychiatry at the University of Wisconsin–Madison, 625 West Washington Avenue, Madison, WI 53703; rjdavids@wisc.edu. His research focuses on the neural bases of individual differences in emotional reactivity and regulation that affect vulnerability and resilience, especially vulnerability toward anxiety and mood disorders.
SIMON B. GOLDBERG is an assistant professor in the PhD, Department of Counseling Psychology and affiliate faculty at the Center for Healthy Minds, University of Wisconsin–Madison, 335 Education Building, 1000 Bascom Mall, Madison, WI 53706; sbgoldberg@wisc.edu. His research focuses on interventions to promote well-being and reduce psychological symptoms, including the delivery of meditation training through digital technology.
Footnotes
Notes: Richard J. Davidson is the founder, president, and member of the board of directors for the nonprofit organization Healthy Minds Innovations, Inc.
All data and code involved in analyses presented in this paper are available by request.
This study was supported by an award from the Chan Zuckerberg Initiative (DAF 2020-218037 [5022]) to RJD & SBG, a National Academy of Education / Spencer Foundation Postdoctoral Research Fellowship (MJH), National Center for Complementary and Integrative Health Grants K23AT010879 (SBG) and U24AT011289-01 (RJD), funding from the Wisconsin Center for Education Research and the University of Wisconsin -Madison Graduate School through support from the Wisconsin Alumni Research Foundation (SBG), REDCap software licensing through the University of Wisconsin -Madison’s Institute for Clinical and Translational Research, and generous individual donations to the Center for Healthy Minds and the University of Wisconsin -Madison’s School of Education. No donors, either anonymous or identified, have participated in the design, conduct, or reporting of research results in this manuscript.
ORCID iDs: Matthew J. Hirshberg 
https://orcid.org/0000-0001-9070-1270
Richard J. Davidson https://orcid.org/0000-0002-8506-4964
Simon B. Goldberg https://orcid.org/0000-0002-6888-0126
References
- Cohen S., Kamarck T., Mermelstein R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24(4), 385–396. [PubMed] [Google Scholar]
- Czeisler M. É. (2020). Mental health, substance use, and suicidal ideation during the COVID-19 pandemic—United States, June 24–30, 2020. Morbidity and Mortality Weekly Report, 69. 10.15585/mmwr.mm6932a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Harding S., Morris R., Gunnell D., Ford T., Hollingworth W., Tilling K., Evans R., Bell S., Grey J., Brockman R., Campbell R., Araya R., Murphy S., Kidger J. (2019). Is teachers’ mental health and wellbeing associated with students’ mental health and wellbeing? Journal of Affective Disorders, 242, 180–187. 10.1016/j.jad.2018.08.080 [DOI] [PubMed] [Google Scholar]
- Hirshberg M. J., Frye C., Dahl C. J., Riordan K. M., Vack N. J., Sachs J., Goldman R., Davidson R. J., Goldberg S. B. (2022). A randomized controlled trial of a smartphone-based well-being training in public school system employees during the COVID-19 pandemic. Journal of Educational Psychology. Advance online publication. 10.1037/edu0000739 [DOI] [PMC free article] [PubMed]
- Kaufman J., Diliberti M., Hunter G., Grant D., Hamilton L., Schwartz H., Setodji C., Snoke J., Young C. (2020). COVID-19 and the State of K-12 Schools: Results and Technical Documentation from the Fall 2020 American Educator Panels COVID-19 Surveys. RAND Corporation. 10.7249/RRA168-5 [DOI] [Google Scholar]
- Pilkonis P. A., Choi S. W., Reise S. P., Stover A. M., Riley W. T., Cella D., & PROMIS Cooperative Group. (2011). Item banks for measuring emotional distress from the Patient-Reported Outcomes Measurement Information System (PROMIS®): Depression, anxiety, and anger. Assessment, 18(3), 263–283. 10.1177/1073191111411667 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pressley T. (2021). Factors contributing to teacher burnout during COVID-19. Educational Researcher, 50(5), 325–327. 10.3102/0013189X211004138 [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material, sj-pdf-1-edr-10.3102_0013189X221142595 for Educators Are Not Alright: Mental Health During COVID-19 by Matthew J. Hirshberg, Richard J. Davidson and Simon B. Goldberg in Educational Researcher