Stress Overload: A Mixed-Methods, Single-Case Exploration of a Principal’s Stress Accumulation, Sleep, and Well-Being over a School Year

Abstract

Work-related stress and well-being in school leaders is an increasing concern in education. This paper explores how stress accumulates in a principal’s practice over time, challenging the assumption that stress is static or uniformly experienced. We conducted a single-case, mixed-methods study and collected quantitative data on sleep quality, mental health, health behaviors, and leadership self-efficacy along with qualitative data from four semi-structured interviews and two observations with a single principal. Our findings show that while work-stress was manageable, and even positive at times, increasing workload and systemic challenges led to an accumulation of stress that negatively affected Mary’s overall health and leadership self-efficacy. Her appraisal of work-stress was influenced by factors such as control, agency, complexity of the stressor, and alignment with job expectations. We recommend that principal preparation programs include education on stress management, health, and leadership sustainability, and we call attention to school districts and their role in supporting principals under increasingly complex and uncertain conditions.

1. Introduction

The role of the principalship can be stressful partly because of the many job responsibilities and work overload associated with the position. Principals experience stressors stemming from numerous job responsibilities and expectations which include standardized testing, student discipline, bureaucratic compliance, budgetary shortfalls, interpersonal conflicts and more (Mahfouz, 2020; Marsh et al., 2023; Oplatka, 2017). Researchers have demonstrated that work-related stressors play a critical role in school leaders’ overall well-being including poor health outcomes, poor mental health, and increased risk of burning out from the position (Chen et al., 2023; Gmelch & Swent, 1984; Riley et al., 2021). For example, Collie et al. (2020) found that staff/teacher shortages and collegial climate significantly predicted principals’ job satisfaction in nearly 6000 principals across more than twenty countries. Recently, we found that work-related stress affected principals’ health through disruption of health-related factors such as exercise, sleep quality, and food and alcohol consumption, which affected perceptions of leadership self-efficacy (E. J. Su-Keene et al., 2024b). Stress in the principalship is a critical educational issue because it impacts leaders’ ability to regulate emotions, build positive relationships and school culture, and maintain healthy cognitive skills related to decision-making (Branson et al., 2024; Marinac et al., 2024; F. Wang, 2024). And though the reasons behind principal turnover are complex, work-related stress and subsequent consequences have been implicated in principals’ decisions to switch schools, leave their position, and retire early (Carr, 1994; Heffernan et al., 2023; Skaalvik, 2023). In a recent paper, Heffernan et al. (2023) stated

The role’s increasing complexity, alongside the overwhelming workload and a perceived lack of support all combine to have detrimental effects on leaders’ health and wellbeing… theorised in this paper as push factors, have had a significant consequence on leaders’ retention within the profession.

(p. 4)

Most studies that focus on stress in the principalship often conceptualize work-stress simplistically. That is, the environment of the school community elicits stressors that have negative impacts on school leadership. However, stress is a complicated, multifaceted construct because it involves the interaction of the individual and the environment, which accounts for the individual’s genetic propensity to stress and resilience, the capacity to evaluate and appraise a stressful situation, and the ability to deal or cope with stress that ultimately determines the degree of stress and consequences on health and profession (Lazarus & Folkman, 1984). As such, stress can have positive and negative effects based on interpretation, intensity, and longevity (Le Fevre et al., 2003). For example, stress can motivate individuals to complete tasks and to increase productivity (Le Fevre et al., 2003; Selye, 2013), or it can accumulate over time and result in negative physical, mental, and occupational outcomes (Guidi et al., 2020; McEwen, 2006). While the field of educational leadership has embarked on critical research exploring the impacts of work-stress on school leaders, there have been few studies that have explored stress as a more nuanced experience in principals’ practice. Further, few have explored the relationship between antecedents, appraisal, and consequences of stress over an extended period of time.

The lack of nuance in research into principals’ stress is largely due to research designs in the field of educational leadership. Most studies involve a snapshot, cross-sectional approach which limits the ability to capture the nuances of work-stress in the principalship. We assume that stress is dependent on job responsibilities, yet responsibilities and demands change over the course of a school year and how an individual understands and responds to those demands is also unique. To date, we have found no studies that have explored the nuances of work-related stress over an extended time frame, and specifically, how stress increases, decreases, and is managed in principals’ practice, though there are examples of such research in teachers (Capel, 1991; Schonfeld, 1992). Thus, we designed a study that could capture the relationship between stressors, appraisal, coping, health, and professional outcomes over the course of a school year. Here, we present a mixed-methods, single-case study of a principal’s work experience to understand her interpretations of stressors, management of stress, and impacts on her personal and professional well-being. Our research questions include the following:

• How does a principal’s personal and professional well-being change over the course of a school year?
• What work duties, responsibilities, and leadership expectations elicit differential stress perceptions and how does stress change over time?
• How does work-stress affect a principal’s personal health and professional well-being over time?

This study integrates a novel design tailored to the long-term exploration of stress in a principal over time with implications for researchers, policymakers, and preparation programs to develop and support leaders who can sustain themselves through the ups and downs of educational leadership careers. In what follows, we outline a brief review of literature on stress and leadership well-being, the conceptual framework that guides this study, our unique methodological approach, our findings, a discussion, and implications for policy and practice.

1.1. Multifaceted Conceptualization of Stress

Stress is a multifaceted concept that varies in definition depending on the field of study. For example, physiological stress is conceptualized as a neural–hormonal response due to a perceived threat—internal or external—that elicits the sympathetic nervous system, also known as the fight-or-flight response (Chu et al., 2024). The response is temporary and typically short term. However, psychological stress is a state of mental or emotional strain due to challenges, circumstances, or worries in a person’s life which can lead to negative personal and occupational consequences (Chu et al., 2024). Psychological stress, especially in the workplace, is a highly subjective experience that seminal researchers like Lazarus and Folkman (1984) and Bakker and Demerouti (2007) have encapsulated into occupational stress frameworks. For example, the transactional model of stress and coping (Lazarus & Folkman, 1984) argues that individuals’ interpretation or appraisal of the stressor, combined with the ability to cope, results in the magnitude of work-related stress. The job-demands resource theory (Bakker & Demerouti, 2007) highlights that the imbalance of job demands compared to job resources creates stress leading to negative work-related outcomes like motivation and performance.

Though stress is often conceptualized as negative, stress can also be positive. Eustress is often, but not always, short-lasting and defined as the optimal amount of stress that supports goal obtainment, motivation, and problem solving (Benson & Allen, 1980; Selye, 1974). In the context of school leadership, principals may experience eustress when accomplishing tasks, helping teachers manage challenging student behaviors, finding additional resources to support learning, and generally enjoying the fast-paced nature of their jobs. While eustress may still feel challenging, it does not negatively impact psychological or physical well-being.

On the other hand, distress is too little or too much stress, typically associated with negative outcomes (Selye, 1974). Unlike eustress, distress is often long-lasting. As described by Le Fevre et al. (2003) the construct of eustress and distress is subjective where perceptions and sensemaking through the individual determines the positive and negative characteristics of various stressors. Further, individuals’ conceptualizations of whether stress is positive or negative can depend on the degree of control or autonomy over the stressor (Spector, 1998). Further, stress has been conceptualized as a “wear and tear” process through McEwen’s (1998) theory of allostatic load where “the cumulative effect of experiences in daily life that involved ordinary events as well as major challenges and also includes the physiological consequences of the resulting health damaging behaviors, such as poor sleep and circadian disruption, lack of exercise, smoking, alcohol consumption and unhealthy diet” (Guidi et al., 2020, p. 12). Unlike Selye’s (1974) conceptualization of stress, McEwen (1998) argues that humans have limited capacity for stress, and chronic accumulation of stress will lead to detrimental outcomes for individuals’ health. These include increased risk of heart disease, diabetes, mood disorders, substance abuse, and sleep disorders.

1.2. Stress and Well-Being in School Leadership

There is a growing interest in educational leadership well-being research as studies implicate school leaders’ (e.g., superintendent, principals, assistant principals, etc.) role in school improvement and student success (Grissom et al., 2021; Hallinger & Heck, 2010). Compared to other subfields like social justice leadership, instructional leadership, and leadership and policy, leadership well-being research is a relatively under-studied area, but a critical issue that has widespread impacts on the individual school leader and the school community (Chen et al., 2023; Doyle Fosco, 2022). Well-being literature in school leadership can be traced to the 1960’s with general interests around work demands and stress (Chen et al., 2023). Following the COVID-19 pandemic, research on leadership well-being saw an exponential increase, suggesting that a peripheral topic had become a focal point (Chen et al., 2023; Doyle Fosco, 2022). Chen et al. (2023) published a review of school leadership well-being literature and found more than 32% of the studies over the last eight decades were published after 2020, catalyzed by deteriorating leadership well-being during the pandemic (S. D. Hayes et al., 2022; Steiner et al., 2022; Urick et al., 2021). Evidence of this deterioration was highlighted in a national study post-COVID-19 where 85% of principals experienced frequent work-related stress and only 19% of principals were coping well under pressure (Steiner et al., 2022).

Historically, there has been a focus on factors that influence leaders’ well-being. This includes individual traits; psychological, physical, and physiological influences from day-to-day work; and systemic educational challenges that influence well-being (Chen et al., 2023; D. DeMatthews et al., 2021; Doyle Fosco, 2022). Examples include studies that examine positive leadership and individual characteristics such as a propensity for gratitude, optimism, and positive reframing in leadership well-being (Beard, 2018; Divoll & Rineiro, 2024; E. Su-Keene & DeMatthews, 2022). In Ireland, researchers found that school facilities, infrastructure, student emotional/behavioral challenges, and issues with discipline were significant predictors of principals’ stress (Darmody & Smyth, 2016). In Australia, Gorrell and De Nobile (2023) highlighted key areas of concern for elementary principals that affected their well-being, which included roles and responsibilities, interpersonal relationships, isolation, stress, and prioritizing the well-being of the school. Recently, the authors of this paper argued for the critical role of mental health, physical health, and sleep quality and it’s impacts on personal and occupational well-being (E. J. Su-Keene et al., 2024a).

In addition to factors that precede leadership well-being, researchers are also interested in the outcomes of principal ill-being. Research in fields such as occupational health and occupational psychology have long-established links between stress, health, wellness, and well-being to occupational outcomes such as productivity, absenteeism, cognitive functioning, and job satisfaction (Bakker, 2015; Dunn et al., 2008; Sparks et al., 2001). This strand of research is still very much in its infancy within educational leadership where a handful of articles have conceptualized and empirically examined the occupational impacts of poor well-being (Chang et al., 2015; Eckman, 2004; Federici & Skaalvik, 2012). For example, Federici and Skaalvik (2012) found that principal burnout was related to lower self-efficacy and greater motivations to leave the job. Some studies have also shown that not all principals are equally at risk. Female principals, principals of color, novice principals, and principals in socioeconomically disadvantaged schools are at greater risk of work-stress and stress-related outcomes like poor mental health, burnout, and intentions to leave (D. E. DeMatthews et al., 2023; Steiner et al., 2022).

The well-being of principals has implications for school communities as well. F. Wang et al. (2018) argued that the well-being of principals plays a critical role in sustainable school improvement through leadership efficacy, job satisfaction, or sustainability in the profession. This notion aligns with Day et al.’s (2014) argument that positive school improvement requires effective AND stable principal leadership which highlights principals’ overall well-being as critical research in educational leadership from a Maslowian perspective. Early exodus of school principals not only disrupts well-laid campus improvement plans and efforts but can also catalyze the exodus of teachers leading to a turnover churn of workplace personnel (D. E. DeMatthews et al., 2022).

To date, there are few intervention studies that have found positive solutions to ameliorate job-related stress. One includes a study by Mahfouz (2018) who found promising outcomes for school leaders who engaged in mindfulness training such as increased self-care and self-compassion that positively impacted leadership skills for relationship-building and reflection. Other researchers found that school leaders engage in a variety of care strategies to support their well-being including exercise, finding positivity and joy, connecting with family and friends, drawing boundaries around work, and finding spiritual supports (Doyle Fosco et al., 2025). Similarly, the authors of this paper have also argued for the practice of positive psychology strategies to mitigate stress and burnout (E. Su-Keene & DeMatthews, 2022).

1.3. Conceptual Framework

In the field of educational leadership, researchers have borrowed conceptual and theoretical frameworks to study stress and well-being from other disciplines like occupational health and psychology. Some dominate frameworks include job demand resource theory (Bakker & Demerouti, 2007), the transactional model of stress and coping (Lazarus & Folkman, 1984), and PERMA (Seligman, 2018) to examine principals’ job demands, resources, stress, motivation, well-being, and the impacts on school organizations (Doyle Fosco, 2022; Marsh et al., 2023; Maxwell & Riley, 2017; Richard, 2024). In this paper, we use Lazarus and Folkman’s (1984) transactional model of stress and coping and McEwen’s (2006) allostatic load to guide the design and analysis of the study. Our research questions aim to qualitatively understand how a principal subjectively interprets stress in the workplace, how she copes with stress, and the impacts on her physical and mental well-being. By using the transactional model of stress, we capture the multifaceted nature of stress encapsulated in Selye’s (1974) theory of eustress and distress and Spector’s (1998) control theory. That is, we consider positive and negative stress and the role that control plays in the principal’s interpretation of stress. Further, the transactional model posits that the individual’s appraisal of the situation combined with their ability to remedy or cope results in the magnitude of stress. Allostatic load further guides this study by acknowledging that the accumulation of stress results in negative health and occupational consequences. These two models allow us to disentangle good versus bad work-related stress, coping strategies and capacities to mitigate stress, stress accumulation over time, and the consequences of chronic work-stress.

2. Methods

We use a mixed-methods, single-case research (MMSCR; Onghena et al., 2019) design to explore a novice elementary principals’ workload and the impacts on her mental and physical well-being. The use of single-case research can be traced historically to the 18th and 19th century by physicians who studied human physiology, disease, and treatments. This includes the creation of the first vaccine by Edward Jenner in 1796 who inoculated James Phillip, an eight-year-old, with cowpox to examine immunological protections against the more virulent smallpox. Single-case studies grew in popularity into the early 20th century in fields like psychology, medicine, and behavioral science (Barlow & Hersen, 1973; Mcleod, 2002; S. C. Hayes, 1981) which has led to important contributions and knowledge generation within respective fields including Pavlov’s seminal work on classical conditioning informed by single animal cases.

In education, single-case studies are of particular interest in special education and human development to examine characteristics, behaviors, and interventions for neuroatypical children and children with special needs. In a review of single-case studies in special education, Cakiroglu (2012) found that defining features included a single subject or participant as the unit of analysis, description of the study setting, repeated measurements of targeted variables, introduction of a treatment or intervention, and a visual analysis of data trends. Though case studies are common in educational leadership, the use of single-case designs has been limited. Of those published, most have explored aspects of school leadership praxis over time. For example, Sanchez et al. (2019) conducted a study that explored a principal’s school improvement efforts across a three-year span. Hoppey and McLeskey (2010) conducted an intensive study on a principal’s ability to build an inclusive school program over a one-year period. Yet, no studies have examined work-related stress in a single principal over an extended period of time.

When researchers employ a single-case design, most take a qualitative or quantitative approach, and the use of mixed methods is infrequent (Onghena et al., 2019). However, mixed-methods, single-case research (MMSCR) designs are distinctive, allowing for innovative opportunities to complement traditionally qualitative or quantitative single-case approaches (Onghena et al., 2019). Taking a mixed-methods approach in a single case allows for more nuanced research questions and a richer, complementary understanding of an individual’s experiences or phenomena. Of the few studies that have integrated an MMSCR, the designs typically lean more towards one paradigm. For example, a quantitative single-case study of a special needs student may integrate qualitative interviews with stakeholders including teachers, parents, and the child to determine the effectiveness of an intervention. Van Ness et al. (2017) combined two common approaches in the field of applied health by integrating a qualitative case study with an n-of-1 clinical intervention to examine preferences and benefits of music therapy. These authors grounded their experimental design in pragmatism informed by philosophers like William James, Charles Pierce, and John Dewey who argued that practical, individualistic, and humanistic orientations can lead to the development of testable conceptualizations, hypotheses, and theories about human experience.

We conducted a mixed-methods, single-case study of a principal to explore perceptions of work-stress and the impact on her personal and professional well-being over the course of a school year. Though studies on school leaders’ well-being are relatively sparse compared to other subfields in educational leadership, there are studies that have examined similar topics among teachers. In a large, longitudinal study of teachers, Capel (1991) noted that the development of burnout over time was highly variable, needing individual examination and remediation. This finding reflects Lazarus and Folkman’s (1984) transactional model of the subjectivity of stress and its manifestations, which includes the appraisal of work as stressful and the ability to cope. Thus, we argue that a mixed-methods, single-case design is well-suited for answering our research questions.

The MMSCR in this paper takes an embedded approach where we use objective and subjective quantitative data to paint health and leadership self-efficacy trends, and we integrate qualitative data for a deeper understanding of work-related stress and how it affects the principal’s overall well-being over time. While many single-case studies employ an intervention, we argue that examining a principal’s perception of stress, health, and leadership throughout the school year—including summers—when a principal’s workload increases and decreases serves as a quasi-intervention. Like other mixed-methods case studies, this MMSCR takes a deductive and constructivist approach aligned with seminal guidance from Yin (2009) and Merriam (1998), respectively.

2.1. Single-Case Study Participant and Context

Mary (pseudonym) is white, female, mid 30’s, second-year principal at a bilingual elementary school in Texas. She was recruited from a group of principals who participated in a prior study that explored stress and sleep and expressed a deep interest in the research topic because of her lived experiences as a first-year principal. As such, we recruited Mary to participate in a year-long study to explore the impact of stress on her personal and professional well-being over the course of a school year. Mary works in a large urban school district that is within the Council of Great City Schools. The district has experienced significant gentrification over prior decades as many low-income students of color have been pushed out of the urban core to suburban fringe communities. Moreover, the district has experienced a significant decline in student enrollment due to charter expansion and a decrease in the city’s school-age population, resulting in significant budget shortfalls for the district and the need to consistently consider and sometimes close campuses. The district also experienced several superintendent transitions and concerns around the district’s special education program, leading to state interventions and increased pressures on principals, teachers, and school staff.

Mary’s campus is on the fringe of the district and serves more than 300 students and a large percentage of low-income, students of color. Given her school’s focus on bilingual Spanish education, most of her students come from Hispanic backgrounds and more than 30% of students are emergent bilingual learners. At Mary’s school, more than 10% of students have special education needs. Both before and during Mary’s principalship, the school has not met the state’s expectations for academic performance, though there have been several pauses and delays in state metrics due to the COVID-19 pandemic, contentious changes to the state standardized test, and criticisms over its alignment with student outcomes, which make pinpointing school performance difficult (Texas AFT, 2024). Even though Mary’s school has had challenges meeting the state’s expectations through standardized tests, her school has received exemplary ratings on other factors, including elements of fine arts, digital learning, wellness, and community involvement from the school district. While the context for this study is unique, it also represents the many commonalities and difficulties that other school contexts are experiencing, such as increased right-wing political influences, shifting economic demographics, increased student and community diversity, and decreased public education funding.

2.2. Research Design, Equipment, and Instruments

Between June 2023 and May 2024, quantitative and qualitative data were collected. To assess Mary’s health, we collected objective sleep actigraphy data and perceptive health data (such as exercise and food consumption). As noted in previous studies, sleep quality is a strong predictor of immediate and long-term health outcomes and is used as a proxy health variable (Arble et al., 2015; Strine & Chapman, 2005; E. J. Su-Keene et al., 2024b). In social science, applied sleep research is often conducted using survey instruments (e.g., the Pittsburg Sleep Questionnaire Index) or objective actigraphy monitoring, both of which have been individually validated and correspond well to one another (Buysse et al., 2008; Forner-Cordero et al., 2018; Zak et al., 2022). Sleep actigraphy is generally preferred as it measures variations in sleep in natural environments that surveys cannot fully capture. Mary wore an Actigraph GT3X-BT (Ametris LLC, Pensacola, FL, USA) sleep watch on her non-dominant wrist for 14 consecutive days, every six weeks (June 2023, September 2023, November 2023, January 2024, March 2024, and May 2024) for a total of six measurements. Actigraph GT3X+ sleep watches track movement and ambient light and are a validated device for measuring people’s sleep in natural settings (de Souza et al., 2003). Users cannot see the data collected in real time. The watch was initiated with the Actilife 6.13 software system and collected data using a standard 30 Hz rate. The first sleep monitoring was conducted in June 2023 during summer working conditions without most teachers and students present, which served as a baseline/control for stress and sleep measurements.

During sleep monitoring periods, Mary completed 14 daily diaries where she recorded perceptions of work and personal stress, work and personal support, perceptions of restfulness, and perceptions of sleep quality on a 7-point Likert scale, along with consumption activity (i.e., caffeine and alcohol). Though we use actigraphy to measure sleep activity, subjectivity and perceptions of sleep can influence a person’s mood, so we have included such questions to determine discrepancies between perceptions and objective measurements. After each monitoring period, Mary also completed a survey that explored perceptions of leadership self-efficacy (PSES; Tschannen-Moran & Gareis, 2004) and symptoms of anxiety (GAD-7; Spitzer et al., 2006) and depression (PHQ-9; Kroenke et al., 2001) across the fourteen-day sleep monitoring period. This resulted in six sets of sleep, daily diary, and post-monitoring quantitative data.

In terms of qualitative data, the research team conducted four semi-structured interviews with Mary throughout a one-year period (i.e., June 2023, October 2023, March 2024, and May 2024). We designed an interview protocol that asked Mary to reflect on current work demands, perceived stressors, and other challenges that impacted her overall personal and leadership well-being. During the interview, we asked Mary to reflect on current challenges in comparison to the last interview to gauge positive or negative changes in stress and well-being over time. Each interview concluded with a reflection on Mary’s commitment to the position to explore her intentions to leave the position. In addition to the interviews, we conducted two school-based observations (in October 2023 and March 2024) to provide additional insights into the relational, physical, and emotional demands of the job. These quantitative and qualitative data not only complemented one another but also assisted with the development of holistic findings.

3. Data Analysis

Sleep data collected on the Actigraph GT3X-BT were downloaded and analyzed using the Cole–Kripke algorithm for scoring adult sleep quality attributes including total sleep time (TST), wake after sleep onset (WASO), sleep efficiency (TST/TST + WASO), and number of nighttime awakenings. Descriptive statistics were performed on sleep and survey data across the six time periods. Repeated measures ANOVA analyses were used to examine significant differences between monitoring periods and Dunnett’s multiple comparison tests were used to examine pairwise differences between the school year and the summer/control. Four of six monitoring periods started on the same day (i.e., Sunday to Saturday nights). For two monitoring periods, the days were shifted so that the day of the week aligned for the analysis. During the summer, there was an outlier data point due to travel, so sleep data from that weekday was excluded throughout the six data sets in the analysis, resulting in 13 instead of 14 daily sleep measurements. We applied Greenhouse–Geisser corrections to analyses where the data violated assumptions of sphericity. We conducted a bivariate correlation analysis of measured sleep quality and daily diary survey data to explore the relationships between Mary’s actigraphy data and perception data. Since there were only six post-monitoring surveys, we did not include the post-monitoring data in the correlation analysis.

Qualitative data which included four interviews and two observations underwent open, thematic coding techniques that focused on the emotional experiences from work and stress within each interview before comparing findings across interviews (Clarke & Braun, 2014). For example, we thematically coded Mary’s response to the question about “current work-related demands” and the interpretation of such demands into stress within each interview independently before comparing across interviews (Clarke & Braun, 2014). After each interview was coded, we conducted a second round of comparative analysis through pattern coding to examine differences in themes between each time period (Saldaña, 2021). We focused particularly on affective codes, or codes that signaled emotions, to gauge positive and/or negative experiences and interpretations of stress and work-stressors (Saldaña, 2021). Analysis of the observation protocols were analyzed in the same procedure but integrated within the chronological order of the interviews to support themes that were emerging over the course of the school year.

4. Findings

4.1. Quantitative Findings: Trends in Sleep Quality, Health, and Leadership Self-Efficacy

To answer the first research question, we provide descriptive and inferential statistics on quantitative data obtained from the actigraphy watch, daily diary surveys, and post-monitoring surveys. The goal of this quantitative portion is to provide an overview and trends in Mary’s perceptions of stress and health prior to the deeper, richer, and explanatory nature of the qualitative findings. Overall, we found that Mary experienced decreased sleep quality, increased work-stress, increased mental health symptoms, and fluctuating perceptions of leadership self-efficacy over the course of a school year.

Sleep quality and fatigue. Data collected from Mary’s Actigraph sleep watch revealed that her overall sleep quality decreased in the first half of the school year and improved towards the end of the school year. During the summer, Mary scored 94.71% for sleep efficiency. However, a repeated measures ANOVA showed that Mary’s sleep quality decreased significantly between the September, November, and January sleep measurements, before showing a positive rebound in March and May (

Table 1. Descriptive statistics of sleep attributes and significance of repeated measures ANOVA.

Sleep Attributes	June 23 (Control)	September 23	November 23	January 24	March 24	May 24	Overall p-Value
	M (SD) p-Value Compared to Control if Significant to Control
Total Sleep Time (min)	440.1	440.1	456.8	460.9	410.3	409.3	0.1345
	(74.97)	(67.62)	(30.11)	(82.19)	(71.04)	(57.01)
Total Time in Bed (min)	464.3	480.5	510.8	510.5	439.2	441.8	0.0355
	(77.44)	(74.59)	(32.69)	(97.14)	(74.97)	(63.60)
WASO (min)	24.23	40.38	54.08	49.62	28.92	32.46	<0.0001
	(10.13)	(12.49)	(20.30)	(18.13)	(10.27)	(10.70)
		0.0249 *	<0.0001 ****	0.0002 ***	0.8744	0.4571
Number of Awakenings	16.15	25.62	30.15	28.54	17.92	18.00	<0.0001
	(5.800)	(8.362)	(7.957)	(9.107)	(5.499)	(5.323)
		0.0053 **	<0.0001 ****	0.0002 ***	0.9526	0.944
% Sleep efficiency	94.71	91.63	89.47	90.50	93.43	92.71	<0.0001
	(2.094)	(2.066)	(3.673)	(2.162)	(2.171)	(1.955)
		0.0138 *	<0.0001 ****	0.0004 ***	0.5785	0.1819

Note. p-value = * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001.

, Figure 1D). The decrease was due to significantly more awakenings and more time awake after sleep onset (WASO) (Table 1, Figure 1B,C). On average, Mary slept 6.5–7.5 h per night during each monitoring period, and we found no significant differences in her total sleep time (Table 1). While total time in bed (i.e., total sleep time and time spent awake) was significant, a Dunnett’s multiple comparison test showed no significant differences compared to the control. Data from daily diary surveys showed that Mary’s perception of sleep quality decreased throughout the year from September to January but improved in March and May, which matches objective actigraphy data (

Table 2. Descriptive statistics of stress, support, and restfulness perceptions and significance of repeated measures ANOVA.

Daily Diary Perceptions	June 23 (Control)	September 23	November 23	January 24	March 24	May 24	Overall p-Value
	M (SD) p-Value Compared to Control if Significant to Control
Work Stress	2.231	3.692	3.154	4.000	4.000	2.769	<0.0001
	(0.9268)	(0.9473)	(1.068)	(1.000)	(0.7071)	(0.7250)
		0.0104 *	0.1376	0.0031 **	0.001 ***	0.2233
Personal Stress	1.538	1.769	2.462	2.154	1.538	2.000	0.0420
	(0.5189)	(0.4385)	(1.330)	(0.3755)	(0.5189)	(0.7071)
Work Support	3.769	5.077	5.462	4.385	3.385	4.385	<0.0001
	(0.7250)	(0.9541)	(0.6602)	(0.5064)	(0.6504)	(0.5064)
		0.0046 **	>0.0001 ****	0.0917	0.4056	0.0203 *
Personal Support	6.846	6.462	6.077	6.308	6.846	6.462	0.1621
	(0.3755)	(0.7763)	(1.382)	(0.6304)	(0.3755)	(0.5189)
Restfulness	5.923	4.923	4.231	4.769	4.077	4.385	<0.0001
	(0.6405)	(1.188)	(1.092)	(0.7250)	(0.9541)	(0.7679)
		0.092	0.0034 **	0.0024 **	0.0006 ***	0.0002 ***
Sleep Quality	5.692	5.154	4.769	4.385	4.615	5.462	0.0816
	(1.032)	(0.9871)	(1.589)	(1.325)	(1.193)	(0.6602)
Caffeine (servings)	0.9231	1.308	1.077	1.692	1.154	1.538	0.0023
	(0.6405)	(0.4804)	(0.4935)	(0.4804)	(0.3755)	(0.5189)
Alcohol (servings)	0.4231	0.07692	0.2692	0.000	0.1538	0.000	0.0549
	(0.5718)	(0.2774)	(0.4385)	(0.000)	(0.3755)	(0.000)

Note. p-value = * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001.

). However, her perception of restfulness steadily and significantly declined throughout the course of the school year (Figure 2C).

Perceptions of stress, support, and mental health symptoms. Mary experienced increasing work-stress, fluctuating perceptions of support, and increasing symptoms of poor mental health throughout the school year. Mary experienced significant increases in work-stress in September, January, and March (Table 2, Figure 2A). Mary’s perceptions of support at work increased during the first two measurements, dipped in January and March, and then rebounded at the end of the year (Figure 2B). The main effect of personal stress during the school year was significant; however, pairwise comparisons to the control showed no significant differences in means. Our analysis of Mary’s mental health symptoms showed that she experienced increased anxiety and depression as the school year continued. Her symptoms increased greatly in January and peaked in March before returning to low levels at the end of the school year (Figure 3). RM ANOVA of GAD-7 and PHQ-9 scores showed significant increases in anxiety (F (5, 30) = 13.09, p < 0.0001) but not in depression symptoms. Interpretations of the scores revealed that Mary experienced minimal to mild symptoms of depression and anxiety in September and November, mild depression and moderately severe anxiety symptoms in January, moderate depression and moderately severe anxiety symptoms in March, and mild symptoms of depression and anxiety in May (Figure 3).

Leadership self-efficacy. Mary’s perceptions of leadership self-efficacy consistently improved for instruction and moral subscales but perceptions of management self-efficacy fluctuated during the school year (

Table 3. Descriptive statistics of leadership self-efficacy scores (SE).

Leadership Self-Efficacy	June 23 (Control)	September 23	November 23	January 24	March 24	May 24	p
	M (SD)
Overall SE	6.056	6.333	7.000	6.167	7.222	6.556	0.0381
	(1.162)	(1.372)	(0.8402)	(1.150)	(1.060)	(1.917)
		0.9272	0.0039 **	0.9987	0.0177 *	0.8288
Moral SE	5.500	5.500	6.667	6.833	7.667	7.500	0.0413
	(1.643	(1.378)	(1.033)	(0.4082)	(1.033)	(1.225)
Instructional SE	6.000	7.167	7.167	6.667	7.333	7.500	0.0241
	(0.6325)	(1.169)	(0.7528)	(0.8165)	(1.211)	(1.049)
		0.1764	0.0417 *	0.0800	0.0800	0.0560
Moral SE	5.500	5.500	6.667	6.833	7.667	7.500	0.0413
	(1.643)	(1.378)	(1.033)	(0.4082)	(1.033)	(1.225)
Management SE	6.667	6.333	7.167	5.000	6.667	4.667	<0.0001
	(0.8165)	(1.211)	(0.7528)	(1.095)	(0.8165)	(1.862)
		0.9319	0.7467	0.0081 **	>0.9999	0.0014 **

Note. p-value = * < 0.05, ** < 0.01.

). There was a significant main effect (F (2.46, 12.31) = 3.93, p < 0.05) in moral self-efficacy and instructional self-efficacy scores (F (2.39, 11.93) = 4.88, p < 0.05). However, management SE scores fluctuated throughout the year but generally trended downwards as the year progressed. There was a significant main effect (F (5, 25) = 8.87, p <0.0001), and a Dunnett’s multiple comparison test showed significantly lower scores in management SE in January and May compared to the summer control (Figure 4D).

Relationships between variables. To examine the relationships between the quantitative variables, we conducted a bivariate Pearson correlation analysis (

Table 4. Correlations of sleep measurements and daily perceptions.

Variable 1	Variable	r	p
Work-stress	Total sleep time	−0.223	p < 0.05
Work-stress	Perceptions of restfulness	−0.41	p < 0.001
Work-stress	Perception of sleep quality	−0.49	p < 0.001
Support at work	WASO	0.224	p < 0.05
Support at work	Personal stress	0.41	p < 0.001
Personal stress	Support in personal life	−0.268	p < 0.05
Support in personal life	Sleep efficiency	0.27	p < 0.05
Support in personal life	WASO	−0.344	p < 0.01
Support in personal life	Awakenings	−0.376	p < 0.001
Perceptions of restfulness	Sleep efficiency	0.28	p < 0.05
Perceptions of restfulness	WASO	−0.233	p < 0.05
Perception of sleep quality	Sleep efficiency	0.273	p < 0.05
Perception of sleep quality	Perceptions of restfulness	0.457	p < 0.001

). Some correlations were expected, such as the clustering between stress, anxiety, and depression variables and variables that define sleep quality. Also, Mary’s perceptions of sleep were correlated to objective measurements on the Actigraph sleep watch which confirms that Mary’s perceptions of sleep were a true reflection of her actual sleep quality measurements. Below, we focus on the unique correlational findings on sleep as proxy for health.

• Work-stress correlates to sleep. Higher work-stress was correlated with less total sleep time (r (83) = −0.223, p < 0.05), less restfulness (r (83) = −0.41, p < 0.001), and poorer sleep quality measurements (r (83) = −0.49, p < 0.001). Surprisingly, support at work was positively correlated with time spent awake at night (WASO, r (83) = 0.22, p < 0.05), though increased time awake at night was correlated with increased personal stress (r (83) = 0.41, p < 0.001). Together, this set of correlations suggests that when work-stress increases, the quality of sleep decreases, and Mary’s positive perceptions of support at work and high personal stress correlate to poorer sleep quality due to an increase in time spent awake after falling asleep—a contradictory finding that needs further exploration.
• Personal support correlates with sleep. As expected, personal stress was negatively correlated to personal support (r (83) = −0.27, p < 0.05). Surprisingly, Mary’s perception of personal support correlated to three of the four sleep quality measurements. High personal support was correlated to higher sleep efficiency (r (83) = 0.27, p < 0.05) and less time spent awake at night (WASO, r (83) = −0.34, p < 0.01) and fewer awakenings (r (83) = −0.38, p < 0.001). That is, personal support was a strong correlate of better sleep quality.
• Alcohol intake correlates to sleep. Mary’s alcohol intake correlated to sleep measurements including sleep efficiency (r (83) = 0.31, p < 0.01), WASO (r (83) = −0.28, p < 0.05), and awakenings (r (83) = −0.25, p < 0.05) though not in the direction expected. Research shows that alcohol intake has a negative impact on sleep quality. For Mary, increase in alcohol consumption was correlated to an increase in sleep quality. It is important to note that Mary consumes alcohol very infrequently. Out of the 83 daily diary entries, she listed alcohol consumption twelve times, with at most 1.5 servings. Furthermore, five out of the 12 days were on the weekend when she presumably experiences better sleep quality than on weekdays.

4.2. Qualitative Findings

We present the qualitative findings for this study chronologically to answer the second and third research questions. For each summer, we present subsections on Mary’s leadership reflections, mental and physical health, and intentional leadership goals for the upcoming school year. For the fall and spring, we present sections on current leadership responsibilities, stressors, and mental and physical health. Additionally, the quantitative findings presented above are weaved throughout to complement the qualitative findings, which provides a more in-depth exploration of Mary’s work-stress and experiences as a school principal.

4.2.1. Summer: Restoration and Reflection

After a hectic first year as principal, Mary spent the summer working, focusing on self-care, and engaging in wellness practices such as spending time with family, finding ways to relax, and finding joy in hobbies. She felt restored from a recent family vacation and swimming in the evenings became “a nice ritual.” She started mornings on a more positive note. “Reading has been really helpful because I’m not like ‘I got to work.’” She also shared that she was happy to be at home with her partner more often.

I spend such long hours at school that if I plan things with friends or anything, it can sometimes mean more time I’m away. So, this weekend I just stayed home with my husband and that was nice to just be in my house and be together.

Mary felt a sense of ease this time of year despite having to go to campus most days in the summer. She said, “There are definitely days I am frustrated with things, but there is just space. Like I kind of exist easier than I do during the school year.” In other words, Mary had the time, space, and energy during the summer to work without being overwhelmed.

This sense of ease and well-being during the summer was reflected in the quantitative findings as well in good sleep quality, low work-stress, and high restfulness. She also expressed high job satisfaction, professional joy, and verbalized personal and professional goal setting. She said, “I love the job in a lot of ways. As hard as last year was, I really enjoy the work. I laugh with teachers and staff and there’s a lot of joy. I’m doing this because I love it and want to be here.” Moving into the second year, she set two personal goals.

One of my goals for this year is to actually eat and drink during the day because I have a desire to hold very still because I don’t want to fully feel what I’m feeling because I don’t want it to seep out on other people.

The importance of eating and hydrating regularly was often deprioritized to address other work-related obligations. Mary reported a “slow wearing down of self” where she didn’t eat, drink, or sleep well for the last ten months. Another goal is to set better expectations for her staff. Specifically, Mary wanted more space from work-related issues during non-working hours by teaching her staff to be more attentive to how and when they send messages and emails. She explained, “I try to tell staff between this time and that time, please don’t message me unless you really need me.” As principal, Mary recognized that there was a tension between needing to be informed and needing a mental break. Overall, the summer was a time when the workload felt manageable, when Mary’s health and well-being was at an all-time high, and when Mary felt very positive about her career.

4.2.2. Fall: Waning Leadership Well-Being

During the fall, Mary reported a range of work responsibilities including transitioning students and staff back into the school year; ensuring students and staff follow existing and new policies and procedures; establishing new relationships with new members of the campus community; managing conflicts and disruptions; submitting paperwork; and handling unanticipated challenges. For example, Mary has been dealing with an increase in student behavior challenges. “I’ve spent a lot of time with kids who are trying to hurt themselves or trying to hurt us. I was holding a girl between my legs and lassoed her arms so she could calm back down.” Another is teacher emergencies. “I had a teacher whose husband had a stroke, and she was out for a week. I didn’t have subs so just trying to cover.” While challenging, Mary felt that these were the normal expectations of the job that “take a long time” to sort out so she cannot attend to other “schoolwork or paperwork”.

Another challenge was the growing concerns around special education and the district’s shortcomings in compliance. Mary said, “our district is under scrutiny for special education so I’m having to sit with people, read their forms, and correct their things” so she spent additional hours training novice special education teachers. Relatedly, Mary was ensuring additional compliance-oriented tasks like state mandated online training. She was constantly asking teachers,

Did you finish reading academies? Are you on track for this new special education training? Did you do your grades and get them out on time? Did you do them correctly? This feels like a big piece of my job right now.

Additionally, Mary was focused on increasing her school’s academic performance and teachers’ pedagogical and curricular capacity. She said, “I’m in rooms and I’m trying to get a sense of where we’re at. How do I need to push people?” so that she can strategically target the needs of both her veteran and novice teachers. Thus, Mary spent time organizing professional development for teachers and provided substitutes during school hours.

While these stressors were challenging, Mary’s did not identify these challenges as negative. Using Lazarus and Folkman’s (1984) transactional stress theory, we noticed Mary feels alignment between the challenge and normal expectations of the principal. Further, Mary generally felt in control and had agency in ameliorating these particular problems (Spector, 1998). Lastly, Mary shared that she liked to solve problems such as developing teachers and improving academic outcomes, which was a “welcoming challenge” that aligned with the concept of eustress.

Though individual responsibilities were manageable, the sheer increase in workload was stressful for Mary. She said, “I don’t feel stressed fully. It’s just the lack of time and the sense that things are building up faster than I can get to”. The quantitative findings reflect a significant increase in work-stress in the fall compared to the summer. Certain work responsibilities negatively affected Mary’s emotional well-being. These include tensions in supporting teachers’ mental health, lack of teacher ownership, and disconnects between district demands and support. Adults in Mary’s school were struggling with their mental well-being because everyone was “under a fire hose.” While Mary was empathetic towards each teacher individually, she was also frustrated.

I’ve had multiple people say, ‘I have to set boundaries. I won’t work outside of school hours. I’m not going to take this with me. I’m not going to X event’ and it’s like, this thing only works with all of us together.

These teachers’ sentiments paired with an increase in “mental health days” have increased the workload for Mary and other teachers. When teachers were absent and disengaged, Mary had to move students to other teachers’ classrooms, find substitutes, and make impromptu lesson plans. When multiple people are out every single day, these tasks feel overwhelming because it is a persistent issue that takes time away from her other responsibilities and burdens other teachers. During an observation, Mary spent the first two hours of the school day managing teachers’ absences. Furthermore, Mary had to step out of her regular tasks, meetings, and responsibilities to monitor the classrooms, halls, and substitutes to make sure everything was running smoothly.

Mary was also increasingly frustrated by teachers’ lack of investment and ownership. For example, several teachers were still out of compliance for state-mandated training in October, even though she provided extra time and personnel support for teachers to finish. She said,

I’m trying to get people to take ownership. I had several people who didn’t finish their reading academies training. It’s like we’ve already talked about this, I’ve already provided you support, but in my head I was like you’re not a child. I’m making good effort to give you the opportunity, but you have to do it… I spent 30 minutes telling people to show up to work on time, and I feel bad about having to treat people like that. They’re adults. I think that kind of task makes you lose faith in people, which kind of sucks.

As Mary worked to reduce teachers’ workload, she felt frustrated by the increasing workload put on her by the district and supervisor. This year, Mary had to lead, rather than support, recruitment efforts for her schools’ dual-language program using the new digital system—a first come, first serve enrollment program was inequitable in her neighborhood for families that have lower social capital, limited language proficiency, and limited technology resources. “That’s now on my job description. I’m having to figure out how to get families who don’t know, to log onto computers first” to ameliorate some of the hidden inequities behind this new process.

Mary used words and phrases like draining, frustrating, and “just okay” to describe physical well-being in the fall. She felt healthy and believed she was sleeping enough, though she was generally tired. Mary said choosing sleep was easy to justify because of narratives around self-care, however she feels conflicted about the tensions between choosing herself and her work.

I am choosing sleep over work because I can tell if I’m tired, I’m not thinking at my best, but it also means that I don’t get caught up with work when I choose sleep. Do I choose to not work this evening, which would be healthier for me but will make more work for me later? Is it really better?

The quantitative data confirmed Mary’s perceptions of feeling less rested and more tired especially in November compared to summer (Table 2, Figure 2). However, a closer examination of Mary’s actigraphy data showed that she was not sleeping as well as she believed. She experienced significantly lower sleep quality in the fall with significantly more awakenings and time spent awake (see Table 1, Figure 1). Simultaneously, this decrease in sleep is significantly correlated with an increase in work stress during this time. Indeed, Mary did sleep more during this time, but the quality of sleep decreased because she was waking up more at night.

An increase in workload meant less time for Mary to attend to her own needs. She shared that she was not exercising as much. Mary’s anxiety and depression scores increased in the fall, though it remained within the “mild” category. Maintaining her food consumption quickly became a challenge.

I’ve been forcing myself to eat at school, but honestly, sometimes it’s just easier not to eat because strangely, I feel judged at work for eating in my office. There’s this perception that principals are in their office and hiding. And if I’m eating in my office, then I’m not handling their problems.

Yes, Mary lacked time, but this quote also reveals socialized constructs around what principals should or should not do. That is, maintaining healthy habits and routines felt at odds with others’ expectations. In the fall, Mary had become increasingly concerned about her heart health. She said, “my heart rate has been really fast, so I went to my doctor and she was like, whoa, your pulse is way too high and that’s a real problem. And then I could feel it throughout the day.” Mary believed that this was a symptom of her work that was different from the year before. Overall, the fall revealed that some stressors were expected because they were part of the job and within her control, while others were more frustrating. Even so, the increase in workload and stress reveals a deterioration in Mary’s well-being throughout the fall academic year.

4.2.3. Spring: Systemic Stressors and Leadership Ill-Being

In the spring, Mary’s main work responsibilities included preparing and executing state standardized testing—including ensuring classrooms were prepared, student attendance was high, and the distribution, collection, and security protocols were in place. However, she was still experiencing many of the same challenges from the fall. Qualitative findings from the spring suggest that systemic stressors outside her control (Spector, 1998), including the ongoing teacher and substitute shortage issues and district policies, politics, and work demands, were negatively impacting her well-being. In January,

So many people were out sick. Every morning, I was telling people you’re going to have to split again and wearing them down. It’s constant problem solving and not knowing how to fix it. Since I started as a principal, I’ve not had a day where I had my entire staff present and all vacancies filled with someone. Not a single day.

In addition to shortages of teachers and still vacant positions, Mary added that the district now had a shortage of substitutes. She said, “If a teacher posted that she’s sick the night before, that post will not get filled” and “I just have to keep solving the same problem over and over again. And there is no solution. I don’t know what to do.” Personnel shortages induced high stress because the root causes, like sicknesses or vacant positions, were not within Mary’s control (Spector, 1998) causing high distress. Mary expressed how she was exhausted, which reflected the negative effects of allostatic load and the chronic stress arising from this Sisyphean task.

District politics and inefficiencies were also causing stress. Mary had become increasingly frustrated with the mismatch between district messaging and district inactions.

I feel like everyone [the district] is saying, you’ve got to make this big change and it’s got to happen so fast. I’m like who’s making this work? Because I don’t think it exists or at what cost, at what cost of children, at what cost to the human beings and your staff. But again, I feel like that’s the dishonesty in the work.

When Mary had asked for support from the district to improve student academic outcomes, they did not meet those needs.

When I made very clear asks and went to the superintendent, they were supposed to meet those requests. They didn’t, and no one came. And so either you don’t believe it can happen or you have set me up to fail. I think the hypocrisy of it is what is going to burn me out.

Mary felt forced to either accept the “way it is” culture or take an agentic approach and risk losing her job. The disconnect between the district messaging and district expectation caused distress that impacted Mary’s emotional well-being. Not only were these sources of stress outside her control, but she also had no solution for issue, letting the stress accumulate and become overwhelming (McEwen, 1998; Spector, 1998). Many of these stressors made her feel like she was just barely maintaining the school so that “things don’t fall apart”. In attending to these issues, she lacked the time and mental energy to develop efforts to make an impact on positive school change. This feeling also caused stress because she lacked the capacity to work on critical aspects of her job responsibilities.

Interestingly, Mary reported feeling that her efforts from the fall were backfiring and negatively affecting her well-being. For example, Mary developed a “live bulletin board” where teachers and staff could have immediate access and live updates to changes throughout the day. She hoped that it could address persistent questions like “Who’s out today? Do we have music? Is there PE today? Who’s subbing for who’s classrooms? What classrooms did kids get moved to?” that took significant time out of her workday. But when some teachers neglected to use it, continued to consult her directly, or wanted a printed version which defeated the “live” purposes of the document, she became more frustrated. “It’s this constant dance trying to do enough for them that they feel supported but not too much for them that they rely on me all the time.” As she reflected upon this, she wondered if she was “enabling her teachers and making it too easy for them.”

As Mary continued to develop teachers’ teaching and learning capacity, she recognized that many of her teachers who were undergoing alternative certification did not have enough information on how to obtain certification. Thus, she asked a district HR team to come to campus and hold an informational to ensure a smooth transition for many of her teachers. During this meeting, she was disappointed that HR did not provide something that teachers could take with them, so she created a one-pager that summarized the information. When she handed these out at the end of the presentation

They [HR] were like, ‘oh, this is so great. We’re going to share it with all the other principals. Is that okay?’ And I was like, ‘of course’ but why don’t you already have this? Isn’t that your job?

Here, we demonstrate that some work responsibilities that Mary is taking on—out of good intentions—may be resulting in unintended stress and frustration further impacting her well-being.

In the spring, the quantitative findings showed that Mary was experiencing significantly greater work-stress in January and March compared to the summer (see Figure 2). In fact, Mary’s perceptions of work-stress were the highest in January and March compared to the rest of the school year (see Table 2) which aligned to the constant teacher absences noted earlier. In addition to her heart rate concerns, Mary had been experiencing other physical symptoms that she attributed to work. She said “it scares me how the job’s impacting my body” which conceptually reflects allostatic load and the “wear and tear” from stress (McEwen, 1998).

I’ve just been in significantly more pain, and I don’t know if that’s partially coming from not sleeping as well or just stress, but I feel like my body just hurts all the time. It scares me sometimes how the job’s impacting my body.

Mary shared she experienced many sleep disruptions in January. When she was interviewed in March she said, “I had been for the past couple months, not sleeping very well and waking up a bunch more.” An examination of Mary’s sleep quality data during the spring showed that her sleep efficiency was significantly lower in January than during the summer (Figure 1); however, Mary’s sleep quality improved during the March sleep measurement. It is important to note that the March measurement occurred at the end of spring break when Mary felt rested and sleep quality may have re-bounded.

In February, Mary made plans to have surgery over the summer to reduce some of the pain she was feeling. Interestingly, her physical health improved. Even though she was stressed, and her mental health symptoms were high, she was more amenable to a diet plan given by her doctor to prepare for the operation. Mary said, “I have been forcing myself to eat in the morning, but mostly I noticed how much I eat when I’m on breaks, which kind highlights how little I’m eating.” Part of the challenge lay within the workload and constant demands of the positions.

However, in March, Mary was accountable to a physician, herself, and her family to maintain good health for a successful surgery. Mary also started to exercise more with a personal trainer and an app that had specific activities she needed to complete day to day.

I have to really think about where my knees are placed, where my hips are placed, and it’s physically hard. I leave those sessions and I feel better. It used to be me just thinking about work when exercising, and when I’m done, I’m just more stressed because I’ve been thinking about it for a whole hour.

For Mary, detaching mentally or “tasks that take brain power and force me to think about other things” improved her well-being. Mary expressed that it took a serious medical concern, an impending surgery, and a study on leadership health to start seriously thinking and taking actions to support her overall health. This personal support may have contributed to the positive correlation between personal support and sleep quality shown by the quantitative findings, as her sleep improved in late spring.

While Mary’s personal well-being improved towards the end of the year, her professional well-being was at its lowest point. Mary shared that she felt unsatisfied with her impact, felt misalignment within her school district, and hinted at intentions to leave. A dominant theme that emerged was a perceived inability to make positive changes at her school.

Before it was like, ‘it’s for the kids and I can make good change and things can happen’ and now I’m doubting that. Now it feels like, ‘well, why am I doing it? It might not be worth it. It might not be worth losing my physical health. I’m already feeling that sense now that I might get there.

Here, Mary shares the tension between her personal well-being, a system set up to induce ill-being, and the uncertainty that she can continue being a principal.

4.2.4. Summer: Reclaiming Well-Being

In the summer, Mary’s personal and profession well-being greatly improved. Her work responsibilities included preparing for the next academic year, hiring for vacant positions, improving curricular alignment and content, and improving aspects of her schools’ physical infrastructure. Some of these tasks were limited because teachers were not contracted to work so many of the curricular decisions and improvements moved slowly. She said, “I’m in the summer right now, and I more easily see how certain requests are unreasonable or it doesn’t really matter if I don’t do that at that time or it will be okay if this doesn’t happen.” She shared that stress from work accumulated over time during the school year, which did not happen during the summer.

Maybe it’s just the culmination of moments in those times that get me. I just thought of that now. Maybe it’s the tidal wave of all those moments that I don’t have the ability to process, and so I end up just taking them on internally and feeling them physically in a way I probably wouldn’t in June where there’s just more space.

In comparison to her first year, Mary’s said that her expectations, vision, and priorities for the school had not changed, but her second year felt a little easier because she had a better handle on the school environment. For example, she hired more teachers and staff aligned to her vision and learned more about each individual.

In the spring, Mary’s physical well-being had improved and now, in the summer, her professional well-being also improved. In regard to the chronic turnover and staffing issues, a predominant stressor, Mary’s tone shifted towards uncomfortable acceptance. She had also decided that she would continue advocating for her school--a compromise she was unwilling to make in her leadership. Moving into her third year, Mary wanted to be more intentional in maintaining the sense of calm that she felt in the summers. She said, “I am wondering how do I carry over a sense of calm into those places and what is it about that time”, in addition to taking care of her health goals from the previous year. In terms of professional goals, Mary planned to enact clearer expectations for her staff. “I feel like I’m very giving and I’m really willing to do a lot for staff and because I’ve been learning and they’ve been learning, I haven’t set really clear expectations.” For example, Mary plans to have clear expectations for how school staff needs to handle interpersonal conflicts “Here’s my expectation. X happens and person A is mad. Give yourself time, don’t say anything while you’re hot, but within 48 hours, you need to go to the person. You’ll use ‘I’ statements and this is what I expect.” Currently, Mary felt that people lean very heavily on her to resolve their issues. To reclaim some of her time, she plans to set clearer guidelines and expectations and encourage conflict resolution as a form of professional development in her school.

5. Discussion

A growing body of research has focused on stress within education systems and among teachers and principals, but few studies focus intensively on one individual longitudinally and rely upon both objective and subjective data. This study adds to existing research by utilizing a novel single-case, mixed-methods approach to study how work-stress changes over the course of a school year. Mary’s stress, health, and leadership self-efficacy evolved over the course of the study and were related to her interpretation of work-stress, her degree of control, alignment with job expectations, and her ability to ameliorate the issue which align with theoretical models of occupational stress (Lazarus & Folkman, 1984; Spector, 1998). Furthermore, not all stress was negative. Some stressors were unique challenges that were exciting and motivating (Selye, 1974). However, increased workload and some stressors, particularly those outside of Mary’s control, caused an accumulation of stress that negatively impacted Mary’s emotional, physical, and mental well-being (McEwen, 1998). We were able to tell this evolving story because we coupled objective sleep data from Actigraphy watches; quantitative, subjective measurements; and interview data collected from different times throughout the course of a year.

The quantitative data showed that Mary’s sleep quality waned during the school year as she woke more often, spent more time awake at night, and experienced lower sleep efficiency. Interestingly, sleep quality seems to rebound immediately after time off, as shown in the March measurement after spring break suggesting that breaks are an important time to de-stress. Mary’s work-stress increased during the school year, and the accumulation of mental health symptoms was stark, especially in the spring semester. Though there is little evidence in educational leadership to compare to these findings, this “wear and tear” of the individual (McEwen, 1998) is reminiscent of some teachers’ perceptions in Capel’s (1991) longitudinal study of teacher burnout. Similarly, another longitudinal study in early career teachers found that working conditions were related to postemployment depression symptoms (Schonfeld, 1992). While Mary’s perceptions of leadership self-efficacy generally improved throughout the year, the improvements were only in instructional and moral self-efficacy. Her perceptions of management self-efficacy fluctuated significantly. Correlation analyses confirmed important links between sleep quality, work-stress, and support. Sleep quality is negatively correlated to work-stress and positively correlated to personal, and not work, support.

Our qualitative findings build upon these quantitative trends by investigating the potential sources for these ups and downs. Specifically, stress from work overload begins shortly after school begins with a “hit the ground running” perception. Mary shared a litany of responsibilities and tasks aligned with management and compliance rather than leadership focused on positive school changes. While Mary perceived the stress as manageable and a result of her expected responsibilities, quantitative health data showed a decrease in her sleep quality, physical health, and mental health. Spring felt more of the same, but stress was accumulating as it emerged from systemic issues where she lacked the agency/control to mitigate or solve (e.g., staff shortages and district politics). This accumulation of stress is evident in her qualitative responses and her anxiety and depression measurements, which peaked towards the end of the school year.

Our findings in this paper align to other studies in the field. The stressors and coping mechanisms that Mary identified and enacted, respectively, matched those identified in a recent paper by Doyle Fosco et al. (2025), although Mary’s concerns were predominately job related rather than behavior related (the predominant stressor cited in Doyle Fosco et al. (2025)). While studies have cited student behavior, parent interactions, and interpersonal conflicts as major sources of stress (Doyle Fosco et al., 2025; Friedman, 2002; Mahfouz, 2020), we found that the appraisal of such stressors was more complicated (Lazarus & Folkman, 1984). For Mary, these work-related stresses were expected and a part of the job, suggesting that some identified stressors may have differential impacts based on perceptions. We also found that there is a pattern during the school year where stress accumulates and summers feel manageable, which aligns with the allostatic load model (McEwen, 1998). Though there are few longitudinal studies in educational leadership, our findings align with Capel’s (1991) study of teachers, which found that burnout was highest in February. Lastly, we argue that breaks are a critical time for school leaders to find reprieve and sustain themselves year after year.

There are a few more findings worthy of noting. First, agency and control over a stressor may not always mitigate the problem. As seen in Mary’s well-intentioned actions, some of her efforts—live bulletins and HR informational—turned herself into part of the problem, because it enabled or made others rely on the principal rather than take ownership. Here, doing things for others can cause more headaches down the road. Additionally, principals cope with and manage stress differently. Unlike other studies that have noted weight gain during principalship, we want to highlight that not eating was a complicated coping mechanism. Partly to hold in “the feelings” and exertion of control, but also because of judgement by others for taking time to oneself. Some counterintuitive findings that we could not explain were the correlations between work support, personal support, and sleep quality. While higher support at work was correlated to higher sleep quality, personal support negatively correlated to work support, suggesting a possible mediating relationship; however, more research is needed to confirm this relationship. Though mixed-methods approaches help minimize the limitations of some methodological paradigms, it is important to note that Mary’s participation in this intensive study (e.g., wearing the Actigraphy watch, daily data entries, multiple interviews, etc.) may have biased her responses, health-related behaviors, or self-care practices.

6. Implications

This research has important implications for the educational system, preparation and practice of school leaders, and educational leadership supervisors. The root cause of the work demands that Mary described can be tied to major systemic educational issues, including teacher and substitute shortages, alternative certification of teachers, accountability demands, and insufficient funding. In a related study on principals’ work-stress and sleep, we found that principals were being asked to do more with less resources and funding (E. J. Su-Keene et al., 2024b). One principal shared that she did not have enough funds to hire an assistant principal, so she absorbed the demands of that role. While this study makes clear that a principal is a unique individual navigating a complex and evolving environment, some of which includes tasks and problems beyond their scope of control, our findings suggest principals may also need more training to understand their boundaries and limitations as well as to help them to identify times within their school year when they may need help or a respite. Based on our study, we urge leaders to be more attentive to their personal health and consider devices like Actigraphy watches or sleep rings to track personal health data.

Currently, professional standards for principal preparation and practice do not include any emphasis on their own health and well-being or how to manage stress, trauma, or other demands of the job. Principal preparation programs can do more to incorporate these topics across coursework and clinical experiences and also partner with school counseling departments and programs to broaden their content expertise. Relatedly, principal supervisors and districts are responsible for creating healthy working conditions for their employees, which include principals. They too can leverage expertise within their district and set principal job expectations that include time and support for proactively managing stress and recovering from stressful events. This may include professional benchmarks for well-being; additional paid time off; programming for mental health, adult exercise classes, or healthy social events; improving HR health benefits; and finding long-term solutions to managing current workforce shortages.

This study also raises several areas of potential future research related to school leadership, stress, and health. First, our study is limited in its focus to one novice principal working in Central Texas. Additional single-case studies with principals from different backgrounds, working in different school contexts, and more years of experience would provide greater insight into how principals experience stress over the course of the school year and how contextual and personal factors are managed by unique individuals. In addition, researchers might consider studying stress within an entire campus and include principals, teachers, staff, and even students. Such studies might leverage Actigraph watches along with other stress and health data tracking to understand how an entire campus community experiences a school year and the social nature of stress within a community and organization. Lastly, we implore researchers to move beyond simply studying the problem of stress in school leadership to identifying and working to implement interventions that help principals improve their health and well-being, which might include training, support, and activities that help them develop enhanced resistance to stress, enable them to recognize when they might need a break or help, and also positive psychology interventions that can help them better appreciate the joys of principalship and working in education.

We conclude this article by urging greater attention to developing principals and principal supervisors who understand the meaning of stress and how it manifests in their work and daily lives. In the foreseeable future, principalship will remain an important but stressful position that will require leaders capable of navigating complexities and uncertainties. They must develop an expanded set of skills to manage the stress from this work. At the same time, policymakers, district administrators, and families must recognize that principals are human and need to be treated as such. We hope the story of Mary can be used to humanize principals and build an awareness that districts and communities are also responsible for creating conditions for principals to thrive. In a popular leadership book Leaders Eat Last, the author argues

the true price of leadership is the willingness to place the needs of others above your own. Great leaders truly care about those they are privileged to lead and understand that the true cost of the leadership privilege comes at the expense of self-interest.

(Sinek, 2014, p. 8)

As former practitioners, we argue that servant leadership may be problematic. Educational leaders have “eaten last” for far too long, and we believe this construct may encourages a toxic culture where leaders are asked to give more and more until they have nothing left. We encourage the field to consider tensions between prevailing leadership models and leadership sustainability.