Sustainable Education Challenges: Structure of Educational Burnout and Associations with Problematic Overstudying

Abstract

Globally increasing educational pressures and associated decreases in well-being among adolescents and young adults make educational burnout a major challenge in sustainable education, which is becoming more systematically investigated. This paper overviews previous studies on the structure of educational burnout. It provides new data on the validity of the School Burnout Inventory (SBI) among 650 Polish university students, including associations between educational burnout and problematic study-related attitudes and behaviors, stress, and anxiety. The original three-factor structure of the SBI was revised to a two-factor model due to the collinearity between cynicism and a sense of inadequacy. The adjusted model demonstrated a good fit and was cross-validated in an independent subsample. The SBI showed good reliability. Exhaustion was strongly and positively correlated with study overload. It showed a more consistent pattern of correlations with problematic study-related attitudes and behaviors than the cynicism/inadequacy component, which showed distinctive associations with learning competencies (learning self-efficacy and study enjoyment). Both components showed relatively strong positive correlations with stress and anxiety. The findings suggest that educational burnout is significantly associated with problematic overstudying and well-being, with exhaustion playing a central role in this construct. This research contributes to understanding educational burnout and its components in a Polish academic context, highlighting the importance of preventing its development to enhance student well-being and educational outcomes.

1. Introduction

In 2015, the United Nations announced a set of 17 global objectives, the Sustainable Development Goals (SDGs), aimed at eliminating poverty, protecting the environment, and ensuring peace and prosperity by 2030 [1]. Although education plays a key role in achieving these goals, serving as the foundation for attaining the remaining objectives and building a future based on sustainable development [1,2,3,4,5], there is still relatively little in the literature focusing specifically on sustainable education [6,7]. Its significance is reflected in SDG 4, which refers to Quality Education and Education for Sustainable Development [1]. Here, it is important to draw attention to a nuanced distinction between, on the one hand, sustainable education referring to educating on sustainability and, on the other hand, education that is sustainable itself, meaning educational systems and practices that enable the healthy and balanced development of knowledge and skills throughout an individual’s lifetime. The first notion is understood as education that fosters sustainability in individuals, communities, and ecosystems by promoting ecological integrity, social justice, and economic well-being [2,3]. This meaning of the term “sustainable education” is most common in the scientific literature, and numerous reviews and meta-analyses on this topic have been published [3,4,5]. It is closely associated with education that is sustainable because it refers to supporting creativity and diversity in the learning process to ensure that everyone has access to high-quality education [8]. It also serves as a foundation for transforming the structures and practices of educational institutions, with an emphasis on personalized learning and a student-centered approach [2].

The second meaning of sustainable education (i.e., education that is sustainable itself) emphasizes the urgent need to transform educational systems and practices towards learning processes that are healthy and balanced. They should minimize the risks associated with excessive educational pressures and non-optimal educational approaches, e.g., the very limited social and emotional support of vulnerable students, that may lead to problematic overstudying and educational burnout with complications such as mental and physical health problems [6,7]. The importance of sustainable education is also reflected in employers’ preferences—the majority of employers require postsecondary credentials [9]. At the same time, with the growing recognition that higher education is one of the most important sectors of our society [10], specialists in education and mental health are expressing growing concern about the mental well-being of young individuals [11,12]. Some of these concerns stem from data indicating that economic downturns, financial shocks, and macroeconomic instability significantly contribute to a decline in mental health [13,14,15]. Periods of recession often amplify uncertainty regarding future employment and financial stability, contributing to chronic stress [13]. However, the alarming rise in perceived stress is also associated with the steadily increasing worldwide educational pressures, overstudying behaviors, and the apparently growing phenomenon of educational burnout among students over the last two decades [16,17]. This phenomenon was originally conceptualized in the context of professional work [18]. Occupational burnout refers to a work-related state characterized by extreme fatigue (exhaustion), a limited ability to regulate emotional and cognitive processes (emotional and cognitive impairments), and psychological distance [19,20,21]. In this context, it represents a barrier to the achievement of SDG 8 (Decent Work and Economic Growth) [1], as employee burnout may not only diminish an individual’s quality of life and well-being (which, in turn, poses a challenge to the achievement of Goal 3, Good Health and Well-Being) but also significantly affect the efficiency and stability of employment [22].

Considering the specific responsibilities that students have, such as taking exams, attending classes, and completing homework [23], it can essentially be said that school is a place where students work [24] (p. 690). Therefore, in the context of education, burnout can be defined as experiencing exhaustion due to study demands, developing a cynical and detached attitude towards studying, and perceiving oneself as an inadequate student [19] (p. 465). Emotional exhaustion pertains to a sense of excessive emotional burden and deficits in one’s coping resources in challenging situations. Cynicism in this context is a distant attitude towards academic tasks in general, a decline in interest in one’s academic work, and a perception of it as lacking meaning. Insufficient school-related efficacy involves diminished feelings of competence, less achievement, and a sense of not accomplishing tasks both in academic work and within the school as a whole [16,19]. Beyond the primary components of burnout, this phenomenon also includes psychological distress, psychosomatic symptoms, and feelings of dejection [19].

Research shows strong connections between burnout and the mental functioning of young people [25,26,27]. For example, it co-occurs with stress [28], anxiety disorders, depression [29], and substance abuse [28]. This shows that educational burnout can hinder the achievement of SDG 3 (Good Health and Well-Being), similar to the case of job burnout. Moreover, in the burnout process, external factors (such as the learning environment, interpersonal relationships, relationships with parents, support systems, or family economic status), as well as internal factors (like personality, self-esteem, and coping abilities), may be involved [30]. Burnout among school students does not differ significantly from that experienced by university students, but the relationship between burnout and social support appears to be stronger among secondary school students than among those in higher education [31].

Some studies also point to gender-related differences in the occurrence of educational burnout, specifically indicating that girls are more prone to burnout than boys [32,33,34,35,36], particularly in the dimension of exhaustion [37,38]. Contrarily, a noteworthy body of research does not confirm this association [39,40,41,42]. However, research on problematic overstudying, strictly associated with educational burnout, consistently shows higher risks of compulsive study behaviors among women rather than men [11,17]. Without addressing healthy and balanced education issues, women will continue to suffer health and productivity inequalities in education and work (SDG 5, Gender Equality). This is particularly important in light of research, including a nine-year six-wave cross-cultural study that followed university students during their studies and several years after they entered the labor market, showing that problematic overstudying (so-called “study addiction”) is an early form of a more established and broadly studied phenomenon of work addiction [43]. A recent global study in 85 cultures on six continents (over 30,000 participants) showed that workaholism is a universal phenomenon worldwide, consistently associated with considerably higher job stress and lower job satisfaction [44]. Global data show that workaholic environments may significantly reinforce it [45]. Current knowledge of work addiction indicates that it is a major burnout risk factor and likely a considerable contributor to the global burden of disease and, consequently, one of the most important global challenges to sustainable work [46].

The field of study can also appear to be a relevant factor, with higher levels of burnout commonly observed among students in medical and caring professions, such as medicine, nursing, or psychology [26,47,48,49]. On the other hand, there is a noticeable lack of studies comparing burnout among medical students with those from non-medical fields, which limits the ability to draw broader conclusions about the role of this field of study in the development of educational burnout. Furthermore, findings suggest that the level of educational burnout may increase as students advance through subsequent years of study, and this can be observed both at the school [39,50] and in the university context [40,51,52]. As a result, burnout may contribute to school dropout, which is highly detrimental within the context of a knowledge-based economy [53] and the SDGs. This is particularly true in the context of SDG 4 (Quality Education, Education for Sustainable Development) and SDG 8 (Decent Work and Economic Growth) [22]. This demonstrates that educational burnout ultimately constitutes a major challenge to sustainable education and, consequently, to sustainable development. Therefore, it is important to take steps to prevent it.

Despite extensive research on both occupational and educational burnout, researchers still do not unanimously agree on the structure of this construct [54,55]. Contentious issues include whether a diminished level of professional efficacy/personal accomplishment is a component of burnout or a factor that can be interpreted as a cause or effect of burnout [55]. Some researchers focus only on the construct of exhaustion as the first stage of educational burnout [56]. There are also models where the components of exhaustion and depersonalization are combined, while personal accomplishment remains a separate component [57]. In other conceptualizations concerning the structure of burnout, researchers suggest that the construct consists of two components: exhaustion and depersonalization/cynicism [58,59]. Others propose that burnout is only a one-dimensional phenomenon [60,61,62].

Taking into account the data suggesting an alarming increase in burnout among students [63], it is crucial to take steps that allow for an efficient screening assessment, facilitating the implementation of preventive measures. One widely used tool for measuring school burnout is the School Burnout Inventory (SBI) developed by Salmela-Aro [16]. This instrument is based on a three-factor approach to burnout (exhaustion, cynicism, and inadequacy). It has been used in Finland [64], Spain [65,66], Germany [67], France [68], Italy [69,70], Chile [71], Serbia [72], and the USA [73,74]. Details on the psychometric properties of these versions of the SBI are provided in

Table 1. Different factorial models of the SBI tested in various countries (by publication date).

Reference	Country	Sample Type	Sample Size	One Second-Order Factor with Three First-Order Factors	Three Correlated-Factors Solution: E, C, I	Two Correlated Factors Solution: E, C/I	One-Factor Solution: Items Representing C and I
[16]	Finland	Secondary school	1418	Yes	Yes	-	-
[66]	Spain a	Lower secondary school	1096	-	Yes	-	-
[69]	Italy	Secondary school	274	-	Yes	-	-
[65]	Spain	University	578	-	Yes	-	-
[68]	France	Lower secondary school	387	-	Yes	-	-
[74]	USA b	University	2364	-	-	-	Yes
[70]	Italy	University	1021	-	Yes	-	-
[71]	Chile	Secondary school	972	Yes	Yes	-	-
[72]	Serbia	University	573	-	-	Yes	-
[67]	Germany	Secondary school	1570	-	-	Yes	-
[73]	USA c	University	151	-	-	Yes	-

^a Item 8 was removed; ^b only four items were used: 2, 3, 5, and 6; ^c item 7 was removed; E—exhaustion; C—cynicism; I—inadequacy.

.

It is worth noting that only two studies confirmed the hierarchical structure of the SBI with one general factor of burnout [16,71]. Due to high correlations between cynicism and feelings of inadequacy, a few other studies [16,69,73] also tested a two-factor model. However, this model showed an inferior fit to the data compared to the three-factor model. Then again, the most recent studies generally found that the two-factor solution best fits the data [67,72,73]. The title of the scale refers to “school burnout”; however, the wording of the items can be readily adjusted so that it allows for its valid usage among undergraduate university students by changing the word “schoolwork” to “academic work,”, e.g., “I feel overwhelmed by my schoolwork” is modified into “I feel overwhelmed by my academic work.”. The scale was adjusted this way and used among undergraduate students in previous studies, showing adequate factorial and criterion validity and reliability [65,73,75].

Another tool for investigating educational burnout has also been developed and comprehensively validated in the Polish and Norwegian contexts, the Multidimensional Inventory-Learning Profile of Students (MILPoS) [76]. It focuses on a broader framework, including the concept of problematic compulsive overstudying, so-called “study addiction,” as a phenomenon strongly associated with educational burnout (analogous to how work addiction/workaholism is closely linked to occupational burnout) and validated this in cross-cultural studies on three continents [17]. The MILPOS measures three groups of variables: (1) components of study addiction, including study compulsion (involving obsession with studying, loss of control, and tolerance and withdrawal symptoms), neglecting social relationships to study more, ignoring health problems, studying despite an individual’s appearance, and study overload meaning overburdening oneself with studying; (2) the key risk factors of study addiction (learning-related dysfunctional perfectionism associated with rigid perfectionism typical for obsessive-compulsive personality disorder, study enjoyment reflecting the pleasure component involved in the development of addictive behaviors, and a tendency to escape from personal problems into learning reflecting escapism tendencies typical for addictive problems); and (3) indicators of academic burnout (low energy and learning-related self-efficacy, and additionally a high study overload as an indicator of the perception that the amount of studying exceeds ones mental and physical resources is somewhat akin to exhaustion) [11]. Conceptually, some of these components overlap when it comes to higher-order constructs. For example, study overload is both an indicator of study addiction and burnout, and learning-related self-efficacy is both a component of burnout and learning competencies (for a detailed discussion, see [11]).

A previous study with the MILPoS identified a profile of burnout students based on cluster analysis. It encompassed students already experiencing burnout, showing the lowest levels of learning-related self-efficacy and energy, a high study overload, low learning enjoyment, and elevated scores on study addiction symptoms [11]. It comprised 11% of the students, and in comparison to other profiles (engaged students, those at risk of study addiction, and those with full-blown study addiction symptoms but still enjoying learning and showing early signs of burnout), they showed the highest exam stress and lowest physical, psychological, social, and environmental quality of life, the highest scores of depression, and lowest academic performance measured with a grade point average. Hence, this inventory can validly identify students showing educational burnout.

According to the authors’ knowledge, there has been no adaptation of the SBI in Poland. Thus, this study aims to investigate the factorial validity and reliability of the SBI among Polish undergraduate students. Furthermore, the associations between burnout, problematic study-related attitudes and behaviors, and well-being (anxiety and stress) were examined. It is hypothesized that the SBI would show good factorial validity and reliability (H1). It was hypothesized that burnout components measured by the SBI would be negatively associated with burnout components measured by the MILPoS, i.e., energy- (H2a) and learning-related self-efficacy (H2b), and positively related to study overload (H2c). Moreover, it was hypothesized that the exhaustion component measured by the SBI would be more strongly and consistently (compared to other components of burnout measured by the SBI) associated with problematic study-related attitudes and behaviors (study compulsion, neglecting social relationships, ignoring health problems, learning-related perfectionism, and a tendency to run away from personal problems into learning, i.e., escapism tendencies; H3). This is because exhaustion is a core consequence of unhealthy attitudes toward studying, resulting in an excessive study load, chronic stress, and, in effect, growing energy depletion [11,76]. On the other hand, it was hypothesized that the components of learning competencies, learning-related self-efficacy (representing the subjective perception of available knowledge, skills, and abilities to learn effectively), and study enjoyment (representing positive attitudes as a component of competencies; for discussion on the overlap between learning competencies and burnout components, see [11]) would be relatively more strongly associated with cynicism and inadequacy than exhaustion (H4). This is because a cynical attitude and detachment from studying, together with feelings of lack of competencies in learning, are, by definition, closely associated with decreased self-efficacy in learning and pleasure derived from learning [11,16,76]. Furthermore, the SBI dimensions were hypothesized to have positive relationships with anxiety (H5) and stress (H6).

2. Materials and Methods

2.1. Procedure

The participants were recruited for the study in 2016 from various fields of study, including social sciences, law and administration, natural sciences, humanities, and technical disciplines at Tri-City Universities. These universities represent a pedagogical model common to higher education institutions in Poland, which largely follow governmental regulations pertaining to the organization of third-level education. The participants completed a paper-and-pencil version of the survey during classes. This method has some advantages over gathering data online. Most notably, it allows access to students who otherwise would not participate in online studies because it could compete with free time or time devoted to studying. Participation was anonymous, with no financial or material incentives provided. Participants were informed that they could withdraw from the study at any point. This study was conducted as part of a larger research project on the potential risk factors and consequences of problematic overstudying. The project received approval from the Research Ethics Committee at the Psychology Department of the University of Gdańsk.

2.2. Participants

The sample comprised 650 university students: 390 females (60%), 256 (39.4%) males, and 4 additional persons (0.6%) who did not report their gender. The response rate was over 90%. The mean age of the participants was M = 20.26 (SD = 2.83). Regarding the year of study, the majority of respondents (534 individuals, 82.2%) were in their first year at the time of the survey. A total of 57 participants (8.8%) were in the second year, 24 (3.7%) were in the third year, two people (0.3%) were in the fourth year, and two participants (0.3%) were in the last year of study. Four individuals (0.5%) were enrolled in multiple years in different fields of study simultaneously, and twenty-seven people (4.2%) did not provide information on their year of study. Moreover, 549 people (84.5%) were full-time students, 74 (11.4%) were part-time, and 27 (4.2%) did not answer. Regarding the average grade obtained for the last semester, 149 participants (22.9%) declared an average of between 4.00 and 4.50, and 38 participants (5.8%) reported an average of above 4.51. This sample mostly represents students at the beginning of their university education, which may be related to somewhat lower levels of academic burnout than the general population, comprising more students from later years of study [40,51,52]. Nevertheless, this study aimed to investigate the relationships among the variables rather than provide prevalence rates of burnout. In this context, the demographic structure of the sample was considerably less likely to introduce bias into the results.

2.3. Instruments

Educational burnout was measured with the School Burnout Inventory (SBI) [16], which includes nine items and three subscales: exhaustion at school (e.g., “I feel overwhelmed by my academic work”); cynicism towards the meaning of school (e.g., “I feel that I am losing interest in my academic work”), and a sense of inadequacy at school (e.g., “I used to have higher expectations of my academic work than I do now”). The responses were provided on a 6-point scale, ranging from 1 (strongly disagree) to 6 (strongly agree).

The Multidimensional Inventory-Learning Profile of Students (MILPoS) [11,76] is a tool used to assess attitudes and behaviors related to learning, particularly associated with a tendency for excessive studying and educational stress. It consists of 36 items. They provide scores on nine scales, including academic burnout indicators (learning-related self-efficacy and energy), study addiction components (study compulsion, neglecting social relationships, ignoring health problems, and study overload), and crucial risk factors (learning-related dysfunctional perfectionism, study enjoyment, and a tendency to escape from personal problems in learning). Example items are as follows: study compulsion (e.g., “I cannot stop thinking about learning”), learning-related dysfunctional perfectionism (e.g., “I feel that if I make a mistake on an exam or test, people will think less of me”), study overload (excessive studying, e.g., “I feel overloaded with learning”), ignoring health problems (e.g., “For me, health problems are not a reason to stop learning”), neglecting social relationships (e.g., “Meetings with friends are less important than learning, and I ignore them”), a tendency to escape from personal problems in learning, i.e., escapism tendencies (e.g., “I learn when I do not want to think about something upsetting me”), learning-related self-efficacy (e.g., “I feel that I can learn even the most difficult material”), energy (e.g., “I feel excited, full of energy”) and study enjoyment (e.g., “Learning is a source of pleasure in my life”). The responses are on a 5-point scale, ranging from 1 (very rarely) to 5 (very often). The degree or risk of academic burnout is estimated based on low energy and learning-related self-efficacy scores, as well as a high study overload. The inventory showed good validity and reliability in previous cross-cultural research in Poland and Norway among over 8500 students [11,76].

The short version of the Perceived Stress Scale was used to measure general stress (PSS-4) [77]. It consists of four items referring to the experiences from the previous month (e.g., “How often have you felt difficulties were piling up so high that you could not overcome them?”). Response options ranged from 1 (never) to 5 (very often). The scale showed good validity and reliability in previous studies on samples of undergraduate students in Poland [11,76,78].

Anxiety was measured using the Short Anxiety Scale (SAS) [79]. It consists of five items (e.g., “I had a fear of the worst happening”), referring to the experiences from the past week. Each item is rated on a four-point scale ranging from 1 (never) to 4 (most of the time). The scale showed good validity and reliability in previous studies among Polish undergraduate students [76,80].

2.4. Statistical Analyses

Confirmatory factor analysis (CFA) is a statistical method used to evaluate the validity of a proposed measurement model [81]. This approach tests the theoretical model’s structure (which is assumed based on the theory of the construct) against empirical (or simulated) data to determine how well it fits. It is a gold standard in establishing the factorial validity of a measure, which is one of the steps demonstrating that the measure measures what it is supposed to be measuring. CFA allows researchers to verify whether specific variables (such as test items or total scores) align with expected patterns that represent theoretical dimensions [82]. It also enables the assessment of how well individual factors account for variability in observed responses, as reflected in participants’ scores on various measures [82]. CFA was performed using R version 4.4.1 [83] with the lavaan package [84], with a Maximum Likelihood Mean-Adjusted (MLM) estimator to assess the structural validity of the SBI. To assess the model fit to the data, the following indicators were utilized: the Root Mean Square Error of Approximation (RMSEA), the Comparative Fit Index (CFI), and the Tucker–Lewis Index (TLI). Cut-off scores for those indexes for an acceptable fit were CFI ≥ 0.90, TLI ≥ 0.90, and RMSEA ≤ 0.08. The investigated factorial model was calibrated and cross-validated in independent subsamples of similar size, derived from the original sample using a random choice of observations via the SPSS algorithm. The convergent validity of both tools was examined using the average variance extracted (AVE) and composite reliability (CR). The values of AVE ≥ 0.50 and CR ≥ 0.70 were good [85]. Moreover, a CFA model including both burnout indicators from MILPoS and burnout components of SBI was investigated to provide estimates of correlations among the latent factors, thus providing estimates of the strength of association accounting for the measurement error (see Figure 1, the measurement pathways are not shown on the figure for the clarity of representation but the factor loadings of the measurement part of the model are reported later). This model provides direct evidence of the level of convergence between specific indicators and components measured by these two psychometric instruments and the expected convergence (e.g., study overload from MILPoS and exhaustion from SBI), as well as associations between other indicators and components. Means, standard deviations, percentages, and correlation coefficients were calculated using IBM SPSS 29.0. All tests were two-tailed, and the significance level was set to α = 0.05.

3. Results

3.1. Confirmatory Factor Analysis

Before proceeding with the analyses, the sample was randomly divided (using the random split function in the IBM SPSS 29.0 software) into two equal subsamples to perform cross-validation. First, analyses were conducted for the first subsample. The model with three factors of burnout provided an inadmissible solution. The investigation of the estimates for the model suggested that the factors of cynicism and a sense of inadequacy are collinear. Consequently, they were merged into one factor. The value of the RMSEA for the model fit for the two-factor solution slightly exceeded the acceptable cut-off: χ² (26) = 105.07, p < 0.001; CFI = 0.94, TLI = 0.91, RMSEA = 0.097, 90% CI [0.080–0.114]. The AVE for exhaustion was 0.47, and for the factor of combined cynicism and sense of inadequacy, it was 0.57. The CR for exhaustion was 0.78, and for cynicism and a sense of inadequacy, it was 0.87. Due to a lack of acceptable model fit based on the modification indices, item 3 was removed because it cross-loaded on both factors. It resulted in the following model fit: χ² (19) = 54.50, p < 0.001; CFI = 0.96, TLI = 0.94, RMSEA = 0.076, and 90% CI [0.055–0.097]. In this case, the AVE for exhaustion was 0.47, and for the combined cynicism and sense of inadequacy factor, it was 0.60; the CR for exhaustion was 0.78, and for cynicism and a sense of inadequacy, it was 0.85. This model was acceptable. Additionally, item 1 (“I feel overwhelmed by my schoolwork”) and item 2 (“I feel a lack of motivation in my schoolwork and often think of giving up”) residuals were allowed to correlate based on modification indices. These two items are phrased in a relatively extreme fashion and share a considerable emphasis on the feelings of being completely overwhelmed with studying. This model had a good fit to the data: χ² (18) = 29.15, p < 0.05; CFI = 0.99, TLI = 0.98, RMSEA = 0.04, 90% CI [0.013–0.069]. The AVE for exhaustion was also 0.47, and for the combined cynicism and sense of inadequacy factor, it was 0.60. The CR for exhaustion was 0.78, and for cynicism and sense of inadequacy, it was 0.85. The correlation between residuals of item 1 and item 2 was 0.37. The second subsample was used as a cross-validation sample. The fit of the model for the final two-factor solution (without item 3 and with correlated residuals) was as follows: χ² (18) = 27.83, p = 0.065; CFI = 0.99, TLI = 0.98, RMSEA = 0.041, and 90% CI [0.001–0.080]. The fit of the final model in both subsamples did not differ statistically: χ² (6) = 4.39, p = 0.623. Detailed values for all models can be found in

Table 2. Fit statistics for confirmatory factor analysis models.

Model	n	χ2	df	CFI	TLI	RMSEA	SRMR
Three-factor model	325	76.23	24	Collinearity between two factors
Two-factor model	325	105.07	26	0.935	0.911	0.097 90% CI [0.080–0.114]	0.060
Two-factor model without item 3	325	54.50	19	0.962	0.944	0.076 90% CI [0.055–0.097]	0.048
Two-factor model without item 3-correlated residuals	325	29.15	18	0.988	0.982	0.044 90% CI [0.013–0.069]	0.037
Cross-validation (two-factor model without item 3)	325	49.22	19	0.967	0.951	0.071 90% CI [0.050–0.092]	0.041
Cross-validation (without item 3-correlated residuals)	325	27.83	18	0.989	0.983	0.041 90% CI [0.001–0.080]	0.030

CFI = Comparative Fit Index; TLI = Tucker–Lewis Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual.

. Standardized factor loadings of the items for both samples were adequate, and most were relatively high. They are presented in

Table 3. Standardized factor loadings for the two-factor model of school burnout in both samples.

Items	Exhaustion Sample 1/Sample 2	Cynicism and Inadequacy Sample 1/Sample 2
Item 1 (exhaustion)	0.70 **/0.65 **	-
Item 4 (exhaustion)	0.74 **/0.74 **	-
Item 7 (exhaustion)	0.57 **/0.60 **	-
Item 9 (exhaustion)	0.74 **/0.75 **	-
Item 2 (cynicism)	-	0.76 **/0.77 **
Item 5 (cynicism)	-	0.86 **/0.86 **
Item 6 (cynicism)	-	0.83 **/0.80 **
Item 8 (inadequacy)	-	0.61 **/0.68 **

** p < 0.001.

. Moreover, the reliability coefficients are presented in

Table 4. Reliability coefficients, means, standard deviations, and correlations between burnout, study overload, study-related attitudes and behaviors, stress, and anxiety for the whole sample (n = 650).

Variable	Reliability α/ω	M/SD	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.
1. Exhaustion	0.78/0.78	13.02/4.67	-
2. Cynicism and Inadequacy	0.85/0.85	14.04/5.14	0.52 **	-
3. Study overload	0.76/0.77	11.99/3.83	0.63 **	0.29 **	-
4. Energy	0.85/0.85	8.79/2.81	−0.31 **	−0.34 **	−0.19 **	-
5. Learning-related self-efficacy	0.79/0.78	15.66/3.93	−0.28 **	−0.44 **	−0.12 **	0.42 **	-
6. Study enjoyment	0.86/0.86	9.18/3.60	0.04	−0.32 **	0.07	0.24 **	0.42 **	-
7. Study compulsion	0.79/0.79	10.99/4.25	0.49 **	0.03	0.49 **	−0.02	0.05	0.42 **	-
8. Neglecting social relationships	0.78/0.79	8.38/3.55	0.48 **	0.03	0.49 **	−0.14 **	0.04	0.32 **	0.59 **	-
9. Ignoring health problems	0.74/0.74	11.07/3.76	0.26 **	−0.07	0.40 **	−0.01	0.28 **	0.25 **	0.47 **	0.43 **	-
10. Learning-related perfectionism	0.74/0.74	10.43/3.73	0.57 **	0.34 **	0.56 **	−0.29 **	−0.22 **	0.07	0.50 **	0.48 **	0.34 **	-
11. Escapism tendencies	0.83/0.83	5.32/2.56	0.36 **	0.01	0.27 **	−0.06	0.05	0.41 **	0.56 **	0.52 **	0.34 **	0.41 **	-
12. Stress	0.69/0.69	11.14/3.00	0.52 **	0.49 **	0.39 **	−0.39 **	−0.39 **	−0.02	0.25 **	0.24 **	0.13 **	0.50 *	0.26 **	-
13. Anxiety	0.74/0.74	10.34/3.35	0.47 **	0.36 **	0.37 **	−0.30 **	−0.21 **	0.07	0.34 **	0.29 **	0.19 **	0.47 **	0.37 **	0.56 **

n = 650, α—Cronbach’s alpha; ω—McDonald’s omega; M—mean; SD—standard deviation. ** p < 0.001, * p < 0.05.

. Both subscales showed good reliability.

3.2. Concurrent and Convergent Validity

Firstly, women differed from men in terms of levels of exhaustion (t (641) = 4.70; p < 0.001; Cohen’s d = 0.38), and for cynicism combined with a sense of inadequacy (t (639) = 2.00; p = 0.046; Cohen’s d = 0.16(. Men had lower levels of both exhaustion (M among men = 11.97, SD = 4.63; M among women = 13.72, SD = 4.60) and cynicism combined with a sense of inadequacy (M among men = 13.56, SD = 5.14; M among women = 14.39, SD = 5.13).

Convergent validity was investigated in the entire sample. The results of the correlation analyses showed that exhaustion is strongly positively correlated with study overload and moderately negatively correlated with energy- and learning-related self-efficacy. Also, the combined cynicism/inadequacy was moderately positively correlated with study overload and negatively correlated with energy and learning-related self-efficacy. Exhaustion, but not cynicism/inadequacy, was positively correlated with study compulsion, neglecting social relationships, ignoring health problems, and a tendency to run away from personal problems into learning (escapism tendencies). Exhaustion was also relatively more strongly positively correlated with learning-related perfectionism than cynicism/inadequacy. The reverse was true for learning-related self-efficacy, which was relatively more strongly correlated with cynicism/inadequacy than exhaustion. Also, study enjoyment was negatively correlated to cynicism/inadequacy but not exhaustion. Both burnout factors were positively correlated with stress and anxiety. The details are provided in Table 4.

Moreover, latent correlations were calculated between burnout components measured by the SBI and burnout indicators measured by the MILPoS of the full sample. All correlations were statistically significant. The exhaustion component was highly convergent with the study overload. The correlation between them was higher than the correlation between exhaustion and the combined dimension of cynicism and a sense of inadequacy, which, in turn, showed moderate convergence with learning-related self-efficacy. The correlation with study overload was moderate and negative for the two burnout indicators from the MILPoS (learning-related self-efficacy and energy). Notably, their correlations with the exhaustion component were somewhat higher. The details of the relationships between the variables are presented in Figure 1, and factor loadings for this model are presented in

Table 5. Standardized factor loadings for the final model with latent correlations.

Items	Exhaustion	Cynicism and Inadequacy	Study Overload	Energy	Learning- Related Self- Efficacy (LRSE)
Item 1 (exhaustion)	0.70 **	-	-	-	-
Item 4 (exhaustion)	0.75 **	-	-	-	-
Item 7 (exhaustion)	0.55 **	-	-	-	-
Item 9 (exhaustion)	0.71 **	-	-	-	-
Item 2 (cynicism)	-	0.77 **	-	-	-
Item 5 (cynicism)	-	0.86 **	-	-	-
Item 6 (cynicism)	-	0.82 **	-	-	-
Item 8 (inadequacy)	-	0.64 **	-	-	-
Item 10 (study overload)	-	-	0.53 **	-	-
Item 11 (study overload)	-	-	0.66 **	-	-
Item 12 (study overload)	-	-	0.71 **	-	-
Item 13 (study overload)	-	-	0.75 **	-	-
Item 14 (energy)	-	-	-	0.77 **	-
Item 15 (energy)	-	-	-	0.84 **	-
Item 16 (energy)	-	-	-	0.82 **	-
Item 17 (LRSE)	-	-	-	-	0.67 **
Item 18 (LRSE)	-	-	-	-	0.62 **
Item 19 (LRSE)	-	-	-	-	0.64 **
Item 20 (LRSE)	-	-	-	-	0.72 **
Item 21 (LRSE)	-	-	-	-	0.61 **

** p < 0.001.

.

4. Discussion

The study results indicate that the model with two correlated factors of the SBI shows good factorial validity and reliability. The original version proposed by Salmela-Aro [16] assumed a three-factor solution. However, in this study, cynicism and inadequacy factors were collinear and had to be merged into one factor. Also, item 3 (“I often have feelings of inadequacy in my schoolwork”) had to be dropped due to cross-loadings on two factors. This could be partially associated with the fact that the Polish translation of the phrase “feelings of inadequacy” is similar to expressions of being unable to face challenges related to learning, which is akin to the feelings of being overwhelmed or thinking of giving up measured by items 1 and 2. These effects show that considerable attention to details of content and the meaning of items need to be paid because they may readily overlap when it comes to measuring exhaustion and other components associated with a decrease in learning self-efficacy. After these modifications, an acceptable fit of the model to the data was obtained, and it was cross-validated in an independent subsample (H1 partially substantiated). Thus, it provided more robust evidence that the two-factor model fits data better than the three-factor model in the Polish sample. The SBI also showed good reliability. The two-factor solution is congruent with the results obtained in previous studies on educational and occupational burnout [58,59], including studies using the SBI, particularly those conducted most recently [67,73,75]. Moreover, some authors reported reliability problems regarding the inadequacy component [86]. Our results add to the discussion about whether burnout is a one-, two- or three-component construct.

Influential burnout conceptualization defines it as “a state of physical, emotional and mental exhaustion that results from long-term involvement in work situations that are emotionally demanding” [87] (p. 501). In line with this definition, recent studies indicate that burnout’s core is emotional exhaustion [88]. Moreover, longitudinal studies also indicate that emotional exhaustion is the most relevant dimension of academic burnout and has the strongest impact on mental health [89,90]. Some researchers have even created separate tools to measure exhaustion [56]. Many researchers also maintain that burnout comprises two primary components, exhaustion, and cynicism, and that the nature of the association between inadequacy and the other two dimensions of school burnout remains uncertain [91].

Some researchers have proposed that inadequacy is caused by cynicism, arguing that disengagement leads to a diminished sense of efficacy. Conversely, others suggest that inadequacy is directly linked to exhaustion, which impairs performance [90,92]. Taris [93] proposed that exhaustion may induce inadequacy both directly and through the intermediary effect of cynicism. Furthermore, it is worth noting that the disagreement over the number of components influencing burnout may also stem from the diverse contexts in which burnout is examined [91]. Nevertheless, exhaustion remains a central aspect across all conceptualizations of burnout, whether it pertains to occupational conditions [94], student burnout [95], or parental burnout [96].

Moreover, this study’s results show that the SBI components and the indicators of burnout measured by the MILPoS show expected associations. Both the exhaustion and combined cynicism/inadequacy components showed negative associations with energy (H2a-substantiated) and learning-related self-efficacy (H2b-substantiated). Importantly, exhaustion and study overload showed high convergence, supporting the notion that they measure the same core aspect of educational burnout. Their latent correlation was higher than their correlations with the cynicism/inadequacy component, which were also positive (H2c-substantiated). This finding provides an important insight into the measurement of the core exhaustion component of burnout. Study overload was initially conceptualized primarily as a component of problematic overstudying because it reflects a tendency to invest excessive time and effort into studying, and previous studies consistently showed that such conceptualization is valid (the CFA results provided a very good fit to the data of a hierarchical factorial model in which study overload was a first-order factor loading on a second-order factor of study addiction) [3,47]. However, feeling overburdened by the study load measured in a subjective self-report manner involves the interaction between two factors: (i) the level of study load and (ii) personal resources available to deal with it. Therefore, it reflects both the tendency to overburden oneself with excessive studying and the feeling of lacking resources and exhaustion from study overload. This suggests a close and intricate association between problematic overstudying and educational burnout, which requires further systematic investigation.

Moreover, learning-related self-efficacy and energy were only moderately associated with study overload and exhaustion. It shows that these burnout indicators assessed by the MILPoS cannot be used as direct measures of educational burnout. Still, consistent with how they were originally conceptualized [3,47], they should be applied as indicators for clustering methods such as latent profile analysis to identify subpopulations of students at high risk of burnout or experiencing full burnout. In such analyses, they provide important advantages of differentiating students who feel exhausted and cynical about their studying but outside the academic context still function relatively well compared to those burnout students who experience considerable functional impairments such as general low energy, helplessness (very low self-efficacy), and associated physical and mental health problems [3].

In relation to other study addiction components, exhaustion, but not cynicism/inadequacy, showed a positive relationship with study compulsion, neglecting social relationships, and ignoring health problems (H3-substantiated). These are components of problematic overstudying, and it further supports the notion that exhaustion, similarly to study overload, is intricately associated with this phenomenon [3,47]. Regarding the risk factors for study addiction, perfectionism was positively associated with both exhaustion and the cynicism/inadequacy dimension. Furthermore, the tendency to escape from personal problems into learning (escapism tendencies) was positively linked to exhaustion, while study enjoyment was negatively associated with the combined cynicism/inadequacy dimension. Also, cynicism/inadequacy was relatively more strongly related to learning-related self-efficacy and study enjoyment, conceptualized as learning competencies (H4-substantiated). This component represents the subjective perception of having available knowledge, skills, and abilities to learn effectively and the component of positive attitudes (pleasure of learning), see [11]. This is consistent with the assumptions that guide the development of the MILPoS [11,76,97]. Notably, exhaustion was relatively more strongly associated with study overload than cynicism/inadequacy combined and even more strongly than energy levels. The strength of the latent correlation of exhaustion and study overload indicates high convergence. It suggests that exhaustion, which is very closely associated with the perception of excessive workload (a subjective assessment that the demands associated with the amount of studying exceed the available mental and physical resources), seems core to the experience of educational burnout. Exhaustion was also related to all components of problematic overstudying, showing that it is closely associated with study compulsion, which may lead to ignoring health problems and neglecting social relationships, contributing to the growing perception of excessive study load and stress, and, in consequence, aggravating burnout [11,76].

The current study found positive associations between exhaustion and cynicism/inadequacy combined with stress and anxiety (H5- and H6-substantiated), which is in line with evidence from previous research. Educational burnout is strongly related to well-being, including stress [98,99] and anxiety [28,100]. These may be both risk factors of burnout as well as its potential consequences because excessive workload may lead to high stress and anxiety.

When considering the issue of educational burnout from the perspective of sustainable education, two dimensions appear to be the most significant: excessive pressure and limited support [7,11,12,17,27,28,31]. These dimensions can hinder the development of students and lead to a decrease in their long-term well-being [7,31]. This occurs because excessive academic demands and insufficient support mechanisms manifest across multiple levels: (i) at the macro level through educational policy and structural expectations, (ii) at the institutional level through school or university practices, and (iii) at the micro level through pedagogical approaches, the availability of resources, and students’ individual vulnerabilities. Moreover, previous studies indicate that support should also be directed at reducing subjective perceptions of threat, not solely at responding to actual difficulties [14]. Only by becoming aware of the specific areas in which the problem lies can we effectively plan targeted interventions aimed at preventing the development of educational burnout, thereby supporting the achievement of the SDGs.

4.1. Strengths and Limitations of This Study

Regarding the strengths of the study, the sample size was relatively large, providing high statistical power, and reliable and valid psychometric tools were used. The CFA results were cross-validated in independent samples, providing robust evidence for the factorial structure of the SBI among Polish undergraduate students. Moreover, latent correlations estimate the strength of associations between components of burnout measured by the SBI and the indicators of burnout measured by the MILPoS accounting for the measurement error. To the authors’ knowledge, this is the first study to investigate the associations of the burnout components measured by the SBI with a wide range of problematic overstudying attitudes and behaviors, including their risk factors.

In terms of the study’s limitations, the sample came from convenience sampling, which means that the results cannot be generalized to other populations without some reservations. Moreover, all data were self-reported, which leaves the results of this study open to the usual weaknesses of such data (e.g., social desirability and recall biases). Future research should involve a more representative (stratified and/or randomly drawn) sample of the general population of students.

4.2. Future Directions

Burnout (both in education and the workplace) is still a debatable construct in terms of its nature and structure, with some researchers suggesting that it could be an early form of depression or depression itself [101,102]. Converging data, including the results of the present study, suggest that exhaustion is its core component. Valid and reliable tools for assessing burnout in various contexts, including education, may help develop high-quality research and advance this debate.

Given that most existing studies focus on students in medical and caring professions, it would be valuable to compare burnout levels across various fields of study, including the humanities, social sciences, and technical disciplines. Additionally, future research should more thoroughly explore educational burnout in the context of neurodivergent individuals whose experiences and stressors may differ significantly from those of the general student population and who are identified as vulnerable populations.

Longitudinal designs would also help to better clarify how burnout evolves over time and how it relates to career outcomes, mental health (other than depression and anxiety), and dropout rates. Future studies should systematically investigate the mechanisms of educational burnout development, taking into account the role of problematic overstudying behaviors, their risk factors, and underlying psychological, social, and institutional mechanisms. This is particularly important due to the close and intricate association between compulsive study behaviors and burnout and their negative effects, including functional impairments and considerable individual, social, and economic harm [72]. In order to understand the scope of this harm, systematic studies into the evaluation of its costs and challenges for sustainable social and economic development are warranted, especially taking into account that these educational phenomena are associated with subsequent functioning in working environments, including analogous work addiction and occupational burnout problems which are known to be major challenges to a sustainable work and family life [73]. Multidimensional tools (such as the SBI or MILPoS) and methods like latent profile analysis (LPA), which allow for the identification of specific profiles of students depending on their burnout risk or status, may prove helpful [103].

Academic institutions should aim to develop strategies and implement evidence-based preventive programs to mitigate the emergence of educational burnout. Moreover, to achieve the highest possible effectiveness of interventions, prevention strategies should be directed both at the individual level (e.g., stress management training) and at the organizational level (e.g., alleviating exam-related pressure) [104]. Creating training programs that help educators recognize and respond to early warning signs of burnout in both students and teaching staff is also important. Policymakers should invest in training teaching staff in student-centered methods, such as tutoring [105], experiential learning [106], and service learning [107,108]. These approaches not only enhance student engagement and motivation but also foster a more supportive and sustainable educational environment.

5. Conclusions

Overall, the results show that the SBI successfully measured two components of educational burnout among undergraduate students in Poland and that exhaustion seems to be its core component. It is consistently associated with problematic and unhealthy study-related attitudes and behaviors, including study addiction components and risk factors. This suggests a mechanism of educational burnout development in which problematic overstudying behaviors (including deprioritization of all other spheres of life, social relationships, and ignoring health problems) lead to chronic and high stress and a growing perception that study load exceeds one’s mental and physical resources analogously to how occupational burnout is currently understood [21]. Accordingly, it may lead to educational burnout and associated consequences such as psychological and physical health problems, a decrease in academic performance, and impairments in social and other areas of functioning [11,25,26,27,28,29]. This directly negatively affects sustainable development in educational and later occupational domains. These conclusions point to the need for reforms in the education system and the labor market organization, including, among others, ensuring greater job stability and fair remuneration. Such reforms are crucial primarily for improving the functioning of society as a whole, including reducing expenditures on social and healthcare services and enhancing economic efficiency [7]. The present study lays the foundation and constitutes a benchmark for further investigation of how problematic overstudying may lead to burnout and its consequences.