Abstract
Teacher burnout and attrition are significant concerns in the United States and globally, particularly in high-needs areas such as world language (WL) teaching. Despite extensive international research on teacher burnout and attrition, few studies have specifically examined how demographic characteristics may influence burnout and intent to quit among WL teachers. To address this gap, this study employed a cross-sectional research design utilizing a factorial multivariate analysis of variance (MANOVA) to explore the relationships between various demographic factors and teacher burnout and intent to quit in high school WL teachers. Analysis revealed statistically significant main, interactive, and between-participants effects for a range of personal characteristics (i.e. gender, age, race, and ethnicity), teacher characteristics (i.e. number of WLs taught, primary language status, years of experience, highest level of education completed, type of certification program, number of professional organizations, and number of additional certifications), and school characteristics (i.e. urbanicity of school, type of school, and region). Findings suggest a need to take these factors into consideration when addressing teacher attrition and burnout through research and practice.
Keywords
I Introduction
In recent years, there has been a growing concern over the burnout and exodus of educators from the teaching profession. Steiner and Woo (2021) reported that nearly one in four teachers in the United States (U.S.) considered leaving the classroom by the end of the 2020–21 school year, with one in six contemplating departure even before the onset of the Covid-19 pandemic. Nguyen et al. (2022) reported more than 36,000 vacant teaching positions in the U.S., as well as a staggering 163,000 U.S. teaching positions being filled by underqualified individuals. Meanwhile, a National Education Association survey revealed that 55% of surveyed teachers expressed their intention to leave the profession earlier than expected (Jotkoff, 2022). Scholars have expressed concern that this problem will likely continue if not addressed (Diliberti et al., 2021; Goldhaber & Theobald, 2022).
Teacher attrition has dire ramifications, with schools grappling with severe staffing shortages, resorting to measures like recruiting central office staff, superintendents, business community members, and even the National Guard (Duncan, 2022). Some schools have resorted to offering increased starting salaries and hiring bonuses to attract and retain teachers (Nguyen et al., 2022). Equity-related concerns have also emerged, with remote and resource-deprived schools, as well as those with high-needs student populations, finding it increasingly challenging to attract and retain teachers, especially teachers of color (Jacobs, 2021). These issues subsequently affect students’ access to effective teachers, as highlighted by Jennings (2020), who warns that teacher burnout and rising attrition rates pose a dire threat to education systems in the U.S. and globally.
In addition to being linked to teacher attrition, teacher burnout also presents detrimental impacts. Teachers who experience burnout may experience adverse mental and physical consequences, such as chronic stress, anxiety, fatigue, and lack of motivation (Bakioğlu & Kiraz, 2019; Mahmoodi-Shahrebabaki, 2019). Teacher burnout may also negatively impact pedagogy and practice, with consequences to preparation and quality of instruction (Jacobson, 2016). Burned-out teachers may also be more likely to exhibit higher rates of absenteeism (Caruso, 2019).
Given the substantial concerns surrounding teacher attrition and burnout, it is imperative to explore potential factors contributing to these issues, particularly in high-needs areas. One field that research has identified as a field of high needs is world languages (WLs) (Carver-Thomas & Darling-Hammond, 2017, 2019). The scarcity of teachers in WLs has been a longstanding issue of significance (P.B. Swanson, 2008a, 2012; P. Swanson, 2022). American Council on the Teaching of Foreign Languages (ACTFL, 2020) reported shortages of WL and bilingual education teachers in 49 of 56 U.S. states and territories for the 2017–18 academic year. They also projected an annual shortage of 100,000 teachers by 2025. Additionally, P. Swanson (2022) identified that 22% of WL teachers surveyed considered leaving the teaching profession during the global pandemic. Most recently, District Administration (2023), which reports that more than half of public schools felt understaffed at the beginning of the 2022–23 school year, positioned WLs as an area of particular concern, more so than other high-needs areas like special education, math, and science. However, there is a gap in research related to the burnout and attrition in WLs (McConnell, 2024; McConnell & Swanson, 2024). This gap includes information on the personal, teacher, and school characteristics that may be of interest when investigating burnout and attrition in this population.
Thus, this article focuses on the possible demographic factors that may be related to teacher burnout and intent to quit in high school WL teachers as a unique population of interest. To do so, the following research question is addressed:
• Research question: Are there significant mean differences in levels of burnout and intent to quit in high school WL teachers based on various demographic factors (i.e. personal characteristics, teacher characteristics, and school characteristics)?
II Literature review
1 Burnout and attrition in world language teachers
Burnout and attrition in WL teachers is a topic of great concern that continues to be met with a lack of focused research. Much extant research focuses on teachers of English as a foreign language (EFL). However, the research that has been conducted on burnout and attrition in WL teachers agrees that they are a unique population of interest meriting a specialized focus. For instance, Powers (2022) investigated the burnout coping strategies of 165 WL teachers in the U.S. during the Covid-19 pandemic. The study revealed that crises like the pandemic can exacerbate burnout by limiting teachers’ access to effective coping strategies, potentially increasing attrition. Similarly, Hlas and Hlas (2024) emphasized the creative strengths of pre-service WL teachers, particularly when they promote interdisciplinary connections and support as well as a sense of autonomy and agency. These creative strengths align with P.B. Swanson’s (2008b) earlier findings on the Holland profile of aspiring WL teachers as being social, artistic, and enterprising. Swanson argued that incorporating these traits into recruitment and admissions decisions could help mitigate WL teacher attrition.
Further research supports the significance of specific factors in addressing burnout and retention in WL teachers. For example, Moser and Wei (2021) surveyed K-12 language teachers impacted by the Covid-19 pandemic, finding that those who stayed and those who left the profession exhibited different profiles (i.e. stayers, conditional stayers, and leavers). Leavers tended to report lower levels of human, social, and structural capital. After analysing interviews with five WL teachers in the U.S. using a four-capital theoretical model of teacher retention (Mason & Matas, 2015), McConnell (2024) noted social capital of particular importance in preventing attrition of WL teachers, followed by structural capital, human capital, and personality/psychological factors. Of particular importance were collaboration within supportive school cultures, access to resources, tailored professional development, opportunities for promotion, a manageable workload, and leadership that involves WL teachers in decisions that impact their classrooms and empowers WL teachers to act autonomously as professionals of their content. McConnell and Swanson (2024) noted relationships between WL teachers’ sense of empowerment and their feelings of burnout and desire to quit. Of particular concern for WL teachers were professional growth, self-efficacy, and autonomy. This resonated with prior research affirming the importance of self-efficacy in preventing WL teachers from leaving the profession prematurely (P.B. Swanson, 2010).
2 Demographic factors in teacher burnout
A number of demographic factors have been explored with respect to teacher burnout, which prior research has conceptualized as a phenomenon characterized by exhaustion (formerly emotional exhaustion), cynicism (formerly depersonalization), and inefficacy (formerly reduced personal accomplishment) (Maslach, 1982, 1998; Maslach & Leiter, 2016). Mazidi et al. (2017) found significant differences in teachers’ sense of powerlessness as a dimension of burnout in elementary school teachers in Iran based on gender. They suggested that males may face more pressures than females both socially and psychologically. Conversely, Jamaludin and You (2019) suggested that among 31 educators surveyed, female educators were particularly impacted by feelings of burnout in terms of reduced personal accomplishment. Ghasemi (2023) found that an empowerment-based intervention alleviated burnout levels of 54 EFL teachers in Iran. Like Jamaludin and You, they also noted that female teachers reported higher levels of burnout and greater benefit from the intervention. Kara (2019) noted gender-based variations in burnout and job satisfaction in 308 visual arts teachers in Turkey. More specifically, they found that mean levels of emotional exhaustion and depersonalization were higher in female teachers, while male teachers demonstrated higher levels of intrinsic and extrinsic job satisfaction. Meanwhile, unlike the previous studies, Răducu and Stănculescu (2022) found that although men demonstrated a higher representation than women in the low, moderate, and high burnout risk categories, the difference was not statistically significant.
In addition to differences based on gender, Jamaludin and You (2019) also found differences in all three burnout dimensions (i.e. emotional exhaustion, depersonalization, and reduced personal accomplishment) based on teaching experience and educational level. Reduced personal accomplishment levels were noted in particular for those with master’s degrees and between six and 10 years of experience. Ghasemi (2023) similarly found that in addition to differences based on gender, teachers with less overall experience reported higher levels of burnout and greater benefit from the intervention in their study. Răducu and Stănculescu (2022) also found a significant impact for professional experience; however, they found that more experienced teachers were more likely to belong to the low burnout risk category than to the no burnout risk category, especially when compared to teachers with less than two years of experience. Kara (2019) found that differences emerged with respect to reduced personal achievement, depersonalization and intrinsic job satisfaction, with those with 21 years or more of seniority experiencing higher levels of those factors than those with less than five years of seniority. Kara found no significant differences with respect to age, however.
Beyond gender and experience, additional factors have been explored. In their work with early education teachers in Kosovo, Hyseni Duraku et al. (2025) found job satisfaction, work motivation, and professional development were all significant negative predictors of burnout. They also noted other professional development as a factor that correlated with burnout. In Kara (2019) noted variations in burnout based on school type, specifically in terms of reduced personal achievement, depersonalization and intrinsic job satisfaction. Burnout levels were higher for those working in private schools in particular. However, Răducu and Stănculescu (2022) found no significant impact of rural/urban teaching environment on teacher burnout.
3 Demographic factors in teacher attrition
Demographic factors have also surfaced in research on teacher attrition, which the present study defines as educators voluntarily leaving the profession before retirement (Ingersoll, 2001). Nguyen et al. (2020) found that teacher turnover was more likely with teachers who taught STEM or special education, as well as those who had a higher academic ability, were younger, and had less experience. They also found a lower likelihood of turnover in minority race teachers compared to White teachers. They did not find a significant impact for teachers based on gender or having a graduate degree. They also found that urbanicity and student characteristics (e.g. proportions of students who identify as Black, Hispanic, eligible for free-or-reduced price lunch, or in need of enhanced supports) were not significant predictors of teacher turnover. Meanwhile, Crouch and Nguyen (2020) found that teachers in rural schools experienced lower levels of turnover than those in urban schools, in addition to being younger and less experienced.
As demonstrated in the literature discussed above, burnout and attrition are global concerns, particularly for specialized content areas like WLs. In addition, although social and organizational factors are crucial to understanding burnout and attrition, demographic variables are also of interest, as they intersect with other factors that influence teachers’ experiences with burnout and attrition. Thus, this study aims to address the gap in research on demographic variables related to burnout and attrition in WL teachers in the U.S. to shed light on which demographic factors should be taken into consideration when researching these phenomena.
III Method
1 Context
To facilitate greater access to high school WL teachers across the U.S. as a specialized population (Dörnyei, 2007), the quantitative survey for this study was administered online using Qualtrics. The researcher obtained Institutional Review Board approval and ensured informed consent. Furthermore, the researcher assured participants that their responses would be anonymous and confidential.
2 Setting and participants
The researcher recruited n = 313 participants across all regions of the U.S. via non-probability convenience sampling for ease of administration and low cost (Dörnyei, 2007). This was accomplished by posting invitations on several social media groups dedicated to WL teachers as well as ACTFL’s special interest research group. Recruitment criteria included being a current U.S. high school WL teacher or a high school WL teacher who had quit teaching within the last two years. 83 percent of participants reported that they were still teaching, while 17 percent indicated that they had quit teaching within the last two years. To increase the representativeness of the data, invitations were distributed to social media groups whose members taught diverse languages. As such, although participants taught only one or two languages, with the most commonly taught languages being Spanish, French, German, or Chinese, participants also reported teaching Russian, Japanese, Portuguese, Italian, Korean, and Arabic. Overall, a majority of participants reported being White, female, 31 to 35 years of age, and not of Hispanic, Latino, and/or Spanish origin. Most reported teaching only one language, with a majority teaching a language that was different from their first language, resonating with literature in which most participants reported teaching Spanish or French (P.B. Swanson, 2008b). A majority reported having at least a master’s degree, completion of a traditional certification program, and no membership in professional organizations or additional certifications. Most participants taught in urban and public schools, with a majority having spent most of their time teaching in the Southwestern U.S. For detailed numerical data on these demographic statistics, please refer to Tables A1 to A3 in the Appendix.
3 Instruments
As part of a larger study, the researcher created two scales – an Intent to Quit Scale and a Perceived Burnout Scale. All four scales included Likert-style questions (1 = strongly disagree; 5 = strongly agree). Items and median levels of agreement are provided in Table 1.
Median levels of agreement among teachers’ perceptions of empowerment, job satisfaction, and burnout.
The researcher then conducted Confirmatory Factor Analysis to determine discriminate validity of the constructs. The results are included in Table 2. All factor loadings were statistically significant (p < .001) and had factor loadings above .7, indicating that the items measured were strong indicators of their respective constructs.
Confirmatory factor analysis.
The researcher also confirmed content validity with the help of content area experts (Escobar-Pérez & Cuervo-Martínez, 2008). More specifically, the researcher enlisted nine current and former WL teachers to act as experts. This number aligned with Yusoff’s (2019) suggestion that the number of experts should be between six and 10. These experts provided ratings on the relevance of each item to its overarching construct. Table 3 shows the resulting Kappa values for each instrument as well as an evaluation of its acceptability (Cicchetti & Sparrow, 1981; Fleiss, 1981).
Content validity analysis for survey instruments.
To further assess content validity, the researcher ran nonparametric bivariate correlation analysis on the items for each scale. All items loaded significantly on their overarching construct (p < .001 in all cases). Additionally, a majority of the items exhibited correlation coefficients lower than .8, indicating that they were related but still distinct. Those which were higher than .8 were only marginally higher, with the highest correlation for burnout at .828 (‘I often feel overwhelmed as a result of teaching’/‘I often feel emotionally exhausted as a result of teaching’) and the highest for intent to quit at .846 (‘I frequently think about voluntarily quitting my job as a teacher’/‘I would voluntarily quit teaching if I had the opportunity’).
4 Data collection
To facilitate greater access to high school WL teachers across the U.S. as a specialized population (Dörnyei, 2007), the quantitative survey for this study was administered online using Qualtrics. The researcher obtained Institutional Review Board approval and ensured informed consent. Participants accessed the Qualtrics online survey management platform using a link provided by the researcher. The first question of the survey asked participants for their current status: ‘Current high school world language teacher in the United States’ or ‘Former high school world language teacher in the United States who recently quit teaching (within the last 2 years)’. Participants who did not meet one of these criteria were unable to continue the survey.
The researcher removed data for participants who had a large amount of incomplete data or failed the validity questions, which asked participants to ‘please select the “Disagree” option for this question’ and ‘please select the “Agree” option for this question’ to increase researcher confidence that participants were thoughtfully reading and processing each question.
5 Analysis
The researcher assessed the reliability for the combined survey as well as each of the scales used (see Table 4). The reliability scores were all considered acceptable or higher (Henson, 2001). After conducting a comprehensive assessment of data normality using skewness and kurtosis values, Kolmogorov–Smirnov tests, and P-P/Q-Q plots for all variables, the researcher used SPSS to conduct factorial multivariate analyses of variance (MANOVAs) to assess for significant mean differences in the constructs of interest (i.e. teacher burnout, teacher intent to quit) based on various demographic factors. The demographic factors assessed included personal characteristics (i.e. gender, age, race, and ethnicity), teacher characteristics (i.e. number of WLs taught, primary language status, years of experience, highest level of education completed, type of certification program, number of professional organizations, and number of additional certifications), and school characteristics (i.e. urbanicity of school, type of school, and region). Follow-up testing for between-participants effects and post hoc analyses were also conducted.
Reliability analysis for survey scales.
IV Results
The researcher performed three different factorial MANOVAs for personal characteristics, teacher characteristics, and school characteristics (as fixed/independent factors). Since Box’s test was significant for all of the analyses, the researcher used Pillai’s trace for its robustness (Ateş et al., 2019; Mertler & Vannatta, 2010; Olson, 1976) and its role as a conservative test statistic (Sheehan, 1995). In addition, the researcher also calculated F ratio statistics, probability values, values for partial η2, and observed statistical power (1 – β).
The researcher considered partial η2 values of 0.01 to indicate a small effect size, partial η2 values of 0.06 to indicate a moderate effect size, and partial η2 values of 0.14 or above to indicate a large effect size (Cohen, 1988; Evans et al., 2020). The researcher considered observed statistical power values of .8 or higher to be acceptable (Norton & Strube, 2001). Together this combination of reported values provided a greater understanding of the strength of the potential impact of the independent factors on the differences in the dependent factors. It also increased the confidence that a statistically significant difference would be found if one existed.
1 Analysis of personal characteristics
Mean levels of burnout and intent to quit were assessed for the subcategories for all four groups of personal characteristics (i.e. gender, age, race, and ethnicity). Tables A4 to A7 in the Appendix present the results.
Analysis of mean differences based on gender indicate that females report higher mean burnout (17.1) and intent to quit (15.3) than males. For age, mean burnout was highest in 51–55 year olds (17.3) followed closely by 26–30 year olds (17.1), while mean intent to quit was highest in 26–30 year olds (16.6). For race, Asian participants reported the highest mean burnout (19.0) and intent to quit (18.7). Those reporting as Black/African American and Multiracial or Biracial also reported higher levels of burnout (15.5 and 14.3, respectively) and intent to quit (15.3 and 15.4, respectively) than those who identified as White/Caucasian. For ethnicity, Latino participants reported the highest mean levels for both burnout (17.9) and intent to quit (17.4).
Factorial MANOVA demonstrated statistically significant main effects for all four factors tested (i.e. gender, age, race, and ethnicity) as demonstrated in Table 5. All of the factors had statistically significant, moderate effect sizes, with age demonstrating the largest effect size. Interactive effects were noted for gender/age as well as gender/ethnicity/age (see Table 6), demonstrating that the impact of these factors combined was more pronounced than when considering them in isolation. The intersection of gender and age demonstrated a large effect size, while gender, ethnicity, and age had a moderate effect size. This implies that although gender, ethnicity, and age had a significant combined influence, the impact of age combined with gender was stronger.
Main effects of personal characteristics.
Interactive effects of personal characteristics.
Between-participants effects for burnout included differences based on gender and gender/age (see Table 7). Between-participants effects for intent to quit included differences based on ethnicity and race/age (see Table 8). These results suggest that burnout levels varied based on gender and gender when considered with age, while intent to quit levels varied based on ethnicity as well as race when considered with age.
Between-participants effects of personal characteristics on burnout.
Between-participants effects of personal characteristics on intent to quit.
Post hoc tests for burnout found statistically significant differences between females and males (p < .001). In terms of race, significant differences were noted between the following racial identifications: Asian and bi/multiracial (p < .05), Asian and other/unknown (p < .05). Differences in ethnicity were found between Hispanic and Latino (p < .01), Hispanic and not of Hispanic, Latino, or Spanish origin (p < .001), and Hispanic and other/unknown (p < .001).
Post hoc tests for intent to quit found statistically significant differences between genders females and males (p < .001). For race, significant differences emerged between Asian and bi/multiracial (p < .05). Differences in ethnicity were noted between Hispanic and Latino (p < .01), Hispanic and not of Hispanic, Latino, or Spanish origin (p < .001), and Hispanic and other/unknown (p < .001).
2 Analysis of teacher characteristics
Mean levels of burnout and intent to quit were assessed for the subcategories for all seven groups of teacher characteristics (i.e. number of WLs taught, primary language status, years of experience, highest level of education completed, type of certification program, number of professional organizations, and number of additional certifications). Tables A8 to A14 in the Appendix present the results.
For number of WLs taught, those who taught four WLs reported the highest mean levels of both burnout (20.0) and intent to quit (16.5). However, those who only taught one WL had the second highest levels of mean burnout (15.6) and intent to quit (14.8). Mean burnout was similar for those who taught the same language as their primary language (15.7) and those who taught a language that was different from their primary language (15.9). However, mean intent to quit was higher in those who taught the same language as their primary language (15.0). In terms of experience and education, those with 16–20 years of experience reported the highest mean levels of burnout (18.0) and intent to quit (18.2). Mean levels of burnout and intent to quit were similar for all levels of education completed, with those who earned a bachelor’s degree emerging as slightly higher for both burnout (16.1) and intent to quit (14.9).
With respect to certification and professional organizations, participants who completed traditional and alternative certification programs reported similar levels of burnout and intent to quit, with those who completed an alternative program reporting slightly higher levels of burnout (15.8) and intent to quit (15.1). Those who were members of three or more professional organizations reported the highest mean levels of both burnout (19.8) and intent to quit (17.2), while those who had three or more additional certifications also reported the highest levels of burnout (17.0) and intent to quit (16.1). When comparing those who had no additional certifications and those who had at least one, since those represented the largest portion of the overall sample, those who had no additional certifications reported higher mean levels of burnout (16.6) and intent to quit (15.0). The same was true for the number of professional organizations reported, with those who were not a member of any professional organizations reporting higher levels of both burnout (16.6) and intent to quit (14.8).
Factorial MANOVA was conducted for teacher characteristics (i.e. number of WLs taught, primary language status, years of experience, highest level of education completed, type of certification program, number of professional organizations, and number of additional certifications). As shown in Table 9, statistically significant main effects surfaced for number of WLs taught, number of additional certifications, and number of professional organizations. Statistically significant main effects were not noted for primary language status, years of experience, highest level of education completed, or type of certification program.
Main effects of teacher characteristics.
Interactive effects emerged for the following combinations: number of WLs taught/primary language status, primary language status/years of experience, primary language status/number of additional certifications, highest level of education completed/years of experience, number of WLs taught/primary language status/highest level of education completed and primary language status/highest level of education completed/years of experience (see Table 10). Thus, although some of these factors did not demonstrate main effects, they did have an impact on burnout and intent to quit when combined with other factors. Specifically, although main effects were noted only for WLs taught, number of additional certifications, and number of professional organizations, other factors that did not demonstrate a main effect (i.e. primary language status, highest level of education completed, years of experience) gained significance through their interaction with other factors.
Interactive effects of teacher characteristics.
As shown in Table 11, between-participants effects for burnout demonstrated significant differences based on the following factors and combinations of factors: primary language status, number of additional certifications, number of WLs taught/primary language, primary language status/years of experience, highest level of education completed/years of experience, number of WLs taught/primary language status/highest level of education completed. Between-participants effects for intent to quit showed significant differences for the following factors and combinations of factors: number of WLs taught, years of experience, number of WLs taught/primary language status, primary language status/years of experience, highest level of education completed/years of experience, and number of WLs taught/primary language status/highest level of education completed (see Table 12). These findings suggest that burnout is influenced by primary language status, number of additional certifications, and multiple combinations of teacher characteristics, while intent to quit is impacted by number of WLs taught, years of experience, and its own combinations of factors from this category.
Between-participants effects of teacher characteristics on burnout.
Between-participants effects of teacher characteristics on intent to quit.
Post hoc tests for burnout illustrated significant differences between those who taught four languages and those who taught an unknown number. There were also significant differences between those who taught both their primary language and another language and those who taught only the same language as their primary language (p < .001), as well as those who only teach one language which is different from their primary language (p < .001). For years of experience, there were significant differences between those who had 1–5 years of experience and those who had six to 10 years of experience (p < . 001) or 16–20 years of experience (p < .01). Differences were also noticed for those with six to 10 years of experience and those with 11–15 (p < .001), 16–20, (p < .001), and 21+ (p < .01) years of experience. Differences emerged between those with no additional certifications and those with one (p < .001) or two (p < .01). Differences were also noted for those who were members of one professional organization and those who were a member of three (p < .001) or none (p < .001)
Post hoc tests for intent to quit found significant differences between those with six to 10 years of experience and those with one to five (p < .001), 11–15 (p < .001), or 16–20 (p < .001) years of experience. Differences also surfaced between those with two additional certifications and those with one (p < .05), three or more (p < .01) or none (p < .001).
3 Analysis of school characteristics
Mean levels of burnout and intent to quit were assessed for the subcategories for all three groups of school characteristics (i.e. urbanicity of school, type of school, and region). Tables A15 to A17 in the Appendix present the results.
In terms of urbanicity of school, those who taught in rural schools reported the highest mean levels of both burnout (18.1) and intent to quit (18.2). For school type, those who worked at parochial schools reported the highest mean levels of burnout (17.4) and intent to quit (19.7). When distinguishing between private and public schools, the levels were similar, with those teaching in public schools reporting slightly higher mean levels of burnout (15.6) and those in private schools reporting slightly higher mean levels of intent to quit (14.5). Concerning region, those who reported teaching in the Mid-Atlantic reported the highest levels of burnout (17.8) and intent to quit (16.6); those who reported teaching in the West reported the same level of intent to quit (16.6) but lower levels of burnout (16.7). Those teaching in the Northeast were not far behind. Those who taught in the northwest reported the lowest levels of both burnout (12.3) and intent to quit (12.2).
As shown in Table 13, factorial MANOVA of school characteristics (i.e. urbanicity of school, type of school, and region) found main effects for type of school (Pillai’s trace, F, p, partial η2, 1 – β). Statistically significant main effects were not found for urbanicity of school or region. As shown in Table 14, interactive effects surfaced for urbanicity/type and urbanicity/type of school/region. Thus, although only type of school surfaced as having a main effect, the other factors studied (i.e. urbanicity, region) may have an impact when considered among the other factors.
Main effects of school characteristics.
Interactive effects of school characteristics.
Between-participants effects for burnout found significant differences for the combined factors urbanicity/type (see Table 15). For intent to quit, significant differences emerged for type as well as the combined factors urbanicity/type (see Table 16). This suggests that burnout and intent to quit are both influenced by a combination of urbanicity and type of school.
Between-participants effects of school characteristics on burnout.
Between-participants effects of school characteristics on intent to quit.
Post hoc analysis for burnout found statistically significant differences between those in urban schools and those in rural schools (p < .05). Regional differences emerged between the northwest and the west (p < .01), the Midwest (p < .001), the southeast (p < .001), the mid-Atlantic (p < .05), and the northeast (p < .05).
Post hoc tests for intent to quit found differences between those in urban schools and those in rural schools (p < .01). Additionally, differences emerged between those in parochial schools and those in private (p < .001), public (p < .001) and other types of schools (p < .01). Regional differences surfaced between the northwest and the west (p = .001) and Midwest (p = .01). Regional differences were also noted between the west and the southwest (p < .05).
V Discussion
As mentioned in the literature review, factors such as age, gender, race, school type, and years of experience have been investigated for their role in teacher attrition and burnout. However, results have been mixed in terms of the extent to which each of these factors had a significant impact. Thus, the present study asked whether significant mean differences in levels of burnout and intent to quit in high school WL teachers exist based on various demographic factors (i.e. personal characteristics, teacher characteristics, and school characteristics). This study provided evidence of main, interactive, and between-participants effects for personal characteristics (i.e. gender, age, race, and ethnicity), teacher characteristics (i.e. number of WLs taught, primary language status, years of experience, highest level of education completed, type of certification program, number of professional organizations, and number of additional certifications), and school characteristics (i.e. urbanicity of school, type of school, and region). A discussion of each set of demographic factors is included below.
1 Personal characteristics
a Gender
This study found statistically significant main effects, as well as interactive and between-participants effects, based on gender. Prior studies conflicted with respect to whether burnout is more likely to impact females (Ghasemi, 2023; Jamaludin & You, 2019; Kara, 2019) or males (Mazidi et al., 2017). The current study found that females reported higher levels of burnout and intent to quit than males. However, upon considering the findings of the current study along with prior studies, it is suggested that school leaders take the complex situations faced by all genders into account when making decisions on policy and practice. Social expectations should be considered based on context. Furthermore, there is a gap in research on teachers who identify as gender fluid or another gender, and a lack of representation in the population of the current study. This illuminates the need for additional research to determine the experiences with burnout in these teachers as a unique population whose needs may otherwise be overlooked in policy and practice.
b Race and ethnicity
Significant main effects for race and ethnicity suggest a need for school leaders and researchers to also take these factors into account when addressing burnout and intent to quit. The present study highlights this concern with the fact that participants who identified as Asian, Black/African American, and Multiracial or Biracial reported higher mean levels of burnout and intent to quit than those who identified as White/Caucasian. Additionally, those who identified as Latino reported higher mean levels of burnout and intent to quit than other ethnicities reported. These findings support the argument that special consideration should be given to addressing turnover in minority teachers, especially given the positive impacts of non-White students having teachers who are also non-White (Nguyen et al., 2020). Given the connections between teacher attrition and identity (Schaefer, 2013), teacher educators and school leaders should also consider identity work as a potential avenue for addressing this issue. Activities such as linguistic portraits (Busch, 2012), identity wheels, and written reflections all may provide WL teachers with a stronger sense of self. More research may shine more light on the impact of these types of interventions.
c Age
Unlike Kara (2019), the present study found statistically significant main effects for age, with mean levels of burnout and intent to quit higher in both younger and older groups of teachers. These results suggest a need to investigate the unique struggles of younger and older WL teachers. Older teachers may struggle more adapting to digital learning environments, particularly when required to do so by adverse situations like the Covid-19 pandemic (Răducu & Stănculescu, 2022). Professional development providing support for digital literacy in language teaching may be beneficial. Meanwhile, younger WL teachers may struggle based on lack of experience, indicating a need to provide mentorship and professional development to facilitate growth and confidence.
2 Teacher characteristics
Statistically significant main effects emerged for number of WLs taught. Additionally, the highest mean levels of burnout and intent to quit surfaced among those participants who reported teaching four languages. Teachers of multiple languages frequently face complex challenges, and their experiences are often overlooked or oversimplified in research and practice (Ku, 2023). Thus, it is crucial for school leaders and researchers to further investigate the unique factors that may lead to burnout and intent to quit in these teachers. Surprisingly, the second highest mean levels of burnout and intent to quit were reported by those who only taught one WL. A larger overall sample size with more representative participants from each category may further clarify this finding in follow-up research.
a Certifications and professional organizations
Main effects also surfaced for number of additional certifications and number of professional organizations. Those who were members of three or more professional organization and had three or more additional certifications reported the highest levels of burnout and intent to quit. However, these participants represented a small portion of the sample. When considering those who represented a majority of the sample, those who were members of at least one professional organization and had obtained at least one additional certification reported lower levels of burnout and intent to quit than those who reported none. This resonates with research suggesting that professional development and membership in professional organizations and professional networks is a factor worthy of exploring with respect to teacher attrition (Kelly, 2004; Moser & Wei, 2021; P.B. Swanson, 2012) and teacher burnout (Hyseni Duraku et al., 2025). Thus, school leaders should provide the support needed for teachers to access these opportunities.
b Primary language status
Interactive and between-participants effects also emerged for primary language status. This finding aligns with research emphasizing the possible contribution of native speaker ideologies (belief systems that prioritize and idealize the linguistic abilities of native speakers of a language over those of non-native speakers) to teacher shortage in dual language immersion programs (Amanti, 2019) that native speakerism may be a contributing factor to teacher attrition and burnout among WL teachers. Thus, it is crucial for schools and school districts to confront fallacies of native speakerism and take steps to create more inclusive and collaborative educational spaces (Rudolph et al., 2015).
c Education and experience
In terms of amount of level of education, the present study diverged from previous scholarship indicating that a higher amount of education was associated with higher levels of burnout (Jamaludin & You, 2019). In terms of amount of experience, the present study found higher levels of burnout in teachers with more experience, while prior research disagreed on burnout being higher in those with less experience (Răducu & Stănculescu, 2022) or more experience (Ghasemi, 2023; Jamaludin & You, 2019; Kara, 2019). This study also found interactive and between-participants effects for years of experience, with post hoc tests noting statistically significant differences between participants with less years of experience and those with more years of experience. In addition, a majority of participants in this study only had one to five years of experience. This aligns with research calling attention to the fact that large numbers of teachers tend to leave the profession within the first five years (Michel, 2020), with a concerning number of teachers leaving within the first year (Ingersoll & Smith, 2003; Ingersoll et al., 2014; Zhang & Zeller, 2016). This demonstrates a need to consider novice teachers as a population at particular risk and in need of tailored interventions. Schools and school districts, as well as teacher preparation programs, should provide support for the transition from pre-service teacher to novice teacher. Additionally, schools and school districts should strive to provide access to resources that enable novice teachers to attend conferences to increase their knowledge, skills, and abilities, and to develop a support network. This may be of particular importance for singleton teachers, both those who are the only WL teacher on campus and those who are the only teacher of their language. Although these teachers may find support in teachers of other content areas on campus, support from others who teach the same content area would provide even greater benefits.
3 School characteristics
a Urbanicity
Main effects surfaced for urbanicity, contrasting with prior research (Nguyen et al., 2020; Răducu & Stănculescu, 2022). This suggests a need for more research in this area. Given that post hoc analysis for both burnout and intent to quit found statistically significant differences between those in urban schools and those in rural schools, and that higher mean levels of burnout and intent to quit were found among those who worked in rural schools, future research should focus on the possible factors which may contribute to that difference. For example, urban schools may have more access to digital resources than rural schools, which may be of particular concern in situations requiring an online learning environment (Răducu & Stănculescu, 2022).
b Type of school
Interactive and between-participants effects were noted for type of school. A possible explanation may be found in research like that of Kara (2019), who found a statistically significant negative relationship between school type and job satisfaction, while McConnell (2023) found a statistically significant negative relationship between job satisfaction and both burnout and intent to quit in high school WL teachers. Thus, more research is needed to investigate the factors related to school type when addressing these concerns.
c Region
Finally, the interactive and between-participants effects based on region serve as a reminder that sociopolitical contexts may impact the likelihood of burnout and intent to quit. Regional differences such as cost of living, teacher pay, and political constraints should all be considered when continuing research in this area. For example, one might consider the differences in these factors between those who teach in the Mid-Atlantic and those who teach in the Midwest, as those two regions reported the highest and lowest levels of burnout and intent to quit, respectively.
VI Conclusions
The findings of this study have important implications for school leaders, policymakers, and researchers. The relationships between the factors studied are complex and multifaceted. These findings align with research emphasizing the need to avoid taking a one-size-fits-all approach (Jiang et al., 2019). Each factor and combination of factors identified should be taken into consideration when conducting research or when planning interventions at all levels.
School leaders should consider the findings of this study when developing policies and programs to support WL teachers, taking into consideration significant demographics when creating policies and programs. In addition, school leaders should invest in programs that support novice teacher retention and provide access to professional development opportunities, professional organization membership, and pathways to certification. As suggested by prior research, these opportunities should be tailored to the unique strengths and needs of WL teachers (McConnell, 2024; McConnell & Swanson, 2024; P.B. Swanson & Mason, 2018). For example, school leaders may provide professional growth opportunities that highlight the creative strengths and professional profile of WL teachers (Hlas & Hlas, 2024; P.B. Swanson, 2008b). Concerns related to equity and inclusivity, such as native speakerism, should also be addressed, especially when considering the importance of supportive school cultures to the retention of WL teachers (Mason & Matas, 2015).
Meanwhile, researchers should continue to investigate the complex relationships between demographic factors, intent to quit, and burnout among WL teachers. The presence of interactive and between-participants effects suggests that certain combinations of factors may have a more pronounced impact on burnout and intent to quit. Further research is needed to explore the real-world situations that could explain these interactions. For example, the interaction between urbanicity and school type might indicate that teachers in rural public schools, who already face challenges in accessing necessary resources, may be particularly susceptible to burnout due to the frequent underfunding of public institutions. However, such hypotheses require further investigation to confirm their validity. Understanding these dynamics could inform the development of targeted interventions to support WL teachers and enhance the retention of this critical workforce.
There are several limitations worth noting. Given that 83 percent of participants were still teaching, those who quit were not as highly represented. Additionally, certain races and ethnicities were underrepresented. As such, future research should make a concerted effort to survey members of underrepresented populations. Furthermore, the possibility of selection bias should also be acknowledged. For example, participants who self-selected to take the survey may have had stronger feelings on the subject, particularly since those who were invited were already members of a research or social media group focused on the concerns of WL teachers. WL teachers with less interest or passion for the topic of burnout and attrition may not have been included, potentially leading to biased findings. As such, future research distributed through other avenues (e.g. a survey of all WL teachers in a specific district or region, regardless of participation in research or social media groups) could present additional insights on the perspectives of individuals potentially missed by the current study.
The current study attempted to address a gap in research on the experiences of high school WL teachers located in the U.S., but more research is needed to investigate the factors of interest in other content areas, educational levels, and contexts. In addition, diverse methodological approaches are needed to address the limitations presented by cross-sectional quantitative research. For example, qualitative research may provide greater depth of understanding as to how these factors emerge in the experiences of WL teachers, and longitudinal research may give insights on participants’ experiences with these factors over time. Ultimately, practitioners and researchers must continue seeking answers to these problems to ensure that learners have access to WL programs that are staffed by teachers who want to stay.
