Support, Mentorship and Well-Being in Canadian and Croatian Faculties of Education: Professor and Student Perspectives

The purpose of this quantitative study was to examine professors’ and Master of Education (MEd) students’ well-being, support, academic self-efficacy and mentorship in Canada and Croatia. Overall, 118 professors and 98 MEd students from three universities in Canada and three universities in Croatia completed the online surveys in English and Croatian, respectively. The frameworks of self-determination theory and relational cultural theory informed interpretation of our findings. Results suggest that for professors in both countries, personal support, professional support and academic self-efficacy predict professional well-being. Only personal support predicts personal well-being in Canadian professors, while personal support and academic self-efficacy predicts personal well-being in Croatia. Personal and professional support was also associated with positive mentorship practices in Canada. Students in both countries, who felt supported professionally and personally, reported greater professional and personal well-being respectively. Self-efficacy may make a difference for Croatian students but seemed to have little unique impact on Canadian students. Studying part-time in Canada was associated with higher personal and professional well-being but was associated with lower personal well-being in Croatia. Mentorship practices seemed to have little effect on well-being in either country. Overall, professors reported higher well-being and support than M.Ed. students. We conclude with recommendations that would be informative for university administrators, graduate programs, and services interested in the well-being of professors and graduate students.


Introduction
Higher education holds a critical role in the global economy (Larson et al., 2019;. Graduate studies are recognized as essential for knowledge production, knowledge mobilization, and ongoing professional training (Rose, 2013). Effective graduate education enhances the development of students' critical, empirical, and creative thinking skills; academic skills (e.g., research and teaching skills); transferable skills (e.g., personal, interpersonal, and career-related skills); and induction into the professions (Rose, 2013). Master's level studies have particularly been recognized as critical prerequisites for pursuing advanced academic and professional careers.

Professor and Graduate Student Well-Being
There is growing recognition that institutions of higher learning are stressful places and that academia is a stressful occupation. Universally, professors' report experiencing stressors related to simultaneously mounting expectations for research and scholarly productivity in context of increased instructional workloads, administrative duties and student expectations (Kinman & Wray, 2013;Kinman, 2019). Graduate students, similarly, report stressors related to demanding and uncertain workloads, work-life imbalances, sense of isolation, financial concerns including restricted access to employment, funding, and scholarship opportunities while on campus, and limited employment opportunities post-graduation (Kim, 2016). Experiences of intense worry, low mood, disturbed sleep, and diminished self-care practices are common (Evans, Bira, Beltran Gastelum, Todd Weiss, & Vanderford, 2018). Collectively, these stressors contribute to increased experiences of stress, emotional exhaustion, depersonalization, and overall sense of poor well-being among professors and graduate students (Jovanovic, 2019;Gulati, Lele Mookerjee, & Rajput, 2018). A 2005 study showed that 42% of university students reported feeling elevated distress, compared with 17% of similarly aged non-students (Adlaf, Demers, & Gliksman, 2005). However, this data was not broken down by undergraduate or graduate program.
Graduate students typically benefit from their supervisors' research and financial support, career guidance, and psycho-socio-emotional support (Woloshyn, Savage, Ratkovic, Hands & Martinovic, 2019). Supervisory and mentoring relationships are critical to the graduate student experience, success and well-being (Rose, 2013). Faculty, in turn, appear to be increasingly aware of the importance of creating inclusive and positive learning environments that promote and sustain graduate student well-being (Woloshyn et al., 2019). While the provision of academic accommodations and on-campus services are recognized as important to these efforts, they are often perceived as secondary to the need to develop authentic and caring relationships, and mentorship in graduate education.
Graduate students who experience a lack of well-being are more likely to report academic difficulties, terminate their studies, or extend their completion rates relative to their peers who report well-being (Neves & Hillman, 2018). This may be particularly important in Croatia where postsecondary graduation rates are low (European Commission, 2019). While professors may not necessarily leave their academic positions because of career-related stressors, experiencing poor well-being may negatively impact their productivity, motivation for work, sense of optimism, and overall levels of engagement (Gulati et al., 2018).
It is realistic to assume that a proportion of graduate students and professors will, at some point, experience a compromised sense of well-being (Evans et al., 2018;Henning et al., 2018). Many higher education institutions across the globe have initiated programs and services to support students' well-being while on campus. These services, however, tend to be tailored for, and almost exclusively used by, undergraduate students, leaving the needs of graduate students and professors potentially unmet (Jaworska, De Somma, Fonseka, Heck, & MacQueen, 2016;Thorley, 2017). The primary purpose of this research study was to explore the relationships and intersections between perceptions of well-being, relational support, and mentorship among professors and MEd students in Canadian (Ontario) and Croatian Faculties of Education.

Theoretical Frameworks
Self-determination theory (Deci & Ryan, 1985;Ryan & Deci, 2000) and relational cultural theory (Miller, 1976;Miller & Stiver, 1997) were used to guide this research study. Both theories help to contextualize mechanisms responsible for graduate student and professor well-being in higher education including the relational activity of mentoring.

Self-Determination Theory (SDT)
According to SDT, individuals are inherently motivated towards self-actualization, with their sense of integrity, growth, and well-being determined through the attainment of three basic and innate psychological needs including autonomy, competence, and connectedness to others (Deci & Ryan, 1985;Ryan & Deci, 2000) Attainment of these three needs has been positively associated with subjective measures of well-being as well as inversely related to mental health disorders such as depression and anxiety (Ng et al., 2012). Functioning and well-being are enhanced to the extent that individuals demonstrate control, choice and autonomy over their career and life choices; experience a sense of achievement, success, challenge and self-efficacy; and engage in caring, supportive and empathetic interpersonal relationships (Ryan & Deci, 2000). In higher education, engagement in meaningful work and interactions with supportive and caring personal relationships promotes and/or moderates academic success, career satisfaction, and sense of well-being (Larson et al., 2019;Seipel & Lawson, 2018).
It is well-established that the contexts and/or environments in which individuals live and work affect their realization of the three basic psychological needs (Deci & Ryan, 1985). In higher education, context or environment is typically reflected through relational experiences within the academic department. These experiences may include support from critical others, such as department chairs, deans, and senior administrators. Environment may also include providing constructive feedback, recognizing contributions and achievements, granting tenure and promotion, and/or renewing contract positions (Larson et al., 2019;Seipel & Larson, 2018).

Relational Cultural Theory (RCT)
Within the RCT framework, growth-fostering relationships include a mutual sense of worth, clarity, action, knowledge and productivity, zest, and desire for ongoing connection lead to increases in well-being (Miller, 1986;Miller & Stiver, 1997). Mentoring has the potential to be one such growth-fostering relational process, providing individuals with opportunities to come to know themselves better, as well as experience a positive sense of self-worth, self-esteem, and/or competency through engagement in genuine discussions in which they feel heard and, in turn, hear others (Jordan, 2010). These critical conversations may also provide motivation to seek out additional connections. In these ways, growth-fostering relationships are characterized by mutual empathy, relational authenticity, and mutual empowerment that serve to deepen learning experiences for students and regenerate professors (Shwartz & Holloway, 2012).
Meaningful, high-quality relationships are associated with well-being and functioning and may even exist in the presence of power and/or role differentials such as in the case of professors and graduate students (Jordan, 2010). In one survey, 511 full-time professors working in public and private institutions ranked the quality of their interpersonal relationships as the primary determinant of their overall sense of well-being (English & Avakian, 2012). In the same way, graduate students reported a positive association between caring formal as well as informal mentoring relationships and learning and academic success. Students also reported enhanced sense of safety, agency, empowerment, as well as enhanced self-esteem, self-worth and connectivity as a function of these relationships (Lucey & White, 2017;Schwartz & Holloway, 2012, 2014.

Purpose
There is relatively little research available exploring the relational and well-being experiences of MEd students and their professors. Given the increasing numbers of student enrollments in MEd programs and the extended impacts on the learning, well-being and success experiences of youth in their care, it seems imperative to study MEd students' and their professors' experiences of well-being within academia. In this study, we explored the relationships between subjective reports of well-being, personal and professional support, and mentoring among professors and MEd students in Canada and Croatia. The following research questions guided this correlational comparative study: 1. What is the relationship between perceptions of well-being, support, academic self-efficacy and mentorship practices among professors and MEd students? Do these relationships differ across Canada and Croatia?
2. What variables are predictive of subjective well-being? Do these predictors differ across professors and MEd students in Canada and Croatia?
3. What variables are predictive of engagement with positive mentorship practices? Do these predictors differ across professors and MEd students Canada and Croatia?
4. Do professors and MEd students differ in their sense of well-being, support, academic self-efficacy and mentorship? Do these patterns differ across Canada and Croatia?

Method
The findings reported here represent the first wave of data emerging from a comparative explanatory mixed-methods investigation of perceptions of well-being and graduate mentorship in Faculties of Education in Canada (Ontario) and Croatia.

Recruitment
MEd students from six participating universities (three in Canada, and three in Croatia) were invited to participate in this study. A multi-layered, email recruitment procedure was used in which letters of invitation and informed consent forms, including a hyperlink to the online survey, were sent out through the MEd student email listservs at participating universities. Similar invitations were extended to professors in both countries using their publicly available email addresses. In Canada, recruitment materials were sent out in English, while in Croatia the materials were sent out in English and Croatian.

Online Survey
The online survey for the MEd students consisted of 78 questions, including 11 demographic questions, 3 standardized instruments and 2 open-ended questions. The online survey for professors consisted of 77 questions, including 10 demographic questions, 3 standardized instruments and 2 open-ended questions. The surveys were translated from English to Croatian by the authors, with demographic questions adjusted to reflect national contexts.

Demographic Questions
Professors were asked to identify their gender, age, whether their position was permanent, academic rank, the number of students they have mentored and whether they supervised MEd theses. We did not ask Croatian professors to distinguish between their roles as thesis supervisors and committee members as all professors are required to serve as research advisors. The MEd students were asked to identify their age, gender, study status (i.e., part-time, full-time) and university affiliation.
2.2.2 Skovholt Practitioner Professional Resiliency and Self-Care Inventory (Skovolt, 2016) This inventory is a 38-item instrument designed to encourage reflection-on-practice. It provides a measure of stress and resilience for practitioners and students involved in the caring professions, including the field of education. The inventory is based on a 4-factor structure including professional vitality, personal vitality, professional stress and personal stress. For the purposes of this study, four new subscales were extracted from the original instrument in order to examine perceptions of well-being and support.
Items that addressed support were isolated, with four items identified as professional support (e.g. "My work environment is like a greenhouse -where everything grows-because the conditions are such that I feel supported in my professional work") and used to develop the professional support subscale (Cronbach's alpha of .68). In the current data this subscale was not normally distributed, with a Shapiro-Wilk score of .98 (p = .011). Examination of histogram plots indicated that this variable had a small negative skew. Five items were identified as personal support, (e.g. "I have a supportive community where I feel connected") with a Cronbach's alpha of .74. In the current data this subscale was also not normally distributed, with a Shapiro-Wilk score of .96 (p < .001). Examination of the histogram plots indicated that this variable had a negative skew. Both of these scales were used primarily as predictor variables, making them robust to violations of normal distribution of scores (Keith, 2019). When the variables were used as dependent variables in regression models, assumptions of those models, including normal distribution of residuals, were met.
The remaining inventory items were used as measures of well-being. The 13 items in the professional vitality (e.g. "I find my work as a professor/student to be meaningful") and professional stress (e.g. "I have found a way to have high standards for my work yet avoid unreachable perfectionism") sections were collapsed to create the professional well-being subscale which had a Cronbach's alpha of .75. In the current data this subscale was normally distributed, with a Shapiro-Wilk score of .98 (p = .134). Similarly, the remaining 16 items in the personal vitality (e.g. "I have plenty of humor and laughter in my life") and personal stress (e.g. "I am able to cope with significant losses in my life") sections were collapsed into the personal well-being subscale, with a Cronbach's alpha of .80. In the current data, this subscale was normally distributed, with a Shapiro-Wilk score of .98 (p = .085). Examination of histogram plots confirmed approximately normal distributions with slight non-significant negative skews. These measures of well-being were used as dependent variables in the data analyses. Greater consideration was given to the residual distribution than the individual variable distributions as hierarchical linear regression comprised the primary analyses. Residual distribution information is presented alongside those analyses in the results section.

Academic Self-Efficacy
This measure consisted of one item ("I am capable of succeeding in my academic career") scored on a 5-point Likert scale. Created by the researchers, it measures how efficacious MEd students and professors feel about their academic and career achievement. (Gullan et al., 2016) This instrument consists of 24 items designed to assess university students' perceptions of the quality of their mentoring relationships with professors, coaches and/or administrative staff across four domains; challenge, authenticity, commitment, and community. Participants were instructed to reflect on either formal or informal experiences when completing the measure. The instrument was modified for use with professors by having them reflect on the experience of being a mentor, as opposed to being a mentee. For example, the item, "My relationship with my mentor is personal" became "My relationship with my mentee is personal." Cronbach's alpha was .94 for the student instrument and .84 for the professor instrument. For the purposes of this study, total scores were used as a measure of positive mentorship practices.

Data Analyses
Descriptive statistics and Pearson correlations were completed to examine the relationships between professor and MEd perceptions of well-being, support, academic self-efficacy, and mentorship practices. T-tests and MANOVAs were used to determine if differences exist between countries and participant groups. Hierarchical regression analyses were completed to examine factors that predicted professional and personal well-being.
Separate analyses for the Canadian and Croatian participants were carried out in order to compare patterns across countries and participant groups. Normal distribution of residuals based on P-plots was found for all reported analyses. Non-significant Glejser tests for heteroskedasticity and Variance Inflation Factor (VIF) values of fewer than 5.0, suggested that the assumptions of multi-collinearity and normality were met. Relationships between professors' involvement in graduate programs and their perceptions of well-being, support and mentorship were also explored. Multiple linear regressions were completed for each of mentorship, personal well-being, professional well-being, personal support, and professional support.
MANOVAs were completed in order to examine potential differences between professors' and students' perceptions of well-being, support, and mentorship. As no differences emerged across formal and informal mentoring experiences, mentoring was analyzed as a collective activity. Participants' personal well-being, professional well-being, personal support, professional support, mentorship practices, and academic self-efficacy scores were used as dependent variables and participant status (student, professor) was used as an independent variable.

Results
Correlation and regression results for professors and MEd students are reported first, with comparisons between these two groups presented last.

Participants
Participants included 118 professors and 98 MEd students in targeted Faculties of Education in Canada (Ontario) and Croatia. Table 1 outlines demographic information for these participants including gender, age, and status. Canadian participants were significantly older than their Croatian counterparts [professors: χ2(4) = 19.55, N = 116, p = .001; MEd students: χ2(5) = 24.92, N = 96, p < .001]. Significantly more Canadian professors with higher ranks completed the study than did Croatian professors, χ2(4) = 22.22, N = 102, p = .001. Significantly more Canadian male MEd students participated in this study than did Croatian male MEd students: χ2(2) = 11.72, N = 98, p = .003. There were no other significant differences across the participant groups (see Table 1).

Professors
Simple correlations were completed to examine relationships between professors' perceptions of support, well-being, academic self-efficacy, and mentorship. All variables were correlated, with the exception of professors' perceived academic self-efficacy, which was not related to their perceptions of professional support or mentorship practices (Table 2). Differences between the two countries were examined (see Table 3). Overall, Canadian professors reported higher personal well-being, t(102) = 2.05, p = .043 and academic self-efficacy, t(102) = 2.62, p = .010, than Croatian professors, who reported higher professional support relative to their Canadian peers, t(94.48) = -2.24, p = .028. These t-test reported a minimum observed power of .65. A hierarchical regression analysis was completed to identify factors that predicted professional well-being. In step one of the regression, gender and permanency were entered. In step two, academic self-efficacy, perceived personal support and perceived professional support were entered. Gender was predictive at the first step with Canadian men reporting higher professional well-being than Canadian women (Table 4). Each step two variable also accounted for a significant amount of variability with respect to Canadian professors' perceptions of professional well-being. Similar patterns emerged for Croatian professors, although there was no significant effect for gender at step one. Overall, personal and professional support and perceptions of academic self-efficacy predicted professors' perceptions of professional well-being across both countries, with gender only predicting professional well-being for Canadian participants. Similar to the identifying factors that predicted professional well-being, a hierarchical regression analysis identified factors that predicted personal well-being. In step one, gender and permanency were entered. In step two, academic self-efficacy, perceived personal support and perceived professional support were entered. Personal support significantly predicted personal well-being for professors in both countries (Table 4). In the Croatian sample, there was a significant effect of academic self-efficacy on personal well-being.
The influences of support and well-being factors for engagement in positive mentorship practices were also explored. In step one of the regression, gender and permanency were entered. In step two, academic self-efficacy, perceived personal support and perceived professional support were entered. A third step was also included in which perceived personal support and perceived professional well-being were entered. While overall change in R 2 between steps one and two was significant for Canadian participants, perceived professional and personal support only marginally predicted engagement in positive mentorship practices (Table 5). In the Croatian sample, a similar pattern emerged with a significant change in R 2 at step two. Neither perceived support nor well-being, predicted mentorship practices. Overall, personal and professional support collectively predicted mentorship practices, but their predictive ability for this variable was intertwined so that they explained more variance when combined than when each variable was uniquely examined. Finally, relationships between professors' involvement in graduate programs and their perceptions of well-being, support and mentorship were explored. Multiple linear regressions were completed for each of mentorship, personal well-being, professional well-being, personal support, and professional support. The approximate number of students mentored and supervisor status (Canadian data only) were simultaneously entered for this analysis. For Canadian professors there was a significant effect of acting as a thesis supervisor on personal and professional well-being (Table 6). For the Croatian sample, the number of students mentored positively predicted engagement in positive mentorship practices, professional well-being, and perceived personal well-being, but not personal support or professional support. Collectively, these findings indicated that for professors, there is a relationship between well-being, personal and professional supports, and mentorship. Personal and professional support was positively associated with the professional and personal well-being of professors in both countries. Personal and professional support was also associated with positive mentorship practices in Canada.
In Canada, male professors reported higher professional support than female professors and academic self-efficacy predicted higher professional well-being for all participants. This was not the case in Croatia. In Canada, acting as a thesis supervisor was predictive of higher personal and professional well-being, while in Croatia, the number of students previously mentored was predictive of engagement in positive mentorship practices, professional well-being and personal well-being.
A hierarchical regression analysis was completed to examine factors that predicted professional well-being. In step one of the regression gender and study status were entered. In step two, academic self-efficacy, perceived personal support and perceived professional support were entered. Finally, in step three, positive mentorship practice was entered. For Canadian MEd students there was a significant effect of study status, with students studying part-time reporting higher professional well-being (Table 8). At step two, personal and professional support predicted professional well-being for Canadian students. There was no significant effect of mentorship, gender, internationality, type of program, or academic self-efficacy. For Croatian MEd students, study status did not have a significant effect on professional well-being, but there was a significant effect of academic self-efficacy and a significant effect of professional support. Personal support, mentorship and gender had no significant effect on well-being.  Vol. 5, No. 1;April, 2021 Another hierarchical regression analysis was completed to examine factors that predicted personal well-being. In step one of the regression gender and study status were entered. In step two, academic self-efficacy, perceived personal support and perceived professional support were entered and in step three, positive mentorship practice was entered. Similar patterns found for professional well-being were also found for personal well-being (Table 8).
For Canadian MEd students, there was a significant effect for study status such that part-time students reported higher personal well-being. There was also a significant effect of personal support, but no effect of professional support, academic self-efficacy, gender, study status, or mentorship practices. For Croatian MEd students, there was a significant effect of study status on personal well-being, with part-time students reporting lower personal wellbeing. There was also a significant effect of personal support, but no effect of professional support, academic self-efficacy, gender, or mentorship.
Collectively these results suggest that for graduate students in Canada and Croatia, the relationship between well-being, support, self-efficacy and mentorship is complex. Students in both countries, who felt supported professionally and personally, reported greater professional and personal well-being respectively. Self-efficacy may make a difference for Croatian students but seemed to have little unique impact on Canadian students. Studying part-time in Canada was associated with higher personal and professional well-being but was associated with lower personal well-being in Croatia. Mentorship practices seemed to have little effect on well-being in either country.

Professors and MEd Students
The final analyses addressed whether there were differences between perceptions of well-being, support, and mentorship among professors and MEd students as a function of country (Table 3). For Canadians, the overall MANOVA was significant, F(6, 96) = 28.63, p < .001. In particular, there was a significant difference between MEd students and professors in personal well-being

Discussion
The findings of this study support and extend previous literature exploring relationships between personal and professional support, autonomy, relatedness and subjective experiences of well-being among professors and MEd students. Canadian professors reported greater experiences of well-being and self-efficacy than did their Croatian counterparts, with this being especially true for male participants. Professors reported higher subjective experiences of well-being than did MEd students. Professional and personal support positively impacted various aspects of subjective experiences of well-being, as did perceptions of positive mentoring relationships.

Professors: Well-Being, Self-Efficacy and Gender
Canadian professors reported higher experiences of personal well-being and academic self-efficacy relative to their Croatian colleagues. These findings, in part, may reflect socio-cultural and institutional differences between the two countries. Croatia is one of the European countries with the lowest index of economic development and national prosperity (Legatum Institute, 2015;United Nations Development Programme, 2014). Economic underdevelopment, income inequality, unemployment, underemployment, social welfare, and political instability, are known to negatively impact individuals' experiences of well-being (Jorm & Ryan, 2014;Ngamaba, Panagioti, & Armitage, 2017).
Canadian professors who participated in this study were significantly older and held higher rank (i.e., associate professor, full professor) than their Croatian colleagues. Attainment of rank is an attestation of scholarly proficiencies and capacities and is likely reflected in these professors' reported sense of self-efficacy (Lumpkin, 2014). Careers in academia can be conceptualized across a continuum of stages extending from early-to mid-and late-career.
In general, early-career professors tend to experience insecurity, isolation, and stress (Lumpkin, 2014). Early-career faculty report concerns related to achieving employment permanency, with these concerns exacerbated by high performance expectations and/or unclear promotion processes and criteria (Ortlieb & Weiss, 2018). These professors also report being intensively focused on developing research competencies and gaining scholarly recognition (Albert, Davia, & Legazpe, 2018). Additionally, early-career professors are often developing instructional curriculum, navigating teaching demands, and negotiating service expectations. Collectively, these requirements may result in competing senses of urgency, demanding workloads, intensive work hours, and overall reduced perceptions of well-being for early-career professors (Lumpkin, 2014).
Stressors associated with the profession are often exacerbated for female professors, who in addition to institutional performance-related pressures, often face additional considerations related to familial responsibilities including raising children, securing partner employment and/or elder-care (Fritsch, 2015). Women are twice as likely as men to report experiencing stress within academia (Hurtado, Alvarez, Guillermo-Wann, Cuellar, & Arellano, 2012). In this study, Canadian female professors also reported experiences of lower professional well-being relative to their male colleagues.
In North America and Europe, universities are recognized as gendered institutions whose organizational, instructional, and funding structures; policies; practices; and cultural presumptions reflect masculine principles, generate gendered divisions of labor, and bolster men's careers while impeding women's professional advancement (Fumasoli, Goastellec, & Kehm, 2015;Leathwood, 2013). While men's roles in academia are well defined, promoting their sense of belonging, security and comfort in the academy, competing demands associated with work and family may increase women's sense of isolation and diminish their abilities to establish and nurture positive collegial relationships (Penney et al., 2015).
Subjective reports of greater well-being by Canadian male professors relative to their female counterparts, may also be reflective of other systemic inequities within academia. Despite the ongoing implementation of legislation, policies, and practices intended to promote equity, and despite considered advancement in the profession, women remain disproportionately underrepresented and underpaid among the professoriate ranks (CAUT, 2018). This is especially true in Faculties of Education where women hold over two-thirds of the available teaching positions. It is well documented that underemployment and precarious employment undermine individuals' sense of subjective well-being (Brand, 2015;Redekopp & Huston, 2019).
Croatian professors however did not differ in their reported experiences of subjective well-being as a function of gender. The Croatian government has created recent legislative frameworks which specifically task Faculties of Education to minimize and eliminate gender inequality in Croatian society (Gender Equality Office of the Government of the Republic of Croatia, 2014). All teachers and future teachers are mandated to be trained in gender equality principles, with the intention that gender stereotypes be eliminated from Croatian society and that gender balance be achieved in the occupation over the next several years. It is thus possible that female professors in Croatia perceive themselves as active agents in promoting gender equality in society, and therefore, do not rank their subjective well-being as any lower than their male colleagues. Consistent with this interpretation, Croatian professors in this study reported significantly greater experiences of professional support compared to Canadian counterparts. Subjective feelings of professional support, in part, may reflect the efforts and outcomes to satisfy mandated legislations and responsibilities.

MEd Students: Well-Being, Study Status and Gender
Professors reported significantly greater experiences of subjective well-being than did MEd students across both countries. This finding is not surprising in context of the demands, uncertainties and challenges associated with graduate studies. Engagement in graduate studies is an intensive, difficult and stressful experience for many students. Graduate students commonly report challenges associated with intensive coursework and demanding timelines. Many graduate students, especially women, face additional challenges related to managing multiple and often competing personal and professional roles that may include student, working professional, caregiver, and partner. Croatian MEd students reported a higher sense of personal well-being relative to their Canadian counterparts. Croatian students were significantly younger than their Canadian peers and thus, potentially more likely to be the recipients of familial support from parents or significant others. At the same time, these younger adults may be less likely to hold familial responsibilities for children, partners or parents.
Post-graduation employment opportunities may have impacted MEd students' sense of well-being. In this study, professional support impacted Croatian students' subjective experiences of well-being. It is likely that receiving support from their instructors and relevant others at the university increased these students' sense of readiness for entry into the profession and optimism for the future. Unlike their Canadian peers, Croatian students will graduate into an era of teaching shortages with demands for skilled educators anticipated to continue over the foreseeable future (European Commission, 2019). Canadian students, on the other hand, enter a workforce where they are likely to experience employment uncertainty and insecurity due to increasingly competitive job markets, limited teaching and postsecondary positions, and increased contract, part-time and precarious positions (Ontario College of Teachers, 2019).
For Canadian students, part-time studies, personal and professional support predicted professional well-being. Engaging in part-time studies is often associated with holding part-time employment which may lessen immediate financial concerns and enhance opportunities for subsequent employment (Woods & Frogge, 2017). On the other hand, the need to balance school, work, and personal needs may be exacerbated for part-time students (Castelló, Pardo, Sala-Bubaré, & Suñe-Soler, 2017). School-work-life balance has been identified as a major challenge for completing graduate studies, with many of these students also reporting a lack of support (Grabowski, 2016;McAlpine et al., 2012). In the current study, we examined part-time status alongside support, which may suggest that engaging in part-time studies while being the recipient of personal and professional support may increase sense of professional well-being for Canadian students.

Professional and Personal Support and Well-Being
Consistent with SDT (Deci & Ryan, 1985;Ryan & Deci, 2000), perceived levels of professional and personal support were positively associated with subjective experiences of well-being for professors and MEd students in both countries, although not in precisely the same ways. Personal and professional support predicted well-being for professors. Professional support predicted well-being for MEd students in Croatia, while personal support predicted well-being for MEd students in Canada. These findings add to existing literature demonstrating that SDT needs, such as personal and professional supports, directly and indirectly predict well-being in tenured, tenure-track and non-tenured professors (Larson, Shelley, & Gahn, 2015;Larson et al., 2019;Seipel & Larson, 2018). Similarly, instructors and peers are known to play a crucial and unique role in creating academic environments that are perceived to be supportive of students' well-being (Schenkenfelder, Frickey & Larson, 2020).

Mentoring
Professors in both countries who felt both personally and professionally supported were more likely to engage in positive mentoring practices. Presumably, professors who feel supported are more likely to engage in growth-fostering relational processes including mentoring. As professors engage in positive mentoring and other relational processes, they are likely to experience a sense of revitalization, which, in turn, works to sustain and motivate their efforts to support others (Schwartz & Holloway, 2012). In this study, Canadian professors who were thesis supervisors, and thus likely engaged in intensive one-on-one mentoring relationships, reported higher personal and professional well-being. Similarly, the number of students mentored positively predicted Croatian professors' engagement in mentoring practices as well as their personal and professional well-being.
For MEd students in both countries, mentoring did not predict professional or personal well-being. A positive correlation was found, however, between professional support and mentoring. Within academia, professional support includes a variety of institutional practices and structures as well as interactions with significant others including peers and instructors. In this way, mentoring may represent one form of professional support available to students. Schmidt and Hansson (2018) found that the relationships that doctoral students formed with peers, supervisors and the scholarly community at large collectively contributed to their sense of well-being. Presumably these experiences would extend to MEd students. General experiences of feeling supported may positively impact MEd students' sense of well-being beyond the relationship with any single faculty member mentor, consistent with findings that graduate students require multiple support structures in order to advance their learning goals and support their well-being (Lorenzetti et al., 2019). The role of professional support is likely to become increasingly important as the nature of master's programs continue to evolve and change to include more course-based, blended and online delivery models. In these instances, the role of the academic mentor, such as the faculty supervisor, may transform as fewer students interact one-on-one and face-to-face with professors to complete research intense projects and course-based activities.

Limitations
In exploring the relationships between professor and MEd student subjective reports of well-being, personal and professional support, and mentoring experiences, the findings of this study provide insights about how well-being can be supported within higher education. There are, however, several limitations associated with this study that need to be acknowledged. The sample size was relatively small, potentially limiting power, and there was an imbalance in the number of professor and student participants in each country. More Canadian professors participated in this study than did Croatian professors, while more Croatian MEd students participated than did Canadian MEd students. Professors were also recruited much more widely than were MEd students, allowing for more varied reflections of institutional experiences within this group. Additionally, our examinations of faculty well-being and support were based on an adjusted version of the Skovholt Practioner Professional Resiliency and Self-Care Inventory, but we were not able to examine the full psychometric properties of our changes with the current data set, future research should examine the relationship between support and well-being with more robust measures. Finally, differences in the educational systems in each country were not widely explored, nor were individual differences of experience based on race, ethnicity, sociocultural status, or years as a student/professor. For instance, Croatia hosts more refuges than Canada (Marsanic et al., 2017) making it more likely that some students enrolled in graduate studies hold complex prior experiences that impact their sense of well-being beyond their postsecondary experiences. In addition, students' experiences with off-campus supports and social services were not explored here leaving questions about the intersectionality between their on-campus and off-campus experiences.

Conclusion and Implications
Overall, this study illustrates the importance of feeling relationally supported within the academy. In both countries, professors and MEd students who felt personally and professionally supported reported greater subjective well-being. This was true despite the interaction with other variables including study status, academic self-efficacy, gender and permanency which had varying effects across the participant groups. Importantly, however, MEd students did report lower levels of subjective well-being than their professors.
Lower levels of well-being reported by MEd students may be rooted in graduate students' need for multiple support structures within the university itself (Lorenzetti et al., 2019). Some of the supports graduate students require that have been identified in the literature include academic, psychological, social and career supports. Support for the acquisition of procedural and disciplinary knowledge, academic thinking and writing skills, time management skills, self-confidence and motivation, coping with stress, academic socialization, interpersonal skills, the establishment of personal and professional networks, self-care and the development of career competencies have all been shown to be important to the well-being of graduate students (Lorenzetti et al., 2019). Further explorations of the provision of these supports by universities and their impacts on graduate students may help reduce the gap between the reported experiences of subjective well-being among professors and MEd students.
As enrolment numbers in master's level programs continue to increase (Wall et al., 2018), it is incumbent on university administrators to consider the unique nature of graduate-level studies and the relational needs of both professors and students (Deci & Ryan, 1985;Ryan & Deci, 2000). Capping class sizes at 15 or fewer students, providing peer-mentoring programs to MEd students and professors, delivering robust orientation sessions that outline on-campus and community-based academic, psychological, social and career supports, and increasing the number of university sponsored communities of learning and social activities are strategies that increase awareness and access to personal and professional support in graduate programs (Benshoff, Cashwell, & Rowell, 2015;Lorenzetti et al., 2019). Additionally, developing university-based counseling services designed specifically for graduate students' unique needs and training academic advisors to be able to connect graduate students with other support services on campus or in the community that are appropriate for them is important to the well-being of graduate students (Benshoff, Cashwell, & Rowell, 2015;Lorenzetti et al., 2019).
Continued efforts to achieve gender equity among the professoriate is another crucial initiative that must be maintained and strengthened to increase the well-being of both faculty members and graduate students. This involves examining, and eliminating, the gendered landscapes in our curriculum, pedagogical practices, career prospects, university administration and the ways we celebrate researchers and scholarship in our fields (Alrashidi, 2017;Angervall, 2018).
Finally, increased formal recognition for engaging in relational practices, such as mentoring, by including them in the promotion and tenure process and in workload considerations for professors would help increase the well-being of the professorate. By including evidence of engaging in relational practices with students as part of the tenure and promotion process and including the mentoring and supervision of graduate students as a formal part of professors' workload these relational practices would be considered important and would increase in universities. In these ways, the university administration, professors and graduate students can each create collective, compatible goals that work to foster individuals' ongoing learning successes and well-being.