How do University Experiences Contribute to Students’ Psychological Wellbeing?

Wellbeing has important implications for students’ success during and beyond university. As such, educators need clear empirical evidence of the aspects of university life that contribute to students’ wellbeing. We use a mixed-methods approach to ask whether and how students’ diverse university experiences contribute to their self-rated wellbeing. In an online survey, 696 students provided accounts of positive and negative experiences at university and self-rated their wellbeing using the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS).   Most of the sample reported high wellbeing, though there was diversity in their reported experiences. Regression analysis revealed that students’ experiences at university significantly contributed to the variation in their wellbeing scores. Social and academic experiences were the strongest positive contributors  to student wellbeing; whereas gender, social experiences and emotional experiences were the strongest negative contributors. Our findings offer guidance for strategies within academic and non-academic university contexts that can effectively and efficiently promote student wellbeing.


Introduction
Student wellbeing is an issue of concern for universities in Australia (QILT Social Research Centre, 2018;TEQSA, 2015). Heightened psychological wellbeing has been associated with persistence, social responsibility, attention and cognition (Deci & Ryan, 2000;Seligman, 2011;Taylor et al., 2014); all of which are important for helping students reach their full potential, including achieving academic success and personal goals, finding social connections and living a meaningful life (Baik et al., 2019). In contrast, prolonged experiences of low wellbeing, emotional distress and mental distress are associated with course attrition and difficulties with learning (QILT Social Research Centre, 2018;Stallman, 2010). Course attrition can have negative impacts on academic pathways and employment opportunities (Orygen Youth Research Centre, 2017). As such, fostering students' wellbeing and mitigating experiences of distress can have important implications for their success during and beyond their university careers.
There are currently strong calls for universities to actively mitigate distress and foster wellbeing (e.g., Orygen Youth Research Centre, 2017;TEQSA, 2015). These calls come with the premise that universities can have a meaningful impact upon students' distress and wellbeing. In support of this premise, educators argue that the university curriculum is the metaphorical glue that holds the student experience together and therefore the most efficient context for fostering student wellbeing (e.g., Baik et al., 2017;McInnis, 2001). They embed new initiatives to support wellbeing within curricula with impressive outcomes (e.g., Johnson et al., 2019). For example, the work by Craig Hassed and colleagues has demonstrated the effectiveness of embedding programs and mindfulness training within tertiary programs (e.g., Hassed et al., 2009;Lo, Francis-Cracknell & Hassed, 2017). Further to the curriculum approach, demographic information such as gender, work and family care commitments (e.g., Larcombe et al., 2016), and psychological factors such as self-efficacy and coping mechanisms (e.g., Deasy et al., 2014;He et Wellbeing has important implications for students' success during and beyond university. As such, educators need clear empirical evidence of the aspects of university life that contribute to students' wellbeing. We use a mixedmethods approach to ask whether and how students' diverse university experiences contribute to their self-rated wellbeing. In an online survey, 696 students provided accounts of positive and negative experiences at university and self-rated their wellbeing using the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS). Most of the sample reported high wellbeing, though there was diversity in their reported experiences. Regression analysis revealed that students' experiences at university significantly contributed to the variation in their wellbeing scores. Social and academic experiences were the strongest positive contributors to student wellbeing; whereas gender, social experiences and emotional experiences were the strongest negative contributors. Our findings offer guidance for strategies within academic and non-academic university contexts that can effectively and efficiently promote student wellbeing.
al., 2018) have been found to contribute to distress and wellbeing. These studies demonstrate that there are aspects of students' lives that contribute to their distress and their wellbeing. However, they do not offer clear guidance whether or to how experiences at university (within and outside the curriculum) directly contribute to variation in students' wellbeing or students' distress. Such evidence could offer explanations for what makes curriculum approaches successful, as well as pointing to other effective means of mitigating distress and fostering wellbeing. We aim in this study to address this gap, by asking: which elements of students' diverse university lives contribute to student wellbeing? Two theoretical lenses offer support for the design of this study. First, self-determination theory (SDT) depicts wellbeing as an indicator of optimal development, growth and mastery. In SDT, a person's wellbeing depends on the extent to which the surrounding environment enhances (or diminishes) three psychological needs: autonomy, competence, and belonging. SDT frameworks have been applied to employment settings (Sears et al., 2013) and childhood education (Vansteenkiste et al., 2004) to successfully increase wellbeing and subsequently, increase productivity, persistence, and prosocial behaviour. Educators are starting to apply SDT frameworks to guide their understanding of student wellbeing (Baik et al., 2017;Field et al. 2014;Jeno et al., 2018). As with all good theory development, evidence of how university environments support or diminish student wellbeing will be invaluable for educators' efforts.
Second, a dynamic systems approach acknowledges that universities comprise multiple "systems" (e.g., student groups, teaching teams, executive leadership teams, professional services) that are so closely related to each other that change within any one system can affect other systems (e.g., Lerner & Overton, 2008). When applied to higher education, this framework can be effective for identifying systems that are most conducive to change, as well as ensuring widespread and sustainable change (e.g., Dooris & Doherty, 2010;Haggis, 2008). For example, Dooris and Doherty (2010) report that university-wide change to improve health behaviour required sustained efforts from students, teaching staff, retail staff, infrastructure and leadership teams. We adopt a dynamic systems approach by: (i) identifying various university systems that can affect student wellbeing; (ii) drawing on students' perspectives of the university systems that affect them; (iii) identifying experiences that are common, for guidance on efficient and impactful methods for improving wellbeing.

The Current Study
The aim of this study is to investigate whether and how experiences at university, from students' own perspectives, contribute to their psychological wellbeing. We ask: which elements of students' diverse university lives contribute to student wellbeing? And, are those contributions positive, or negative, or both? We take a mixed-methods approach by drawing on students' accounts of their experiences and their self-ratings of psychological wellbeing.

Participants
Participants were 696 university students studying at a large metropolitan university in Australia, with gender distributions that were representative of the broader university student cohort (71% female, 27% male, 0.6% Not-male-or-female). Participants were recruited as part of a larger study about psychological distress and wellbeing (contact first author for details). They were diverse in age, with the majority under 21 (range: 17 to 54, M = 19.7, SD = 3.5). Most were undergraduate students studying psychology subjects as part of a Bachelor of Science (43.1%), Arts (34.2%), Commerce (1.0%) or Biomedicine (0.9%). A small proportion were studying a graduate diploma (2.4%) or postgraduate studies (0.5%).

Procedure
Students were invited to participate via online advertisement methods within the University. Participation involved completing an online survey that took approximately 20 minutes to complete. Students could complete the survey at any time, at their own pace, and anywhere that had internet access. They completed the survey between May and September of 2017.

Measures and Analysis
The current study focuses on two tasks within the online survey: (i) students' accounts of their university experiences, and (ii) students' psychological wellbeing.
We asked students to provide accounts of their experiences by describing a time at university "when they felt good", followed by a time "when they felt bad". We asked students to include details about what happened, how they felt, and why they felt that way. We coded their responses using a bottom-up approach: identifying elements that were important factors in the student's account and keeping a coding schedule of the various experiences. There were 80 experiences in total (29 positive, 51 negative). Binary scores were created for each experience, in which 0 = not mentioned by the student, 1 = mentioned by the student. Cross-coding of the first 40 responses revealed excellent inter-rater reliability (for positive: 92% agreement, Cohen's = .90; for negative: 81% agreement, Cohen's = .76).
Experiences were coded into themes. Themes were identified in a "bottom-up" process, in which similar experiences were grouped together, rather than using pre-determined categories. Positive and negative experiences were each represented by the following five themes: (i) social experiences (e.g., family, meeting or making friends), (ii) academic experiences (e.g., class activities, lectures, studying, assessments), (iii) non-academic experiences (e.g., being outside, attending university events, meals), (iv) emotional experiences (e.g., pride, excitement, inspiration), and ideographic experiences (e.g., experiences that were mentioned by no other student or very few other students). Frequency scores for each theme were calculated as the number of experiences each student described that related to that theme.
Psychological wellbeing was measured with the Warwick Edinburgh Measure of Psychological Wellbeing Scale (WEMWBS; Tennant et al., 2007). The scale includes 14 items about events over the past two weeks, each rated from 1 = none of the time, to 5 = all the time. Ratings are summed to give a total score from 14 to 70. The internal consistency for WEMWBS is excellent, ranging from a = .89 to a = .91 (Tennant et al., 2007).
Preliminary analysis included investigating any differences in wellbeing and experience for students with different demographic experiences, as such differences might affect the main analysis. For the main analysis, we regressed students' wellbeing onto the 84 experiences from students' accounts, as well as demographic information. Categorical demographic variables with more than one group (e.g., gender: male, female, not-male-or-female, prefer not to say) were transformed into binary variables so that 1 = one specific group, 0 = all other groups. This allowed inclusion of categorical demographic variables in the regression. The demographic information was entered together in Stage 1 of the regression and the 80 experiences were entered in Stage 2.

Results
Students' ratings of their wellbeing were diverse, ranging from 16 to 70 (M = 46.86, SD = 9.38). Most of the sample (71%) reported high wellbeing (WEMWBS score > 38), with 13.9% reporting very high wellbeing (> 57), 11.6% reporting low wellbeing (< 37) and 3.2% reporting very low wellbeing (< 28). They described up to nine experiences within their positive accounts (M = 2.74, SD = 1.54) and up to ten experiences in their negative accounts (M = 2.81, SD = 1.83). For positive accounts, the most common theme was "academic experiences" (587 mentions), followed by social (519), emotional (422), non-academic (334), and ideographic (48) experiences. For negative accounts, the most common theme was "academic experiences" (864 mentions), followed by emotional (662), social (244), ideographic (134) and non-academic (55) experiences. The overall number of mentions made for each theme is shown in Table 1.  Table 2 shows the significant univariate associations between students' demographic information and their wellbeing scores. Three significant associations were found: male students reported higher wellbeing than students with other genders, F (3, 695) = 3.67, p = .01; older students reported higher wellbeing than younger students, F (3, 688) = 2.85, p = .04; and students living with others reported higher wellbeing than students living alone, F (1, 695) = 4.80, p = .03. These three demographic variables have been included in the regression model (main analysis) in order to account for these effects. In Stage 1 of the regression, the demographic experiences (age, living alone, and gender [male, female, not-male-or-female]) contributed significantly to the wellbeing regression model, F (5, 683) = 3.69, p = .003, explaining approximately 2% of variance in students' wellbeing (R 2 = .03, Adjusted R 2 = .02). Living alone and the 'not-male-or-female' gender dummy variable made significant negative contributions, indicating that those who were not-male-or-female or who were living alone had lower wellbeing than others.

Preliminary Analysis: Gender, Age, and Living Situation were Associated with Wellbeing
When the 80 experiences were added at Stage 2 of the regression, the variance explained in the model increased to 8% (R 2 = .19, Adjusted R 2 = .08), F (80, 603) = 1.56, p = .002. Among variables that made a positive contribution to the model (i.e., increased students' wellbeing scores), the positive-academic experience "lectures" made the largest contribution, followed by the positive social experience "like-minded people", the positive non-academic experience "university clubs/events", and the negative ideographic experience "overcome challenges". Among the variables that made negative contributions to the model, the 'not-male-or-female' gender variable made the largest contribution, followed by the negative academic experience "exams/tests", the negative social experience "friends", the negative emotion experience "self-doubt", the positive emotion experience "confidence", and the negative ideographic experiences of "health" and "weather". Living alone also continued to make a significant negative contribution to variance in wellbeing. The raw and standardised regression coefficients and structure coefficients are shown in Table 3.

Discussion
The aim of this study was to investigate whether and how students' university experiences contributed to their psychological wellbeing. Most students reported high wellbeing, however there was diversity in their ratings. Specific accounts of academic (lectures, exams), social (like-minded people, issues with friends), non-academic (university clubs/events), emotional (selfdoubt, confidence), ideographic (overcoming challenges, health), and demographic experiences (gender, living alone) contributed to the diversity in students' wellbeing. Lectures, like-minded people, university clubs/events, and overcoming challenges enhanced student wellbeing; whereas exams, issues with friends, self-doubt, confidence, poor health, identifying as not-male-or-female, and living alone diminished student wellbeing.

Academic, Social and Emotional Experiences Support SDT Frameworks of Student Wellbeing
The social and academic experiences that contributed to wellbeing are consistent with an SDT framework. Finding like-minded people and joining social clubs can both be aligned with SDT's "belonging", as both indicate meaningful relationships built upon the individual student's own interest (e.g., Deci & Ryan, 2000). Conversely, issues with friends can affect a sense of belonging by creating feelings of mistrust and isolation. This suggests that helping students to establish a sense of belonging with like-minded friends, rather than a sense of belonging to the wider university community, might be an effective strategy for enhancing student wellbeing and success at university.
Positive experiences in lectures, overcoming challenges, and negative experiences with exams, can be aligned with SDT's concept "competence" (e.g., Deci & Ryan, 2000), in that they represent opportunities for the student to build and confirm (or diminish) their competence in the discipline. A key part of SDT's competence is that a person sees potential to achieve mastery in the future (Deci & Ryan, 2000). Students' experiences overcoming challenges and engaging meaningfully in lectures are both opportunities for students to build this type of potential competence. In contrast, negative exam experiences (e.g., anxiety about upcoming exams, or doing poorly on exams) diminish a student's sense of competence or potential competence. The implication here is not to avoid negative exam experiences, but instead to help students see such difficult experiences as challenges to be overcome. This echoes research regarding student resilience, which advocates that university activities such as regular assessment and feedback can help students develop strategies to thrive in difficult academic tasks (e.g., Holdsworth et al., 2018). The recommendations from both SDT and resilience frameworks suggest that resilience is promoted by highlighting students' potential for competence and mastery of their discipline.
The negative contributions of two emotions (confidence and self-doubt) to wellbeing offer indirect support for SDT frameworks of student wellbeing. Well-established theories of wellbeing conceptualise psychological wellbeing as having more depth and complexity than can be explained by positive affect alone (e.g., Deci & Ryan, 2000;Seligman, 2011;). Our findings are a reminder that academics invested in student wellbeing are not aiming to make students happy or avoid student's difficult emotions; instead, they are aiming to help students develop resilience and social responsibility, and to reach their full potential.

Diversity of Experiences Provide Support for a Dynamic Systems Approach to Higher Education
Our study has revealed a variety of experiences within academic, non-academic, social, and emotional contexts that (i) were readily recalled as part of students' positive and negative experiences at university; and (ii) contribute to their wellbeing. Each of these contexts represents at least one system within the university. For example, academic contexts represented a variety of systems that relate to the curriculum (e.g., learning spaces, activities, and relationships); whereas non-academic contexts represented a variety of systems that sit alongside and support the curriculum (infrastructure, university-lead programs, and services). This speaks to the diversity of systems that students encounter as part of their day-to-day experiences of university and supports Dooris and Doherty's (2010) argument that addressing a single issue at university requires addressing many systems simultaneously.
We were also interested in identifying the systems that are most common and therefore have wide impact across student cohorts. The academic theme was the most commonly cited context, which supports the depiction of the curriculum as the "glue" that holds the university students' experience together (McInnis, 2001). The substantially high number of negative academic experiences warrants attention, as it is a reminder of the difficulties embedded within students' learning experiences. During their university experience, students are likely to encounter pressure to meet academic demands, social and identity issues, financial strain, and stressors of transitioning to an unfamiliar environment (DeRosier et al., 2013). As mentioned above, approaching these difficulties as challenges to be overcome can be a rewarding experience for students. The task for the academic educator is to be aware of these common academic difficulties and to help students develop strategies to overcome them.
One experience that was negatively associated with wellbeing was gender identity, namely, those who did not identify as male or female reported the lowest wellbeing. The number of students identifying as non-binary in our study is admittedly too low to make any strong conclusions; however, the negative association echoes research of the dangers, discrimination, and vulnerabilities reported by gender minority students (Roffee & Waling, 2017;Woodford et al., 2014). Victimisation can negatively impact student mental health, leading to distress and poorer academic outcomes (Woodford et al., 2014). A dynamic systems approach suggests that various university systems have potential to better support and protect minority students. For example, enacting university policies that recognise sexual and gender minority students communicates institutional LGBTIQ+ norms and values (Pitcher et al., 2018). Access to resource centres and the presence of student organisations can also offer safety and support from peers and staff . Putting such support in place will go a long way toward supporting the wellbeing of minority students.

Limitations, Future Directions, and Conclusion
The fact that students were not asked directly whether their positive and negative experiences affected their wellbeing might be a limitation, as this might have provided clear direction on how well SDT and a dynamic systems approach can be applied to student experiences. However, by not asking directly about how their experience affects their wellbeing, we remove bias in students' opinions about wellbeing. Instead, we have a variety of real experiences that happen to students at university, and we are able to deduce whether and how these experiences contribute to wellbeing. Further, information about university life was volunteered by the students in how they spoke about their experiences. When describing a negative experience, 83 students opted to include that they had managed to overcome that experience. Their experiences made significant contributions to the overall regression, and their wellbeing was significantly higher than students who did not describe this type of experience (as shown by the beta-weights in Table 3). Had we asked about their experiences in any other way, we might not have captured this perspective. Future studies would be advised to follow our approach and reduce bias and misunderstanding about wellbeing.
The students in our study described diverse experiences, and specific experiences within academic, non-academic, social, emotional, and demographic contexts affected variation in their wellbeing. Our findings offer guidance for how universities can best support student wellbeing, including: helping students to find like-minded people in their classes; designing lectures that challenge and build students competence, and; embedding policies that support and protect more vulnerable students, especially gender minority students and students living on their own. This study provides clear support for SDT and dynamic systems approaches to student wellbeing. We anticipate that initiatives that draw principles from both theories will find success in their enhancement of student wellbeing.