College Student Well-Being: Explaining Academic and Behavioral Outcomes from a Representative College Student Sample

Positive psychology's growth has impacted many fields, including the field of education. In recent years, higher education scholars and practitioners have paid increasing attention to the importance of student well-being (Donohue & Bornman, 2021). This has become especially important given the rise in mental health concerns, the impact of COVID-19 and isolation on well-being, and the growth of online learning opportunities (Nurunnabi et al., 2022; Douwes et al., 2023). Recent studies on college student well-being have demonstrated the potentially long-lasting effects that COVID-19, in particular, has in a longitudinal examination of college student well-being (Lanza et al., 2022). The U.S. Department of Education has made mental health and well-being a key goal: “invest in every student's mental health and well-being by increasing school-based health services for students and building schools that support students' overall well-being” (DOE, n.d.). Promoting college student well-being is an important goal that higher education scholars, policy-makers, and practitioners share.

At the same time, there is also increasing interest in the benefits of well-being, particularly in subjective well-being (SWB). The idea that SWB can generate positive life outcomes is based on the Broaden-and-Build theory within positive psychology (Fredrickson, 2001). The theory proposes that positive emotions broaden an individual’s thoughts and actions, which leads to more positive behaviors and experiences and, in the longer run, this subsequently builds the individual’s personal resources that are not only essential to coping and survival but create positive outcomes (Fredrickson, 2004). Indeed, several reviews have shown that SWB is related to better life outcomes across different life domains, including health, work, and relationships. De Neve et al. (2013) summarized the benefits of SWB to a wide array of different outcomes, including “(i) health & longevity; (ii) income, productivity, & organizational behavior; and (iii) individual & social behavior” (p. 2); De Neve et al.’s review discussed different theoretical mechanisms for these relationships, such as evidence that SWB promotes good behaviors and practices, positive and fulfilling social relationships, self-control and long-term time orientation, and a feedback loop where SWB begets positive experiences that begets SWB. Diener et al.’s (2018) review of SWB research further provided specific validity estimates, including r = 0.18 (health and longevity), 0.24–0.27 (job satisfaction), and 0.32 (resilience and stress recovery).

Within research in higher education however, there has been less work examining the link between well-being and student outcomes. Studies have suggested a relationship between student well-being and academic outcomes such as GPA (Bücker et al., 2018; Swanson et al., 2020; Tesfaye, 2020), but there is less research examining how well-being may be related to and predictive of educational outcomes beyond GPA. Most notably, Bücker et al.’s (2018) meta-analysis focused on 47 studies of SWB and academic achievement (defined as grades and scores on standardized tests) to find an overall correlation of 0.16. They identified key future research directions as including other academic outcomes beyond grades and tests, investigating other measures of SWB beyond the single-item scale or unitary dimension used by many studies, and more longitudinal research. A few other relevant studies since the meta-analysis include Swanson et al. (2020), which focused on a sample of around 1,500 students to show how campus belonging and self-efficacy (which are related concepts to SWB) predict GPA over the course of three years, and Tesfaye (2020), which conducted a deeper study of 123 Ethiopian college students to show how multiple dimensions of well-being (e.g., self-acceptance, purpose) correlated with academic achievement. Thus, there is a call for more research on the educational outcomes of well-being, especially as it relates to the measurement and conceptualization of well-being and other academic outcomes in higher education, and especially in light of recent changes and shifts in how student well-being is affected by academic grades post-COVID and with new online learning programs (e.g., increased exam anxiety shown by Arora et al., 2021).

Within positive psychology, the notion of well-being has expanded from SWB to a broader vision of well-being that comprises multiple dimensions. This includes psychological well-being (six dimensions including autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, and self-acceptance; Ryff & Keyes, 1995) and PERMA (i.e., positive emotions, engagement, relationships, meaning, and accomplishment; Seligman, 2011). These approaches conceptualize multiple dimensions as “building blocks” of well-being that help identify specific domains or areas where interventions or public policy can help promote overall well-being (Seligman, 2018; Kansky, 2017). As such, the measures of well-being have similarly included multiple dimensions. For example, Lui and Fernando (2018) developed a multidimensional scale of well-being that assesses the various “physical, financial, social, hedonic, and eudaimonic domains of this construct” (p. 135). Nevertheless, these psychological approaches to understanding well-being as a multidimensional construct have been more limited in their application in higher education. At the same time, other studies have focused on only one or two dimensions of well-being related to college students. For example, several meta-analyses show the importance of grit for academic outcomes (Credé et al., 2017; Lam & Zhou, 2019), while others have focused on loneliness and belonging (Nicpon et al., 2016; Swanson et al., 2020).

Given this, scholars have proposed a multidimensional approach to investigating holistic well-being among students. Early on, Kern et al. (2015) proposed a multidimensional approach to measuring well-being based on the PERMA model of well-being. They tested this model among a sample of 516 male high school-aged students, finding good factor analytic results that correlate with other constructs, such as physical activity and school engagement. They argued that “directly assessing subjective well-being across multiple domains offers the potential for schools to more systematically understand and promote well-being” (p. 262). Other similar models of student well-being, all focusing on adolescents under the age of 18, include Huebner and Gilman’s (2002) Multidimensional Students’ Life Satisfaction Scale – which includes dimensions such as Family and Friends and Self – and Renshaw et al.’s (2015) Student Subjective Wellbeing Questionnaire – which includes dimensions such as school-specific relationships and affective experiences. Of note, none of these measures focus on college student well-being. While related, college student well-being is notably different from adolescent well-being in the increased influence of peers, decreased influence of parents and family, increased opportunities and access to a range of positive and negative activities, and focus on self-discovery and exploration (Baldwin et al., 2017).

The present article is based on McCuskey and Zhang’s (2021) framework for holistic campus well-being dubbed Steps to Leaps, “a community-wide approach to improving the well-being of students broadly… designed to foster growth in the areas of well-being, leadership, impact-making, networks, and grit” (p. 648). They argued for approaching well-being as made up of many different “pillars” that support college student well-being. Such a model seeks to capture a variety of influences and challenges that college students draw from or face, and providing a framework through which higher education scholars and practitioners can understand—and target interventions for—improving college student well-being. While other models of college student well-being exist – for example, Renshaw et al.’s (2015) College Student Subjective Wellbeing Questionnaire, which measures academic efficacy, college gratitude, school connectedness, and academic satisfaction – the present model focuses on college students’ self-examination of well-being independent of their academic outcomes. In contrast, other models include academic outcomes as a measure of well-being. Under the Steps to Leaps model, academic outcomes are instead treated as a correlate to student well-being, and, as such, are presented as evidence of convergent validity in this study.

Given the significant gap in understanding how well-being is related to student outcomes, and the need for a holistic multidimensional approach, the present study focuses on five campus well-being “themes” proposed by McCuskey and Zhang (2021) to determine how they are related to key college student academic and behavioral outcomes. These “themes”, which they called “pillars”, are: (1) subjective well-being, based on traditional understandings of SWB comprising in part high positive and low negative affect (Diener, 1984); (2) positive relationships, based on key studies on the importance of social networks and belonging in higher education (Thomas, 2000; Zhu et al., 2013); (3) leadership, based on studies showing the unique value of college student development in leadership and teamwork (Cress et al., 2001), (4) grit, based on the meta-analyses showing the value of grit in predicting academic outcomes (Credé et al., 2017; Lam & Zhou, 2019), and finally (5) purpose and meaning, based on studies surrounding purpose and meaning in life as predictors of academic outcomes (Sharma & Yukhymenko-Lescroart, 2018; Tesfaye, 2020).¹

Prior work has illustrated how specific components of campus well-being predict academic outcomes. For example, Thomas (2000) used a sample of 322 freshmen to show how a broader social network led to student integration and persistence. Cress et al. (2001) focused on leadership activities as a predictor of developmental outcomes (e.g., multicultural awareness, personal and societal values) in a sample of 875 college students. Sharma and Yukhymenko-Lescroart (2018) found in a sample of 1,010 college students that a sense of purpose predicted self-reported degree commitment (B = 0.24), while Tesfaye’s (2020) study of 123 Ethiopian college students found a correlation of 0.59 between purpose in life with academic achievement. By far, the most-studied component of campus well-being is grit, with multiple meta-analyses showing correlations of 0.16 to 0.26 with academic achievement (Credé et al., 2017; Lam & Zhou, 2019). Put together, these individual studies provide evidence that each of the five proposed dimensions of well-being relates to academic outcomes. To our knowledge, however, our study is the first to (a) study these dimensions in relationship with one another, and (b) investigate multiple academic outcomes beyond GPA. This is critical to help positive psychologists and practitioners understand the relative importance of these dimensions for different types of student outcomes within higher education.

Finally, there is a need for more robust investigations of the psychometric properties of a campus well-being measure. Diener et al.’s (2018) review of the research on SWB (in adult and general populations) discussed psychometric properties, including reliability over time (r = 0.79), convergent validity with others-report measures (r = 0.42), and convergent validity with other positive social behaviors (r = 0.26 to 0.47). While different holistic measures of campus well-being have been proposed and tested in past smaller studies, there has yet to be a larger-scale investigation of the psychometric properties, including reliability over time, factor structure, and construct validity as it relates to multidimensional theory (i.e., five dimensions), convergent validity with academic behaviors such as class attendance and shared meals in social settings. Our investigation of these properties seeks to develop and promote a psychometrically sound, valid, and useful multidimensional measure of campus well-being that can be applied to college campuses in order to promote and assess future interventions or policies that support college student well-being.

In this paper, we seek to use a holistic approach to assess well-being in higher education and examine how it is associated with key educational outcomes. Based on McCuskey and Zhang’s (2021) model of five “pillars” of well-being, a 49-item measure of Campus Well-Being was created and assessed across two large samples of college students across two semesters. Critically, we examine how the various dimensions of well-being are related to different behavioral outcomes (e.g., class attendance, participation in student organizations, meal card swipes in in-person dining halls, and time spent in dorm rooms) and academic outcomes (e.g., GPA). Our results serve to demonstrate the importance of holistic assessments of well-being in higher education and its relation to key student outcomes.

The Campus Well-Being measure comprised 49 items based on McCuskey and Zhang’s (2021) Steps to Leaps model organized into five themes or pillars: subjective well-being (e.g., “I felt happy”, “I felt stressed”), grit (e.g., “I am passionate about what I study”, “I persist even in setbacks I face”, “I recover from stressful events quickly”), purpose and meaning (e.g., “I have a clear sense of purpose”, “My studies help me feel capable in the work I can do”), relationships (e.g., “I have been treated with less respect than others”, “Others value diversity of beliefs”, “I look for ways to help others in need”, “I can trust faculty and staff”), and leadership (e.g., “Others would say I am able to start and lead changes in direction in my groups”, “Others would say I am aware of others’ cultural backgrounds”, “Others see me as able to motivate those I work with to work toward group goals”). These five themes are comprised of 14 distinct dimensions, with three to five items per dimension. An overview of the measure is depicted in Fig. 1, and the full measure can be found in Appendix A.

Bild vergrößern

Other correlates were collected through the institutional data office. These include: (a) GPA in the previous semester; (b) cumulative GPA as of the previous semester; (c) percentage of all course grades in the previous semester that were a C, D, F, or W; (d) percentage of all scheduled classes that the student attended in the first eight weeks of the semester; (e) percent of the student’s allotted meal swipes that they used in the semester so far; (f) number of co-swipes (i.e., swiping their meal card with other people, implying that they dined together) in the week leading up to the survey; (g) average time spent in minutes per day in on-campus academic buildings during the semester; (h) average time spent in minutes per day in on-campus residential buildings during the semester; (i) average change in time spent on campus from weekdays to weekends (i.e., identifying students who frequently left campus during the weekend vs. staying on campus); (j) whether or not the student is a registered member of a student organization; and (k) whether or not the student is a registered leader or officer of a student organization. Additionally, we assessed the three academic outcome variables (GPA, cumulative GPA, and CDFW rate) normed within colleges. For example, we took the GPA in the previous semester and computed how far above or below it was to the average GPA within the students’ college. Since some colleges are more academically challenging than others, this was meant to help us assess academic outcomes relative to the students’ peers in their college.

First, we assessed the reliability of each sample individually and correlations to assess test–retest reliability over time. We also conducted factor analyses on each sample to assess the factorial validity of the measure of Campus Well-Being. Next, we examined the convergent validity of the campus well-being with the academic and behavior-related correlates. Finally, we conducted a relative importance analysis of the dimensions of the Campus Well-Being assessment in predicting the key academic and behavior-related correlates, to examine which of the dimensions are relatively more important in predicting desirable outcomes. Although an a priori power analysis was not possible, due to the use of an archival dataset provided by the Institutional Data office, we conducted post-hoc power analyses reported in Appendix B, Tables 14 and 15. Overall, the power for each CFA test (using the semPower package and following recommendations by Moshagen & Bader, 2023) was greater than 0.999. Power for each of the convergent validity tests (using the pwr package) was more variable: power was strongest for the academic outcomes and for subjective well-being, grit, purpose, and leadership; however, it was notably weaker for relationships and for the number of dining partners, time spent on campus, and student organization participation.

Not only does our paper add further evidence to the importance of well-being in college student life and outcomes, it also demonstrates the importance of studying it holistically – including both hedonic and eudaimonic approaches. Earlier studies, as noted previously, have generally focused on what is traditionally considered the hedonic model of well-being: SWB, other dimensions that focus on happiness and positive affect (Ryan & Deci, 2001). Our research demonstrates that SWB is only part of the picture; for many academic and behavioral outcomes, eudaimonic models of well-being such as purpose/meaning are just as important, if not more so. These eudaimonic approaches focus on meaning and self-realization and are captured in Steps to Leaps dimensions such as purpose and leadership in our present study (Ryan & Deci, 2001). In fact, the relative importance analyses point towards grit as the most important theme, especially for academic outcomes. This aligns with prior meta-analyses showing the importance of grit, but it adds insight into how important grit is, relative to other components of well-being (e.g., Credé et al., 2017; Lam & Zhou, 2019). As such, it suggests that when investigating campus well-being, a variety of different constructs or components should be studied holistically. Both the hedonic and eudaimonic approaches are necessary to understand how college students thrive both in their positive affect and in their self-realization. Capturing this variety of dimensions would allow future studies to understand, for their unique sample or university, which are most relevant to their desired outcomes and worth targeting for future interventions, training, and other opportunities to enhance student well-being.

These findings have important practical implications as well. First and foremost, we offer a new instrument that can be used to holistically measure campus well-being, above and beyond traditional measures that focus solely on subjective well-being or grit. As a case study, we demonstrate the value of this instrument for students at the institution from which we collected data, including a set of policy initiatives and programming based on this instrument (see earlier footnote on McCuskey & Zhang, 2021). Secondly, we link campus well-being based on our instrument to a set of academic and behavioral correlates using a rich set of institutional data that thus far has been difficult to obtain in traditional studies of well-being (e.g., actual time spent on or off campus, use of meal card swipes). In doing so, we show the value of student well-being to “objective” indicators that other university administrators might be interested in, thus creating opportunities for buy-in from multiple campus stakeholders to provide resources and support to improving student well-being. Lastly, our holistic approach to well-being considers other aspects that have received less attention in the literature thus far, such as the value of relationship networks – including both positive and negative experiences with diversity – and a sense of purpose. The use of relative importance analysis helps put these aspects of well-being in comparison with one another, further emphasizing the importance for university administrators, student affairs professionals, and instructors to consider student well-being in light of and from the perspective of a number of different factors.

It is worth noting that our study was limited to two semesters of data collection and focused on covariates obtained from data within the same semester. Moreover, a key limitation of our study was the reliance on the five-dimensional framework for well-being used at the campus from which we collected data (McCuskey & Zhang, 2021). This five-dimensional framework was developed through an iterative process with faculty, staff, and students at the specific campus from which data were collected; thus, it is uniquely calibrated to studying well-being at a large public research university campus in the Midwest. In doing so, our archival dataset focused on the items displayed in Fig. 1 and thus does not allow us to compare this framework of student well-being with other existing frameworks (e.g., Renshaw et al., 2015) nor to investigate if this framework applies equally as well with other institutions’ student bodies. Our study does not attempt to promote this framework of well-being above and beyond existing frameworks. However, future studies might compare and contrast the incremental validity of the different instruments – included concepts related to well-being such as life satisfaction and general health – to offer advice for future researchers and practitioners on the preferred framework for student well-being.

Future studies can address some of our limitations by expanding to multiple institutions, collecting alternative data sources (e.g., student retention, other survey constructs), or obtaining longitudinal data to follow changes in well-being over time. As an example for a post-hoc analysis, we examined the degree to which each well-being dimension predicted next semester retention (i.e., did the student re-enroll in courses in the next semester or graduate?). In total, of the 3109 respondents to the survey from the Spring 2023 sample, 2667 (85.8%) were retained (i.e., re-enrolled in Fall 2023 or graduated in Spring 2023); 442 (14.2%) were not retained (i.e., did not re-enroll in Fall 2023 and did not graduate). Appendix B (Table 10) depicts the odds ratios (OR) for each dimension when all 14 were used to predict retention in a logistic regression. Purpose (external) and relationships (prosocial) were significant and positive (OR = 1.22 and 1.32 respectively), indicating that higher scores on these dimensions of well-being predicted higher likelihood of student retention. Interestingly, leadership (skills) was significant and negative (OR = 0.79), suggesting that higher scores predicted lower likelihood of student retention. Future studies can build on these initial findings through additional data and longitudinal data collection to better assess the impact of well-being on other outcomes such as retention, career outcomes, etc. Such longitudinal data would be especially important to track changes in college student well-being in the post-COVID era as higher education continues to react and “recover” from the major disruptions of COVID and online learning.

Finally, our study also raises several interesting research questions that can be explored in future studies. First, as a post-hoc analysis, we investigated how our results might differ by gender and race, given the extensive research showing how well-being functions differently based on such demographics (e.g., Batz & Tay, 2018; Woody & Green, 2001). Here, we focused on the convergent validity estimates from Spring 2023 (Table 5). Of the estimates that were statistically significant in the combined full sample, we examined if any were significantly different when separated by genders or race-ethnicity. Appendix B (Table 9) compares the convergent validity correlations when separated by demographic groups. There were a few notable patterns of differences in convergent validity. For example, the convergent validities of well-being dimensions with behavioral covariates such as student organization participation and meal swipes were notably stronger among females than males (difference in correlation of up to 0.33), while the convergent validities with academic outcomes were stronger among males than females (difference in correlation of under 0.10). Similarly, the convergent validities of well-being dimensions with CDFW rate were slightly stronger among URM students than non-URM, but the convergent validities with cumulative GPA were notably stronger among non-URM students (difference in correlation of up to 0.37). Lastly, we also conducted measurement invariance testing of each of the five well-being dimensions between genders; we found that configural invariance held between genders for the grit, purpose, and leadership scales, but scalar invariance did not hold for relationships and subjective well-being (Appendix B, Tables 11, 12 and 13). This suggests that further research is necessary to better understand how campus well-being functions differently based on different demographic characteristics.

Secondly, our holistic approach to studying college student well-being raises interesting questions on which dimensions can and should be included in a measure of well-being. While we elected to use McCuskey and Zhang’s (2021) model, which included dimensions such as grit and leadership that were pertinent to the institution from which data were obtained, other dimensions might also be of interest (e.g., the traditional “life satisfaction” dimension from Diener’s SWB framework). Moreover, while the evidence is clear that leadership can be developed (Day & Dragoni, 2015), an anonymous reviewer rightfully pointed out that students might not be interested in developing or have the capacity to develop as a leader; thus, would such students never reach “high” levels of well-being? Additionally, there may be “dark side” effects of some of these dimensions of well-being; for example, there is evidence that too much focus on grit can lead to overworking and burnout (Czerwinski et al., 2023). Thus, we encourage future scholars to dive deeper into the specific dimensionality of college student well-being to compare and contrast different models and their validity, especially concerning academic and behavioral outcomes.

As much of the landscape within higher education changes in the coming years, from the rise of online education to the ever-increasing concerns of mental health, the necessity of caring for student well-being continues to be a focal point for higher education research and practice. Drawing from the robust literature in positive psychology, we offer unique insights into the value of a holistic measure of campus well-being and demonstrate its implications for student academic outcomes, behaviors, and other variables of interest. We hope that these findings spur future investigations into the multidimensional nature of well-being and its ongoing implications for college student life and outcomes.