Strengths-Based Models of Resilience in Tertiary Education Students

Abstract

Tertiary education students experience higher levels of distress than the general public and their age-matched peers. Preventive health programs targeting tertiary education students are needed to combat high levels of distress in this population. This article investigates strengths-based theories of resilience which state that improvements in coping with stress also improve resilience. It utilized hierarchical regression to determine the amount of variance in mental health and resilience outcomes associated with the stress management resources, namely paying attention to the present moment (mindfulness) and engaging in personally meaningful activities (engagement), in a sample of 1,072 Australian tertiary education students. The results found that mindfulness and engagement are negatively associated with symptoms of depression and anxiety and positively associated with psychological well-being and resilience in tertiary education students. We also found that the amount of variance in mental health outcomes associated with engagement and mindfulness is additive. When a second stress management resource was added to the modeling, the amount of variance in the measured mental health outcomes explained by the coping resources increased. These findings help to provide a conceptual framework that can be used to design resilience training programs in tertiary education students to address the high levels of mental health disorder in this population.

Keywords:

• Distress
• disorder
• mental health
• resilience
• tertiary education students

Introduction

The mental health of Australian tertiary education students is a problem of growing concern (Larcombe et al., 2016). Mental health disorders are the leading cause of premature mortality in Australia (Australian Bureau of Statistics, 2024). Onset of mental illness takes place by age 25 in close to two-thirds (62.5%) of mental illnesses onset by age 25 (Solmi et al., 2022). Suicide is the most common cause of death for people aged 14 to 44 years in Australia (Australian Institute of Health and Welfare, 2024) and is the fourth most common cause of death for people aged 15 to 29 years worldwide (World Health Organisation, 2024a). An individual who dies by suicide loses an average of 35.6 potential years of life (Australian Bureau of Statistics, 2024). In addition to suicide-related mortality, mental health disorders in early life are associated with reduced functioning and well-being, including educational attainment, social and familial relationships, workplace productivity and career outcomes, physical health, and life expectancy (Productivity Commission, 2020). Vigo et al. (2016) estimated that mental illnesses are responsible for 25% of the global burden of disease. However, this is likely an underestimation, with more recent work demonstrating that the negative global impact of mental disorders is second only to cardiovascular disease (Vigo et al., 2016). Arias et al. (2022) further estimated that the worldwide burden of mental health disorders cost USD 5 trillion in 2019 alone.

As a group, tertiary education students experience higher levels of distress (comprising depression and anxiety) and mental illness than both the general population (Larcombe et al., 2016) and their age-matched peers not attending university (Rickwood et al., 2017). This is because tertiary students are subjected to stressors, including transitioning from high school to tertiary education, financial pressure from time out of work to study, increasing cost of tertiary degrees and student debt, and competitive job markets and limited jobs for tertiary graduates (Pidgeon et al., 2014; Auerbach et al, 2016). Additionally, the COVID-19 pandemic has exacerbated levels of distress and mental health disorder in tertiary students worldwide (Aristovnik et al., 2020). Therefore, addressing the high prevalence of distress and mental illness in tertiary students is an important avenue for improving population health and reducing economic and social costs of mental disorders.

There is growing recognition that mental health systems place too much emphasis on treating mental illnesses and too little emphasis on preventive mental health care (Productivity Commission, 2020). Medical health services in developed countries typically combine preventive efforts with clinical treatment in order to address the burden of disease on society (World Health Organisation & United Nations Children’s Fund, 2020). In comparison, however, mental health systems devote few, if any, resources toward preventive mental health care (World Health Organisation, 2013). Health literacy is a key determinant of health outcomes across the lifespan (Berkman et al., 2011). Therefore, teaching mental health literacy to individuals at a young age, including tertiary students, would be expected to provide large benefits in improving mental health outcomes (Productivity Commission, 2020). In line with this, the World Health Organization (World Health Organization, 2024b) has called for research and development of appropriate action plans to promote health literacy.

An emerging field of study in preventive mental health is the area of psychological resilience. Resilience researchers attempt to understand why some people exposed to potentially traumatic events experience mental health disorders, while others do not (Bonanno, 2004). Resilience research is hampered by lack of both an agreed-on definition and a gold-standard measure to test levels of resilience (Fisher & Law, 2020). These factors make it hard to measure resilience to compare results across research (Fletcher & Sarkar, 2013). Additionally, validation studies of existing resilience measures show that the factors that promote resilience differ across varied populations (Green et al., 2014; Gucciardi et al., 2011). These differences make it difficult to find a universal model of resilience that can be utilized across varied contexts. Therefore, this study will investigate existing theories of resilience in an attempt to better understand how resilience in tertiary education students may be promoted.

One theory of resilience often investigated in previous research is Fredrickson’s (2001) “broaden and build” theory. Fredrickson’s theory stems from research into the evolutionary value of positive emotions. Broaden and build theory is an extension of approach/avoidance theory, which states that emotions motivate people to avoid threatening situations (Roth & Cohen, 1986). Broaden and build theory expands on approach/avoidance theory by explaining that the purpose of positive emotions is to “undo” the effect of negative emotions. This theory states that positive emotions counteract the avoidance/withdrawal effect of negatively valanced emotions and promote approach/engagement orientation. Fredrickson’s theory further explains that the engaging effect of positive affect promotes exploration with the environment in ways that help individuals learn to deal with challenges. This learning is believed to help to build knowledge of behaviors, or Fredrickson’s so-called thought-action repertoires, which individuals can use to overcome future potential stressors. Tugade and Fredrickson (2004) later demonstrated an association between positive affect and resilience, which was replicated in a recent study (Tuck et al. 2023).

A more recent model of resilience proposes that coping resources can be utilized to manage distress. Similarly to Fredrickson’s (2001) concept of thought-action repertoires, Hamby et al.’s (2018) poly-strengths theory states that resilience depends on the coping resources available to an individual and their ability to effectively employ them in order to manage potential stressors. In this model, however, coping resources have an additive effect on mental health outcomes, meaning that strengths accumulate to improve an individual’s mental health and resilience. This theory evolved independently of Fredrickson’s model. Instead, it is based on trauma research showing that the accumulation of multiple stressors have impacts on functioning beyond a single potential stressor (Finkelhor et al., 2011). This work was then extended to demonstrate that the opposite is also true (i.e., that multiple strengths improve mental functioning more than one or no protective factors). However, the similarity between Fredrickson’s and Hamby et al.’s models demonstrates how the concept of utilizing resources to manage potential stressors is central to both of these models.

Hamby et al.’s (2018) study used poly-strengths theory to determine factors that would promote mental health and resilience in a sample of younger adults (M = 30 years) from a low–socioeconomic status region of the United States. The strengths investigated in Hamby et al.’s study focused on protective factors that could be targeted by interventions rather than those that cannot be easily changed. The study found that emotional awareness and regulation, sense of purpose, optimism, and psychological endurance were all related to improvements in mental health outcomes. A study conducted by Canada’s Research Chair in Interpersonal Traumas and Resilience (RCITR; Moisan et al., 2019) utilized poly-strengths theory to inform a study investigating preventive mental health factors in a sample of Canadian students (M = 15.85 years) taken from the Quebec Youth’s Romantic Relationships Survey (Hébert et al., 2017). This study found that self-esteem, academic achievement, optimism, parental support and attachment, and social support were all related to lower levels of psychological distress.

A previous systematic review of intervention studies that promote mental health outcomes in tertiary education students (Tuck et al., 2022a) demonstrated that paying attention to the present moment had the largest impact on reducing symptoms of anxiety and that taking part in personally meaningful and enjoyable activities had the largest impact on reducing symptoms of depression compared to the other intervention types included in the review. However, this systematic review utilized the above mental health outcomes as a proxy for resilience due to the fact that resilience was not used as an outcome in any of the reviewed interventions (see Tuck et al., 2022a for a review). Therefore, the current study will investigate the relationships between paying attention to the present moment and taking part in personally meaningful and enjoyable activities with symptoms of depression and anxiety and psychological well-being and resilience to determine whether depression and anxiety can be utilized as proxies for resilience. Additionally, there are no published studies investigating the utility of poly-strengths theory for informing health promotion programs designed for tertiary education students. If the poly-strengths theory of resilience (Hamby et al., 2018) is correct, then the amount of variance in the relationship between taking part in personally meaningful and enjoyable activities and paying attention to the present moment with symptoms of depression and anxiety, psychological well-being, and resilience will be additive. Therefore, this study will also test whether two of these factors together explain more variance in mental health and resilience outcomes than either of the factors alone.

Previous research has uncovered a number of factors that influence the mental health of tertiary education students. A review of research investigating distress in tertiary students by Orygen (2017) found that academic pressure, socioeconomic status, international student status, and beginning tertiary studies were all risk factors for student distress. A more recent review by Klepac Pogrmilovic et al. (2021) identified socioeconomic status, international student status, and sexual orientation as risk factors for distress in tertiary students. Finally, a study investigating tertiary student distress during the COVID-19 pandemic (Tuck et al., 2022b) found that age, gender, previous diagnosis of mental illness, and mental health care status were all associated with student distress during this period. Therefore, age, gender, previous diagnosis of mental illness, mental health care status, socioeconomic status, sexual orientation, and international student status will be utilized as control variables in the current study.

Objectives

Goals of this research were to determine whether paying attention to the present moment and taking part in personally meaningful and enjoyable activities are associated with well-being, resilience, and the absence of distress in tertiary education students and to determine whether multiple resilience factors explain more of the variance associated with mental health outcomes in tertiary education students compared to a single resilience factor.

Method

Participants

Participants were 1,072 students enrolled in tertiary education institutions across the eight states and territories of Australia. The online survey for the study was hosted by the Qualtrics survey platform and required participants to report their age, gender, any previous diagnosis of mental illness, mental health care status, socioeconomic status, sexual orientation, and international student status along with completing the psychometric questionnaires. Recruitment for the study was conducted online via advertisements on the Monash Sona Systems participant recruitment webpage, Twitter, a dedicated webpage for the study, and multiple student-run Facebook pages for tertiary students. Participants self-reported their age (M = 25.54, SD = 8.18), gender (69.7% women, 29.2% men, 0.7% trans/nonbinary). Participants were all residing in Australia and enrolled in an Australian tertiary education institution. Participant demographics are displayed in Table 1.

Table 1. Participant Demographics.

Variable	No.	%
Total	1,072	100
Age
18-24	976	91.7
25-34	56	5.3
35-44	24	2.3
45+	8	0.8
Gender
Man	312	29.2
Woman	746	69.7
Trans/nonbinary	7	0.7
Previous diagnosis^a
Yes	490	46
No	576	54
Care status^b
Receiving care	250	24.5
Not receiving care	771	75.5
Socioeconomic status^c
High	578	54.5
Low	482	45.5
Sexual orientation
Heterosexual	767	71.5
LGBTQ	305	28.5
Student status
Local	729	68
International	343	32

Note. ^aPrevious diagnosis of mental illness. ^bMental health status. ^cSocioeconomic status.

Measures

Life Engagement Test

The Life Engagement Test (LET; Scheier et al., 2005) was designed to measure the extent to which an individual engages in activities that they personally value. This concept of engagement measured by the LET most closely resembles the kind of engagement that was also involved in the behavioral activation intervention studies contained in the systematic review informing the current study (Parra et al., 2019; Robatmili et al., 2015; Takagaki et al., 2016, 2018; Tuck et al., 2022a). The LET comprises a 5-point Likert scale with responses ranging from “strongly disagree” to “strongly agree.” It contains three positively scored and three negatively scored items. The LET was originally validated with two samples of tertiary undergraduate students, demonstrating Cronbach’s alpha coefficients of α = .79 and α = .72, respectively.

Mindful Attention Awareness Scale–Short

The Mindful Attention Awareness Scale–Short version (MAAS-S; Osman et al., 2016) is a brief version of the 15-item Mindfulness Attention Awareness Scale (Brown & Ryan, 2003). The MAAS-S was designed to measure the concept of paying attention to the present moment (Osman et al., 2016). It was developed to reduce participant fatigue in research studies. The MAAS-S was chosen for use in the current study because it most closely represents the concept of paying attention to the present moment, which was utilized in the attention training interventions reported in the systematic review informing this study (Callinan et al., 2015; Myhr et al., 2019; Tuck et al., 2022a). The MAAS-S contains a 6-point Likert rating scale with responses ranging from “almost always” to “almost never.” The original validation of the MAAS-S was conducted with tertiary undergraduate students and resulted in a validity measure (Pearson’s rho) of 0.88.

Kessler 10-item Psychological Distress Inventory

The Kessler 10-item Psychological Distress Inventory (K-10; Kessler et al., 2003) was designed as a short screening instrument for nonspecific symptoms of psychological distress. The K-10 consists of two subscales measuring symptoms of depression and anxiety, respectively. The K-10 utilizes a 6-point Likert scale ranging from “none of the time” to “all of the time” and asks participants to choose the response indicating how they have felt over the past 4 weeks. The K-10 was shown to be the best short screening instrument for use with Australian adults (Furukawa et al., 2003) and is therefore widely used in Australia by researchers and clinicians as a short screening instrument for this purpose (Beyond Blue, 2024). One of the original validation studies of the K-10 utilized an Australian Government National Survey of Mental Health and Well-Being, revealing a Cronbach’s alpha coefficient of α = .92.

Brief Resilience Scale

The Brief Resilience Scale (BRS) is a short scale designed to measure the concept of an individual’s ability to bounce back or recover from stress (Smith et al. 2008). A review of existing resilience measurement scales found that of the existing scales, the BRS, the Connor-Davidson Resilience Scale (CD-RISK; Connor & Davidson, 2003), and the Resilience Scale for Adults (RSA; Friborg et al., 2003) had the best psychometric ratings (Velickovic et al., 2020). However, the CD-RISK and RSA rely on measuring subfactors that were shown to promote resilience in the populations they were validated with and may not translate to other contexts (Green et al., 2014; Gucciardi et al., 2011). This indicates that, of these three instruments, the BRS is the most appropriate for use with Australian participants. The BRS utilizes a 5-point Likert scale asking participants to rate their ability to bounce back from stress, ranging from “strongly disagree” to “strongly agree.” The BRS contains three positively worded and three negatively worded items. Items 2, 4, and 6 are reverse-scored. The BRS was originally validated using two samples of undergraduate students, resulting in Cronbach’s alpha coefficients of α = .84 and α = .87, respectively.

5-Item World Health Organization Well-Being Index

The 5-Item World Health Organization Well-Being Index (WHO-5, Staehr-Johansen, 1998) was developed from the 10-Item World Health Organization Well-Being Index (WHO-10, Bech et al., 1996). The WHO-5 contains only the positively worded items from the WHO-10 in order to measure only the concept of positive mental health (Staehr-Johansen, 1998). The WHO-5 has been proven to be the most useful existing instrument to measure subjective well-being because it is unrelated to symptoms of any specific disorder (Hall et al, 2011). The WHO-5 uses a 6-point Likert rating scale ranging from “all of the time” to “at no time.” A validation study of the WHO-5 conducted with North American tertiary education students resulted in a Cronbach’s alpha coefficient of .86 (Downs et al., 2017).

Procedure

Ethical clearance for the current study was provided by Monash University Human Research Ethics Committee (Project ID: 25241). Participants completed the survey online via the Qualtrics survey platform. The explanatory statement on the first page of the online survey included a fully detailed explanation of the study followed by information about consenting to participate, potential benefits and risks, and support services if adversely affected by participation in the study. Eligibility criteria indicated that participants must be aged 18 years or older, living in Australia, and enrolled in a tertiary education institution in one of the eight states and territories within Australia. Clicking the link to begin the survey indicated consent to participate. Data were collected for 5.5 months, beginning on September 6, 2020, and ending on February 22, 2021. A total of 1,262 participants enrolled in the study and 1,075 completed the entire survey, for a completion rate of 85.10%. Participants could opt to enter a prize draw for one of one hundred $20 shopping vouchers once they completed the survey.

The current study follows Strengthening the Reporting of Observational Studies in Epidemiology (strobe.org, n.d.) reporting guidelines. The study was registered on the Centre for Open Science’s (n.d.) forum website. The participant questionnaire and explanatory statement for the study are available at https://osf.io/8wc7s/?view_only=68b52e9d0bdd4ceeaae1e1dc7cb7f4b9. Data for this study are available upon reasonable request.

Data cleaning and analyses

Participants who did not complete at least the demographic questions and first inventory and were omitted from the analysis, resulting in the loss of 189 participants. A further three participants were omitted for repetitive response styles. Additionally, the trans/nonbinary category was automatically deleted by SPSS from analyses involving gender categories due to its small sample size. All data analyses were performed using SPSS version 27.0 (IBM Corp, 2020).

Scores of all of the measures were calculated according to author instructions. Reliability scores for all measures are given in Table 3. The SPSS explore procedure was utilized to assess normality of the data and indicated nonnormal distribution for all variables. However, skewness and kurtosis values were between 1 and −1 for all variables, which should not negatively impact the results of the analysis. Additionally, analyses with more than 30 cases are generally robust to nonnormal data (Tabachnick & Fidell, 2018). A scatterplot matrix revealed no bivariate outliers for any variable, resulting in the retention of the 1,072 participants.

To address the research question of whether paying attention to the present moment (mindfulness) and taking part in personally meaningful and enjoyable activities (engagement) are associated with well-being, resilience, and the absence of distress in tertiary education students, correlation analyses were used to determine relationships between engagement and mindfulness and symptoms of depression and anxiety, well-being, and resilience, along with the demographic variables age, gender, previous diagnosis of mental illness, mental health care status, socioeconomic status, sexual orientation, and international student status.

To address the research question of whether multiple resilience factors explain more of the variance associated with mental health outcomes in tertiary education students compared to a single resilience factor, hierarchical multiple regression analyses were performed to demonstrate the amount of variance associated with engagement and mindfulness separately with symptoms of depression and anxiety and psychological well-being and resilience. Age, gender, previous diagnosis of mental illness, mental health care status, socioeconomic status, sexual orientation, and international student status were entered into the hierarchical regression analyses as control variables.

Results

Bivariate correlations

Correlations were calculated to determine the relationships among all of the study variables. These indicate significant relationships between engagement and mindfulness separately with symptoms of depression and anxiety as well as well-being and resilience. Age was significantly correlated with all of the study variables apart from well-being. Gender, previous diagnosis of mental illness, mental health status, and sexual orientation were significantly correlated with depression (r_pb(1045) = .126, p < .001), (r_pb(1046) = .307, p < .001), (r_pb(1003) = .289, p < .001), (r_pb(1052) = .071, p = .021); anxiety (r_pb(1048) = .100, p = .001), (r_pb(1049) = .293, p < .001), (r_pb(1005) = .266, p < .001), (r_pb(1055) = .075, p = .015); well-being (r_pb(1048) = −.137, p < .001), (r_pb(1049) = −.282, p < .001), (r_pb(1005) = −.291, p < .001), (r_pb(1055) = −.073, p = .018); and resilience (r_pb(1062) = −.194, p < .001), (r_pb(1063) = −.283, p < .001), (r_pb(1018) = −.234, p < .001), (r_pb(1069) = −.117, p < .001). Socioeconomic status was significantly associated with anxiety (r_pb(1043) = −.071, p = .022).

The strength of the relationships among all numerical study variables is displayed in Table 3.

Hierarchical multiple regression analyses

Hierarchical multiple regression analyses were performed to determine whether the amount of variance attributable to the relationships between engagement and mindfulness with symptoms of depression, anxiety, well-being, and resilience were additive. Bivariate correlations between the variables used in the analysis are shown in Table 2, and the regression statistics for the four hierarchical regression analyses are displayed in Tables 3, 4, 5, and 6.

Table 2. Scale Reliability and Bivariate Correlations Among All Study Variables.

Variable	α	1	2	3	4	5	6	7
1. Age	NA	1
2. Eng.	.85	.149**	1
3. Mind.	.87	.145**	.397**	1
4. Depress.	.90	−.101**	−.568**	−.444**	1
5. Anx.	.81	−.134**	−.334**	−.439**	.657**	1
6. Well.	.77	.022	.601**	.333**	.665**	−.473**	1
7. Res.	.88	.061*	.394**	.290**	−.509**	−.416**	.481**	1

Note. *Significant at p < .005. **Significant at p < .001.

Eng. = engagement; Mind. = mindfulness; Depress. = depressive symptoms; Anx. = anxiety symptoms; Well. = Well-being; Res. = resilience.

Table 3. Summary of Hierarchical Analysis for Engagement and Mindfulness on Depression.

Variables	B	CI	SE	β	t	R^2adj
Control variables						.15
Age	−.08	−.12, −.04	.02	−.12	−4.08**
Gender	.86	.18, 1.55	.35	.07	2.47*
Diagnosis^a	.11	−.54, .76	.33	.01	.33
Status^b	.59	−.12, 1.29,	.36	.05	1.65
Socio.	2.53	1.82, 3.23	.36	.23	7.00**
Orient.	2.39	1.58, 3.21	.42	.19	5.76**
Student^c	1.09	.39, 3.21	.35	.09	3.08*
Step 2						.39
Engagement	−.63	−.69, −.57	.03	−.51	−19.66**
Step 3						.43
Engagement	−.54	−.61, −.47	.03	−.44	−16.14**
Mindfulness	−.23	−.29, −.17	.03	−.21	−7.57**

Note. **p < .001, *p < .05.

^aPrevious diagnosis of mental illness. ^bMental health status. ^cInternational student status.

CI = confidence interval; Socio. = socioeconomic status; Orient. = sexual orientation.

Table 4. Summary of Hierarchical Analysis for Engagement and Mindfulness on Anxiety.

Variables	B	CI	SE	β	t	R^2adj
Control variables						.14
Age	−.06	-.09. −.04	.01	−.16	−5.27**
Gender	.43	.01, .85	.21	.06	2.02*
Diagnosis^a	1.46	1.03, 1.89	.21	.22	6.67**
Status^b	1.28	.79, 1.78	.25	.17	5.08**
Socio.	.24	−.15, .64	.20	.04	1.21
Orient.	.46	.03, .89	.20	.06	2.12
Student^c	.12	−.30, .54	.21	.02	.54
Step 2						.20
Engagement	−.19	−.23, −.15	.02	−.26	−8.59**
Step 3						.27
Engagement	−.11	−.16, −.07	.02	−.15	−4.93**
Mindfulness	−.19	−.23, −.15	.02	−.30	−9.44**

Note. **p < .001, *p < .05.

^aPrevious diagnosis of mental illness. ^bMental health status. ^cInternational student status.

CI = confidence interval; Socio. = socioeconomic status; Orient. = sexual orientation.

Table 5. Summary of Hierarchical Analysis for Engagement and Mindfulness on Well-Being.

Variables	B	CI	SE	β	t	R^2adj
Control variables						.12
Age	.03	−.01, .06	.02	.05	1.53
Gender	−.86	−1.49, −.23	.32	−.08	−2.68*
Diagnosis^a	−2.05	−2.70, −1.40	.33	−.20	−6.16**
Status^b	−2.15	−2.90, −1.40	.38	−.19	−5.61**
Socio.	−.10	−.70, .493	.30	−.01	−.34
Orient.	−.49	−1.14, .16	.33	−.05	−1.48
Student^c	−.51	−1.15, .13	.33	−.05	−1.57
Step 2						.43
Engagement	.65	.59, .70	.03	.58	23.01**
Step 3						.44
Engagement	.62	.56, .68	.03	.55	20.50**
Mindfulness	.07	.02, .13	.03	.07	2.70*

Note. **p < .001, *p < .05.

Table 6. Summary of Hierarchical Analysis for Engagement and Mindfulness on Resilience.

Variables	B	CI	SE	β	t	R^2adj
Control variables						.13
Age	.06	.02, .09	.02	.10	3.16*
Gender	−1.48	−2.11, −.86	.03	−.14	−4.65**
Diagnosis^a	−2.17	−2.81, −1.52	.03	−.22	−6.62**
Status^b	−1.47	−2.21, −.73	.04	−.13	−3.89**
Socio.	.29	−.30, .88	.03	.03	.97
Orient.	−1.00	−1.65, −.36	.03	−.09	−3.06*
Student^c	−.60	−1.23, .02	.03	−.06	−1.89
Step 2						.24
Engagement	.38	.32, .45	.03	.34	11.87**
Step 3						.26
Engagement	.34	.27, .41	.04	.30	9.75**
Mindfulness	.11	.05, .17	.03	.11	3.60**

Note. **p < .001, *p < .05.

^aPrevious diagnosis of mental illness. ^bMental health status. ^cInternational student status.

CI = confidence interval; Socio. = socioeconomic status; Orient. = sexual orientation.

The first hierarchical multiple regression revealed that the control variables age, gender, previous diagnosis of mental illness, mental health status, socioeconomic status, sexual orientation, and international student status contributed significantly to the model, F(7, 967) = 25.57, p < .001, and accounted for 15.4% of variance in depressive symptoms. Age, t (973) = −4.08, p < .001; gender, t (973) = 2.47, p = .014; previous diagnosis of mental health disorder, t (973) = 7.00, p < .001; mental health care status, t (973) = 5.76, p < .001; and international student status, t (973) = 3.08, p = .002 all significantly predicted depressive symptoms. Adding engagement to the model at step two contributed significantly to the regression model, F (8, 960) = 79.03, p < .001, and accounted for an additional 24.3% of the variation of depressive symptoms. Introducing mindfulness to the model at step three explained an additional 3.4% of variation in symptoms of depression, and this change in R² was significant, F (9, 951) = 80.71, p < .001. Engagement and mindfulness were both significant predictors of symptoms of depression when included together in the model and accounted for 28% of the variance in depressive symptoms. Calculating Bayesian information criterion (BIC) to compare models with the coefficients for engagement and mindfulness constrained to be equal or unconstrained resulted in a BIC difference of 54.75, favoring the unconstrained model. Thus, very strong evidence (BIC difference > 10) indicates that engagement and mindfulness do not have the same association with depression symptoms (Raftery, 1995). Regression statistics for the first hierarchical multiple regression are displayed in Table 3.

The second hierarchical multiple regression revealed that the control variables age, gender, previous diagnosis of mental illness, mental health status, socioeconomic status, sexual orientation, and international student status contributed significantly to the model, F (7, 969) = 23.52, p < .001, and accounted for 14.5% of variance in depressive symptoms. Age, t (975) = −5.27, p < .001; gender, t (975) = 2.06, p = .044; previous diagnosis of mental health disorder, t (975) = 6.67, p < .001; mental health status, t (975) = 5.08, p < .001; and sexual orientation, t (975) = 2.12, p = .034, all significantly predicted anxiety symptoms. Adding engagement to the model at step two contributed significantly to the regression model, F (8, 962 = 73.84, p < .001, accounting for an additional 6.1% of the variation in symptoms of anxiety. Introducing mindfulness to the model at step three explained an additional 6.8% of variation in symptoms of anxiety, and this change in R² was significant, F (9, 953) = 89.18, p < .001. Engagement and mindfulness were both significant predictors of symptoms of anxiety when included together in the model and accounted for 12.9% of the variance in anxiety symptoms. Calculating BIC to compare models with the coefficients for engagement and mindfulness constrained to be equal or unconstrained resulted in a BIC difference of 44.10, favoring the unconstrained model. Very strong evidence (BIC difference > 10) indicates that engagement and mindfulness do not have the same association with symptoms of anxiety (Raftery, 1995). Regression statistics for the second hierarchical multiple regression are displayed in Table 4.

The third hierarchical multiple regression revealed that the control variables age, gender, previous diagnosis of mental illness, mental health status, socioeconomic status, sexual orientation, and international student status contributed significantly to the model, F (7, 969) = 20.79, p < .001, and accounted for 13.1% of variance in well-being. Gender, t (975) = −2.68, p = .008; previous diagnosis of mental health disorder, t (975) = −6.16, p < .001; and mental health care status, t (975) = 5.61, p < .001, all significantly predicted anxiety symptoms. Adding engagement to the model at step two contributed significantly to the regression model, F (8, 962) = 529.57, p < .001, and accounted for an additional 31% of the variation in level of well-being. Introducing mindfulness to the model explained an additional 0.4% of variation in level of well-being, and this change in R² was significant, F (9, 957) = 7.30, p = .007. Engagement and mindfulness were both significant predictors of level of well-being when included together in the model and accounted for 31.4% of the variance in well-being. Calculating BIC to compare models with the coefficients for engagement and mindfulness constrained to be equal or unconstrained resulted in a BIC difference of 21.10, favoring the unconstrained model. This indicates that engagement and mindfulness do not have the same association with well-being, with very strong evidence (BIC difference >10; Raftery, 1995). Regression statistics for the third hierarchical multiple regression are displayed in Table 5.

The final hierarchical multiple regression revealed that the control variables age, gender, previous diagnosis of mental illness, mental health status, socioeconomic status, sexual orientation, and international student status contributed significantly to the model, F (7, 982) = 22.70, p < .001, and accounted for 13.9% of variance in depressive symptoms. Age, t (988) = 3.16, p < .001; gender, t (975) = −4.65, p < .001; previous diagnosis of mental health disorder, t (988) = 6.62, p < .001; mental health status, t (988) = 3.89, p < .001; and sexual orientation, t (988) = −3.06, p = .002, all significantly predicted anxiety symptoms. Adding engagement to the model at step two contributed significantly to the regression model, F (8, 975) = 140.57, p < .001, and accounted for an additional 10.8% of the variation in level of resilience. Introducing mindfulness to the model at step three explained an additional 1% of variation in level of resilience, and this change in R² was significant, F (9, 957) = 12.93, p < .001. Engagement and mindfulness were both significant predictors of level of resilience, and together they accounted for 11.8% of the variance in resilience. Calculating BIC to compare models with the coefficients for engagement and mindfulness constrained to be equal or unconstrained resulted in a BIC difference of 95.17, favoring the unconstrained model. Very strong evidence (BIC difference > 10) indicates that engagement and mindfulness do not have the same association with resilience (Raftery, 1995). Regression statistics for the final hierarchical multiple regression are displayed in Table 6.

Discussion

The current project attempted to determine how to improve psychological resilience in tertiary education students by investigating whether engaging in personally meaningful and enjoyable activities (engagement) and paying attention to the present moment (mindfulness) are associated with symptoms of depression and anxiety and well-being and resilience. Existing theories of psychological resilience state that stress management resources help to build resilience by improving coping with potentially distressing events (Fredrickson, 2001; Hamby et al., 2018). It is further believed that improvements in stress coping afforded by individual stress management resources have compounding effects on psychological functioning. Hamby et al. (2018) state that, similar to the way in which multiple stressors have cumulative negative effects on functioning, coping resources also have cumulative effects on improving mental health and resilience outcomes that go beyond the effect of single resources. Therefore, the current project investigated the relationships between engagement and mindfulness with symptoms of depression and anxiety along with well-being and resilience in tertiary education students.

The results of the bivariate correlations demonstrated significant relationships between the coping resources of engagement and mindfulness separately with symptoms of depression and anxiety along with well-being and resilience. These results provide support for the two resilience theories discussed in the introduction, which state that coping resources promote improved resilience and coping with distress (Fredrickson, 2001; Hamby et al., 2018). The results of the current study also align with the previously mentioned systematic review by Tuck et al. (2022a), which found that both engagement and mindfulness are useful for promoting mental health and resilience outcomes in tertiary education students. Therefore, the current study supports the claim of existing resilience theory (Fredrickson, 2001; Hamby et al., 2018) as well as previous empirical findings (Tuck et al., 2022a) that interventions aimed at improving these coping resources may reduce distress and improve levels of psychological well-being and resilience in tertiary students.

In the current study, the bivariate relationships between symptoms of depression, well-being, and resilience were stronger with engagement than with mindfulness. Conversely, the relationship between mindfulness and symptoms of anxiety was stronger than that between engagement and symptoms of anxiety. These results again align with the findings of the previously mentioned systematic review (Tuck et al., 2022a) in which interventions that improved paying attention to the present moment (mindfulness; Callinan et al., 2015; Myhr et al., 2019) had the greatest impact on symptoms of anxiety and engaging in personally meaningful and enjoyable activities (engagement; Parra et al., 2019; Robatmili et al., 2015; Takagaki et al., 2016, 2018) had the greatest impact on symptoms of depression of any of the reviewed intervention types. Therefore, the combined results of Tuck et al. (2022a) and the current study indicate that engagement may be useful for improving symptoms of depression, well-being and resilience in tertiary education students and that mindfulness may be useful for eliminating symptoms of anxiety.

The results of the hierarchical multiple regression analyses demonstrated that the amount of variance in mental health and resilience outcomes associated with coping resources is additive. When engagement and mindfulness were added to the regression equations, the amount of variance explained by the coping resources increased for all outcomes being investigated in each of the models (i.e., depression, anxiety, well-being, and resilience) increased. Additionally, when added to each model, engagement and mindfulness both contributed significant amounts of variance in addition to that accounted for by the control variables. These results provide further support for Hamby et al.’s (2018) poly-strengths theory, which states that coping resources have additive effects on relieving distress and improving functioning after potentially distressing events. Taken together, the results suggest that interventions that provide tertiary students with opportunities to learn multiple ways of managing stress may have greater effects on relieving distress and improving well-being and resilience than those that rely on a single intervention type.

Mirroring the results of the bivariate correlations, the hierarchical regression analyses indicated that symptoms of depression, well-being, and resilience have a stronger relationship with engagement than with mindfulness and that mindfulness has a stronger relationship with symptoms of anxiety than engagement. The relationships may be explained by the results of previous studies that demonstrated that engaging in personally meaningful and enjoyable activities (engagement) improves positive affect, and positive affect reduces symptoms of depression and improves resilience (Young et al., 2018). Young et al. (2018) explain that this occurs because engaging in personally meaningful activities promotes a greater sense of goal achievement and meaning in life, which in turn improves mental health outcomes. Therefore, it is likely that a similar process occurs when tertiary students engage in activities that are personally meaningful to them, given that the engagement interventions informing this study involved activities based on an individual’s goals in life (Tuck et al., 2022a).

The results of the current study demonstrating that paying attention to the present moment (mindfulness) is associated with symptoms of anxiety may be explained by the results of a previous systematic review investigating the impact of mindfulness on mental health outcomes (Boyd et al., 2018). The review found that the amount of attention being paid to potential stressors is negatively associated with both peri- and post-event distress (Thompson et al., 2011) as well as levels of posttraumatic stress symptoms (Gallegos et al., 2015, Goldsmith et al., 2014, Possemato et al., 2016). The authors of the systematic review concluded that this was possibly due to an increased ability to process potentially traumatic events when greater amounts of attention are being paid to them (Boyd et al., 2018). Therefore, higher levels of mindfulness are likely to reduce distress and posttraumatic stress disorder symptomology in tertiary students by promoting more effective processing of threatening events.

Theoretical and practical implications

The finding that improvements in coping resources are associated with lower levels of distress and greater levels of well-being and resilience in tertiary education students adds support for Fredrickson (2001) and Hamby et al.’s (2018) resilience theories. This is an important finding because preventive mental health literature has long been hampered by the lack of a unifying conceptual framework. This in turn has made it difficult to conduct preventive mental health research, which has led to much of the work in this field being conducted atheoretically (Hamby et al., 2018). The findings of the current study support the claim of existing resilience theory (Fredrickson, 2001; Hamby et al., 2018) as well as previous empirical findings (Tuck et al., 2022a) that interventions aimed at improving coping resources in tertiary students may reduce distress and improve levels of well-being and resilience. The additional finding that the amount of variance in mental health and resilience outcomes associated with coping resources is additive provides further support for Hamby et al.’s (2018) poly-strengths theory. This demonstrates that poly-strengths theory is a useful framework for designing preventive mental health interventions for tertiary education students.

The existence of a useful framework for understanding how to prevent the development of disorder in tertiary students may be utilized to improve preventive mental health efforts by educational institutions and other stakeholders working in this area. Stress management interventions that aim to improve resilience in tertiary education students are an obvious target for preventive mental health campaigns. Considering that the coping resources investigated in the current study were obtained from a systematic review of intervention studies investigating how to reduce distress among tertiary students (Tuck et al., 2022a) and that the current study found that these practices are also associated with resilience, studies investigating the impact of practices designed to improve mental health outcomes are potentially useful for promoting resilience in this population. The current study also provides a starting point for exercises that could be included in resilience interventions for tertiary education students by demonstrating the aforementioned relationships between engagement and mindfulness with mental health and resilience outcomes.

Limitations and future directions

The main limitation of this study is that it utilizes cross-sectional data. This leaves the relationships between the coping and outcome variables included in the study open to questions of causality. However, the fact that the relationships between these variables is similar to those found in the studies from the systematic review informing the current study (Tuck et al., 2022a) does provide evidence that the coping variables most likely are impacting the stated mental health outcomes in this study. Additionally, the sample in the current study is composed 100% of tertiary education students studying in Australia. Therefore, while this is a strength when investigating the practices that will improve resilience and mental health outcomes in this population, it is not certain whether they are applicable to other populations. As stated in the introduction, previous research has demonstrated that the factors that promote resilience may not generalize across different populations (Green et al., 2014; Gucciardi et al., 2011). Therefore, further research in this area could utilize experimental studies to investigate the impact of interventions designed to improve coping with stress in tertiary education students as well as in other target populations. It would also be useful if these intervention studies included multiple mental health outcomes in the same way as the current study measured distress, well-being, and resilience due to the fact that there is currently no gold-standard measure of resilience (Fisher & Law, 2020). Therefore, measuring symptoms of distress, well-being, and resilience would give a more accurate picture of the impact of intervention programs compared to relying on resilience measures alone.

Conclusion

The findings of the current study support the assertion made by the broaden and build (Fredrickson, 2001) and poly-strengths (Hamby et al., 2018) theories of resilience that improvements in coping with stress also improve resilience. As demonstrated in the results of the study, engaging in personally meaningful and enjoyable activities (engagement) and paying attention to the present moment (mindfulness) are negatively associated with symptoms of depression and anxiety and positively associated with psychological well-being and resilience. The current study also found that, in support of Hamby et al.’s (2018) poly-strengths theory, the amount of variance in mental health and resilience outcomes associated with coping resources is additive. These findings support that poly-strengths theory provides a conceptual framework that can be used to design resilience training programs for tertiary education students in order to address the high levels of mental health disorder in this population. Addressing these high levels of mental disorder in tertiary students would in turn help to reduce dropout rates and financial losses incurred from noncompletion of courses and prevent escalation of mental health problems and potential suicides among students. These findings provide a starting point for tertiary education institutions looking to introduce preventive mental health programs by demonstrating the relationship between engagement and mindfulness with mental health outcomes in a university sample. Further work in this area could investigate the impacts of these and other activities on mental health outcomes in various other populations via experimental studies in order to build greater understanding of the types of interventions that should be included in programs tailored for particular groups.

Declaration of interest

There are no conflicts of interest to declare in the preparation of this manuscript.