Well-Being amongst College Students during COVID-19 Pandemic: Evidence from a Developing Country

College students face unique challenges that the consequences of COVID-19 might aggravate. To explore the pandemic’s consequences on college students’ well-being, we conducted an online survey with 634 students from a private university in Cali, Colombia. The study sought to assess students’ well-being due to COVID-19, and to explore the mediating effects of optimism, gratitude, and emotional closeness on college students’ well-being. Results showed that COVID-19 affected students’ mental health and well-being. Being optimistic and grateful mediated with life satisfaction and happiness. Optimism, emotional closeness, and gratitude also mediated the negative effect of fear of infection and the pandemic’s impact on students’ academic performance. The results of this analysis will promote discussion of the implementation of coping strategies to help students thrive, promote resilience, and contribute to students’ well-being and better mental health.


Introduction
The restrictions associated with the containment of COVID-19 are, without a doubt, the "largest physiological experiment even conducted" [1]. With over two million deaths worldwide and more than 100 million confirmed cases [2] the pandemic, the measures taken to control the spread of the virus, and the global economic consequences, are imposing a heavy toll on our mental health. This crisis and the pervasiveness of feelings such as fear and uncertainty have affected the entire global population to a viable extent.
Quarantine, the most used measure to contain the pandemic contagious rate, has a negative psychological effect: producing stress, confusion, and anger. It also adds stressors, such as fear, frustration, boredom, and financial concerns [3]. For some population groups, such as children and young adults, the pandemic may have more severe consequences on mental health, even more significant than the consequences on their physical health [2].
College students face unique challenges that negatively affect their mental health. Before COVID-19 unfolded, there were reports of the prevalence of poor mental health worldwide amongst this population. The rates of anxiety and depression are steadily increasing. Information from the US indicates a spike of 63% of young adults (18 years and over) reporting symptoms associated with significant depression between 2005 and 2017, mainly affecting girls [4]. Worldwide, one in five college students experiences one or more diagnoses of mental disorders [5].
College students can be subject to stressful situations. The pressures of establishing a career path, academic demands, the transition from adolescence to adulthood, peer pressures, and many other factors affect their general well-being and mental health [6]. Before the pandemic, there was a significant number of reports showing that the increase in stress, depression, and reduction in the overall well-being amongst college students was prevalent, and was the subject of different interventions from the public health system

Optimism
Optimism is defined as an attitude associated with expectations of the future, which individuals regard as positive to his/her advantage [25]. Optimism is fundamental in times of uncertainty because optimistic people tend to cope better with difficult situations [26]. Dispositional optimism helps people to cope with stress because optimistic people manage, in a better way, stressful and challenging events, accepting them instead of trying to wish them away [27]. Optimistic people also try to actively resolve problems compared to pessimists that tend to avoid coping strategies [28][29][30][31]. Among college students, optimistic individuals are more committed to social, academic, and personal activities, and positive psychological functioning [32]. Optimism can help college students to adjust in a better way during stressful situations [33], reduces feelings of loneliness [32], and is a predictor of life satisfaction and positive emotions [34]. It has also been found that students with higher levels of optimism have higher retention rates and higher grade point averages because they have higher levels of motivation and are better at adjusting and adapting to stressful situations [27,35]. Optimism contributes to subjective well-being, and college students with high optimism tend to experience positive emotions and higher life satisfaction [36,37]. Optimism also proved to be pivotal during the lockdown for college students. Evidence from Spain shows that optimism moderated educational achievement and improved grades during the pandemic for the population analyzed [38].

Gratitude
Most authors define gratitude as a state where grateful individuals have a high level of appreciation for life and a subjective feeling of thankfulness [36,37]. Gratitude is also defined as a trait that can vary among individuals [22,39]. Wood [40] suggested that gratitude should be conceptualized as a life orientation, where grateful people appreciate the world's positive things. Gratitude is an emotion that is felt towards another person or thing, and it is a disposition to express and feel the emotion of thankfulness across situations [36]. Gratitude brings positive emotional states [41], is related to prosocial behavior [42], satisfaction and acceptance of difficult circumstances [43], happiness [44], and long-term subjective well-being [39,40]. In college students, being grateful contributes to their happiness [45], improves retention and academic performance [46], and promotes social support [47]. In the context of the pandemic, the evidence available shows that gratitude lessened mental health difficulties and promoted resilience [48], and that interventions fostering gratitude were effective in improving the mental health of college students during the crisis [49].

Emotional Closeness-Relationships
The presence and quality of supportive relationships are necessary for every aspect of the life of any individual. It is crucial to build relationships with people who can offer advice, help, or assistance during different life cycle episodes [50]. Supportive relationships facilitate coping [24], moderate the stress-depression relationship [51], and predicts a reduction in loneliness [52]. Relations with family, friends, and romantic partners are related to subjective well-being [53]. Among college students, those who spent time with family and friends have higher perceived happiness levels [54]. Social support and social connections help students adjust better to college [55] also relate to positively perceived happiness and well-being [56,57]. Emotional closeness and social connectedness were pivotal for wellbeing during the pandemic, providing a strong sense of being in control and improving emotional well-being [58]. In the higher education context, strategies aimed at grouping students showed that group support provided opportunities for improving academic achievement and outlets for emotional support [59].
This study aimed to contribute to the bulk of the literature focusing on students' wellbeing and life satisfaction by analyzing the mediating effect of optimism, gratitude, and emotional closeness, on well-being during the pandemic caused by COVID-19. Additionally, this study contributes to the literature by providing evidence from a developed country (Colombia), where information about college students' well-being and the pandemic are scarce. This analysis is composed of six sections, including this introduction. Section 2 discusses the analytical framework and research questions that guide our analysis. In the Section 3, we present the dataset's details to answer the research questions and the methods. The Section 4 presents the results and ends with the discussion and conclusions of this research.

Analytical Framework and Hypothesis
Our analytical framework assumes that COVID-19 represents a stressful event for college students, which reduces their overall well-being (life satisfaction) and increases the prevalence of emotions such as worry and depression. However, coping strategies such as a positive attitude, being grateful, and having close relationships with peers and family can mediate college students' negative emotional states during the current crisis. Figure 1 presents the analytical framework guiding this study.

Research Questions and Aims
Coping strategies such as practicing gratitude, staying close to loved ones and being optimistic exert protective effects against worry and depression. The literature's significant bulk shows a positive correlation between those coping approaches and better mental health [60]. This research aimed to evaluate college students' coping strategies to assess

Research Questions and Aims
Coping strategies such as practicing gratitude, staying close to loved ones and being optimistic exert protective effects against worry and depression. The literature's significant bulk shows a positive correlation between those coping approaches and better mental health [60]. This research aimed to evaluate college students' coping strategies to assess how those strategies help them cope during confinement. This analysis has two purposes: first, to assess students' well-being and mental health during COVID-19; second, to explore the mediating effects of optimism, gratitude, and supportive relationships on college students' well-being and mental health. Three hypotheses guided our analysis: Hypothesis 1 (H1): Some of the negative consequences of COVID-19 on students' psychological well-being will be attenuated for grateful students insofar as grateful students cope better with the aftermath of COVID-19.

Hypothesis 2 (H2):
Some of the negative consequences of COVID-19 on students' psychological well-being will be attenuated for optimistic students because optimistic students tend to have a positive outlook towards difficult circumstances.

Hypothesis 3 (H3):
Some of the negative consequences of COVID-19 on students' psychological well-being will be attenuated by closer relationships with parents and friends, since students who have at their disposal supportive relationships are better at coping with stress and difficulties.
To answer these questions, we conducted an online survey in April 2020, after one month of quarantine in Colombia. The respondents were 634 students from a private university in Cali, Colombia. The study asked students about their general well-being and the prevalence of feelings of worry and depression. The survey also inquired about students' attitudes toward optimism and gratefulness.

Data and Methods
This analysis was derived from an online survey conducted between mid-April and early May of 2020, one month after the quarantine unfolded in Colombia. The survey was uploaded into "Typeform," a web server for online polling and surveys. For distributing the survey, researchers used a convenience sampling strategy and disseminated the online survey through four channels: (i) the university's social networks; (ii) emails sent from professors to their students; (iii) distribution lists; and (iv) students' associations. Survey participation was voluntary, and before starting the survey, the researchers provided a complete description of the research aims and data use, students participating consented to use the aggregated data for academic purposes. The survey was anonymous, and no personal information was collected to ensure complete anonymity. A total of 19% of students participating in the study were 16 or 17 years old (minors under Colombian legislation) at the moment of the survey. However, given the anonymous nature of the survey and their consent, we included their responses in the analysis. In total, 634 students between 16 and 24 completed the survey, 10% of the university's total undergrad population, making a satisfactory rate compared to the median web survey participation [61]. At the survey time, students had online synchronous classes, most living at their parent's homes. Their classes had the same content as face-to-face classes. One significant change was that the final evaluation was qualitative: 'approve or not approved' instead of numeric.
The survey competition took about ten minutes, and before the final questionnaire was released, the survey was piloted with 20 students. The ethics committee of Universidad Icesi approved the study (code # 278). Figure 2 presents the study design. the survey time, students had online synchronous classes, most living at their parent's homes. Their classes had the same content as face-to-face classes. One significant change was that the final evaluation was qualitative: 'approve or not approved' instead of numeric. The survey competition took about ten minutes, and before the final questionnaire was released, the survey was piloted with 20 students. The ethics committee of Universidad Icesi approved the study (code # 278). Figure 2 presents the study design. After survey competition, students downloaded a gratitude journal and a stress management diary designed for this study at the end of the survey as a reward for participation. We included the gratitude journal as a reward for participation due to the evidence indicating that writing about gratitude has a positive effect on immediate thoughts of participants [62].
The survey inquired about subjective well-being, mental health, and concerns about the consequences of COVID-19, optimism, gratitude, and emotional closeness with parents and friends. In total, the survey had 17 questions. Appendix A presents the questionnaire, and data is available with complete description of the metrics used at Mendely data repository [63].

Well-Being and Mental Health
In this study, we adhered to the World Health Organization's overall concept of mental health. This idea refers to mental health as a state of well-being in which individuals can cope with the stress of life and is not restricted to the diagnosis of a mental disorder [2]. To measure well-being and the prevalence of the most common negative emotions affecting mental health (worry and depression), we used the standardized and validated scale of core well-being measures [64].
Well-being is measured using subjective and experienced well-being variables. Measures of subjective well-being come from self-reported life satisfaction. We asked students, "how satisfied are you with all aspects of your life? Answers ranged from 0 to 10, with 0 referring to completely dissatisfied and a 10 to completely satisfied.
The estimation of experienced well-being comes from using one positive emotional state: happiness and two negative states: worried and depression that students experienced the day before of the survey. The following questions were asked: Overall, how happy did you feel yesterday? How worried did you feel yesterday? And how depressed did you feel yesterday? Students' answers were also reported on a scale from 0 to 10, with 0 the lowest and ten the highest [64]. After survey competition, students downloaded a gratitude journal and a stress management diary designed for this study at the end of the survey as a reward for participation. We included the gratitude journal as a reward for participation due to the evidence indicating that writing about gratitude has a positive effect on immediate thoughts of participants [62].
The survey inquired about subjective well-being, mental health, and concerns about the consequences of COVID-19, optimism, gratitude, and emotional closeness with parents and friends. In total, the survey had 17 questions. Appendix A presents the questionnaire, and data is available with complete description of the metrics used at Mendely data repository [63].

Well-Being and Mental Health
In this study, we adhered to the World Health Organization's overall concept of mental health. This idea refers to mental health as a state of well-being in which individuals can cope with the stress of life and is not restricted to the diagnosis of a mental disorder [2]. To measure well-being and the prevalence of the most common negative emotions affecting mental health (worry and depression), we used the standardized and validated scale of core well-being measures [64].
Well-being is measured using subjective and experienced well-being variables. Measures of subjective well-being come from self-reported life satisfaction. We asked students, "how satisfied are you with all aspects of your life? Answers ranged from 0 to 10, with 0 referring to completely dissatisfied and a 10 to completely satisfied.
The estimation of experienced well-being comes from using one positive emotional state: happiness and two negative states: worried and depression that students experienced the day before of the survey. The following questions were asked: Overall, how happy did you feel yesterday? How worried did you feel yesterday? And how depressed did you feel yesterday? Students' answers were also reported on a scale from 0 to 10, with 0 the lowest and ten the highest [64].

Concerns and Feelings about COVID-19 Consequences
The study inquired about the economic, health, and academic consequences of the coronavirus and the government's perception of social distancing measures. We used three questions in which students were asked about how much they agreed or disagreed, measured on a 0-10 scale: • I am concerned about the financial consequences of the coronavirus; • The probability of that a family member or I, acquire the virus is high; • I am concerned that my academic performance will be affected by the coronavirus.

Optimism
The survey included a short version of the originally ten item-long Life Orientation Test-Revised (LOT-R). Students were asked: (i) In uncertain times, I usually expect the best; (ii) I am always optimistic about my future; and (iii) Overall, I expect more good things to happen to me than bad. The score was on a scale from 0 (completely disagree) to 10 (completely agree). The Cronbach reliability for the scale was 0.82 (Table 1), similar to the Cronbach alpha of the original inventory 0.82 [28]. The mean of the score of these three items was the measure of students' optimism. Measures for gratitude come from a short version of a gratitude self-reported questionnaire, validated to Spanish [65]. The questions assessed individual differences in experiencing gratitude in daily life. Two items (3 and 5) in this scale had reversed scores. This measure included the following items: • I have so much in life to be thankful; • If I had to list everything that I felt grateful for, it would be a very long list; • When I look at the world, I don't see much to be grateful; • I am grateful to a wide variety of people; • Large amounts of time can go by before I feel grateful for something or someone.
Students' responses ranged from 0 to 10. The reliability of gratitude items was 0.90, showing good internal consistency for the instrument (Table 1).

Emotional Closeness-Relationships
The survey asked students to score from 0 to 10 "how emotionally close do you feel towards (1) your parent(s)/legal guardian(s) and (2) your friends; your significant other, and your college classmates. For this analysis, we defined two dichotomic variables: emotional closeness to parents; and emotional closes to friends.

Mediation Analysis
For analysis, we used a modifying Baron and Kenny's [66] approach to establish the mediation of each coping strategy on students' wellbeing. Following Mehmetoglu [67], the mediation effect was estimated following these steps. First, we fit a structural equation model (SEM) of coping strategy on students' wellbeing by controlling for pandemic consequence to estimate simultaneously direct (c) and indirect paths (a alpha, b beta) ( Figure 3). If either one path was not significant (or both were not significant) there was no mediation. Evidence of mediation was provided when both Pandemic Consequences → Coping Strategy and Coping Strategy → Wellbeing outcome coefficients were statistically significant. After that, Sobel's z test was calculated to estimate the relative sizes of the indirect (mediated) vs. direct paths. The results of no, partial or complete mediation were determined as follows: (i) if the Sobel's z test was significant and the direct path of Pandemic consequence → Wellbeing outcome was not, the mediation was full or complete; (ii) If both the z and the direct path of Pandemic consequence → Wellbeing outcome were significant, the mediation was partial; (iii) If the z was not significant but the direct path of Pandemic consequence → Wellbeing outcome was, the mediation was partial in the presence of a direct effect, (iv) If neither the z nor the direct path of Pandemic Consequence → Wellbeing outcome were significant, the mediation was partial in the absence of a direct effect. We also tested the coping strategy mediation hypothesis by using 634 Monte Carlo replications. Additional explanations of the testing mediation hypothesis can be found in Mehmetoglu [67].

Mediation Analysis
For analysis, we used a modifying Baron and Kenny's [66] approach to establish the mediation of each coping strategy on students' wellbeing. Following Mehmetoglu [67], the mediation effect was estimated following these steps. First, we fit a structural equation model (SEM) of coping strategy on students' wellbeing by controlling for pandemic consequence to estimate simultaneously direct (c) and indirect paths (a alpha, b beta) ( Figure  3). If either one path was not significant (or both were not significant) there was no mediation. Evidence of mediation was provided when both Pandemic Consequences → Coping Strategy and Coping Strategy → Wellbeing outcome coefficients were statistically significant. After that, Sobel's z test was calculated to estimate the relative sizes of the indirect (mediated) vs. direct paths. The results of no, partial or complete mediation were determined as follows: (i) if the Sobel's z test was significant and the direct path of Pandemic consequence → Wellbeing outcome was not, the mediation was full or complete; (ii) If both the z and the direct path of Pandemic consequence → Wellbeing outcome were significant, the mediation was partial; (iii) If the z was not significant but the direct path of Pandemic consequence → Wellbeing outcome was, the mediation was partial in the presence of a direct effect, (iv) If neither the z nor the direct path of Pandemic Consequence → Wellbeing outcome were significant, the mediation was partial in the absence of a direct effect. We also tested the coping strategy mediation hypothesis by using 634 Monte Carlo replications. Additional explanations of the testing mediation hypothesis can be found in Mehmetoglu [67]. For modeling, we used a medsem package in Stata 15 [67], which employs structural equation modeling to estimate the statistically significance of each path, controlling for age and gender. Coefficients were standardized. The effect of size of indirect effect of each coping strategy was estimated by the ratio of the indirect effect to the total effect ( * * ) and the ratio of the indirect effect to the direct effect ( * ).

Results
Students who participated in the study were, on average, 19.  For modeling, we used a medsem package in Stata 15 [67], which employs structural equation modeling to estimate the statistically significance of each path, controlling for age and gender. Coefficients were standardized. The effect of size of indirect effect of each coping strategy was estimated by the ratio of the indirect effect to the total effect (RIT = a * b (a * b)+c ) and the ratio of the indirect effect to the direct effect (RID = a * b c ).

Results
Students who participated in the study were, on average, 19.4 years old (std = 2.0), and 56% of the responders were female. To the question about how satisfied you with all aspects of your life are, students reported on average 6.8 (std = 1.8). Although it is not strictly comparable, yearly life satisfaction measures in Cali report constant life satisfaction scores of 8.5, almost two units above students' life satisfaction scores during the quarantine. This score of 8.5 has not changed, importantly, between 2014 and 2019 and is the same as national statistics [68].
Students reported being happy the day before on 5.9 (std = 2.3). This number contrasted with being worried, which scored high among students, 6.4 (std = 2.6) on average. Being depressed scored 4.5 (std = 2.9). As expected, correlations between life satisfaction and happiness were positive, and negative with feelings of worry and depression (Figure 4).
Taken all together, students were highly optimistic and grateful. Gratitude was the coping strategy with the highest score (7.8, std = 2.6). These strategies were correlated among them, suggesting that these measures were related to similar feelings that enable people to overcome unpleasant situations. In contrast, optimism and gratitude correlated positively with well-being variables. Table 2 presents the correlation scores for the variables of this study.
Students reported being happy the day before on 5.9 (std = 2.3). This number contrasted with being worried, which scored high among students, 6.4 (std = 2.6) on average. Being depressed scored 4.5 (std = 2.9). As expected, correlations between life satisfaction and happiness were positive, and negative with feelings of worry and depression ( Figure  4). Taken all together, students were highly optimistic and grateful. Gratitude was the coping strategy with the highest score (7.8, std = 2.6). These strategies were correlated among them, suggesting that these measures were related to similar feelings that enable people to overcome unpleasant situations. In contrast, optimism and gratitude correlated positively with well-being variables. Table 2 presents the correlation scores for the variables of this study.

The Mediating Effect of the Coping Strategies
Thes mediation effect of coping strategies was mainly observed for the relationship between the financial concerns raised by COVID-19 and students' life satisfaction (Table 3). Table 4 present the results of mediation analysis of coping strategies on positive and negative students subjective well-being outcomes. About 197% (RIT) of this relationship was mediated by optimism, 368% by gratitude, 213% by emotional closeness to parents and 230% by emotional closeness to friends. The ratio of the indirect effect of the coping strategy on the effect of financial consequences of COVID-19 (RID) was higher than 100%.  Regarding the mediating effect (Table 4) of the coping strategies in the relationship of life satisfaction and the fear of being infected by coronavirus, optimism and emotional closeness to friends turned out to be statistically significant. Gratitude played a partial mediating effect in this relationship, and emotional closeness to parents does not have a mediating effect at all. Concerning the mediating role of coping strategies in the relationship of the stress affecting student academic performance and life satisfaction, none of the coping strategies played a role in this relationship.
Turning to the relationship of financial pandemic consequences and happiness, optimism, and emotional closeness to parents and friends played a mediating role in this relationship. For the fear of being infected by the coronavirus and the stress affecting students' academic performance, none of the strategies showed a complete mediating contribution to the relationship. Similar results were obtained for the mediating effect of the coping strategies for being worried, except for the role of optimism on the pandemic's financial consequences relationship. In contrast, all coping strategies showed a mediating effect on the relationship between students' financial concerns and depression. For the fear of being infected by the coronavirus, and the stress affecting student academic performance, none of the coping strategies played a complete mediating effect.
These results may suggest that optimism, emotional closeness, and gratitude allow young people to include in their cognitive bandwidth the new reality imposed by the pandemic, in particular, for the financial concerns related to the pandemic. Nevertheless, the mediating role of the coping strategies on the fear of being infected by coronavirus, and the stress affecting student academic performance, was partial or non-existent.

Discussion
In line with the available evidence regarding the consequences of the pandemic on students' well-being and mental health [5,[10][11][12][13], we found that the pandemic had negative consequences on student life, associated with their subjective well-being. The pandemic caused by COVID-19 affected students by reducing their life satisfaction and feelings of happiness, and by increasing depression and worry [69][70][71][72]. However, correlations between well-being variables were positive and positively correlated to gratitude, emotional closeness, and optimism. Similar to other studies conducted during the pandemic [48,49,59], we found that being optimistic and grateful helped students reduce the negative effect of fear of infection and the pandemic's impact on students' academic performance, after adjusting for age and gender. Results showed that students who are more optimistic, closer to their parents and friends, and grateful can cope better with a crisis. Likewise, these positive attitudes and emotions increased well-being and reduced the prevalence of depression and worry.
This study faces several limitations. An important one is the lack of measures of well-being and mental health before the quarantine unfolded, limiting the possibility of better capturing the effects attributed to the pandemic or other factors. Another limitation is the study's cross-sectional nature, which only captured measures at one point during the pandemic. Likewise, this study does not represent the college population in the context studied. Lastly, the original survey did not include variables to control students' socioeconomic conditions, which are pivotal in this crisis. Despite its limitations, this study aims to promoted a broader discussion of students' well-being.
Results suggest that coping strategies are a pivotal component for overcoming difficult circumstances. In the case of optimism and gratitude, there is significant evidence showing their positive correlation with good mental health and well-being [30]. Students who practice gratitude have better results in achieving their goals, report less physical complaints, and are more optimistic [44]. Optimistic students have a higher probability of ending their studies than dropping out because they tend to adapt, expect better results, and confront problems better than pessimistic students [35]. There is also evidence that implementing gratitude-enhancing techniques in college can reduce dropout rates; one of the major concerns of this pandemic. Graduating from college gives students more opportunities in the long run, particularly in a middle-income country like Colombia. A spike in dropout rates not only harms students' futures but also the accumulation of human capital in society. There is a need to foster gratitude interventions at the college level particularly, given the mounting evidence of its effectiveness for helping students coping with difiicult situations [49,58,73,74].
Gratitude, optimism, and happiness levels among college students positively influence their academic performance, college adaptation, and retention. The challenge is to implement interventions effectively. College programs can significantly contribute to issues such as students' adaptation and the promotion of students' activities to enhance their roles beyond academic responsibilities. Online interventions have been an option for college students who do not seek formal help [75][76][77]. Moreover, fostering interventions to promote gratitude, optimism or emotional closeness can reduce the financial concerns raised by the pandemic and allow students to manage their emotions and their expectations for the future.

Conclusions
This paper explored the mediating effects of optimism, emotional closeness, and gratitude on well-being during the pandemic caused by COVID-19. This analysis used information from an online survey collected in Cali, Colombia, among college students after one month of quarantine. Similar to other studies conducted with the same population, we found a negative impact of the pandemic on students' well-being. Our results also showed that coping strategies such as optimism, gratitude, and emotional closeness reduced the negative consequences of a crisis. An important contribution of our study is contributing to the ongoing discussion of college students' well-being by providing evidence from a context with little available data.
This study also provides arguments for the call to universities and the educational system to foster positive interventions, such as optimism and gratitude, by offering and promoting programs within the academic curricula. A substantial body of research from positive psychology shows that attitudes such as those studied in this analysis can have positive returns. Aspects of how to implement those strategies or transmit them to the student population are beyond the scope of this analysis. Along with the benefits of the programs, it is important to promote social connectedness in students. These activities promote optimism and gratitude at the individual level and should also provide an opportunity for students to improve their social relationships with other students.

Conflicts of Interest:
No potential conflict of interest is reported by the author(s).