Student Learning Experiences and Well-Being during the COVID-19 Pandemic at the University of Free State

South African universities were closed nationally during the COVID-19 pandemic. On-campus learning was suspended, and on-campus student accommodation was closed from March 2020, with phased returns to campus continuing until early 2022. The study intended to identify categories of students whose learning and well-being were most affected by the COVID-19 pandemic and provide empirical evidence to inform initiatives to support these students better. An online cross-sectional survey completed by 1 562 respondents studying at a South African university explored students’ experiences of the COVID-19 pandemic, focusing on students’ learning and well-being, subjective social status, sense of coherence, and future anxiety. Multivariable regression models showed younger age (18 to 21 years), female gender, low subjective social status, and isiZulu and English home language were significantly associated with lower well-being. Regression analysis also showed a significant association between the dimensions of manageability and meaningfulness of sense of coherence and well-being and that high levels of future anxiety were

This study intended to identify categories of students whose mental health was most affected by the COVID-19 pandemic by investigating the impact of the pandemic on the psychological well-being and learning experiences of university students.These students, in particular, need support when traumatic, uncertain, or disruptive situations arise.This was done using a quantitative research method.The survey tool was employed to quantitatively assess opinions about the student experience during the pandemic in 2020.The study covered all three University of the Free State (UFS) campuses in South Africa.Given that difficulties with mental health impact not only learning and academic participation but all facets of a student's life (Baik, Larcombe, and Brooker 2019;Eisenberg, Golberstein, and Hunt 2009;Wyatt and Oswalt 2013), a better understanding of the mental health, well-being, and learning experiences of students is of significant importance to universities.

Background
Extraordinary measures were implemented globally to halt the COVID-19 pandemic and reduce transmission of the novel coronavirus.South Africa went into a hard lockdown at the end of March 2020, with severe social restrictions continuing in May 2020 and again in June and July 2021 (Lebuso 2022).The country remained in a state of national disaster from March 2020 until April 2022 (Government of South Africa 2022), with the changing levels of lockdown continually influencing how the higher education space operated.South African universities were closed nationally, on-campus learning was suspended, conferences, workshops, and other social events were postponed or cancelled, and on-campus student accommodation was closed.Students had to rapidly transition from face-to-face to online lectures to ensure the academic year would be completed.The closure period lasted from 18 March 2020 to 21 February 2022; a blended teaching and learning approach was adopted, where 67% of modules offered were face-to-face at UFS (University of the Free State, 2022).
According to the UFS Students' Access to and Use of Learning Materials Survey Report (Centre for Teaching and Learning, 2020), 90% of UFS students had modules that moved to remote learning during the first semester of 2020.Other students mostly engaged with their studies through self-study.Face-to-face teaching and learning continued only in extraordinary cases, for example, for students in the health sciences.Most UFS students used smartphones to engage in academic activities, although this was also considered the device that students found most difficult to use for academic purposes.Most (54%) students reported not having a quiet place to study, and only half had good connectivity from their accommodations.The following were recorded as challenges from an online/blended learning perspective: frustrations with technology; a lack of appropriate skills to engage with technology; difficulties engaging with lecturers or peers; feelings that some learning was lost due to less engagement with lecturers or work; and being distracted when at home.It is, therefore, evident that UFS students had to adapt to a new way of learning very quickly due to the pandemic.
As noted in the introduction, previous studies have found a substantial influence on mental health due to the pandemic.Common psychological responses reported include depression, anxiety, and post-traumatic stress (Casagrande et al. 2020;Chew et al. 2020;Odriozola-González et al. 2020;Rajkumar 2020;Wang et al. 2020;Xiong et al. 2020).Chew et al. (2020) also found that susceptibility to the disease, changes in habits, financial uncertainty, and concerns about the well-being of loved ones were all factors that contributed to increased anxiety.Regarding university students specifically, Dodd et al. (2021) found that first-year students experienced more challenges than senior students for various reasons, such as underdeveloped coping mechanisms to deal with adversity.

Research objectives
This study aimed to provide empirical evidence to inform interventions or initiatives intended to better support the mental well-being of UFS students by exploring the experiences of local and international undergraduate and postgraduate students at the UFS during the COVID-19 pandemic.The objectives of the study were as follows: • Identify and profile UFS students who are most affected by the pandemic.
• Investigate the relationships between students' learning and their well-being, specifically psychological well-being, social well-being, emotional well-being, subjective social status, sense of coherence, and anxiety during the pandemic.
• Investigate the relationship between students' well-being and sociodemographic factors, specifically gender, age, race, language, nationality, economic status, and social capital.
To achieve the above research objectives, the following research questions were asked: Research method

Study population
The unit of observation in this study was registered UFS students of all categories across all three UFS campuses aged 18 years and older.

Design and sampling
Probability sampling was adopted, and participants were invited to participate in the study on the grounds of being registered as a UFS student.

Data collection procedure
An online survey questionnaire was distributed via email to all registered UFS students.
The survey accepted responses from 1 December 2021 to 5 March 2022.During this time, two reminders were sent to potential participants, resulting in a 14% response rate.This study was approved by the General/Human Research Ethics Committee of the UFS (HSD2021/1325).

Data collection instruments
The survey questionnaire collected sociodemographic data [for example, age and gender, level of study, sources of university funding, and subjective social status (SSS)], data about the impact of COVID-19 (in general, and on students' learning experiences and living arrangements specifically), reactions to the pandemic, sense of coherence, future anxiety, and well-being.As explicated in Table 1, SSS, sense of coherence, future anxiety, and well-being were measured through the MacArthur Scale of Subjective Social Status (Adler et al. 2000), the Work-related Sense of Coherence questionnaire (Vogt, Jenny, and Bauer 2013), the Dark Futures Scale (Zaleski 1996), the World Health Organization Well-Being Index (WHO-5) (Topp et al. 2015), and the Mental Health Continuum Short Form (MHC-SF) (Keyes 2009).The researchers compiled additional survey questions to collect data for the remaining variables, following the example of Dodd et al. (2021) and Visser and Law-van Wyk (2021).

Data analysis
Data were analysed using R Programming.Scale reliability was measured through Cronbach's alpha.Descriptive analyses were performed for the sample profile, the impact of COVID-19 on living and learning, and the psychological variables.Independent t-tests were performed to compare psychological factors.The chi-squared test was used to assess if there was an association between well-being and various sociodemographic factors (i.e., age, gender, qualification type, and subjective social status).For all analyses, p-values < .05were deemed statistically significant.Several binary logistic regression models were performed to identify the predictors of low wellbeing.Model 1 included the sociodemographic variables of age, gender, qualification level, subjective social status, and home language; model 2 expanded on model 1 by adding a sense of coherence scale; and model 3 expanded on model 2 by including future anxiety.

Sample profile
Of the 2,064 survey responses received, 502 that did not answer all the survey questions were removed from the dataset, resulting in 1,562 responses in the analysis.
[2] In order of prevalence, National Student Financial Aid Scheme (74%), Other, or National Research Foundation.
[3] In order or prevalence: from private sponsors/businesses or parastatals (e.g., Eskom) or for academic merit.

Impact of COVID-19 on living and learning
Most respondents reported that the COVID-19 pandemic disrupted their lives (79%) and had a negative financial impact (73%; see Table 3).The respondents were more fearful of a loved one contracting COVID-19 (83%) than themselves (66%).The majority felt isolated (54%), vulnerable/not in control (59%), and as if life had been on hold (67%) since the start of the pandemic.However, only half of the respondents reported that COVID-19 negatively impacted their student experience.The respondents preferred online to face-to-face learning (52%).They found it less time-consuming (64%), with many respondents reporting that they did not find it more difficult to learn online than face-to-face (49%) or to interact with lecturers (48%) or other students (45%) online.However, similar proportions of respondents agreed (40%) and disagreed (43%) that their unreliable internet disrupts online learning, and 41% felt distracted from their learning due to COVID-19.Most agreed that their home environment supports online learning (57%).The respondents were living mostly in townships (38%) or urban suburbs (25%) during lockdown, with family members (65%) in their family homes (51%).[1] agree includes strongly agree and agree [2] disagree includes strongly disagree and disagree

Future anxiety, sense of coherence, and well-being
Future anxiety was high for two scale items, i.e., fear about facing life's crises or difficulties (61% true; see Table 4) and fear about changes in the economic and political situation threatening the future (71% true).Conversely, 46% of respondents were not afraid that their lives would change for the worse in the future, and 44% were not disturbed by the thought that they would not realise their goals in the future.The chi-squared test identified significant differences in age, gender and subjective social status (SSS) (Table 5).However, when examining the strength of association using Cramer's V, it was identified that all the significant variables had weak to moderate associations.Only gender and SSS had Cramer's V above .1 (Gender, Cramer's V = .109;SSS, Cramer's V = .12).This signifies a moderate association compared to Age.Table 6 presents a correlation matrix.As shown, there is a weak positive correlation between well-being and the different measures of sense of coherence (SoC) (comprehensibility = .36;manageability = .29;meaningfulness = .37).Moreover, wellbeing and future anxiety demonstrate a moderate negative correlation of -.44, indicating that as well-being increases, future anxiety tends to decrease.Measures of SoC are moderate to strongly positively correlated with each other (comprehensibility and manageability = .61;comprehensibility and meaningfulness = .71;manageability and meaningfulness = .52).Lastly, all measures of SoC are moderately negatively correlated with future anxiety (comprehensibility = -.27;manageability = -.19;meaningfulness = -.28).

Recommendations and Conclusion
The current study aimed to explore the learning experiences of local and international undergraduate and postgraduate students at the UFS during the COVID-19 pandemic.Furthermore, the study aimed to determine the relationship between well-being and subjective social status, as well as the sense of coherence and anxiety during the pandemic.The study further investigated the association of sociodemographic variables (i.e., age, gender, race, home language, education, and social and financial capital), students' learning and well-being, subjective social status, sense of coherence, and future anxiety.Multivariable regression models showed younger age (18 to 21 years), female gender, low subjective social status, and isiZulu and English home language were significantly associated with lower well-being.Regression analysis also showed a significant association between the dimensions of manageability and meaningfulness of sense of coherence and well-being and that high levels of future anxiety were associated with lower well-being.The findings concur with those of Dodd et al. (2021), which also found that younger adult students (aged 18-24 years) had more symptoms of anxiety and depression during the COVID-19 pandemic than older adult students (≥2 years), supporting our findings of differences between undergraduate and postgraduate students.Half of the respondents (49.9%) reported being first-generation students, already considered a vulnerable population, and more than half (70.8%) agreed with the statement that they fear changes in the economic and political situation will threaten their future.This result indicates that students often feel their financial situations have no hope of changing even though they are currently obtaining a degree.Therefore, this indicates that younger adult students require targeted support and resources, particularly at the undergraduate level.These students may require additional assistance to cope with psychological challenges brought on by the pandemic.This highlights the notion that age is a strong determinant of how individuals cope with anxiety and depression (Maulana, 2021).Higher education institutions (HEIs) may also want to consider a flexible learning environment post-pandemic, which may give students greater control over their learning environment and reduce stress.Peer support networks and mentorship programmes can also be beneficial as postgraduate students can guide their younger peers through trusted relationships.Additionally, universities must encourage students to use existing services.Naidoo and Cartwright (2022) argue that student counselling services should strive to reflect and develop a social justice approach that recognises how greater socioeconomic injustices and oppression that stem from disparities perpetuated during apartheid affect student academic performance and mental health.There is an increased need for these services to be made available virtually to accommodate students in an environment they feel most comfortable in.Similar findings to this study were found in a large-scale international study done by Aristovnik et al. (2020), where the mental health of students relied substantially on the amount of change in their daily routine, whether they received social support, being an undergraduate student, and living at a lower standard (subjective social status).However, the studies differed regarding gender.The international study found that the pandemic had a stronger impact on male students.The study also revealed that African students (Egypt and South Africa) were less satisfied with online lectures than North America and Europe, possibly due to the difference in developed ICT infrastructure.However, 67% of respondents in our study agreed that their home environment was conducive to online learning.Interestingly, the least anxious students in the international study were African students.
These findings highlight the need for HEIs' continued support of students regarding well-being and mental health and the need for improved interventions when change occurs suddenly in the sector that impacts students.Findings indicate that students with a lower SoC and a high level of future anxiety also experienced lower well-being.Initiatives to improve students' coping s ills, resilience, and positive mindset are ways HEIs can strengthen students' capacity to navigate changes throughout their learning careers.The opportunity for students to participate in continued career guidance and skills development is also a way universities can help reduce student anxiety.Continuously improving the online learning space through digital literacy and support is vital to ensure equity among students in the online learning space.The study also highlighted the interdependency of variables contributing to well-being and student success.Therefore, interventions with a holistic approach are essential, given the complexity of this interdependency.However, some limitations of the study included surveying students at only one South African university.Therefore, comparisons could not be drawn in this way.Because the study only adopted a quantitative approach, rich data was not gathered in the form of the thoughts and feelings of respondents.In conclusion, the study provides valuable insights and evidence on the experiences of UFS students during the COVID-19 pandemic and the impact on their mental health during this challenging time.The findings highlighted the most vulnerable groups of students in higher education and encouraged continuous targeted support for student well-being.

Table 1 :
Data collection instruments

Table 5 :
Well-being of university students differentiated by sociodemographic characteristics (N=1433)

Table 7
provides the result of the multivariable regression models for low/very low beings.In model 1, we found significant associations between low/very low well-being and student age (18-21, OR = 1.579, 95% CI: 1.140-2.203),being female (OR = 1.732, 95% CI: 1.384-2.714),low subjective social status (SSS) (OR = 1.844, 95% CI: 1.323-2.587)and isiZulu home language (OR = 1.257, 95% CI: 0.979-1.615).After adding a sense of coherence in model 2, we found that student age was no longer significant while the other relationships (Female, low SSS and isiZulu) remained significant., resulting in low SSS and low SoC, with comprehensibility no longer a significant predictor of low/very low well-being.The OR for low SSS had decreased from model 1 to model 3.The association between isiZulu and English home languages increased, whereas being female lessened slightly.In addition, we found a significant association between low/very low well-being and high future anxiety (OR = 3.961, 95% CI: 2.910-4.696).The respective OR was the highest in the model.