A Cross-country Investigation of Perceived Stress During COVID-19 Pandemic: The Impact of Situational Factors and Personality Traits

The study investigates the public mental health cross-national during the global COVID-19 pandemic. Analysis of data collected from people across different countries shows that people are experiencing moderate stress level and demonstrates that stress level is affected by internal genetic factor and external conditions. Specifically, perceived stress is affected by personality traits and contextual settings namely residence location, personal finance situation, and usage of social media during the pandemic.


Introduction
In the end of 2019, beginning of 2020, the Coronavirus COVID -19 took the world for a storm as it quickly spread out to most nations within only a few months, and was declared a global pandemic, with nearly 8 millions infected and nearly half a million dead (and counting) (worldometers, 2020). Many nations declared quarantine order, cancel large gatherings, and even close off borders in an attempt to slow the virus spread. Starting as a health crisis, it rapidly escalated to creating negative social, economic and political impact. People's lives are disrupted when schools are closed, adults work at home, businesses recorded huge losses and even bankruptcy, many lose their jobs and are left hungry and homeless. With such challenging conditions, it is predictable that people would experience emotional distress, especially when many nations place a compulsory social distancing and quarantine that bounds people at home. This study aims to investigate mental health during the Covid-19 by examining stress levels of people in relation to external factors (contextual settings) and genetic factors (personality traits).
Literature Review Lazarus (1990) viewed stress response as the result of the interaction between humans and environment, which means that individuals would differ in their perception of stress. In the transactional model of stress developed by Lazarus and Folkman (1984), individuals experience two appraisal processes, one to evaluate the external stressors and personal stake and the other evaluates personal resources to cope with stressors; stress occurs when imbalance between stressors' demands and personal resources to cope with such demands. When judging a situation to perceive threats, individuals differ in terms of perception and cognitive appraisal, which leads to varied interpretation of situations. People differ in terms of personal resources to cope with external threats, which means how they react to perceived threat is not the same as well.
Even though personality was not speci cally included in the transactional model of stress, it clearly impacts individuals' cognitive appraisal processes. Trait theory views personality traits as patterns of thoughts and action, and there are characteristics that are stable and consistent across situations (Costa and McCrae, 1992). However, there are studies that proved personality traits can uctuate and be changed and even in a short time frame (see for example, Wilson et al., 2016;Wrzus and Roberts, 2017;Sherman et al., 2015). Whatever the case it is, the relationship between perceived stress and personality has been widely accepted as previous research has proved the relationship between personality and subjective perception of stress. Vollrath (2001) believed that personality affect the descriptive representation of situation and evaluative perception of situation. Cross-sectional and longitudinal studies showed a signi cant association between personality and perceived stress (Ebstrup et al., 2011). Shields et al. (2016) discovered that change in perceived stress is strongly related to change in pessimism on a weekly basis. Kondratyuk and Morosanova (2014) found out that high level of neuroticism and introversion can lead to high chronic stress. Şahinn and Çetin (2017) conducted a weekly assessment study and con rmed that neuroticism and extraversion predict perceived stress. The cross-sectional population-based research by Feizi et al. (2014) proved that people with high neuroticism tend to have high level of stress and anxiety, and those with high extraversion and consciousness have lower the chance of high psychological stress. Borkoles et al. (2018) linked Type D personality to increased level of perceived stress and discovered that people with Type D personality tend to search the environment for threats.
As de ned, stress response is the interaction of environment and individuals, debate on the role of environmental factors and personal factors on perceived has attracted attention. The study by Luo et al., (2017) found that genetic factors largely explained for the concurrent relation and the continuity between the two; meanwhile, environmental factors played a more signi cant role in the association between changes in personality and in perceived stress. Similarly, the research on students by Schmidt et al. (2013) showed that structural conditions explained perceived stress better than personality. Lebois et al. (2016) proposed that situations of certain features (such as containing self-threat, negative affect, certainty elements, etc.) can predict the level of stress that people perceive. For example, the work by Onah (2003) about pregnant Nigerian women's perception of environmental stressors con rmed that economic, futuristic, health and social factors caused increase in stress. Or the research by Kallio et al. (2020) found that people can be more stressed out by the poor quality of the environment. Knight et al. (2013) demonstrated that contract-related factors (such as having contracts or not, working hours) in uence stress level of coaches in Canada. Similarly, the examination on team work by Guznov et al. (2010) found an association between high neuroticism and higher stress and workload for individuals in the position of having direct control but not in other positions, which suggests the effect of external environmental demands (social interactions required in a team role).
To provide a comprehensive picture, this study proposes that perceived stress is in uenced by personality traits and context but does not attempt to identify nor compare the magnitudes of such impacts. Methodology Procedure A survey questionnaire was administered and distributed online on randomly selected Facebook groups that have members living in Vietnam and foreign countries. The questionnaire was open for three weeks, from mid April to the beginning of May, when the virus started to spread uncontrollably, the number of infected cases skyrocketed worldwide, chaos and system overload happened, many countries h declared national lock-downs. After three weeks, a total of 219 responses were collected on Google Form, 3 were not valid, leaving only 216 eligible responses. The survey comprises three sections: the rst one asks about personal information and personal circumstances during Covid-19 pandemic, the second section and the third section measure personality and perceived stress scale respectively.

Instruments
Personality is measured on ve domains, using a 30-item short form questionnaire tested and re ned by Soto and John (2017). The questionnaire is developed based on the Big Five Model, one of the most popular frameworks (McCrae and Costa, 1999); many have praised the model in examining stable personality traits in threatening situations (see Besser and Shackelford, 2007;Hojat et al., 2003). The Big Five Model examined ve personality traits: Extroversion, Agreeableness, Neuroticism, Conscientiousness, and Openmindedness. In this study, the term "Neuroticism" is replaced by the term "Negative Emotionality" to be consistent with the original questionnaire by Soto and John. The level (high or low) of each trait is calculated by adding up the scores of the questions that measure the trait.
To measure stress, the Perceived Stress Scale (PSS) developed by Sheldon Cohen and his colleagues (1983) was used. It is one of the most popular psychological instruments used to measure individuals' perceived stress. The PSS is a self-reported scale that asks about individuals' experience of perceived stress in the last 30 days. Among the three versions of the scale: PSS-14, PSS-10, and PSS-4, the 10-item version was selected to use in this study. There are 10 questions in the scale, each asks the interviewee to chose the option that best describes their frequency of experiencing the stressful situations. The original scale was 0 -Never, 1 -Almost Never, 2 -Sometimes, 3 -Fairly Often, 4 -Very Often. The scale used in this study kept the same 5 levels of frequency, but changes the numbering of each option as followed: 1 -Never, 2 -Almost Never, 3 -Sometimes, 4 -Fairly Often, 5 -Very Often. Score for each question is added up to calculate the level of perceived stress; scores ranging from 0-13 is considered as low stress, 14-26 is considered as moderate stress, 27-40 is considered as high stress.

Data
The research sample consisted of 216 people living in 21 countries affected by Covid -19 pandemic. Speci cally, 44% of the respondents are living in Vietnam, 39% are living in European countries, and the rest are residing in other nations (see Table 1). Around 90% of the respondents are living in quarantined areas (see Table 2), more than 80% of them believe that quarantine helps to prevent the spread of the virus. 60% of the respondents are female (131 people), around 39% are male (85 people), and only 1% identi ed themselves as others (2 people). Nearly half of the participants (48%) are from age 25 to 34, 29% are from age 35 to 49, around 10% are from 18 to 24 and over 50, only a minority are under 18. In terms of employment status, half of the respondents are full-time employed, 6% are part-time employed, 17% are self-employed, 19% are students (including those on scholarship), and the rest are other types of employment (such as housewives, unemployed, etc.

Results
Perceived stress level and Personality The relationship between personality and perceived stress are determined by Pearson Correlation. The results showed a relationship between personality and perceived stress level.
To be more speci c, with r = 0.417, p < 0.01, Negative Emotionality was proved to have a positive moderate relationship with perceived stress level. This could be understood as the higher level of negative emotionality is, the higher the stress level is.
With r = -0.184, p = 0.007 < 0.01, it can be concluded that there is a weak negative relationship between Conscientiousness and perceived stress level, which means the lower level of conscientiousness a person has, the higher level of stress he/she perceives.
The ANOVA test was run to investigate whether different groups recorded different personality. With Sig. = 0.001 < 0.05 and Sig. = 0.043 < 5, the result showed that different age groups vary in Negative Emotionality and Conscientiousness respectively. There are no differences in personality by Gender and Nationality.
Then, means of the groups were compared to further investigate the differences in personality.
The result shows that the older people are, the lower level of negative emotionality they experience, and that people in the age group from 35 to 49 have the highest level of conscientiousness. And Sig= 0.018 < 0.05 shows a difference in stress level of people living in different countries.
Then, means of groups are compared to look more speci cally into the differences in stress levels. Table  11 shows that people living in Vietnam experienced less stress than those from Europe. Not surprisingly, the result in Table 4 demonstrates that people whose nance is negatively affected by the pandemic experienced higher stress compared to those whose nance is stable. Interestingly, based on Table 5, it seems that the more hours that people spend on social media, the higher level of stress they experience.

Discussion
This study is the rst one to examine people's stress in different countries during Covid-19. The research aims to investigate the relationship between stress level and personality and how contexts may affect people's stress.
The rst nding con rmed the relationship between personality traits and perceived stress level. Analysis of data showed that perceived stress is positively related to negative emotionality (or neuroticism), and negatively related to conscientiousness. This result of negative emotionality and perceived stress is in line with previous research by many authors, such as the ones by Fornés-Vives, et al. (2012), Dalia, Lucia and Nathalie (2017), or Conard and Matthews (2008). Individuals with high scores of negative emotionality have more di culties in controlling emotion and are more likely to experience negative emotions; therefore, it is logical that they tend to have higher level of stress than those that have low negative emotionality. On the other hands, trait of conscientiousness is found to have a negative relationship with perceived stress, which means that the more conscientious a person is, the lower level of stress he perceives. This outcome concurs with existing literature, such as the study by Luo and Roberts (2015) who stated that people who increased in conscientiousness experienced less stress over time.
Because conscientiousness is related to positive thinking and positive reappraisals (Watson and Hubbard, 1996), it is comprehensible that high level of conscientiousness is associated with low level of stress. Even though the result of data agrees with existing research on the impact of genetic factor (personality traits) on perceived stress, the validity of this nding may be questioned and calls for further investigation due to the small sample size.
The second nding of the paper illustrated the impact of situational factors on stress level. National and global contexts have been proved to in uence people's experience of stress. For example, Ragnarsdóttir, Bernburg, and Ólafsdóttir (2013) hypothesized and inspected different outcomes (caused by the global nancial crisis) that affect emotional distress and found out that Icelandic people's comparison of past outcomes (to the present) has the most signi cant impact on their distress. On a similar note, the research by Abu-Ras, Suárez, and Abu-Bader (2018) during the 2016 presidential election in America (where one potential candidate frequently underlined his intention to restrict immigrants from entering the country) discovered that people of certain ethnicities suffer from discrimination stress, which makes them more likely to have higher rate of mental health problems than those who are not discriminated.
In this research, the analysis showed that people residing in Vietnam have less stress compared to those living in Europe during Covid-19. This could be explained by the country's successful containment of the virus compared to others (Quach and Hoang, 2020;Dinh et al., 2020, Huynh, 2020Ha et al., 2020), which make their citizens feel more secured than those living in nations where the number of infected and death toll is much higher. Even though Vietnam imposes strict quarantine and social-distancing order compared to Western countries, the amazingly lower number of infected cases and death toll in Vietnam implies the undeniable effectiveness of restriction to contain the virus. This country case may have a crucial implication for future treatment and containment of contagious viruses. Furthermore, people whose nance is negatively affected by the Covid-19 perceive higher stress than those whose nance is stable. The Covid-19 has severely disrupted businesses worldwide, causing many people to either lose parts of their regular incomes, or become unemployed. This result suggests that mental health can be severely affected by personal nances. This nding corresponds to the study on families by Wei and Chen (2014) who con rmed that nancial stress is signi cantly related to mental health problems. Sweet et al. (2013) also discovered that high nancial debt is associated with higher perceived stress and depression in the US. Similarly, Faresjö (2013) compared the stress level between Greek and Swedish young adults and found out that the Greek had higher perceived stress as a result of living in deteriorating national economic conditions. Finally, the result showed that people who use social media more frequently reports higher perceived stress. The nding seems contradictory to the study by Bland et al. (2012) who claimed that people used internet social networks as a way to reduce stress. The need to stay connected is perhaps even more imperative during the turbulent time, especially when the majority of respondents lived in quarantined areas at the time this study was conducted. Higher stress from using social media can be explained as even though social networks established through social media is a means of having connectedness and perceived support, such support is super cial, and thus, the more people are connected, the more lonely they can get (Ventriglio and Bhugra, 2017). Besides, people can be stressed out by receiving negative information about the virus daily, especially when social media is used as a channel for the government to communicate with the public concerning health issues (see Coiera, 2013;Neiger et al., 2012, or Mehta andAtreja, 2015). Because in addition to the primary use of social media for keeping in touch with friends and families or for entertainment purpose during the quarantined period, people also rely on this channel for updating and sharing information related to the pandemic. In the case of Vietnam, the government integrated this channel into other initiatives to combat Covid-19, where citizens are regularly updated on the situation and are encouraged to stay connected for timely information. Social media proves to be effective in propaganda; therefore, agenda should be developed to maximize the bene ts of this channel (such as connecting people and sharing information) in the time of terror and panic.

Limitations
The study was not conducted without limitation. First of all, small size of sample may affect the precision of ndings. Secondly, the research was conducted at a speci c time during the global health crisis, which can only describe the fact at that moment and cannot generate a clear understanding of the impact of situation on stress. In the future, perhaps a larger sample size and a longitudinal study at different times of the pandemic may better explain the effect of situational factors on perceived stress.

Conclusion
The research is conducted cross-country during a severe global pandemic to investigate the impact of personality and situational factors on perceived stress level. Analysis of data con rms the association between personality traits of negative emotionality and conscientiousness and perceived stress, which is consistent with existing literature. The study also shows that people experience moderate stress during the global Covid-19 epidemic and that perceived stress is affected by situational factors including residing country, personal nance, and usage of social media. Future research could inspect stress level through different times of a situation to provide a more comprehensive understanding of how context in uences stress.

Declarations
Authorship: the work is processed by a single author.
Originality and plagiarism: the author has written this entirely original work and appropriately cited or quoted the works of others.
Data access and retention: the author has entire access to raw data and is committed to provide public access to such data.
Funding Statement: this research received no speci c grant from any funding agency, commercial or notfor-pro t sectors.
Con ict of Interest: the author has no con ict of interest.
Ethical Standards: this work does not involve chemicals, procedures or equipment that have any unusual hazards inherent in their use, or the use of animal or human subjects. All human participants in the survey featured in this work are voluntary. Wrzus, C., and Roberts, B. (2017). Processes of Personality Development in Adulthood: The TESSERA Framework. Personality and Social Psychology Review,21(3), pp. 253-277. Figure 1 Stress level during Covid-19 pandemic