key: cord-0856732-vxv0rpx6 authors: Ikizer, Gözde; Kowal, Marta; Aldemir, İlknur Dilekler; Jeftić, Alma; Memisoglu-Sanli, Aybegum; Najmussaqib, Arooj; Lacko, David; Eichel, Kristina; Turk, Fidan; Chrona, Stavroula; Ahmed, Oli; Rasmussen, Jesper; Kumaga, Raisa; Uddin, Muhammad Kamal; Reynoso-Alcántara, Vicenta; Pankowski, Daniel; Coll-Martín, Tao title: Big Five traits predict stress and loneliness during the COVID-19 pandemic: Evidence for the role of neuroticism date: 2022-01-25 journal: Pers Individ Dif DOI: 10.1016/j.paid.2022.111531 sha: 49091bc794ea15433360242a6a6f4bba82cf376d doc_id: 856732 cord_uid: vxv0rpx6 The rapid outbreak of the coronavirus disease (COVID-19) has affected citizens' daily lives in an unprecedented way. To curb the spread of the pandemic, governments have taken numerous measures such as social distancing and quarantine, which may be associated with psychological consequences, namely stress and loneliness globally. To understand differential associations of personality traits with psychological consequences of COVID-19, we utilize data from a sample of 99,217 individuals from 41 countries collected as part of the COVIDiSTRESS Global Survey. Data were analyzed using multigroup confirmatory factor analysis and multilevel regression models. Findings showed that while some of the associations were rather weak, Big Five personality traits were significantly associated with perceived stress and loneliness during the pandemic. Our study illustrates that neuroticism especially can be a vulnerability factor for stress and loneliness in times of crisis and can contribute to detection of at-risk individuals and optimization of psychological treatments during or after the COVID-19 pandemic. As a global and threatening stressor by itself, the current COVID-19 pandemic has set the ground for an unprecedented crisis in the social, emotional, psychological, and economical sphere. COVID-related stress and negative mental health outcomes are found to be widespread indiscriminately across different groups of people (e.g., Bao et al., 2020; Brooks et al., 2020; Minahan et al., 2021) . Moreover, as a result of lockdown and social distancing measures, an alarming increase in the levels of loneliness also exists according to recent studies (Horigian et al., 2020; van der Velden et al., 2021) . As can be expected, loneliness was found to be closely associated with worry, depression, and anxiety during the COVID-19 pandemic (Hoffart et al., 2020; Palgi et al., 2020; Tso & Park, 2020) . A remarkable factor associated with stress and loneliness during COVID-19 pandemic might be personality traits of individuals (Taylor, 2019) . One of the key taxonomies that has been used to classify and operationalize personality traits is the five-factor model (or Big Five), which describes personality as a system with Neuroticism, Extraversion, Openness to experience, Agreeableness, and Conscientiousness as higher-order domains (McCrae & Costa, 1999) . The Big Five Model is a very often used approach to conceptualize personality, and the five-dimensional structure of this model was found to show high replicability across all the major regions of the world (Rolland, 2002; Schmitt et al., 2007) . The link between personality traits and mental health has been extensively studied under normal circumstances outside a pandemic. However, when considering this relationship, contextual or situational factors, such as the COVID-19 pandemic, should also be considered. Personality is one of the factors which may explain individual differences in response to challenging situations (Han et al., 2021) . As Modersitzki et al. (2020) have J o u r n a l P r e -p r o o f Journal Pre-proof underlined, psychological consequences of the COVID-19 pandemic have strong associations with person-level variables such as personality characteristics. Especially individuals' mood in extremely stressful situations are suggested to be driven more by personality compared to the pandemic experience (Anglim & Horwood, 2021) . Besser and colleagues (2020) argue that some personality factors could act as a risk factor during the pandemic, whereas others could play a protective role in terms of adaptability to the current situation and coping with it. Previous studies during the pandemic have studied personality factors in relation to engagement in preventive behaviors and compliance to measures (e.g., Carvalho et al., 2020; Han, 2021; Zajenkowski et al., 2021) , subjective well-being (Anglim & Horwood, 2021; Modersitzki et al., 2020) , depressive and anxiety symptoms and suicide risk (Han et al., 2021) , stress-related (Zacher & Rudolph, 2021) and global pandemic-related appraisals (Modersitzki et al., 2020) , COVID-19 anxiety (Nikčević et al., 2021) , and behavioral and emotional responses including relaxation and emotional improvement or negative emotional response (Kohút et al., 2021) during the pandemic. Therefore, although there are several studies investigating the implications of personality for mental health outcomes during the pandemic, to the best of our best knowledge, no studies have examined the link between personality, pandemic-related stress levels, and loneliness simultaneously during the pandemic. In the current study, we aimed to understand whether personality traits explain a degree of between-subject variance in pandemic-related stress and loneliness during the COVID-19 situation. Understanding these associations would help to design targeted interventions for undesirable psychological outcomes. dataset contains 125,306 participants (see https://osf.io/f8h9w for the cleaning procedures). The data were gathered during the beginning of the COVID-19 pandemic, between March 30th, 2020 and May 30th, 2020. The project received a waiver to proceed from Aarhus University's Board of Research Ethics Office. In compliance with General Data Protection Regulation standards, all data were anonymous. An organic snowballing recruitment strategy was adopted for sampling. Participation in the study was voluntary. Participants received information on the aims of the study, confidentiality, and right to withdraw at any phase of the survey. We used the cleaned dataset and further applied additional cleaning procedures (see Data Analysis section). These steps resulted in the final sample used in this study consisting of 99,217 respondents from 41 countries. In the sample, 72.6% of participants reported to be female, 26.3% are male, while 1.1% responded "other/would rather not say". The mean age was 39.23 (SD = 13.95), ranging from 18 to 110 years. Sociodemographic characteristics of the sample per country are presented in Table S1 . Participants were asked about age, gender, education level, marital status, and country of residence. Personality traits were assessed using the shortened version of Big-Five Inventory (BFI-S; Lang et al., 2011) . It is a 15-item 6-point Likert-type self-report inventory, ranging from 1 (strongly disagree) to 6 (strongly agree). Responses within each personality subdomain were averaged. According to Hahn et al. (2012) , the scale has internal consistency with mostly acceptable Cronbach alpha coefficients ranging between .44 and .76, convergent (correlations average = .60) and discriminant validity (coefficients ranging between -.01 and J o u r n a l P r e -p r o o f Journal Pre-proof .35), and a significant test-retest stability within 18 months (Neuroticism = .74, Extraversion = .80, Openness to experience = .72, Agreeableness = .57, Conscientiousness = .67). We obtained similar Cronbach's alpha coefficients for the whole dataset across the five personality traits measured by BFI-S (Neuroticism = .69, Extraversion = .75, Openness to experience = .66, Agreeableness = .54, Conscientiousness = .59). Participants' perceived stress level for the past week was assessed using the Perceived Stress Scale (Cohen et al., 1983) , 10-item version (PPS-10; Cohen & Williamson, 1988) , with a 5-point Likert scale (ranging from 1 -"never" to 5 -"very often"). Reliability of the scale is reported as .84 (Taylor, 2015) , and mean scores were used with higher scores indicating higher stress levels. In this study, the PSS-10 had an acceptable internal consistency (α = .88). Loneliness was assessed with a three-item version of the UCLA Loneliness Scale (Hughes et al., 2004; Russell, 1996) . Participants reported how often they feel left out, isolated from others, and lack companionship in the last week on a 5-point Likert scale, ranging from 1 -"never" to 5 -"very often". We obtained Cronbach's alpha coefficient of .77, similar to previous studies (e.g., Hughes et al., 2004) . Statistical analyses were conducted using Jamovi (Version 1.6.11; The jamovi project, 2021) and R (Version 4.0.3; R Core Team, 2021). In the first step of data cleaning, we retained only complete cases on the variables of interest (N = 101,558). Then, we excluded countries with less than 200 respondents (N = 99,217). We transformed gender by coding the responses "Other/would rather not say" as missing values. Next, we assessed internal consistency (using Cronbach's alpha) and then proceeded with testing the cross-cultural equivalence via the measurement invariance analysis of the J o u r n a l P r e -p r o o f Journal Pre-proof study instruments. Since the aim of our study was the meaningful comparisons of factor variances and covariances, we only focused on the metric level of measurement invariance (Millsap, 2011; Lacko et al., 2021) . The scalar level that allows meaningful comparisons of latent means across countries was, therefore, not reported. In multi-group confirmatory factor analyses, we compared the models across 41 countries (configural invariance), with a model with factor loadings constrained to be equal (metric invariance). We evaluated the configural models fit relying on the usually recommended criteria (Hu & Bentler, 1999) . As these models are characterized by a growing complexity (each subsequent model is nested within the previous one), while assessing models' superiority we relied on the cut-off criteria recommended by Chen et al. (2008) and Cheung & Rensvold (2002) for testing measurement invariance. We then proceeded with testing multilevel regression models, in which participants were nested within countries, and continuous variables were centered. We regressed mean scores of stress and loneliness on participants' gender, age, and mean scores of Big Five traits. In these models, we set random intercepts on a country level. Descriptive statistics of the sample per country and mean scores of study instruments across countries are given as supplementary information (Table S1 and S2). Table 1 shows the Pearson correlation coefficients for each pair of variables of interest. regulation strategies such as avoidance, suppression, rumination, and worry (Barańczuk, 2019) , which maintain the inability to tolerate the distressing emotions that arise with the uncertainty about the threats related to pandemic. Our findings also suggest that while all estimates for these associations of other four Big Five personality traits with stress and loneliness in multilevel models were statistically significant, some of the observed effect sizes were too small to carry reliable meaning and significance, and therefore, should be interpreted with caution. Firstly, our findings showed that that higher extraversion accounted for explaining more stress and more loneliness. A recent meta-analysis by Buecker et al. (2020) provided evidence for a strong negative link between extraversion and loneliness, which highlights the fact that -in normal conditionsthose high in extraversion feel less lonely, possibly due to their frequent engagement in various activities (Lai & Qin, 2020) . However, beneficial effects of extraversion on subjective well-being may be particularly attenuated during the pandemic (Anglim & Horwood, 2021) . Hence, this finding further provides evidence that this finding may be due to the uniqueness of the current pandemic and the impact of preventive measures. Moreover, higher levels of openness were related to the higher levels of stress during the pandemic. Openness has been previously shown to be associated with depressive and anxiety symptoms and suicide risk (Han et al., 2021 , Blüml et al. 2013 , and less fear of infection in the context of COVID-19 (Martin, 2020) . Hence, openness to experience as a personality trait can be taken as a vulnerability factor during the COVID-19 pandemic as it can possibly increase the involvement into various types of risk behaviours, which in turn can be related with experiencing higher levels of stress. Agreeableness is found to be positively associated with stress and loneliness in the (Wilkowski et al., 2006) . However, due to pandemic related restrictions, they might be facing certain stress and loneliness as those limit their ability to carry out helping behaviors and protect others (McCrae & Costa, 1999; Penner et al., 2005) . Zajenkowski et al. (2020) also showed that agreeable individuals comply more with rules and regulations therefore have been perceiving more duty, and negativity. Conscientiousness, in our study, comes out to be negatively associated with stress and loneliness. Studies have depicted that conscientiousness is related to increased positive affect (Bartley & Roesch, 2011) , use of active problem-solving strategies (Connor-Smith & Flacshbart, 2007) , and lessened distress and loneliness (Buecker et al., 2020; Luo & Roberts, 2015) in general. It has been shown that resilience, support seeking, appraisal of self-efficacy, and coping and self-regulation abilities mediate the link between conscientiousness and distress (Hoyle, 2006; Kocjan et al., 2021; Nikčević et al., 2021) and suggested that conscientious individuals are perceived more positively by others and feel less lonely (Schermer & Martin, 2019) . Therefore, the possibility that people high in conscientiousness benefit from close relationships, cope better, and feel less stressed and lonely during the pandemic could explain our findings. Our data also showed that women and younger individuals were more stressed and feeling lonelier during the pandemic, as consistent with Kowal et al. (2020 ), Luchetti et al. (2020 ), and Luo et al. (2020 . In sum, using a worldwide dataset, Big Five personality traits, especially neuroticism, have been shown to be associated with stress and loneliness during the pandemic. Although the effect sizes for associations of stress and loneliness except for neuroticism were quite small, our findings contribute to the literature of personality traits by exploring their role in shaping responses to feelings of stress and loneliness in a crisis such as the COVID-19 pandemic. In times of crisis, stress and loneliness can have a detrimental effect on the way individuals cope and manage their everyday life and responsibilities. Mapping the personality profiles of those who are more vulnerable to stress and loneliness may be crucial in identifying those at risk and can help prevent severe psychological problems. However, this study has several limitations. Given the timeframe of our study, we used single source cross-sectional data, raising concerns about prohibiting inferences of causality. Our sample was not nationally representative and hence, the generalizability of our findings is not given across gender or age. In future studies, non-self-report measures such as fine-grained daily diaries and/or experience-sampling designs would be used to better understand the associations between personality and psychological well-being during crises. Future studies would also focus on associations of sociodemographic variables including gender and age with personality and psychological outcomes during the pandemic. Finally, because of the limited length of our survey, future studies should investigate other aspects of personality that could be surveyed to understand the psychological toll of the pandemic in more detail. J o u r n a l P r e -p r o o f Journal Pre-proof Tables Table 1 Correlations Between Study Variables Variable 1 2 3 4 5 6 7 8 Age (1) -Gender (2) <0.001 -Neuroticism (3) -0.213 *** 0.077 *** -Extraversion (4) -0.011 *** 0.111 *** 0.11 *** -Openness (5) 0.038 *** 0.063 *** 0.111 *** 0.267 *** -Agreeableness (6) -0.054 *** 0.023 *** 0.088 *** 0.219 *** 0.21 *** -Conscientiousness (7) -0.097 *** 0.064 *** 0.163 *** 0.15 *** 0.181 *** 0.147 *** -Stress (8) -0.145 *** 0.152 *** 0.292 *** 0.15 *** 0.157 *** 0.089 *** 0.096 *** -Loneliness (9) -0.17 *** 0.099 *** 0.191 *** 0.081 *** 0.036 *** 0.051 *** 0.044 *** 0.403 *** Effect of the COVID-19 pandemic and Big Five personality on subjective and psychological well-being 2019-nCoV epidemic: Address mental health care to empower society The five factor model of personality and emotion regulation: A metaanalysis Coping with daily stress: The role of conscientiousness Personality, cognition, and personality traits Loneliness matters: A theoretical and empirical review of consequences and mechanisms Loneliness and social distancing during the COVID-19 pandemic: Risk factors and associations with psychopathology Loneliness, mental health, and substance use among US young adults during COVID-19 Personality and self-regulation: Trait and information-processing perspectives Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives A short scale for measuring loneliness in large surveys: Results from two population-based studies Self-regulation and the Five-Factor Model of personality traits Statistical approaches to measurement invariance The COVID-19 pandemic and psychosocial outcomes across age through the stress and coping framework Who is impacted? Personality predicts individual differences in psychological consequences of the COVID-19 pandemic in Germany Advance online publication Modelling the contribution of the Big Five personality traits, health anxiety, and COVID-19 psychological distress to generalised anxiety and depressive symptoms during the COVID-19 pandemic The loneliness pandemic: Loneliness and other concomitants of depression R: A language and environment for statistical computing. R Foundation for Statistical Computing Let's talk about stress: History of stress research The cross-cultural generalizability of the Five-Factor model of personality UCLA Loneliness Scale (version 3): Reliability, validity, and factor structure Is perceived stress of French university students related to personality traits A behavior genetic analysis of personality and loneliness The geographic distribution of Big Five personality traits: Patterns and profiles of human selfdescription across 56 nations Assessing measurement invariance in cross-national consumer research Psychometric analysis of the Ten-Item Perceived Stress Scale The psychology of pandemics: Preparing for the next global outbreak of the infectious disease The jamovi project (2021). jamovi (Version 1.6) [Computer Software Alarming levels of psychiatric symptoms and the role of loneliness during the COVID-19 epidemic: A case study of Hong Kong The role of neuroticism and extraversion in the stress-anxiety and stress-depression relationships Anxiety and depression symptoms, the recovery from symptoms, and loneliness before and after the COVID-19 outbreak among the general population: Findings from a Dutch population-based longitudinal study Agreeableness and the prolonged spatial processing of antisocial and prosocial information Who complies with the restrictions to reduce the spread of COVID-19?: Personality and perceptions of the COVID-19 situation The joint effect of personality traits and perceived stress on pedestrian behavior in a Chinese sample