Introduction

Over the past years, COVID-19 has spread rapidly around the world and millions of people have lost their lives in the pandemic. A key aspect of public health epidemiology is how individual and community actions can help us to mitigate and manage the costs of a pandemic. Limiting close human interactions is an effective measure to contain transmission1, since decreasing the frequency of interpersonal contacts can reduce the COVID-19 circulation2. However, especially during the initial phases of the COVID-19 outbreak, social distancing remained a voluntary behavior to which the general population did not always conform3. To convince people to comply with health recommendations, many governments decided to make heavy use of mass media advertisements4. In the most extreme cases, they have turned to the use of military forces, such as in the case of pandemic drones5 or robots used to monitor the respect of social distancing in Singapore6. From a psychological perspective, these solutions did not adequately leverage on well-established knowledge showing that compliance with preventive behaviors (and more in general with behaviors that benefit the society or the environment7,8) represents the output of a complex mental process through which individuals weigh the positive and negative implications of these behaviours2,9. In fact, evidence from both the COVID-19 emergency and previous pandemics (e.g., SARS and Ebola) suggest that adherence to quarantine protocols was directly proportional to the extent to which individuals perceived the expected benefits of these measures in reducing the transmission of viruses to be higher than the associated costs10,11. Thus, in this paper, we sought to understand whether personality factors and emotional intelligence (EI) contributed to the evaluation of the benefits and costs of social distancing behaviors during the COVID-19 pandemic.

According to Maharaj and Kleckzkowski12, when facing an epidemic outbreak, the authorities can either choose to adopt a strict control, by drastically reducing social contacts as soon as the disease is detected (as done by China, for example), or give up control and let the epidemic run its course (as done by Sweden, for example). Regardless of the choice made by governments, the analysis of the benefits and costs of social distancing is a key element in the development of the public health protection strategies of individual countries. Christner and colleagues13 proposed that social distancing can be conceptualized according to two different approaches: (a) a selfish approach, in which fear of infection or punishment are motivations to adhere to governmental recommendations14, or (b) a prosocial approach, in which social distancing represents a form of altruistic behavior aimed at preserving the well-being of other people15. In their study, the authors showed that factors such as moral judgement, moral identity and empathy for loved ones were positively correlated with the respect of social distancing norms13,16.

Since emotional intelligence (EI) has been positively linked to prosociality17,18, it seems reasonable to hypothesize that this ability might likewise play a crucial role in shaping protective health behaviors. One of the first definitions of EI has been advanced by Mayer and Salovey19, who identified four components: the ability to accurately perceive, appraise, and express emotion; the ability to access and/or generate feelings when they facilitate thought; the ability to understand emotions and emotional knowledge; and the ability to regulate emotions to promote emotional and intellectual growth. Since then, two different conceptualizations of EI have emerged in literature: while some researchers refer to EI in terms of an ability that can be assessed through performance tasks20, other scholars define it as a trait (i.e., a personal disposition involving the perception of one’s own emotional self-efficacy) that can be evaluated through self-report questionnaires21,22. According to Bandura23 social cognitive theory, perceptions of one’s own abilities (broadly known as perceived self-efficacy) stem directly from the experiences and opportunities that individuals have to test their skills. Most crucially, these perceptions of self-efficacy have been proposed as key self-regulatory mechanisms that allow people to effectively deal with interpersonal situations23. Thus, they may influence, among other things, the objectives that are selected, the persistence in reaching these objectives, the perception of the obstacles and opportunities available in the environment, and the degree of resilience in the face of adversities. Social distancing can be conceptualized as an effective way of regulating social interactions during health-crisis periods and is critically dependent on the ability to sustain personal, societal, and economic costs: therefore, it should be modulated by perceptions of emotional self-efficacy. Studying the relations between EI and social distancing is also important from a practical perspective, since researchers have now developed effective training programs to increase individuals’ EI levels24 and these interventions have proven to be useful in reducing depression, suicidal ideation, and state anxiety at the onset of the COVID-19 pandemic25.

In line with the trait model, high levels of EI have been associated with higher scores in a range of variables, including empathy, extroversion, openness to feelings, self-esteem, and life satisfaction17,26. Emotionally intelligent individuals are better prepared to perceive the need for help27 and to offer prosocial responses28. For instance, in a study conducted on 134 adolescents in a military training camp, Charbonneau and Nicol17 found a positive correlation between EI and prosocial behaviors, indicated by organizational citizenship behaviors such as altruism, conscientiousness, and civic virtue. Along the same lines, in the context of the COVID-19 pandemic, individuals with high EI might be more likely to adopt an altruistic approach to social distancing, by valuing social benefits (i.e., the protection of people’s health) more than personal costs (i.e., the restrictions to one’s own freedom). Despite the plausibility of the above hypothesis, previous studies have been primarily focused on the prediction of preventive behaviors29 and have therefore disregarded the assessment of the links between EI and the way in which participants evaluate the benefits and costs of social distancing.

In addition to the link with prosociality, the multifaceted nature of the EI construct suggests that its impact might be due to many different factors. For example, individuals who are high in EI might be better equipped to understand the nature of their own emotional reactions to the COVID-19 pandemic and identify appropriate strategies to cope with them25,30,30,31,32,33. Previous research demonstrated that EI has a fundamental importance in the successful management of health-related crises, such as that produced by the rapid spreading of COVID-1934,35. Furthermore, there is evidence showing that people with high scores in the other-emotion appraisal and use of emotion dimension of EI were more likely to adopt proactive (task-oriented and emotion-oriented) coping strategies during the COVID-19 pandemic29, whereas people with lower abilities of emotion regulation (as indicated by a higher use of expressive suppression and a lower use of cognitive reappraisal) were more likely to use maladaptive coping strategies (avoidance and self-punishment coping) in a study focused on suicidal behavior36. Higher EI levels may allow individuals to find better ways to cope with COVID-19 restrictions, such as dedicate themselves to hobbies and/or use the internet to stay in contact with friends and familiars37. Moreover, maintaining a correct social distance may be considered as an appropriate strategy to cope with the COVID-19 diffusion (just like washing hands frequently or wearing face masks): the evaluation of its costs and benefits should be therefore related to multiple EI dimensions.

Personality is another factor to consider when trying to unravel the mechanisms that might explain why people in at-risk contexts continue to engage in unsafe behaviors. In fact, both the studies conducted during previous pandemic emergencies and the more recent studies conducted during the COVID-19 outbreak showed that there is a relationship between personality traits and adherence to social distancing. For example, in a survey conducted during the A/H1N1 pandemic, it was found that personality factors such as general activity, interest levels and impulsivity contributed to the prediction of the use of protective and avoidance behaviors in the later stages of the epidemic38. Most of the research performed during the COVID-19 pandemic has however referred to the Big Five model of personality39, which encompasses five broadly-defined traits—namely, extroversion, agreeableness, conscientiousness, emotional stability, and openness to experience. In the last decades, this model has been validated and applied in different cultures and contexts, becoming one of the leading approaches to personality40. Previous studies have explored the role of the five personality traits in relation to the use of protective behaviours41,42. In particular, in the present study we focused our attention on two personality traits that appear to have opposite effects on the use of preventive behaviors—namely, emotional stability and extroversion. In fact, the common pattern was that emotional stability was positively associated with the adoption of social distancing behaviors43, whereas extraversion was negatively associated (i.e., high level of extraversion predicted a failure to respect social distancing recommendations44). Yet, there is no data directly assessing the impact of personality traits on the evaluation of the benefits and costs of social distancing.

Given the paucity of relevant evidence, the present study aimed at investigating the relations between personality traits, EI and the subjective judgment of the benefits and costs of social distancing during a specific period of the COVID-19 pandemic (i.e., between March and April 2021). To this purpose, we used a series of questions devised by Xie and colleagues2, intended to evaluate the way in which people perceive the positive and negative aspects associated with COVID-19-related restrictions to interpersonal contacts. In addition to ascertaining the independent roles of EI and personality traits, we were also interested in determining whether EI mediated the effects of personality traits on informed decision processes. As mentioned above, trait EI refers to a broad array of emotion-related dispositions and self-perceptions (typically measured via self-report questionnaires) and have been conceptualized as a distinct, compound personality construct located at the lower levels of the Big Five taxonomy45,46. Petrides and colleagues21, for example, found that trait EI emerged as an oblique factor that was significantly associated with the classical Big Five factors. Given these findings, we reasoned that the impact of personality dimensions on the subjective assessment of the benefits and costs of social distancing might be, at least in part, mediated by the way in which people perceive their abilities to recognize and regulate emotions.

Our predictions were as follows: (1) we expected that EI and emotional stability should be positively associated with the ability to prioritize the benefits of social distancing over its costs, whereas extroversion should be negatively associated; (2) we expected that the associations between personality traits (emotional stability and extroversion) and the evaluation of the benefits and costs of social distancing should be partially mediated by EI. Less centrally for the present purposes, we also assessed the potentially confounding role of age. Based on previous findings showing that age predicted the adherence to COVID-19 preventive behaviors47,48,49,50, we expected that older adults should be better able than younger adults to prioritize the benefits of social distancing; however, the effects of EI and personality traits should be independent from those of age.

Method and materials

Participants

The required sample size was established on the basis of current methodological standards. For regression analyses with six or more predictors, Tabachnick et al.51 suggested that a minimum of 10 participants per predictor is appropriate. However, to achieve sufficient power to detect even small effect sizes, approximately 30 participants per predictor are recommended52 (Wilson & Van Voorhis, 2007). With 7 predictors included in the regression analysis (see later), these guidelines suggest a minimum sample size of 70 participants for detecting high or medium effect sizes and 210 participants for detecting small effect sizes.

To fulfill these requirements, we collected data from March to August 2021. However, for the present purposes, we only analyzed the data of participants who took part in the study between March and April 2021. This choice was justified by the fact that, according to the epidemiological data reported by the Italian National Institute of Health, the circulation of the COVID-19 virus during these two months (as indicated by incidence rates), and hence the perceived importance of social distancing measures, were considerably higher than those observed in the following months (i.e., from May to August)53. In addition to having completed the survey during the selected period of interest, two other eligibility criteria were a) being an Italian resident and b) having responded to all the questions (i.e., no missing data).

At the end, 223 participants were recruited—of which 192 females and 31 males. The mean age of our participants was 30.78 ± 9.97 (range: 19–58 years). Education levels were distributed as follows: 98 (43.9%) participants had a high-school degree, 105 (47.1%) had a bachelor’s degree, 14 (6.3%) had a postgraduate degree and 6 (2.7%) had a specialization or PhD degree.

Instruments

Benefits and costs of social distancing

To assess the benefits and costs of social distancing, we translated and adapted to the Italian language the nine items used by Xie et al.2 Two of the Authors of the present paper (AS and PS) independently translated the items of the English version into Italian. Their translations were then compared to discuss discrepancies and reach an agreement on discordant translations, leading to the generation of a preliminary Italian version of the questionnaire. Next, one of the Italian-English bilingual authors (CRA), blind to the original English version, checked the translated items through a back-version procedure. The original and the back-translated versions were finally reviewed and modified based on comments by all the authors, reaching the final consensual version (which is freely available online at the following DOI: 10.5281/zenodo.13774770). Comprehension was ensured by presenting the statements to a small group of Italian university students who attended one of the authors’ (CRA) classes (N = 15, students’ comprehension tested by AS).

Five statements investigated the perceived benefits of social distancing (e.g., “Social distancing stops coronavirus from spreading around”; Cronbach’s α = 0.62), whereas two statements investigated perceived costs (e.g., “ Social distancing makes people lose their jobs.”; Cronbach’s α = 0.68). The choice of using a subset of the four items devised by Xie et al.2 to assess perceived costs was based on evidence indicating that the excluded items showed small, non-significant correlations with the selected items (which were instead highly correlated to each other). Participants responded to each statement on a four-point Likert scale (from “I don’t think it is true” = 0 to “It is very true” = 3). We firwhich is freely available online at the followingst summed the scores within each category and then calculated the differences between perceived benefits and costs. The greater the difference, the higher the participants’ ability to prioritize benefits over costs.

Emotional Intelligence

Emotional intelligence was measured with the Italian version of the Emotional Intelligence Scale (EIS54,55). This 24-item scale includes three subscales: (a) Evaluation and expression of emotions in relation to others (e.g., “I know what other people are feeling just by looking at them”; Cronbach’s α = 0.83), (b) Evaluation and expression of emotions in relation to self (e.g., “I am aware of the emotions I feel”; Cronbach’s α = 0.77), and (c) Regulation and use of emotions (e.g., “I present myself in a way that makes a good impression on others”; Cronbach’s α = 0.76). Participants responded on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Personality traits

Personality traits were assessed with the Italian version of the 10-item Big Five Inventory (BFI256,57). This instruments allowed us to assess five personality dimensions: Agreeableness (e.g., “I see myself as someone who is generally trusting”; Cronbach’s α = 0.20), Conscientiousness (e.g., “I see myself as someone who does a thorough job”; Cronbach’s α = 0.30), Emotional stability (e.g., “I see myself as someone who is relaxed, handles stress very well”; Cronbach’s α = 0.84), Extroversion (e.g., “I see myself as someone who is outgoing, sociable”; Cronbach’s α = 0.66) and Openness (e.g., “I see myself as someone who has an active imagination”; Cronbach’s α = 0.44). Participants were asked to indicate the extent to which they agreed with each statement, using a five-point Likert scale ranging from “strongly disagree” to “strongly agree”. Given the low values of internal consistency reported above, we also performed an exploratory factor analysis (principal axis factoring, oblimin rotation) to verify whether the factorial structure reported by Guido et al.56 was replicated in our data. To establish the number of factors, we adopted a parallel analysis58. The results indicated a four-factor solution (see Table 1 in the Supplementary File). The first, second and fourth factors replicated the Emotional stability, Extroversion and Openness factors illustrated by Guido et al.56. The third factor was saturated only by item 3 (“I see myself as someone who tends to be lazy”; reverse-scored) and could therefore represent the Conscientiousness dimension56. Items 7, 2 and 8 had high uniqueness values and did not load on any factor (using the 0.4 threshold level). Considering these results, only the Emotional stability and Extroversion factors were included in the following statistical analyses.

Table 1 Descriptive measures for the total sample and separately for males and females, together with the results of Mann–Whitney U tests. Values in brackets are standard deviations.
Full size table

Procedure

The survey was prepared using Psytoolkit59 and disseminated through different social media (including Facebook, Instagram, Twitter, LinkedIn, Telegram, and WhatsApp), in line with the Italian government’s recommendations on limiting face-to-face interactions. Data were collected between March and August 2021, although for the present study we limited our analyses to the first two months (i.e., March and April 2021). We used a snowball sampling technique: hence, the links were initially shared with a sample of university students, who were encouraged to pass them on to others, with a focus on recruiting the general population. The present study adhered to the recommended standards for conducting and reporting web-based surveys—i.e., the Checklist for Reporting Results of Internet E-Surveys (CHERRIES60). In particular, the questionnaire was open to each visitor of the site, was advertised as research aimed at studying the relations between personally, emotional intelligence and social distancing, participation was voluntary, and no financial incentives were offered. The research was approved by the Ethical Committee of the University Sapienza of Rome (Protocol N.0002194) and all respondents signed an informed consent before participating.

Data analysis

A preliminary inspection of data showed that all variables were normally distributed since asymmetry values ranged from − 0.42 to 0.21. Gender differences were assessed with the non-parametric Mann–Whitney U test, since the two groups were not comparable in size. Pearson’s correlations were used to examine the associations between the variables. To determine the concurrent predictors of the ability of prioritizing the benefits of social distancing, we ran a hierarchical regression analysis. Age and education were entered in the first step to remove their potentially confounding effects; then, personality traits were entered in the second step and EI was entered in the third step. We followed this order because we wanted to determine whether EI accounted for a unique portion of variance, over and above that explained by personality. Lastly, to determine whether EI mediated the associations between personality traits and the benefits-and-costs evaluation of social distancing behaviors, we used the mediation analysis module available in the JASP software61, which is based on the Lavaan package62 and allows the simultaneous consideration of multiple mediators. The significance of total, direct and indirect effects was estimated by applying bootstrapping and the 95% confidence intervals were calculated using the bias-corrected percentile method (1000 resamples; Maximum Likelihood estimator), following Biesanz et al.63. The consensus is that, if the confidence intervals do not contain zero, then the indirect effect can be considered significant64.

To quantify the evidence in support of the alternative hypotheses, we computed Bayes factors (BF10). According to van Doorn et al.65, Bayes factors between 1 and 3 indicate weak evidence, Bayes factors between 3 and 10 indicate moderate evidence, and Bayes factors greater than 10 indicate strong evidence. For correlational analyses, the alternative hypothesis was that two variables were significantly associated (evaluated against the null hypothesis of no correlation); for regression analyses, the alternative hypothesis was that a given variable should enter the final equation (evaluated against the null hypothesis of exclusion).

Results

Table 1 reports descriptive statistics for the measures examined in the present study, separately for males and females. Significant differences were only observed for the Emotional stability subscale of the BFI. As compared to females, males were emotionally more stable and were better able to regulate and use their emotions.

Table 2 reports Pearson’s correlations between the main variables. Differences in the evaluation of the benefits and costs of social distancing were positively and significantly correlated with age, the Emotional stability subscale of the BFI and the Emotions in relation to self and Emotion regulation subscales of the EIS. Thus, the ability to prioritize the benefits of social distancing over its costs was higher in participants who were older, had greater levels of emotional stability, were better able to evaluate and express their own emotions, and had greater competence in regulating and using emotions. The BF10 Bayes Factors were 1.77 for age, 35.50 for Emotional stability, 4.23 for Emotions in relation to self and 18.17 for Emotional regulation. In addition, the differences between the benefits and costs of social distancing were negatively associated with the Extroversion subscale of the BFI. Thus, participants who had higher levels of extroversion were less able to prioritize the benefits of social distancing (compared to participants who had lower levels of extroversion). The BF10 Bayes Factor for this correlation was 7.98, indicating moderate evidence in support of the alternative hypothesis.

Table 2 Pearson’s correlations in the whole sample (N = 223).
Full size table

Table 3 illustrates the results of the hierarchical regression analysis. The whole model was significant [F(6, 213) = 6.67, p < 0.001] and all steps explained a significant (or marginally significant) portion of variance. Specifically, age explained 3.5% of the variance [F(1, 218) = 7.83, p = 0.006], Emotional stability and Extroversion explained 9.5% of the variance [F(2, 216) = 11.83, p < 0.001], and the three EI dimensions explained 2.8% of the variance [F(3, 213) = 2.38, p = 0.070]. As can be noted from Table 3, the differences between the benefits and costs of social distancing were negatively predicted by the Extroversion subscale of the BFI and positively predicted by the Emotion regulation subscale of the EIS. In addition, the contribution of the Emotional stability subscale of the BFI was marginally significant (p = 0.075). The Bayesian analysis (performed with JASP42) confirmed that the best model was that including the Extroversion and Emotional stability subscales of the BFI and the Emotion regulation subscale of the EIS—BF10 = 6402.25 (relative to a null model). The Bayes Factors for the β coefficients, averaged across all models, were 182.63 for Extroversion, 7.57 for Emotion regulation and 1.97 for Emotional stability, suggesting very strong evidence in favor of the inclusion of the first variable in the final model, moderate evidence in favor of the inclusion of the second variable, and weak evidence in favor of the inclusion of the third variable.

Table 3 Hierarchical regression predicting the differences between the benefits and costs of social distancing.
Full size table

Lastly, Fig. 1 illustrates the results of the mediation analyses. Both the total and the direct effects of Emotional stability and Extroversion on the differences between the benefits and costs of social distancing were significant (for the total effects: β = 0.10, SE = 0.02, z = 3.83, p < 0.001, 95%CI [0.04, 0.15] for Emotional stability and β =  − 0.11, SE = 0.03, z = − 3.39, p < 0.001, 95%CI [− 0.16, − 0.04] for Extroversion, respectively; for the direct effects: β = 0.06, SE = 0.03, z = 1.98, p = 0.047, 95%CI [− 0.01, 0.13] for Emotional stability and β = − 0.13, SE = 0.03, z =  − 3.94, p < 0.001, 95%CI [− 0.019, − 0.06] for Extroversion, respectively). More interestingly, the total indirect effect of Emotional stability was also significant (β = 0.04, SE = 0.01, z = 2.56, p = 0.010, 95%CI [0.01, 0.07]), whereas the total indirect effect of Extroversion approached the significance level (β = 0.02, SE = 0.01, z = 1.88, p = 0.060, 95%CI [0.00, 0.05]). Looking at the individual effects (see Fig. 1), it turned out that the indirect effects of Emotional stability and Extroversion on the differences between the benefits and costs of social distancing were both mediated by the Emotion regulation dimensions of EI (β = 0.03, SE = 0.01, z = 2.03, p = 0.041, 95%CI [0.00, 0.06] for Emotional stability and β = 0.02, SE = 0.01, z = 1.82, p = 0.067, 95%CI [0.00, 0.04] for Extroversion).

Fig. 1
figure 1

Mediation models assessing the indirect effects of extroversion and emotional stability on the benefits-and-costs evaluation of social distancing. Path values represent standardized coefficients.

Full size image

Discussion

The respect of preventive behaviors such as social distancing has been constantly emphasized during the COVID-19 emergency, because of their importance in limiting the spread of the disease. Thus, health recommendations have been widely and intensively disseminated through traditional and online media, with the intention of reinforcing people’s compliance66. In this context, we focused on the analysis of the psychological mechanisms which influenced the adherence to and the implementation of social distancing behaviors in Italy between March and April 2021. In previous pandemics, the degree of adherence to quarantine was primarily determined by the expected benefits of precautionary measures and the perceived risk of contracting the disease67. For example, during the past SARS and Ebola outbreaks, it was seen that people were more likely to adhere to quarantine protocols when they judged this behavior as beneficial in reducing transmission10. More recently, Hansen and colleagues11 found that the perceived susceptibility towards COVID-19 (the subjective probability of getting the SARS-CoV2) and the perceived benefits of social distancing were the most significant predictors of compliance with governmental recommendations in U.S. adults. It is therefore evident that the respect of protective behaviors arises from complex judgments in which different aspects are weighted, including the perceived risk to one’s own health, the potential danger of contracting the disease, the benefits to others’ health, and the costs in terms of subjective freedom. Thus, investigating the psychological processes associated with the evaluation of the benefits and costs of social distancing represents a necessary starting point toward understanding individual differences in the adoption of preventive behaviors9.

As expected, our results showed that the tendency to prioritize the benefits of social distancing over personal costs was positively associated with participants’ age. That is, older participants were better able to judge the benefits of social distancing as more important than its costs. This result provides a reasonable theoretical explanation for the well-known finding that younger individuals were less likely to comply with preventive measures and avoid social closeness during the COVID-19 pandemic—as compared to older individuals47. Thus, while previous studies have already demonstrated that age predicted the adoption of protective behaviors48,49,50, our data move a step forward in showing that the impact of this demographic variable might depend on individual differences in the way in which people judge the benefits and costs of social distancing.

Turning to the focus of the present study, personality traits are known to influence the perception of external stressors, including those conveyed by health crises, contributing to different behavioral responses to epidemics68. Our results support this proposal by showing that the tendency to prioritize the benefits of social distancing was positively influenced by emotional stability. In agreement, Han and colleagues43 reported that higher levels of emotional stability were associated with a closer adherence to social distancing, and more generally with a higher engagement in preventive behaviors. In addition, we found a strong negative correlation between extroversion and individual differences in the evaluation of the costs and benefits of social distancing, such that highly extroverted participants were less likely to evaluate the benefits of social distancing as more important than its costs. This result could explain why in previous studies individuals with high levels of extroversion exhibited a lower compliance with social distancing recommendations during the COVID-19 epidemic44. These individuals might judge the personal costs implied by the reduction of social closeness too high, when compared with the benefits, and too difficult to accept.

While the role of personality traits in facing the COVID-19 pandemic has been widely investigated in literature, empirical data concerning the impact of EI on the acceptance of preventive behaviors are scant. In this respect, our study revealed that several EI dimensions were associated with the way people judged the benefits and costs of social distancing. Specifically, participants who were better at understanding their own emotions and the situations that triggered them, and those who were better at regulating their emotions, were more likely to prioritize the benefits of social distancing. The significant correlation with the EI dimension assessing knowledge of one’s own emotions may be explained by suggesting that individuals with high EI may be better able to recognize the negative emotions and feelings triggered by the COVID-19 outbreak and, as a result, more likely to use appropriate coping strategies. Consistent with this proposal, previous research documented that awareness and acceptance of one’s own emotions were positively related to the use of adaptive coping strategies both in adults and in children32,33. In agreement, Prentice et al.29 showed that higher levels of self-emotion appraisal predicted a more frequent adoption of task-oriented and emotion-oriented coping strategies during the COVID-19 pandemic. A similar explanation might be proposed to account for the correlation with the EI dimension assessing emotion regulation. Ong and Thompson36, for example, found that the ability to regulate one’s own emotions by reevaluating the situation represented an adaptive method to cope with stressful circumstances and reduced the risk of suicidal behavior. Cognitive reappraisal might likewise be useful to deal with COVID-19-related stress and appreciate the potential benefits of social distancing behaviors. In agreement, a large multi-country study by Wang et al.37 showed that two brief reappraisal interventions, based on reconstrual or repurposing, were effective in reducing the negative emotions induced by the COVID-19 pandemic, without reducing intentions to practice preventive health behaviors.

Contrary to expectations, there was no significant association between the benefit/cost evaluations of social distancing and the EI subscale that measures the evaluation and expression of emotions in relation to others. This result is surprising, because high levels of empathy were theoretically expected to increase participants’ ability to adopt an altruistic approach to social distancing and contribute to a closer adherence to the recommended behaviors13. A potential explanation might be provided by the well-known distinction between cognitive empathy (i.e., the process of adopting another’s psychological point of view, which is closely related to theory of mind) and affective empathy (the capacity of experiencing affective reactions to the observed experiences of others69). Self-report questionnaires such as the one employed in the present study are better suited to evaluate the cognitive understanding of others’ emotions. However, the affective component of empathy might influence more closely the evaluation of the benefits and costs of social distancing. Previous studies have indeed demonstrated that affective empathy predicted self-isolation behaviors, social distancing and wearing of face masks during the COVID-19 pandemic16,70.

Importantly, we also found that the EI dimensions assessing knowledge and regulation of one’s own emotions mediated the associations between personality dimensions (emotional stability and extroversion) and the cost/benefit evaluations of social distancing. Thus, individuals with high levels of emotional stability and extroversion were better able to understand and regulate their own emotions, and this in turn allowed them to prioritize the benefits of social distancing over its costs. These results are fully consistent with the view that EI can be considered as a distinct but compound construct that lies at the lowest levels of the Big Five hierarchy21. In fact, when entered after personality factors in the regression analysis, EI accounted for an additional (albeit marginally significant) portion of variance in the prediction of the costs-and-benefits difference scores, suggesting a partial differentiation between the two domains. As discussed above, EI was originally operationalized as a form of intelligence that should be distinguished from personality traits. However, the widespread use of self-report instruments for its assessment, including the EIS scale used in the present study, led researchers to propose a distinction between trait and ability EI71. In this respect, our findings align with previous conclusions indicating a close relationship between trait EI and personality dimensions21,72.

How can we account for the fact that the impact of personality traits on the judgement of the benefits and costs of social distancing was mediated by emotion-related dispositions and self-perceptions? We believe that a key suggestion is provided by recent evidence indicating that EI is significantly related to the use of both task-oriented and emotion-oriented coping strategies29,36. Lockdown and social distancing measures caused fear and anxiety in the general population, because they were unprecedented and imposed a strong limit on social activities. However, emotionally stable individuals who are high in EI may be better able to appreciate the situation and identify appropriate tasks to cope with restriction. For example, they may take preventive measures as an opportunity to engage in hobbies that cannot be pursued in normal situations (a form of task-oriented coping). Along the same line, extroverted individuals who are high in EI may use the lockdown period to spend more time with family or relatives (a form of emotion-oriented coping), or they may find different ways to stay in touch with their friends, for instance by using social media or platforms of video communication. Additional longitudinal studies are needed to assess the validity of this serial mediation hypothesis.

Limitations and practical implications

The present study has some limitations to bear in mind. First, the cross-sectional nature of the design precluded us the possibility to determine the causal nature of the reported associations. Second, only self-report measures were used, which are potentially affected by social desirability71, and all data were collected online, to comply with the recommendations to avoid direct contact with respondents during the COVID-19 pandemic. Furthermore, the questionnaire borrowed from Xie et al. (2020) was not fully validated in the Italian population and three of the five subscales of the Big Five Inventory exhibited low internal validity (as estimated with Cronbach’s αs), resulting in the exclusion of the corresponding factors from statistical analyses. Third, the study was conducted between March and April 2021. Although we chose these two months because the data reported by the Italian National Institute of Health indicated the presence of a slight peak of infections, the overall circulation of the COVID-19 virus was significantly reduced in Italy during this period and risk perception was considerably lower73. In these less cogent conditions, the evaluation of the benefits and costs of social distancing might have been quite different, as compared with the initial phases of the pandemic, in the years 2019–2020. Finally, our sample was not representative of the general Italian population and was not balanced in terms of gender, with a strong prevalence of females. The available data indicate that, overall, men were less likely to adopt preventive behaviors against COVID-19, including handwashing, social distancing, and use of face74. It would be therefore interesting for future studies to determine whether the direct and indirect relations between personality traits, EI and social distancing are apparent and equally strong in males and females.

Despite the above limitations, the present study reached important conclusions about the factors that may determine how people perceive the adoption of preventive behaviors during a pandemic. Our results highlighted the critical role of personality traits and EI in prioritizing the benefits of social distancing and provide useful indications for developing appropriate communication strategies tailored to the general population. From a practical perspective, these findings are important because they suggest that EI trainings aimed at increasing emotional awareness and regulation could be an effective way to increase people’s adherence to social distancing recommendations. A recent meta-analysis identified 24 studies aimed at increasing individual-level EI among healthy adults and concluded that trainings were moderately effective in producing significant changes between pre- and post-measurements24. Furthermore, Persich et al.25 examined whether training in EI skills, provided before the pandemic, could serve as a protective intervention against mental illness during the COVID-19 pandemic. To this purpose, they compared the effectiveness of a web-based EI training program versus a non-emotion-focused placebo program. The EI program took approximately 9–11 h to complete (distributed across several days) and targeted key emotional skills such as perceiving emotions in oneself and others, understanding emotions, using emotions in adaptive ways, and regulating one’s own and other’s emotions. The results showed that, relative to individuals who were assigned to the placebo program, those who participated to the EI training program reported lower levels of depression, suicidal ideation, and state anxiety in a follow-up assessment scheduled after the onset of the COVID-19 crisis. Although these findings need replication and extension, it appears that EI can be easily trained and that, in the near future, these interventions may represent a promising way to deal with the negative emotional impact of novel pandemics.

However, the implications of our findings are not necessarily limited to the management of health crises. People engage in the evaluations of the costs and benefits of their behaviors in a wide variety of contexts. In the contemporary debate, an important area of study is represented by pro-environmental choices and behaviors, which are almost always directed at benefiting other people or the environment, at the expense of personal interests7. De Groot and Steg8, for example, discussed two strategies to promote pro-environmental behaviors, by either increasing the saliency of altruistic and biospheric values (thereby reducing the strength of egoistic values) or by making anti-environmental, egoistic values more compatible with pro-environmental, altruistic values. Interestingly, individual differences in personality traits and EI have been implicated in the adoption of pro-environmental attitudes and behaviors. Specifically, it has been demonstrated that high levels of trait EI have positive influences on both climate change perception and pro-environmental behaviors, through the mediation of locus of control75 and connectedness to nature76. Similarly, a recent meta-analysis found that openness, agreeableness, conscientiousness, and, to a lesser extent, extraversion emerged as the strongest correlates of pro-environmental attitudes and behaviors77. It would be therefore interesting to ascertain whether EI mediates the influence of personality traits on the evaluation of the costs and benefits of pro-environmental actions. If this should be the case, then the introduction of EI training programs in educational settings might provide a novel way to help the young generations to prioritize the benefits of ecological behaviors—one of the most important challenges imposed by climate changes. (Supplementary Information).