Personality regulation of decisions on physical distancing: Cross-cultural comparison (Russia, Azerbaijan, China).

The present study investigated cross-cultural comparison of the personality variables (rationality, risk readiness, empathy, Dark Triad traits, implicit theories of emotions) in predicting decisions on physical distancing during the COVID-19 pandemic. The sample included 1077 participants from Russia, Azerbaijan, and China. After reporting if they trust the media, participants chose from different reasons why they wear or don't wear a mask: care for self vs others, risk for oneself vs others, autonomy for oneself vs others, risk estimation, law-abidingness; then participants completed questionnaires. We expected people from collectivistic countries to make decisions based on care for others and people from more individualistic countries - on care for self and autonomy. The results revealed a different trend: participants from all countries chose care for self more frequently than other reasons. This was most prevalent in China, less - in Azerbaijan and less so - in Russia. Rationality and empathy were positive predictors of decisions to wear a mask, risk readiness and psychopathy were negative predictors, the role of narcissism depended on the country. Implicit theories of emotions correlated with empathy in China and Azerbaijan. These two measures predicted the choice of "care for others" over "care for self" in all countries.

Introduction

The COVID-19 pandemic has significantly changed people's living conditions. Wearing a face mask has become a mandatory requirement of physical distancing and health protection. China was the first country to face this new threat and to introduce state regulation to control the spread of infection.

For this cross-cultural comparison of decision to wear a mask or to remain without it when going to public places we chose three countries that differ in tradition, a dominant religion and a degree of state regulation on the implementation of the restrictions – Russia, Azerbaijan and China.

The study took place from the end of April till the end of May 2020 when China had announced overcoming pandemic but wearing masks remained mandatory. In Azerbaijan the pandemic just started and in Russia it was reaching its peak.

It is reported that on the collectivism-individualism scale where higher scores refer to individualism China has the score of 20, Azerbaijan – 22, Russia – 39 out of 100 (Web Reference 1, 2020). The patterns of personality regulation of decision-making could be more similar in countries with closer indices.

Azerbaijan and Russia as the former republics of one country can be considered more similar in context variables, although the difference is multiconfessionality in one country and the predominance of Islam in another. Azerbaijani men showed higher intolerance of uncertainty than Russian men (Kornilova, Chumakova, & Izmailova, 2015).

The main research question in this study is to what extent prosocial behavior is guided by individual differences in a number of personality characteristics.

Individual differences in prosocial behavior

The threat of infection with an unknown disease was studied in psychology using the “Asian Disease” problem, where one should choose between the two programs of saving people from an epidemic on a ship. Framing effects in this situation were discussed within the framework of the cognitive approach (dual process theories). Later Kahneman introduced the concept of emotional framing (Kahneman, 2011).

Jordan, Yoeli, and Rand (2020) showed that prosocial framing was more effective than self-focused framing in promoting physical distancing intentions only early on in the pandemic.

Emotional response to the coronavirus, not moral or political orientation, predicts compliance with the norms (Harper, Satchell, Fido, & Latzman, 2020). The role of emotions and their interaction with attitudes towards uncertainty and risk in decision making is in its early stages of exploration under the intelligence-affect unity framework (Anderson, Carleton, Diefenbach, & Han, 2019; Kornilova, 2016).

In everyday life, decisions related to taking care of one's own health and health of others can be based on different reasons: prosocial or selfish behavior, autonomy for oneself or others, risk underestimation.

Implicit theories (IT)

The roots of the concept of IT go the works of G. Kelly, R. Tagiuri, and others. It emerged in the domain of intelligence (Sternberg et al., 2000) and as “lay theories” (A. Furnham). Functioning implicitly, they influence interpretations of situations and behavioral strategies. C. Dweck has made a distinction between fixed IT and malleable IT based on the perceived nature of attributes (Dweck & Leggett, 1988; etc.).

IT of emotions (Tamir, John, Srivastava, & Gross, 2007) as beliefs about the controllability of emotions might have important influence when emotional regulation is essential for decision making. Fixed IT of emotions correlate with avoidance of negative experiences (Kappes & Schikowski, 2013).

Nationwide research in China (Qiu et al., 2020) and meta-analysis of data from all over the world (Luo, Guo, Yu, Jiang, & Wang, 2020) have shown a prevalence of depression and anxiety during the pandemic. Presumably, ability to manage one's emotions might improve adaptation to a changed reality.

Risk readiness

The categorization of risks doesn't determine the direction of behavior by themselves, neither do personality dispositions. But the interaction of situational and personality factors influences judgment and risk-taking. In recent years, researchers work on improving risk literacy, in particular, in medical risks (Gigerenzer, 2015).

The concept of risk intelligence (Evans, 2012) emphasizes cognitive control of judgments on risk. Perception of threat is influenced by its representation in the media (Slovic, 2000), and individual differences (Chauvin, 2018; Kornilova, Pavlova, Bogacheva, Kamenev, & Krasavtseva, 2020). Moreover, the presence of affect impairs judgment on the probability of events (Slovic & Peters, 2006).

The Dark Triad (DT)

The DT traits include Machiavellianism, subclinical psychopathy, and narcissism which have different etiologies and closeness of interconnections in different countries (Jonason et al., 2017). Аll three aversive traits characterize the so-called “unstable” emotional core and are related to the need for social domination and utilitarian solutions of moral dilemmas.

Narcissism might be the brightest from all DT traits because it is positively associated with intelligence and tolerance of uncertainty (Krasavtseva & Kornilova, 2019). Some data suggest that narcissism is a more “distinctive” trait (Savard, Simard, & Jonason, 2017).

Dark Triad traits predict lower adherence to COVID-19-related restrictions (Zajenkowski, Jonason, Leniarska, & Kozakiewicz, 2020; Zettler, Schild, Lau, & Böhm, 2020).

Empathy

Emotional regulation is studied less in eastern cultures. Indirect conclusions stem from the emotional attitude to another person and how empathy changes when moving along the collectivism-individualism scale. Empathy differs depending on whether the person we experience empathy for belongs to our in-group or not (Abu-Akel, Fischer-Shofty, Levkovitz, Decety, & Shamay-Tsoory, 2014). Assuming that relationships between people in Chinese society are closer than in individualistic cultures, we can expect that during the pandemic Chinese people will show high emotional empathy.

Empathy to the most vulnerable to the coronavirus was related to compliance with social distancing in Western countries (Pfattheicher, Nockur, Böhm, & Sassenrath, 2020).

Empathy, risk readiness, and rationality regulate decision making through the interaction with a person's world view which is largely influenced by the media. Contextual variables of cultural traditions influence behavior, and since health threats are probabilistic by nature, the role of emotional and personality factors in decision making is important to investigate.

Hypotheses

In countries with higher collectivism (China, Azerbaijan), choices based on caring for others (not just caring for oneself) will be frequent while autonomy-based choices will be less frequent.

Personality measures of empathy, rationality, and malleable IT of emotions are positive predictors of physical distancing compliance while risk readiness and DT traits are negative predictors.

Methods

Participants

The overall study sample included 1077 online participants recruited using snowball technique.

Russian sample: 308 participants 18 to 80 y.o. (M = 32.3, SD = 11.71), 80% female.

Azerbaijani sample: 352 participants 17 to 74 y.o. (M = 30.5 SD = 10.40), 74% female.

Chinese sample: 417 participants 17 to 52 y.o. (M = 25.1, SD = 7.16), 59% female.

Procedure

After providing informed consent to participate in the study, participants were asked if they trust information from the media that they might be asymptomatically infected with coronavirus. Then they made choices in verbal tasks-vignettes and filled out questionnaires.

Measures

Verbal tasks

We developed 5 verbal tasks – situations where decisions on whether to wear a mask (reasons A1 and A2) or not (B1 and B2) were made by choosing from four given reasons. We used a quasi-experimental design: all participants received one of four tasks with different framing in reasons for and against wearing a mask and one other tasks that was as a control task for every participant (see example Task 2, Appendix 1). In different tasks, reasons for wearing a mask represented the following categories: 1, 2 and 3) Care for Self vs Others, 4) High risk for Self vs Law-abidingness, 5) High risk for others vs Law-abidingness. Reasons for choice not to wear a mask also varied across tasks: 1 and 4) Autonomy for Oneself vs Others, 2) Risk for Oneself vs Others, 3) Risk Underestimation vs Autonomy for Others, 5) Autonomy for Oneself vs Risk Underestimation.

Questionnaires

Implicit theories of emotions (ITE) scale (Tamir et al., 2007) includes four items, two of them refer to fixed ITE and two – to malleable ITE. Higher scores correspond to malleable ITE. Cronbach's alpha for this scale are: Russia (α = 0.692), Azerbaijan (α = 0.633), China (α = 0.630).

Questionnaire of Cognitive and Affective Empathy (QCAE) consists of 31 items divided into 5 subscales and 2 scales: perspective taking and online simulation (cognitive empathy scale), emotion contagion, proximal responsivity, and peripheral responsivity (affective empathy scale) (Reniers, Corcoran, Drake, Shryane, & Völlm, 2011). We used Chinese (Wang & Su, 2019) and Russian versions.

Personality Factors of Decision-making Questionnaire measures risk readiness and rationality (Kornilova, Chumakova, Kornilov, & Novikova, 2010), and was validated on Russian and Azerbaijani samples.

The “Dirty Dozen” questionnaire (Jonason & Webster, 2010) measures DT traits: subclinical narcissism, subclinical psychopathy, and Machiavellianism. It was validated on Russian and Azerbaijani samples (Kornilova, Kornilov, Chumakova, & Talmach, 2015; Kornilova, Zirenko, & Guseynova, 2017).

The two last measures weren't administered in China. All measures in Azerbaijan were presented in Russian language.

Results

Descriptive statistics

Descriptive statistics on personality measures and results of Kruskal-Wallis Test are presented in the table below.

Cultural groups showed significant differences in all measures except for affective empathy and narcissism (Table 1 ).

Table 1

Descriptive statistics for personality variables.

	Russia		Azerbaijan		China		χ2	p
	M	SD	M	SD	M	SD
1. ITE	3.50	0.87	3.61	0.81	3.43	0.72	10.70	.005
2. Perspective taking	28.16	4.25	29.34	3.89	28.64	4.77	11.75	.003
3. Online simulation	25.27	3.37	26.07	3.20	27.39	3.68	68.92	.000
4. Emotion contagion	10.03	2.26	10.09	2.04	11.43	2.28	99.76	.000
5. Proximal responsivity	11.16	2.00	11.85	1.81	11.12	1.93	26.67	.000
6. Peripheral responsivity	11.48	2.13	11.09	1.78	10.28	2.21	62.42	.000
7. Cognitive empathy	53.42	6.29	55.41	5.93	56.03	7.55	33.03	.000
8. Affective empathy	32.66	4.88	33.03	4.03	32.83	4.90	0.70	.704
9. Empathy	86.09	9.33	88.44	8.09	88.87	9.86	20.82	.000
10. Risk readiness	1.20	4.33	2.25	3.93	–	–	9.52	.002
11. Rationality	3.49	3.69	4.75	3.46	–	–	21.70	.000
12. Machiavellianism	9.64	3.97	8.90	4.12	–	–	6.65	.010
13. Psychopathy	7.72	3.53	7.18	3.33	–	–	4.25	.039
14. Narcissism	13.16	3.98	12.78	4.03	–	–	1.43	.232

Note. Dashes indicate variables for which we haven't collected data.

Frequencies of choices on trusting the media

Participant groups from different countries differed in theirs answer on trusting the media (χ 2=176.55, p < .001) (Table 2 ). The answer “I don't trust the media” was more frequently observed in Azerbaijan where the pandemic was just starting (hundreds of cases daily), while “I trust the media, but I couldn't be infected” – in China where the pandemic has decreased.

Table 2

Trusting the media about the possibility of being asymptomatically infected with COVID-19.

	I don't trust the media	I trust the media and I could be infected	I trust the media but I couldn't be infected
Russia	9.7%	85.8%	4.5%
Azerbaijan	18.8%	75.6%	5.6%
Chinа	5.0%	77.3%	17.2%

Frequencies of choice alternative preferences to wear a mask or remain without it

Using chi-square test, we examined the differences between choices of reasons for wearing or not wearing a face mask in cultural groups (see Fig. 1 ).

Fig. 1

Frequencies of choice alternative preferences among three groups of participants.

In Tasks 1, 2, and 3 in all countries the choice to wear a mask for a reason “care for self” prevailed over“care for others”, more often in the Chinese sample (79.5 vs 19.6%, 72 vs 28% and 87.3 vs 11.8%) compared to the Russian and Azerbaijani samples (χ 2=62.488, p = .001 for Task 1, χ 2=103.262, p = .001 for Task 2, χ2=44.377, p = .000 for Task 3). The choice to remain without a mask was very rare in the Chinese data (less than 1% in Tasks 1–3) while in Azerbaijani and Russian samples it was made with the reasons of autonomy for oneself over autonomy for others (Task 1), low fear of getting infected (Task 2) and risk underestimation over autonomy for others (Task 3).

In Task 4, as reasons for wearing a mask high risk assessment and law-abidingness were chosen with almost equal frequency in Russia (38.6 vs 40.4%), in Azerbaijan prevailed law-abidingness (22.2% vs 46.7), and in China – high risk assessment (71.6 vs 28.4). Not wearing a mask was explained predominantly by autonomy for oneself (not for others) in both Russia (15.8 vs 5.3%) and Azerbaijan (24.4 vs 6.7%). Among the Chinese participants, no one chose to remain without the mask.

In Task 5, law-abidingness prevailed over the risk of infecting others in all groups. When choosing not to wear a mask, Russians more often chose personal autonomy over risk underestimation (13.3 vs 9.7%), Azerbaijani participants showed the opposite trend (10.2 vs 17.9% of personal autonomy). Only 0.2% of Chinese participants chose not to wear a mask.

Intercorrelations of personality traits in cultural samples

Correlations between personality traits for Russian and Azerbaijani participants are presented In Table 3 ; for Chinese participants – in Appendix 2 (see Table 5).

Table 3

Partial correlations for the Russian and Azerbaijani samples (adjusted for sex and age).

	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1.IT of emotions	1	0.07	0.07	0.02	−0.01	0.00	0.07	0.11*	−0.14**	0.03	−0.04	0.11*	−0.07	0.04
2.Risk readiness	0.10	1	−0.18***	0.14**	0.07	0.16***	0.26***	0.04	−0.12*	0.15***	−0.05	0.19***	−0.01	0.14**
3. Rationality	−0.002	−0.33**	1	−0.21***	−0.25***	−0.09	0.26***	0.27***	−0.06	0.15**	0.004	0.31***	0.04	0.25***
4. Machiavellianism	0.04	0.17**	−0.18***	1	0.49***	0.36***	0.03	−0.15**	0.02	−0.08	−0.12*	−0.06	−0.08	−0.08
5. Psychopathy	−0.03	0.23***	−0.11	0.38***	1	0.16***	−0.11*	−0.22***	−0.08	−0.21***	−0.24***	−0.19***	−0.25***	−0.26***
6. Narcissism	0.04	0.12	−0.09	0.44***	0.06	1	0.08	0.00	0.21***	0.08	0.05	0.05	0.16***	0.12*
7. Perspective taking	0.04	0.24***	0.14*	0.09	0.07	0.09	1	0.40***	−0.13*	0.36***	0.15***	0.87***	0.17***	0.72***
8. Online simulation	0.003	−0.10	0.26***	−0.22***	−0.31***	−0.06	0.31***	1	0.09	0.43***	0.25***	0.80***	0.35***	0.76***
9. Emotion contagion	0.01	−0.30***	0.10	−0.01	−0.21***	0.20***	−0.10	0.12**	1	0.34***	0.11*	−0.04	0.71***	0.32***
10. Proximal responsivity	−0.02	−0.04	0.14*	−0.07	−0.32***	0.21***	0.37***	0.28***	0.43***	1	0.33***	0.47***	0.77***	0.73***
11. Peripheral responsivity	0.02	0.06	−0.07	0.02	−0.23***	0.27***	0.31***	0.29***	0.21***	0.44***	1	0.24***	0.65***	0.49***
12. Cognitive empathy	0.03	0.12	0.23***	−0.06	−0.12	0.03	0.86***	0.75***	0.01	0.41***	0.37***	1	0.30***	0.88***
13. Affective empathy	0.007	−0.14*	0.07	−0.02	−0.33***	0.29***	0.24***	0.30***	0.75***	0.81***	0.72***	0.33***	1	0.71***
14. Empathy	0.03	0.005	0.20***	−0.05	−0.26***	0.18**	0.72***	0.68***	0.39***	0.71***	0.63***	0.87***	0.75***	1

Note. 1) *p < .05, **p < .01, ***p < .001; 2) correlations on Russian sample are presented below the diagonal, correlations on Azerbaijani sample – above the diagonal.

Table 5

Partial correlations for the Chinese sample (adjusted for sex and age).

	1. ITE	2. PT	3. OS	4. EC	5. PR	6. PerR	7. CE	8. AE
2. PT	0.09	1
3. OS	0.21***	0.59***	1
4. EC	0.007	0.15***	0.16***	1
5. PR	0.02	0.36***	0.29***	0.66***	1
6. PerR	-0.07	−0.05	−0.04	0.27***	0.16***	1
7. CE	0.16***	0.92***	0.86***	0.17***	0.37***	−0.05	1
8. AE	−0.02	0.19***	0.17***	0.85***	0.78***	0.65***	0.20***	1
9. E	0.11*	0.80***	0.75***	0.56***	0.67***	0.28***	0.87***	0.66***

Note. 1) *p < .05, ***p < .001.

On the Azerbaijani sample ITE showed positive associations with online simulation and cognitive empathy, and negative – with emotion contagion (Table 3). On the Chinese sample, ITE was positively associated with cognitive empathy, and overall score of empathy.

Risk readiness was positively associated with Machiavellianism and perspective taking in both samples, with proximal responsivity and cognitive empathy in Azerbaijan, with affective empathy and narcissism in Russia; and negatively – with rationality and emotion contagion in both samples.

Personality predictors of choices to wear or not to wear a mask

We ran individual logistic regression analyses for each of the personality measures as independent variables and a binary variable “wear a mask” or “remain without a mask” as dependent variable (regardless of reasons for the decision). The Chinese participants rarely chose “remain without a mask” therefore regressions weren't performed.

Rationality was a positive predictor of the choice to “wear a mask” in Russia (Tasks 2 and 5) and in Azerbaijan (Task 1) (see Table 4 ). Narcissism was a positive predictor in Russia (Task 3) and a negative predictor in three out of five tasks in Azerbaijan. Risk readiness and psychopathy were negative predictors in three cases.

Table 4

Personality variables as predictors of wearing or not wearing a mask.

Task	Russia	Azerbaijan
	Significant predictors
Task 1	1. Proximal responsivity (B = 0.22*)2. Psychopathy (B = −0.13*)3. Cognitive empathy (B = 0.09*)4. Empathy (B = 0.06*)	1. Online simulation (B = 0.26**)2. Rationality (B = 0.19**)3. Cognitive empathy (B = 0.11*)
Task 2	1. Proximal responsivity (B = 0.34*)2. Rationality (B = 0.21*)	1. Narcissism (B = −0.12*)
Task 3	1. Psychopathy (B = −0.22*)2. Narcissism (B = 0.21*)	No significant predictors
Task 4	No significant predictors	1. Narcissism (B = −0.12*)
Task 5	1. Proximal responsivity (B = 0.19**)2. Rationality (B = 0.13***)3. Online simulation (B = 0.12**)4. Psychopathy (B = −0.11**)5. Perspective taking (B = 0.08**)6. Cognitive empathy (B = 0.08***)7. Affective empathy (B = 0.08**)8. Risk readiness (B = −0.07*)9. Empathy (B = 0.06***)	1. Narcissism (B = −0.07*)2. Risk readiness (B = −0.06*)

Note. 1) *p < .05, **p < .01, ***p < .001.

Different empathy subscales and scales were positive predictors for Russian and Azerbaijani participants.

Personality predictors of reasons to wear a mask

Using a mixed logistic regression, we investigated personality variables as predictors of the probability of making the choice “care for others” vs the reference level “care for self” across Tasks 1–3. The baseline model included fixed effects of sex, age, and group (Russian as the reference level and dummy-coded Chinese and Azerbaijani group status). All continuous predictors were mean-centered (age) or scaled to have M = 0, SD = 1. Task effects were modeled as random effects. Analysis was focused on establishing the main effects of personality variables on choice as well as the interaction effect between personality variables and the group that would reflect the moderation of the relationship conditional on group status.

Two sets of analyses were performed: 1) using variables measured in both Russian and Azerbaijani samples, and 2) using variables measured in all three samples (see Fig. 2 ). First, the baseline model revealed lower probability of this choice in the Chinese group overall (B = -1.03, SE = 0.21, Z = -4.83, P = .0000014).

Fig. 2

Marginal effects for the mixed logistic regression models predicting choice “care for others” (vs “care for self”) with confidence intervals.

Second, we found that several variables positively predicted the probability of choice “care for others”: using the Russian sample as the baseline, we established that this choice was positively predicted by ITE (B = 0.30, SE = 0.15, Z = 2.06, P = .039), perspective taking (B = 0.57, SE = 0.17, Z = 3.32, P = .001), online simulation (B = 0.44, SE = 0.17, Z = 2.60, P = .009), overall empathy (B = 0.46, SE = 0.16, Z = 2.88, P = .004) and, most strongly, cognitive empathy (B = 0.62, SE = 0.18, Z = 3.53, P = .0004). Notably, for two of these variables, we established a significant interaction with the group status: for cognitive empathy and perspective taking, the two strongest predictors of the choice, the effects were significantly smaller in the Chinese group (for the interaction, B = -0.46, SE = 0.22, Z = -2.10, P = .03533 for cognitive empathy, and B = -0.45, SE = 0.22, Z = -2.07, P = .0389 for perspective taking, respectively). No other main or interaction terms reached nominal significance.

Discussion

In Russia and Azerbaijan we observed similar relationships between empathy and risk readiness, rationality, and DT traits. Previous cross-cultural study showed that DT traits and emotional intelligence do not correlate in these cultural samples (Kornilova, Chumakova, & Gadjieva, 2016), therefore, the observed relationship between empathy and DT traits is not likely to be mediated by emotional intelligence which we didn't measure in this study.

The Chinese sample was characterized by higher cognitive empathy, but the three groups didn't differ in affective empathy. On Azerbaijani and Chinese samples IT of emotions correlated with cognitive empathy, suggesting IT of emotions is rather a cognitive than affective component of emotion regulation.

Since Chinese sample didn't receive the measures of risk readiness, rationality, and DT traits, we couldn't build a more distinctive personality profile. And since extremely low percentage of Chinese participants chose to “remain without a mask” we couldn't identify predictors of this choice.

The choice to wear a mask was most often made for a reason of care for self in all countries. This reason prevailed mostly among the Chinese participants, less – among Azerbaijanis and less so – among Russians. For Russians, as representatives of a more individualistic culture, the gap between the frequency of choosing this reason and care for others was the smallest.

This result contradicts hypothesis 1 and suggests that the concepts of collectivism and individualism as characteristics of cultures may not correspond to the individual reasons of decision making. It also contradicts the idea that in relational contexts prosocial behavior is guided by interpersonal responsibilities and in autonomous contexts – by personal choice and autonomy (Köster, Schuhmacher, & Kärtner, 2015). At the same time, the Chinese participants showed higher trust in the media message that they may be asymptomatically infected, therefore these results on choice of reasons could be due to varying risk representation in different countries.

Contrary to hypothesis 1, Azerbaijani participants chose “not to wear a mask” most often and for the reasons of risk underestimation and autonomy for themselves (the latter is contrary to showing concern for others, which we expected to observe in collectivistic cultures).

We assumed that the risks of contracting a virus with and without symptoms are equal, therefore, this factor couldn't be the one influencing the asymmetry in choice reasons.

Risk readiness influenced the decision not to wear a mask for Russians and Azerbaijanis which allows us to accept hypothesis 2. Rationality was a positive predictor of complying with the norms which corresponds to the US data (Stanley, Barr, Peters, & Seli, 2020). Our findings on the positive role of empathy are also consistent with data from the USA, UK, and Germany (Pfattheicher et al., 2020). ITE wasn't a significant predictor of decision to wear a mask, contrary to what we expected.

In line with the Danish study (Zettler et al., 2020), we assumed the negative role of the DT traits in adhering to the restrictions. The difference was that in the Russian sample psychopathy was a negative predictor of3 wearing a mask while narcissism was a positive predictor, corroborating the idea of it being “the lightest” trait in DT (Krasavtseva & Kornilova, 2019) while in Azerbaijani sample narcissism was a negative predictor. It could be possible that cultural context mediates whether narcissism leads to protesting restrictions (seen as an ego threat) and disregarding for caring for others.

Limitations

The results of this study might be influenced by the different severity of the epidemic in three countries. Moreover, we found that strict normative regulation of behavior makes it almost impossible to reveal personality regulation of choices.

IT of emotions and subscales of cognitive empathy emerged as the variables that predict caring for others over caring for self as reasons for wearing a mask suggesting these components of emotional regulation play a role in prosocial behavior. But further studies are needed to understand why IT of emotions didn't influence decision to wear a mask.

We also need to study further why in comparison with Russia collectivist countries (Azerbaijan and China) have shown higher indices of cognitive, but not emotional empathy which could be expected in cultures with closer interpersonal relationships.

CRediT authorship contribution statement

Maria Zirenko: Methodology, Investigation, Data curation, Writing - original draft, Writing - review & editing. Tatiana Kornilova: Conceptualization, Methodology, Writing - original draft, Writing - review & editing, Funding acquisition. Zhou Qiuqi: Investigation. Ayan Izmailova: Investigation.