Cross-cultural differences in beliefs about emotions: A comparison between Brazilian and British participants

ABSTRACT Objective: Perfectionist beliefs about emotions impact the experience and expression of emotions, being linked to increased levels of depression and anxiety. Given the influence of culture in the representation and expression of emotion, it is possible that beliefs vary across countries, but few empirical studies have been conducted on the theme. This study aims to compare Brazilian and British samples regarding their beliefs about emotional experience and expression. Methods: The current study compared a total of 960 Brazilian and British participants, with the samples having a similar profile in terms of age, gender and ethnicity. Participants answered online the Generalized Anxiety Disorder (GAD-7), Patient Health Questionnaire (PHQ-9), and Beliefs about Emotions Scale (BES). This study aims to compare Brazilian and British samples with regard to beliefs about emotional experience and expression. Results: Significant differences between samples were found for beliefs about emotions. As hypothesized, Brazilians scored lower on unhelpful beliefs about emotions, except for beliefs about experiencing negative feelings and emotional control. Differences in total BES scores remained even after the inclusion of depression and anxiety as covariates. Conclusions: Results suggest higher endorsement of perfectionist beliefs in a European versus a Latin American context, but highlight that this pattern depends on the specific beliefs being studied. These differences should be considered when working with people from different cultural backgrounds and developing cultural adaptations for clinical interventions and psychopathology models.


INTRODUCTION
Personal beliefs that the experience and expression of negative emotions are unacceptable have been associated with higher levels of fatigue, depression and anxiety 1 and can be described as a transdiagnostic vulnerability factor.As suggested by cognitive models 2,3 , unhelpful beliefs about emotions can have a deleterious effect on symptoms, as well as on treatment adherence and outcomes.Such beliefs may be considered a form of negative perfectionism 1,4,5 due to the high standards imposed towards one's emotions and the fear of negative evaluation by peers if this ideal is not reached.
Identifying unhelpful beliefs about emotions is important to reduce maladaptive coping and compensatory strategies, for instance understanding their contribution to potentially harmful behaviors such as emotional suppression 6 .Considering that there are cultural differences in the expression of emotions 7 , with different societies showing distinct display rules 8 , it is likely that there are cross-cultural and contextual differences in relation to these beliefs.Nevertheless, this topic has been seldom explored empirically.
Culture can be understood as shared and socially transmitted ideas (e.g., values, beliefs, attitudes) that are reflected in and reinforced by institutions, products, and rituals 9 .Culture can modulate whether emotions are perceived as harmful or beneficial, and there are differences in the profile of emotional responses to specific contexts, consistent with socially rewarded cultural values 10 .Sociocultural factors may also impact emotional regulation strategies 11 .For instance, emotional suppression, which is typically associated with negative outcomes in individualistic Western cultures, has been shown to differ in its association with psychopathology across cultures 12 .Nevertheless, most comparisons have focused on the contrast between Western individualistic and Eastern collectivist cultures 11 .Investigating beliefs about emotion in different cultural value systems may have important implications for the understanding of behavior and social interaction, as well as the way mental disorders are defined and treated in different contexts.
Comparing African American and European American mothers, Nelson and colleagues 13 investigated whether cultural differences in emotion socialization are associated with maternal beliefs concerning their children exhibition and management of negative emotions.Their results indicate that African American mothers were less prone to endorse their children showing negative emotions in public or private settings than European American mothers.One potential explanation is that minority mothers can be more apprehensive regarding their children being ill judged by the predominant ethnic group, reinforcing prejudice.
Specifically in Latin America, Halberstadt et al. 14 investigated parents and teachers' beliefs about emotions in children.Their results suggest an overestimation of independence and capability in controlling and regulating emotions in children, and the belief that positive emotions can have undesirable consequences if they are not moderated.Morelen and Thomassin 15 explored family processes in relation to emotion, contrasting Latin American and European American families.Due to values such as individualism, the latter believed emotions were less important compared with families who followed Latin American values, like "familismo", i.e., the critical emotional bond that ties families.This suggests the possibility of enriched emotional experiences in Latin American families 15 .Despite these initial efforts in the region, there are no studies comparing beliefs about emotions between Latin and European cultures with samples collected from their respective regions.
Accordingly, the current study aims to compare Brazilian and British participants in relation to beliefs about emotional experience and expression.To the best of our knowledge, this is the first study contrasting a European and a Latin American culture in beliefs about emotion.In order to isolate the effect of culture on beliefs about emotions, data were collected with samples matched for age, gender and ethnicity.It was hypothesized that the Brazilian sample would endorse fewer unhelpful beliefs about emotions.

Participants
The sample was composed of 960 participants aged between 18 and 69 years, with half of them being recruited in the UK and half in Brazil.The British data had been collected previously, so data collection in Brazil aimed to include participants with a similar profile in terms of age, gender and ethnicity (classified as a binary White/non-White variable); samples were not matched in terms of educational achievement because these data were not available in the British sample.Data were collected between 2018-2019.
UK participants were recruited via opportunity sampling through university e-mails, online research recruitment sites and social networking sites.Recruitment in Brazil was carried out through electronic advertisements (social networks and e-mails) and word of mouth.Inclusion criteria were being 18 years or older, being literate and possessing the means to access and respond to the electronic survey (e.g., through a computer or cell phone).All participants provided informed consent before taking part in the study.

Procedures
Data was collected anonymously online without time limit to complete the questionnaires.Participants received specific instructions about each instrument and had the option of stopping the survey and withdrawing from the study at any point.

Instruments
For all instruments, validated versions in the local language, with established psychometric properties, were used.

Generalized Anxiety Disorder (GAD-7)
The GAD-7 is a 4-point Likert scale with 7 items measuring general anxiety levels, organized in a single factor, showing good psychometric properties 16,17 .Individuals evaluate each item based on how often the symptoms have caused distress in the past two weeks (not at all = 0; several days = 1; more than half the days = 2; nearly every day = 3).The total scores can range from 0 to 21, where higher scores indicate greater severity of GAD symptoms 16 .For the Brazilian scale, Cronbach's alpha coefficient (α = .92)and rho composite reliability coefficient (ρ = .91)were suitable 16 .

Patient Health Questionnaire (PHQ-9)
This scale was developed to measure depression and response to treatment based on the 9 diagnostic criteria for major depression present in DSM-IV, with 9 items scored with a 4-point Likert scale 17,18 .The total score ranges from 0 to 27.If an item is missing, ordinary mean substitution is employed, but only if fewer than three items are missing.It has solid psychometric properties 17,18 , with good Cronbach's alpha coefficient (α = 0.87) 18 .

Beliefs about Emotions Scale (BES)
This scale evaluates how much people endorse beliefs about the unacceptability of experiencing or expressing emotions through 12 questions scored in a 7-point Likert scale.Good psychometric properties have been reported for the scale 1,19 .For the Brazilian full scale, Cronbach's alpha was appropriate (α = .86),suggesting adequate internal consistency; a twofactor solution for the BES-BR presented a well-defined structure, with all items having salient loadings in a single factor exclusively, and without any hyperplane items.The two factors referred to: (1) seeing emotions as signs of weakness and inferiority and not expressing them in front of others; (2) emotional control 19 .

Ethical procedures
The project was approved by the King's College London (KCL) Research Ethics Committee (PNM/13/14-50) and by the Psychology Department Ethics Committee PUC-Rio (018/2014).
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
All participants signed the Informed Consent Form formalizing compliance with the Guidelines and Regulatory Standards for research involving human subjects, as outlined in Resolution 196 of the National Health Council (CNS), which was ratified on October 10, 1996.

Data analysis
Descriptive statistics were generated to illustrate sociodemographic and clinical characteristics of the samples.Chi-square tests were used to investigate differences between samples regarding gender and ethnicity, with t-tests for independent samples being calculated for age, anxiety (GAD-7), depression (PHQ-9) and total BES scores.Analysis of total BES scores were repeated as analysis of variance models to allow inclusion of anxiety and depression as covariates (ANCOVA).For individual BES items, considering the ordinal characteristics of the variables, Mann-Whitney tests were used to compare the samples; considering the need to correct for multiple testing, in this analysis α was set at .001.All data were analyzed with version 26 of the Statistical Package for the Social Sciences (SPSS).
Table 2 shows the mean and standard deviation for each item and total scores of the Beliefs about Emotions scale.Almost all items showed significant differences between Brazilian and British participants (p < .001),with the exception of items #4 ("I should be able to control my emotions.")(p = .175)and #8 ("I should be able to cope with difficulties on my own without turning to others for support.")(p = .202).For items #7 ("I should not let myself give in to negative feelings.")(p < .001)and #12 ("Others expect me to always be in control of my emotions.")(p < .001) the Brazilian sample showed higher scores, while the British sample showed higher scores in all other items and in the total scale score.Additionally, ANCOVA indicated that although anxiety (F (1, 948) = 6.25, p = .013,η p 2 = .01)and depression (F (1, 948) = 21.64,p < .001,η p 2 = .02)are significantly associated with total BES scores, differences between groups in total scores remain unchanged after their inclusion as covariates (p < .001,η p 2 = .12).

DISCUSSION
This study aimed to investigate beliefs about the unacceptability of experiencing or expressing emotions cross-culturally, exploring these in Brazilian and British participants.There were no significant differences between samples for sociodemographic variables, potentially isolating the impact of culture in beliefs about emotions.Overall, British participants endorsed more unhelpful beliefs, but for two items, exploring negative feelings and emotional control, scores were higher for the Brazilian sample.Although anxiety and depression were significantly linked to total BES scores, their inclusion as covariates did not eliminate group differences in total BES scores.
Confirming the main hypothesis of the study, the Brazilian sample generally showed lower endorsement of unhelpful beliefs about emotional experience and expression.The groups differed in anxiety and depression scores, and the current research also confirmed the association between anxiety and depression on beliefs about emotions, in line with previous results 20 .However, controlling for anxiety and depression as covariates did not eliminate group differences, so it is unlikely that differences in BES scores were driven by these variables.It is plausible that differences in the BES scores reflect cultural differences about how emotions should be modulated in everyday life.
For instance, Clark et al. 21, researching pain management by contrasting samples from European and Latin American populations, found that the latter rated anxiety and depression related to pain as more disruptive to their daily lives than participants from Europe.This was interpreted in terms of a greater "expressive" nature of Latin American participants.The finding that an expressive culture can show increased anxiety and depression in relation to a more stoic culture converges with our results, suggesting that these variables may be correlated to beliefs about emotion but without providing a full explanation to the phenomenon.
In this line, Wagstaff and Rowledge 22 , in a study on stoicism, found that British participants can show traits such as denial, suppression, and control of emotions accentuated, particularly in men.This stoicism may lead to beliefs that sharing emotions is a sign of fragility or lack of control that would result in a poor evaluation by peers, while other cultures can be less judgmental about the expression of emotions.Support for this notion comes from a study by Okougha and Tilki 23 , in which the experiences of nurses from Ghana and Philippines caring for British families were analyzed.Their results suggest that British stoicism dealing with grief can lead to communication difficulties and others being judgmental about a perceived lack of distress.Their findings may also indicate reduced perceived usefulness of emotions, regardless of them being seen as signs of weakness 24 .
Contrary to expectations, and in contrast with overall results, the Brazilian sample showed higher scores for item #12 ("Others expect me to always be in control of my emotions.").One potential explanation for that is that Brazilians may believe that others expect emotional control, but if this control does not occur, they will not necessarily be evaluated as inferior.This is reinforced by item #3 ("If I lose control of my emotions in front of others, they will think less of me.") having higher scores in the British sample, suggesting that Brazilians may perceive their peers as less judgmental about emotional expression.This is in line with the factor structure found by Mograbi et al. 19 for the BES in Brazil, indicating emotional selfcontrol and emotional expression as a sign of weakness as independent factors.
A similar reasoning can be used in relation to findings with item #7 ("I should not let myself give in to negative feelings.")and #9 ("To be acceptable to others, I must keep any difficulties or negative feelings to myself."), with Brazilians showing higher scores for the former but lower for the latter.While item #7 represents emotional self-control, item #9 refers to emotional expression.This suggests that Brazilians might think that they cannot give in to negative feelings, although this would not necessarily interfere with social acceptance.The emphasis of item #7 on negative feelings may also explain higher scores in the Brazilian sample, suggesting that participants from this country may show difficulties to accept negative emotions.This is supported by Senft et al. 25 results about the concept of "simpatia", which expresses a warm ambience on interpersonal interaction found in Usonians with Latino heritage, suggesting that the experience and expression of negative emotions are less desirable for this group.This is also partially supported by the results of the cross-cultural adaptation of the Negative Mood Regulation Scale 26 , in which Brazilians seem to have less capacity to regulate unpleasant emotions 27 in relation to samples from other countries from the same region 28 , although there was no direct comparison between samples.This effect can also be linked to fear of judgement between social groups 13 , with developing countries having many layers of social divisions and low levels of social mobility compared to developed regions 29 .Future studies should investigate further potential differences in the expression and acceptability of negative emotions in Brazil.
The current study has a few important limitations.Firstly, data were collected online, which may have led to sampling

Table 2 .
Mean score for each BES item and total scale BES: Beliefs about Emotions Scale.

Table 1 .
Socio-demographic and clinical characteristics of participants mv: missing values; PHQ: Patient Health Questionnaire; GAD: Generalized Anxiety Disorder.