Other-Oriented Empathetic Concern as the Underlying Factor of Extrinsic Emotion Regulation Motives and Its Positive Application for Social Well-Being

Introduction

The prevalence of depression and anxiety has been exacerbated by the lockdowns, isolation, and health threats resulting from the COVID-19 pandemic.^1–3 Throughout the pandemic, in response to these challenging circumstances, many individuals have engaged in emotional management processes, either for themselves or to assist others in regulating their emotions. On a global scale, nearly everyone has experienced interpersonal emotional support at various times, whether from family and friends, local community members, or even strangers encountered online. While numerous existing studies have underscored the significance of intrapersonal regulation of one’s own emotions, it’s important to recognize that individuals not only regulate their own emotions but also frequently endeavor to influence the emotions of those around them. A recent study showed that individuals attempt to help others regulate their emotions with more conscious effort than they put into the management of their own emotions.⁴ The process by which people attempt to regulate their own and others’ emotions through social interaction is known as interpersonal emotional regulation or the social regulation of emotions.^5,6

The process model of interpersonal emotion regulation proposed by Zaki and Williams is currently the most systematic theoretical structure in the field.^5,7 It divides interpersonal emotion regulation into intrinsic and extrinsic regulation according to whether this regulates one’s own or others’ emotions. Intrinsic interpersonal emotion regulation (IER) refers to the regulatory process dominated by the regulation of one’s own emotions in social interaction; the regulatory process dominated by the regulation of other people’s emotions is called extrinsic IER. Most studies investigating interpersonal emotion regulation have focused on the intrinsic regulation aspect.^8–11 Although people regulated others’ emotions nearly twice as often as they turned to others to help with regulating their own emotions, there are fewer studies on extrinsic IER than on intrinsic IER.⁴ Nozaki and Mikolajczak developed an extended process model of extrinsic IER, which modeled how the extrinsic IER occurs and how individual differences in extrinsic IER abilities may arise.¹² Zaki integrates the study of empathy and prosocial behavior with interpersonal emotion regulation, pointing out that behaviors such as empathy, help, and support are characteristics shared by prosocial behavior and extrinsic IER.⁵ Both prosocial behavior and extrinsic IER aim to decrease the target’s negative emotions, increase their positive emotions, and help the target achieve their goals. Extrinsic IER helps the regulation of other people’s emotions and has been associated with crucial relationship and well-being outcomes.^13,14

Although we have found extrinsic IER to be useful, research that explores the reasons why people engage in extrinsic IER is still lacking. According to previous literature, people who engage in extrinsic IER are motivated by different goals, whether these are hedonic, non-hedonic (impression manage motivation), selfish or prosocial.^5,14–16 To assess the motives of extrinsic IER, Cloonan developed the Extrinsic Emotion Regulation Motive Scale.¹⁷ The EERMS consists of 28 items that describe regulation motives covering four higher-order goals: other-orientation to help reduce others’ negative emotions; self-orientation to regulate others’ emotions to benefit the self; obligation as a sense of responsibility to help others feel better; and reciprocation with receiving or expecting the same help from others. The factor correlations between other-orientation and reciprocation and between self-orientation and obligation were moderately positive, whereas other between-factor correlations were low or insignificant. The EERMS is a self-reporting tool that is simple to implement. However, these descriptions and factors of EERMS were originally identified based on responses from western-American participants. Thus, careful validation is essential before applying these elements in non-Western cultures.

Although research on cultural disparities in extrinsic emotion regulation is limited, prior studies have highlighted differences in intrinsic emotion regulation between Eastern and Western cultures.^18–20 For instance, Western cultures tend to encourage open emotional expression as a means of social connection, while non-Western cultures often prioritize emotional suppression and the preservation of emotional harmony in social settings.^21,22 In non-Western cultures valuing emotional control and emotional harmony, it is anticipated that regulators may use gentler and indirect methods to influence other’s emotions.²³ These approaches may help preserve the “face” of individuals receiving extrinsic emotion regulation and maintain interpersonal harmony while influencing others’ emotional states.²⁴ These cultural factors may have profoundly influenced people’s motivations for engaging in extrinsic emotion regulation. Nevertheless, we maintain that, regardless of cultural background, there are common motivations underlying extrinsic emotion regulation behaviors. Exploring these shared motivations alongside cultural distinctions in extrinsic emotion regulation will provide valuable insights into the diverse applications of emotion regulation in different cultural contexts.

Previous studies have shown that extrinsic IER significantly correlates with individual differences and results in positive implications for social relationships.^14,25–29 To test the convergent validity of the EERMS, we investigated the associations of the motives of extrinsic IER with several well-being indicators (moral self-identity, peer relationships, and perceived social support). Moral identity has been described as the degree to which one holds moral concerns as important to one’s identity.³⁰ This may increase prosocial intentions and is especially effective in accounting for prosocial behaviors.^31,32 Regarding the emotional aspect, moral identity (both internalization and symbolization) was positively related to emotional prosocial behaviors among adolescents.³³ Moreover, previous studies have found that extrinsic IER could predict perceived couple’s relationship, quality of friendship over time, female friends’ psychopathology.^15,34,35 Given the importance of extrinsic IER in social domain, we thus hypothesized that other-oriented motives of extrinsic IER would report higher moral identity, peer relationships and perceived social support.

The Present Study

Existing literature has shown notable cultural contingencies in the norms of regulating and expressing emotion. Although we would like to adapt the EERMS to a different culture, we did not expect that the original factor structure would explain the current data best. We took an exploratory factor analysis (EFA) approach to understand the construct of extrinsic IER motives by specifying a series of EFA models. Moreover, as proposed by Zaki, the other-oriented concern including the element of empathy may be a superordinate prosocial response that motivates people to engage in the improvement of others’ welfare, which includes the regulation of other people’s emotions.⁵ To check this proposition, a bifactor exploratory factor analysis (bi-EFA) model with one general factor and several specific factors was specified. As people are more conscientious when regulating others’ emotions, and because other-oriented empathetic concern may be a common reason as to why people engage in this regulation, we hypothesized that a bifactor model would yield more interpretable results, relative to a multidimensional EFA structure without a general factor.^4,5

Participants and Methods

Participants and Procedure

A total of 464 young adults (73.1% females, 95.5% Han-Chinese) aged 18–54 years old (M = 21.67, SD = 2.09) participated in this online survey. Participants were recruited from college classes, social media, and online forums. It was found that 5.6% had a “low” social economic status, 18.5% “below middle”, 35.6% “middle level”, 4.3% “above middle”, and 0.2% “high”. Of these participants, 166 subjects (59.6% females) answered a short questionnaire containing the EERMS and demographic questions, and 298 subjects (80.2% females) answered all measures. Participants were first introduced to the purpose of the survey and the online consent letter and then were required to answer all the questions and submit their responses. After their submission, participants received an electronically generated red envelope with varying amounts of electronic money for online payments and shopping. The protocol and procedures used in this study were reviewed and approved by the local University Review Committee.

Measures

Extrinsic Emotion Regulation Motives Scale

The 28-item EERMS first underwent a translation and back-translation process to develop a Chinese version.¹⁷ The scale has four specific domains: other-orientated motives, self-orientated motives, reciprocal motives, and obligated motives. It was rated on a 6-point scale ranging from 1 (strongly disagree) to 6 (strongly agree). The Cronbach’s α coefficient for the total scale was 0.94.

Moral Self-Identity Scale

The Chinese version of the Moral Self-identity Scale was used to measure the self-importance of moral identity.³⁶ It includes two dimensions: internalization and symbolization. The scale was rated on a 6-point scale ranging from 1 (strongly disagree) to 6 (strongly agree). The Cronbach’s α coefficient for the scale was 0.73 for internalization and 0.74 for symbolization.

Peer Relationship Scale

The 4-item Peer Relationship Scale was adapted to assess the positive peer relationship of young adults.³⁷ The scale was rated on a 5-point scale ranging from 1 (extremely disagree) to 5 (completely agree). The structure of this scale was estimated through confirmatory factor analysis (CFA). The one-factor CFA model resulted in a good fit: χ²(df) = 4.902 (2), p < 0.001, CFI = 0.989, RMSEA = 0.070, SRMR = 0.019. The Cronbach’s α coefficient was 0.84.

Perceived Social Support Scale

The Chinese version of Perceived Social Support Scale was used to assess perceptions of support from friends, family, and significant others.³⁸ This 12-item scale includes three dimensions: friends’ support, family support, and significant other support, which was rated on a 7-point scale ranging from 1 (very strongly disagree) to 7 (very strongly agree). The Cronbach’s α coefficient was 0.93 for the total scale, 0.91 for friend support, 0.88 for family support, and 0.87 for significant other support.

Data Analysis

We first examined the originally designed four-factor CFA model with orthogonal rotation. Second, we ran several EFA models with oblique rotation to ascertain the number of factors to retain. Following this, we examined exploratory bifactor models with one general factor (bigeomin rotational rotation) and three (S-1) specific factors according to the suggestion from Heinrich et al³⁹ All models were conducted in Mplus 8.3 using the robust maximum likelihood (MLR) estimator which did not assume the multivariate normality of data. Items with significant factor loadings of ≥ 0.30 were considered relevant in all models. All models were considered acceptable when they achieved the following model fit cut-offs: a comparative fit index (CFI) ≥ 0.90, a root mean square error of approximation (RMSEA) ≤ 0.10, and a standardized root mean square residual (SRMR) ≤ 0.08.⁴⁰ For model comparisons, a change ≥ 0.010 in CFI supplemented by a change of ≥ 0.015 in RMSEA would be viewed as a significant change in the model fit.⁴¹

Results

Factor Analyses

CFA Model Results

We estimated a CFA model based on the originally designed four factors, resulted in a poor model fit, χ²(df)= 1680.32 (344), p< 0.001, CFI= 0.809, TLI= 0.790, RMSEA= 0.091 (0.087, 0.096), SRMR= 0.070. The bivariate correlations among the four factors were high (rs = 0.62–0.84, M_r = 0.73), implying a hierarchical factor underlying the different dimensions.

EFA Models

We then performed an EFA by estimating the model fit of one to four factors (see Table 1). The four-factor EFA model outperformed the one-, two-, and three-factor models and showed the best fit: χ²(df)= 778.57 (272), p< 0.001, CFI= 0.907, RMSEA= 0.063 (0.058, 0.069), SRMR= 0.034. However, after screening the items, we found that there were many cross-loaded items (with factor loadings above 0.30) that were explained by more than one factor (see Table 2). The correlation among these latent factors was also high, indicating a possible general factor underlying these items.

Table 1 Model Fit Indices for Bifactor Exploratory Factor Analysis

Table 2 Standardized Factor Loadings for the Four-Factor EFA Model and the Best-Fit Bi-EFA Model

Bi-EFA Models

Bi-EFA models with one to four factors were first estimated with bigeomin orthogonal rotation. The model with one general factor and three (S-1) specific factors best fit the data: χ²(df)= 904.59 (272), p< 0.001, CFI= 0.909, RMSEA= 0.071 (0.066, 076), SRMR= 0.034. However, we found that Item 27 did not load on the general factor or any specific factor (with factor loadings below 0.30), and Item 27 was thus deleted. Bi-EFA models were reexamined with the remained 27 items (see Table 1). Among these models, the bi-EFA model with one general factor and three specific factors (M3) outperformed M1 and M2 and fit the data best, Thus, M3 (see Figure 1) was retained as the best-fit model for the current data.

Figure 1 The best-fit bifactor exploratory factor analysis model.

Table 2 presents the factor loadings for M3. All items were significantly loaded on the general factor, with loadings ranging from 0.47 to 0.76 to indicate a strong general factor. The first specific factor self-orientation included four items (Item 3, Item 12, Item 21, and Item 25) from the originally designed self-oriented dimension. The second specific factor obligation comprised of three items (Item 4, Item 6, and Item 15) and the third specific factor reciprocation contained three items (Item 9, Item 10, and Item 11). The remaining items substantially loaded on only the general factor, supporting a hierarchical factor behind these motives. The Cronbach’s α coefficient was 0.94 for the total scale, 0.77 for self-orientation, 0.77 for obligation, and 0.83 for reciprocation.

Associations with Social Well-Being

Table 3 shows the correlations between the total EERMS and moral self-identity, peer relationships, and perceived social support from friends, family, and others. Consistent with our hypotheses, the overall EERMS score showed moderate to high positive correlations with moral self-identity and relationship indicators. The three specific dimensions of EERMS also showed positive correlations with moral self-identity, peer relationship, and perceived social support, aside for self-orientation, which was found to only correlate with moral identity.

Table 3 Means, Standard Deviation, and Correlation Between Factors of EERMS and Moral Identity, Peer Relationship, and Perceived Social Support (N=298)

Discussion

Using an exploratory factor analysis approach and based on the EERMS, the study investigated the structure of the motives of extrinsic emotion regulation. The current findings provide further insights into the conceptualization of extrinsic IER motives and to understand people’s extrinsic IER behaviors during COVID-19 pandemic. Results indicated that a bi-EFA model with one general factor and three specific factors best fit the data, consistent with Zaki’s statement that the common motive, namely other-oriented empathetic concern, may be the dominant motive that motivates people to regulate others’ emotions.⁵ Moreover, previous studies on the relationship between interpersonal emotion regulation and well-being primarily focused on intrinsic emotion regulation,^10,11,42,43 whereas this study explores the correlation between extrinsic emotion regulation and young adults’ social well-being.

To investigate the factor structure of EERMS, a series of EFA models were estimated to reveal that many items were cross loaded on more than one latent factor. The bi-EFA modeling indicated that one general and three specific factors best fit the current data. In this model, the general factor, namely other-oriented empathetic concern, was represented by all items. Most items were loaded on the general empathetic concern factor only. Results indicated that the variation in an EERMS total score is mostly attributable to a general factor and that variances in self-oriented, reciprocation, or obligation motives may still be found between individuals. These findings are unique compared to prior studies. The reason could be that during the COVID-19 pandemic, a common emergency, people consistently experienced urgency and insecurity, which potentially increased cooperation and pro-social inclinations in individuals.^44,45 Additionally, due to pandemic-induced lockdown measures, people spent more time engaging in online activities. By breaking down temporal and spatial barriers, online pro-social behaviors incurred lower costs compared to traditional pro-social support, likely protecting individuals’ intuition to assist others.⁴⁶ Moreover, inconsistent with the previous study,¹⁷ reciprocation was significantly correlated with self-oriented motives. This is understandable in a culture with a tradition of reciprocity.⁴⁷ When people attempt to regulate others’ emotions, they naturally consider the possible benefit of such attempts, including making themselves feel better both in the present and future.⁴⁸

We found that the general factor of other-oriented empathetic concern and three specific factors significantly correlated with moral identity, and the strongest association was found for the general factor. Previous studies have presented similar links between emotional social support and moral identity.^25,30,32 As previous studies found, the general factor of other-oriented empathetic concern and two specific factors (obligation and reciprocation) significantly correlated with peer relationship and perceived social support.^33,34,49 However, the self-oriented motives were not associated with peer relationships or perceived social support. This is reasonable, as when people engage in extrinsic IER to achieve their self-benefit goals, the regulator-benefit goals may be well understood by targets. The target would be more likely to see the regulator as unsupportive and unhelpful, which in turn decreases the social connection between the individuals.

The results of this study have several potential implications. Given the relatively scarce research on extrinsic emotion regulation compared to intrinsic emotion regulation, our findings suggest that the EERMS may be an ideal research and practical tool with which to assess the motives of extrinsic emotion regulation in diverse cultures. This study contributes to our understanding of the underlying factors driving extrinsic emotional support behaviors, which holds great value for psychologists, therapists, and counselors seeking insight into the motivations behind these altruistic behaviors. Moreover, the study demonstrates that these extrinsic emotion regulation motives are interconnected with moral identity and social relationships, particularly in the context of the COVID-19 pandemic. These insights can be practically applied in therapeutic settings, assisting in the design of targeted programs aimed at enhancing interpersonal emotion regulation ability and social emotional well-being.

This study has following limitations. Firstly, the collection of data occurred during the pandemic and lockdown policies differed for participants from different locations, which may affect participants’ responses. For example, participants in regions with lockdowns might have had different emotional experiences and responses compared to those in regions with no restrictions.⁵⁰ Secondly, Participants’ responses could also vary before and after the implementation and removal of these restrictions.^51,52 Thirdly, we did not distinguish the individuals’ COVID-19 infection statuses when they responded to the questionnaire.⁵³ Participants’ experiences with infection may have had the potential to influence the results of the current research. Finally, this preliminary study lacks a confirmation for the bifactor model, and future studies are needed to fully replicate the current study using large and diverse samples. In summary, the current findings should be explained by considering the impact of the COVID-19 pandemic.

Conclusion

In conclusion, this study contributes to our knowledge in understanding the essence of multiple motives behind regulating other people’s emotions. We hope to ignite researchers’ and practitioners’ interest in the investigation of extrinsic emotion regulation. Future studies from other Western and non-Western cultures would be useful for further examining the construct validity of the EERMS and its usefulness in diverse cultures.