Cultural Divergence in Emotion Regulation: U.S. and Korean Relationship Dynamics During Social Crises

Abstract

(1) Background: In times of social crisis, the way individuals manage their emotions undergoes substantial changes. As emotion regulation strategies also vary across cultures, how they impact relationship dynamics between U.S. and Korean populations may differ. This study explores these cultural differences, shedding light on the interplay between culture, emotion, and social dynamics. (2) Methods: Through analysis of 210 U.S. and 197 Korean participants, the research investigated three relationship categories: spousal/couple relationships, family relationships, and friendships during social crisis. (3) Results: Using moderated mediation analysis, results revealed distinct patterns in ER’s mediating effects across these relationship types, with notable cultural variations. In spousal relationships, ER partially mediated the NA-relationship quality association among American participants but showed no mediation effect among Korean participants. Family relationships demonstrated complete ER mediation across both cultural groups, while friendships showed partial mediation effects regardless of cultural background; (4) Conclusions: These findings highlight how relationship type and cultural context jointly shape the role of emotion regulation in relationship quality, advancing our understanding of emotional dynamics across different cultural and relational settings. The study provides implications for developing culturally sensitive approaches to relationship maintenance and intervention strategies.

1. Introduction

Periods of social upheaval—such as economic downturns, political unrest, or global pandemics—are often accompanied by heightened psychological stress that strains interpersonal relationships (Bonanno et al. 2010; Prime et al. 2020). A key mechanism by which such crises undermine relationship quality involves emotional disruption. Empirical findings suggest that a decline in positive affect and a surge in negative affect threaten the stability and satisfaction of close relationships (Goodwin et al. 2020). These emotional disturbances may alter communication patterns (Cameron and Mascarenas 2020; Campione-Barr et al. 2024), leading to ruptures in family cohesion (Eales et al. 2021; Rogers et al. 2020) and deterioration in close friendships (Philpot et al. 2021).

These shifts underscore the critical role of emotion regulation (ER) in sustaining interpersonal relationship quality during periods of crisis. ER refers to strategies used to reduce, enhance, or maintain emotional experiences (Gross 2014). Within family systems, ER is crucial for fostering harmony, resolving conflicts, and maintaining emotional balance in interactions (Lopes et al. 2005). For instance, while anger is a natural emotion, an uncontrolled outburst can significantly reduce relationship satisfaction (Slep et al. 2021).

The ability to regulate and control emotional reactions serves as a core psychological process essential for individuals to manage everyday life situations. Individuals typically utilize various strategies like cognitive reappraisal, distraction, and acceptance to maintain emotional equilibrium (Kozubal et al. 2023). However, during societal crises, ER patterns often shift, influenced by elevated stress and uncertainty. These changes can compromise the efficacy of conventional regulation approaches (Congard et al. 2023; Brenning et al. 2023; Farmer and Kashdan 2012).

Research shows that during crises, people adopt more constructive regulation techniques like positive reframing and acceptance (Congard et al. 2023; Brenning et al. 2023), but also tend toward suppression and distraction, especially when emotional demands intensify (Kozubal et al. 2023). These shifts in strategy can directly affect relationship dynamics. Interpersonal support, empathy, and attentive listening become even more vital to relationship quality in socially strained contexts (Chen et al. 2023; Tustonja et al. 2024; Yuan et al. 2025).

Understanding how ER operates across different types of close relationships during times of crisis is therefore crucial. We hypothesize that ER can act as a buffer for the negative emotions evoked, which may otherwise decrease the quality of interpersonal relationships:

Hypothesis 1:

Emotion regulation mediates the association between negative affect and relationship quality during social crises.

1.1. Cultural Differences in Managing Negative Emotions in Close Relationships

Culture is a broader factor that influences individuals’ emotional regulation. Cultural differences affect how emotions are expressed and regulated in social interactions (Ford and Mauss 2015; Liddell and Williams 2019; De Leersnyder et al. 2013; Mesquita and Boiger 2014). For example, in collectivist cultures such as those in East Asia, maintaining family harmony is often prioritized. As a result, individuals may suppress or reinterpret negative emotions to avoid conflict. In contrast, Western cultures, which value individualism and self-expression, encourage open emotional communication.

Research has shown that these differences may stem from variations in emotional complexity within family interactions (Shiota et al. 2010). Emotional complexity refers to the simultaneous experience of both positive and negative emotions in response to the same event (Larsen et al. 2001). Shiota et al. (2010) found that love and negative emotions are less negatively correlated among Asian Americans than among European Americans. This suggests that emotional complexity during family interactions may be more common among East Asians than Westerners, making negative emotions less harmful to relational outcomes. Hence, East Asian family members may feel less compelled to regulate these emotions.

While the individualism–collectivism framework is widely used to characterize cultural tendencies in emotion expression and regulation, it has also been criticized for its oversimplification and limited explanatory power. For instance, Oyserman et al. (2002) argued that cultural values vary significantly within national groups and that the individualism–collectivism distinction lacks consistent predictive validity. Similarly, Takano and Osaka (1999) challenged the assumption that Japanese individuals are more collectivist than Americans, suggesting that cultural behaviors are not always aligned with presumed value orientations. In this study, we employ the framework as a heuristic rather than a strict binary, while acknowledging its conceptual limitations.

Hypothesis 2:

The mediating role of emotion regulation in the link between negative affect and relationship quality is weaker in East Asian cultures compared to Western cultures.

1.2. Current Study

To explore these hypotheses, this study investigates the role of ER in mediating the impact of negative affect on relationship quality during the COVID-19 pandemic. We analyze this process across three distinct relationship types: spousal/romantic, familial (non-spousal), and close friendships. The study draws on data from 210 U.S. participants and 197 Korean participants, using moderated mediation analysis. Age, gender, and family structure (i.e., living with or without family) were included as covariates.

Romantic relationships, such as spousal partnerships, are often characterized by passionate love that transitions into deeper commitment over time (Montgomery and Sorell 1997). In Korean culture, it is commonly said that a spouse is simultaneously the closest and most precarious relationship. This paradox suggests that ER may either be critically necessary or altogether bypassed depending on how the partner is perceived in terms of shared identity (c.f., Kraus and Kitayama 2019).

In contrast, familial relationships such as those between siblings or parents and children tend to be more stable and grounded in evolutionary obligations (Buss 2000; Curry et al. 2013). Friendships, on the other hand, are non-kin relationships governed by mutual affinity rather than duty, and emotional regulation may function differently in such contexts (O’Gorman and Roberts 2017).

By comparing these three relationship categories and examining cross-cultural variations, the study aims to provide a more nuanced understanding of how emotional processes are shaped by both relational and cultural contexts during times of social crisis.

2. Materials and Methods

2.1. Participants

The US sample comprised 210 participants who were born or resided in the US for more than 20 years (M_age = 36.47, SD_age = 10.68; 34.8% female). The relatively broad age range ensures diversity in participants’ experiences, which is critical for capturing a wide spectrum of perspectives on emotional regulation and relationship dynamics. Ensuring gender balance is equally important, as gender differences in emotional expression and regulation could influence the findings. Addressing these representativeness concerns would enhance the generalizability and applicability of the study’s conclusions. Amazon’s Mechanical Turk “www.mturk.com (accessed on 21 August 2020)” was used for recruiting, and all participants received monetary compensation ($1.00) after the survey. The Korean sample comprised 197 participants who were born and resided only in South Korea (M_age = 23.71, SD_age = 7.14; 69.0% female). They were recruited from Daegu, South Korea, and were compensated with course credit, or a monetary reward of 1000 Korean won.

2.2. Measures

2.2.1. Negative Emotions Under COVID-19

Participants were asked to indicate the degree to which they experienced anger, hatred, depression, frustration, helplessness, and anxiousness and compare it to the time before the pandemic, using a 5-point scale from 1 (“much less than before”) to 5 (“much more than before”). These specific emotions were chosen because they represent a broad spectrum of common negative emotional reactions during crises, as noted in prior research (e.g., Rogers et al. 2020). The average of the five emotions was used to create an index of negative affect (NA).

2.2.2. Emotion Regulation

The Difficulties in Emotion Regulation Scale–Short Form (DERS-SF) (Kaufman et al. 2016) was used to assess individuals’ ER tendencies. DERS includes six subdomains: nonacceptance of emotional responses, lack of emotional awareness, limited access to ER strategies, difficulties engaging in goal-directed behavior when emotionally aroused, impulse control difficulties, and lack of emotional clarity (Gratz and Roemer 2004). Participants responded to the degree to which the statement described their emotional life using a 5-point scale ranging from 1 (“almost never [0–10%]”) to 5 (“almost always [91–100%]”). For ease of interpretation, the scores were reserved to ensure that higher scores indicate less difficulty in ER. The average of the 18 items was used in the analysis as an index of difficulties in ER. The scale’s original version was translated and validated for the Korean sample (Cho and Hong 2013).

2.2.3. Relationship Quality

Three types of interpersonal relationship qualities were assessed, including those with a romantic partner, family members, and friends. Participants were asked about the degree to which their interpersonal relationship was negatively impacted by COVID-19, using a 5-point scale from 1 (“not at all”) to 5 (“very much”).

2.2.4. Family Construct

Participants were asked to report whether they live alone (coded as 1) or with family.

2.2.5. Demographics

Demographic information including age, gender, country of birth, and residence were assessed.

2.3. Procedures

At their convenience, participants completed an online survey designed using Qualtrics, a web-based survey and research software, by clicking on a link to access an information page and an online consent form. The online survey procedure was as follows. First, the study was introduced to the participants as a survey on the COVID-19 outbreak and subsequent life changes. Second, participants signed the online consent form before beginning the survey. Third, as this study required both English and Korean versions, original items developed in English were translated into Korean and reviewed and approved by bilingual experts using forward- and back-translation processes. To ensure cultural relevance and semantic equivalence, bilingual experts carefully compared the translated items with the original English versions, resolving discrepancies through collaborative discussions. Specific challenges included maintaining the subtle nuances of certain terms that carry different connotations in Korean compared to English. For example, terms related to emotional expressions had to be adjusted to align with cultural expectations in Korean society. These challenges were addressed by involving additional experts and conducting iterative reviews to ensure both accuracy and cultural appropriateness. Additionally, pilot testing was conducted to confirm that the translated items were comprehensible and contextually appropriate for Korean participants. Finally, an ethical committee for human research approved the study protocols.

3. Results

To investigate the association between participants’ NA during COVID-19 and their relationships, descriptive statistics and correlation analysis were conducted using SPSS Statistics version 20.0.

Table 1. Descriptive Statistics and Correlation among Variables.

Variables	M	SD	1	2	3	4	5	6
US
1. NA	3.26	0.84	-
2. Emotion regulation	3.07	0.79	−0.405 ***
3. Relationship 1	4.32	1.93	0.362 ***	−0.600 ***
4. Relationship 2	4.15	1.93	0.181 *	−0.565 ***	0.620 ***
5. Relationship 3	4.33	1.83	0.250 ***	−0.543 ***	0.709 ***	0.720 ***
6. Social distancing	7.45	2.10	−0.015	0.181 *	0.048	−0.035	0.075
7. Age	36.47	10.68	−0.072	0.175 *	−0.023	−0.099	−0.100	0.094
Korea
1. NA	3.58	0.74
2. Emotion regulation	3.44	0.66	−0.211 ***
3. Relationship 1	3.88	2.18	0.424 ***	−0.160
4. Relationship 2	2.75	1.94	0.276 **	−0.308 ***	0.444 ***
5. Relationship 3	3.58	1.99	0.285 **	−0.208 **	0.525 ***	0.541 ***
6. Social distancing	7.43	1.91	0.068	0.064	0.081	−0.039	0.115
7. Age	23.71	7.14	0.030	−0.188 **	−0.028	0.118	0.001	0.155 *

Note. p < 0.001 ***, p < 0.01 **, p < 0.05 *. Relationship 1 = relationship quality with partner, Relationship 2 = relationship quality with family members, Relationship 3 = relationship quality with friends, NA = negative affect.

presents descriptive statistics and correlations among the study’s main variables. Key insights include a significant negative correlation between negative affect (NA) experienced during COVID-19 and relationship quality across all relationship types. Specifically, higher NA was associated with greater negative impacts on relationships with romantic partners (r_US = 0.362, r_KR = 0.424), family members (r_US = 0.181, r_KR = 0.276), and friends (r_US = 0.250, r_KR = 0.285), supporting the hypothesized adverse effects of NA on interpersonal relationships. The results are summarized in terms of cultural groups based on birth and residence. Categorical variables, such as family construct and gender, are not included. Among the samples, 12.9% of US and 20.3% of Korean participants lived alone.

3.1. Mediation Model

To test the mediation model, SPSS Process Macro Model 4 (Preacher and Hayes 2008) was used with a 95% confidence interval based on 5000 bootstrap samples. This approach was chosen to separately examine how NA impacts different relationship types (spouses/couples, other family members, and friends) through ER. The mediation analysis was conducted in three independent models with gender, age, and family structure were controlled.

Table 2. Unstandardized Coefficients for the Mediation Model.

(a) Couples
Predictor	DV = ER (R* = 0.148; F (4, 284) = 12.682; p < 0.001)
	B	SE	t	p	LLCI	ULCI
Constant	11.693	1.491	7.843	0.000	8.759	14.628
Corona NA	1.736	0.271	6.395	0.000	1.202	2.271
Age	0.067	0.024	2.750	0.006	0.019	0.114
Gender	0.202	0.477	0.424	0.672	−0.737	1.142
Family construct	−0.358	0.179	−1.994	0.047	−0.711	−0.005
Predictor	DV = Relationship quality (R* = 0.270; F(5, 283) = 20.889; p < 0.001)
	B	SE	t	p	LLCI	ULCI
Constant	−0.650	0.733	−0.887	0.376	−2.092	0.793
Corona NA	0.604	0.129	4.670	0.000	0.350	0.859
Emotion regulation	0.179	0.026	6.764	0.000	0.127	0.231
Age	−0.002	0.011	−0.192	0.848	−0.024	0.019
Gender	−0.273	0.213	−1.284	0.200	−0.692	0.146
Family construct	−0.059	0.081	−0.736	0.462	−0.218	0.099
(b) Family members
Predictor	DV = ER (R* = 0.116; F(4, 380) = 12.503; p < 0.001)
	B	SE	t	p	LLCI	ULCI
Constant	12.489	1.378	9.062	0.000	9.779	15.198
COVID-19 NA	1.657	0.253	6.541	0.000	1.159	2.155
Age	0.041	0.021	1.953	0.052	0.000	0.082
Gender	−0.016	0.416	−0.038	0.970	−0.834	0.802
Family construct	−0.299	0.147	−2.03	0.043	−0.588	−0.009
Predictor	DV = Relationship quality (R2 = 0.300; F(5, 379) = 32.417; p < 0.001)
	B	SE	t	p	LLCI	ULCI
Constant	−0.356	0.651	−0.546	0.585	−1.636	0.925
COVID-19 NA	0.124	0.114	1.081	0.280	−0.101	0.349
Emotion regulation	0.222	0.022	10.089	0.000	0.179	0.265
Age	0.012	0.009	1.387	0.166	−0.005	0.030
Gender	−0.329	0.178	−1.847	0.066	−0.680	0.021
Family construct	−0.185	0.063	−2.923	0.004	−0.310	−0.061
(c) Friends
Predictor	DV = ER (R* = 0.120; F(4, 388) = 13.174; p < 0.001)
	B	SE	t	p	LLCI	ULCI
Constant	12.381	1.372	9.026	0.000	9.684	15.078
COVID-19 NA	1.715	0.251	6.839	0.000	1.222	2.208
Age	0.039	0.020	1.820	0.070	−0.003	0.077
Gender	−0.073	0.421	−0.178	0.859	−0.883	0.037
Family construct	−0.277	0.146	−1.896	0.059	−0.564	0.010
Predictor	DV = Relationship quality (R* = 0.224; F(5, 387) = 22.375; p < 0.001)
	B	SE	t	p	LLCI	ULCI
Constant	0.906	0.643	1.409	0.160	−0.358	2.17
COVID-19 NA	0.274	0.113	2.426	0.016	0.052	0.497
Emotion regulation	0.176	0.022	8.124	0.000	0.133	0.218
Age	−0.003	0.009	−0.311	0.756	−0.020	0.014
Gender	−0.486	0.176	−2.767	0.006	−0.831	−0.141
Family construct	−0.092	0.062	−1.47	0.142	−0.215	0.031

presents the coefficients of all paths in the mediation model of the three relationships and the amount of variance explained in each model.

The results showed that, overall, NA under COVID-19 influenced all three types of relationship qualities through ER. Regarding the relationship with spouse/couples, the direct effect of NA on the relationship quality was significant (B = 0.604, BootSE = 0.129, p < 0.001, 95% CI [0.350, 0.859]). The indirect effect of NA on the relationship quality with spouses/couples through ER was also significant (B = 0.310, BootSE = 0.074, 95% CI [0.181, 0.469]), which suggest partial mediation effect of ER.

Regarding the relationship with family members, the direct effect of NA on the relationship quality was not significant (B = 0.124, BootSE = 0.114, p = 0.280, 95% CI [−0.101, 0.349]). The indirect effect of NA on the relationship quality with family members through ER was significant (B = 0.367, BootSE = 0.070, 95% CI [0.242, 0.515]), which suggests a complete mediation effect of ER.

Regarding the relationship with friends, the direct effect of NA on the relationship damage was significant (B = 0.274, BootSE = 0.113, p = 0.016, 95% CI [0.052, 0.497]). The indirect effect of NA on the relationship quality with friends through ER was also significant (B = 0.301, BootSE = 0.062, 95% CI [0.191, 0.435]), which suggests a partial mediation effect of ER. Overall, these findings support Hypothesis 1.

3.2. Moderated Mediation Model

To test the moderated mediation model, SPSS Process Macro Model 14 (Preacher and Hayes 2008) was used with a 95% confidence interval based on 5000 bootstrap samples. This method was chosen because it allows for the simultaneous examination of direct, indirect, and interaction effects, aligning well with the study’s goal of exploring how cultural differences moderate the mediating role of emotion regulation on relationship quality during COVID-19. The analysis for moderated mediation was conducted in three independent models. This approach was chosen to separately examine how NA impacts different relationship types (spouses/couples, other family members, and friends) through ER. By using independent models, the analysis could validate whether the mediation effects and cultural differences vary depending on the type of relationship, thereby ensuring a comprehensive test of the hypotheses. The results were summarized in

Table 3. Conditional Indirect Effects Across Cultures.

DV	Group	B	BootSE	BootLLCI	BootULCI
Couples	US	0.328	0.079	0.181	0.493
	Korea	0.102	0.085	−0.046	0.287
Family members	US	0.321	0.072	0.187	0.467
	Korea	0.209	0.07	0.091	0.363
Friends	US	0.285	0.065	0.169	0.42
	Korea	0.149	0.068	0.03	0.301

Note. SE = standard error, LLCI = lower limit confidence interval, HLCI = higher limit confidence interval.

.

3.2.1. Results on Spouses/Couples

Regarding the relationship with spouses/couples, the direct effect of NA on the relationship quality was significant (B = 0.605, BootSE = 0.132, p < 0.001), and the indirect effect of NA on the relationship quality through ER was also significant among participants from the US (B = 0.328, BootSE = 0.079, 95% CI [0.181, 0.493]). These findings suggest the partial mediation of ER regarding the relationship between NA and relationship quality between US couples. However, the indirect effect was not significant among Korean participants (B = 0.102, BootSE = 0.085, 95% CI [−0.046, 0.287]).

The moderated mediation test shows that the difference in indirect effects across cultural groups was statistically significant (B = 0.226, BootSE = 0.096, 95% CI [0.051, 0.426]), suggesting that the size of mediation coefficients in spouse/couple relationships differ across groups (

Table 4. Index of Moderated Mediation: Difference between Conditional Indirect Effects Across Groups).

	Index	Boot SE	Boot LLCI	Boot ULCI
Spouses/Couples	0.226	0.096	0.051	0.426
Family members	0.111	0.066	−0.017	0.242
Friends	0.136	0.075	−0.003	0.289

Note. SE = standard error, LLCI = lower limit confidence interval, HLCI = higher limit confidence interval.

).

3.2.2. Results on Family Members

Regarding the relationship with family members, the direct effect of NA on the relationship damage was not significant (B = 0.219, BootSE = 0.118, p = 0.063). This suggests that NA alone does not have a direct influence on relationship quality with family members. However, the mediation effect through ER indicates that NA’s impact on relationship quality is indirect, with ER serving as a crucial mechanism for maintaining relationship quality. This finding highlights the importance of ER in buffering the negative effects of NA on family relationships. Its indirect effect on the relationship quality through ER was significant among US (B = 0.321, BootSE = 0.072, 95% CI [0.187, 0.467]) and Korean participants (B = 0.209, BootSE = 0.070, 95% CI [0.091, 0.363]).

The moderated mediation test shows that the differences in indirect effects across cultural groups were not statistically significant (B = 0.111, BootSE = 0.066, 95% CI [−0.017, 0.242]), suggesting that the size of mediation coefficients in family members relationship does not differ across groups.

3.2.3. Results on Friends

Regarding the relationship with friends, the direct effect of NA on the relationship damage was significant (B = 0.356, BootSE = 0.118, p = 0.003). The indirect effect of NA on the relationship quality with friends through ER was significant among both US (B = 0.285, BootSE = 0.065, 95% CI [0.169, 0.420]) and Korean participants (B = 0.149, BootSE = 0.068, 95% CI [0.030, 0.301]). This difference in effect size may reflect the extent to which cultural norms influence the role of emotional regulation in friendships. For instance, individuals in Western cultures may be more accustomed to using explicit emotional regulation strategies in friendships, while individuals in East Asian cultures may rely on implicit strategies such as emotional suppression, leading to a smaller observed effect size.

The moderated mediation test shows that the differences in indirect effects across cultural groups were not statistically significant (B = 0.136, BootSE = 0.075, 95% CI [−0.003, 0.289]), suggesting that the size of mediation coefficients on friend relationship does not differ across groups

Overall, Hypothesis 2 was supported only in spouse/couple relationships showing the differences in the mediation coefficients across cultures. The coefficients of all paths in the moderated mediation model of the three relationships and the amount of variance explained in each model are summarized in Supplementary Material (Table S1).

4. Discussion

This study investigated the psychological mechanisms through which unprecedented conditions of psychological and social isolation impacted relationship quality among spouses/couples, family members, and friends, with a particular focus on emotion regulation (ER) processes and their mediating effects across different relationship types. Our analysis revealed that ER exerted varying mediating effects depending on the relationship type: partial mediation in spouse/couple relationships, complete mediation in family member relationships, and partial mediation in friendships.

Our analysis also uncovered that ER played distinct mediating strength of ER across different relationships by cultural groups. When cultural groups were considered, a more nuanced understanding of how emotion regulation impacts relationship quality. The cultural differences in spousal relationships, in particular, showed interesting patterns. While Americans managed their relationship quality through emotion regulation despite negative experiences during the pandemic, Koreans’ emotional states directly influenced relationship quality without ER mediation. This distinction might be attributed to East Asian cultural perspectives that view close relationships as extensions of the self (Brewer and Gardner 1996). In Korean contexts, where motivation for emotion regulation of self tends to be lower, individuals might feel less compelled to regulate emotions toward spouses whom they consider part of themselves (c.f., Kraus and Kitayama 2019). Future research would benefit from measuring the degree of self-identification and motivation for emotion regulation to examine whether relationships with higher self-identification show similar patterns of emotion regulation motivation.

These findings expand upon earlier research by Congard et al. (2023) and Brenning et al. (2023), who demonstrated the general shift in ER strategies during social crises. However, those studies did not differentiate between types of close relationships or explore cultural differences. Our results provide a more nuanced perspective by showing that the strength and pattern of ER’s mediating effect vary significantly across relationship types and cultural contexts. In particular, the absence of mediation in Korean spousal relationships contrasts sharply with the partial mediation observed in U.S. samples, thus refining the broad conclusions of earlier crisis-focused studies.

Regarding family relationships, our findings showed that ER fully mediated the effect of NA on relationship quality across both cultural groups, highlighting its crucial role regardless of cultural background. This difference contrast with spousal relationships might stem from the unique characteristics of family relationships governed by kinship principles (Buss 2000). Not only the investment of physical resources but also of emotional resources, might follow a hierarchical pattern based on kinship proximity: the closer the kinship tie, the greater the resource investment (c.f., Curry et al. 2013). This suggests an adaptive strategy aligning psychological resource allocation with evolutionary kinship priorities, where individuals invest more emotional regulatory effort in closer kinship relationships.

In the analysis of friendships, representing close but non-familial relationships, ER demonstrated partial mediation between NA and relationship quality across both cultural groups. This positioning—between the absence of mediation in spousal relationships and complete mediation in family relationships—is particularly notable in the Korean group, in particular. It aligns with existing research that places friends in an intermediate position between close and distant kin in social proximity. Research indicates that individuals typically consider family most central to their in-group, followed by friends, and then distant kin (O’Gorman and Roberts 2017). While this study did not specifically assess individuals’ personal definitions of distant kinship boundaries, future research could investigate whether emotional closeness corresponds to genealogical distance and how the role of ER changes accordingly.

This study makes contributions to our understanding of how cultural differences and relationship types jointly influence emotion regulation’s role in maintaining relationship quality. The findings indicate that the effectiveness of emotion regulation strategies varies not only across cultures but also across different relationship categories within cultures. This variation suggests that relationship maintenance strategies should be tailored to both cultural contexts and specific relationship types, as identical emotional regulation approaches may yield different outcomes depending on these combined factors.

These insights have important implications for clinical practice, relationship counseling, and cross-cultural communication, particularly in developing culturally sensitive interventions for relationship enhancement and maintenance across different relationship types. For instance, during the COVID-19 pandemic, several European countries implemented culturally adapted interventions such as mindfulness-based stress reduction (MBSR) programs and emotion-focused family therapy (EFFT) to support families experiencing emotional disconnection (e.g., Foroughe et al. 2022; Witarto et al. 2022). These approaches emphasized emotion labeling, acceptance, and cognitive reappraisal—strategies that align with the regulatory mechanisms discussed in our study. Integrating such culturally attuned techniques into clinical and community-based interventions may enhance their relevance and effectiveness, particularly in contexts where emotional suppression or ambivalence is more culturally normative (e.g., East Asian populations).

This study also builds on existing research highlighting the diminished role of emotion in various aspects of life among Asian populations, including psychological well-being (Suh 2000), partner selection (c.f., Blair and Madigan 2016), and other psychological processes (Ramzan and Amjad 2017). The weaker or no mediating effect of emotion regulation among Korean participants, especially in spousal relationships, suggests that relationship maintenance in East Asian cultures may rely more heavily on other mechanisms beyond emotional processes, such as social obligations or role fulfillment. This understanding can inform culturally adapted relationship interventions that consider alternative pathways to relationship quality beyond emotional regulation strategies.

Several methodological limitations warrant consideration. As a preliminary cross-cultural investigation, we collected US and Korean samples, which demonstrate dissimilarities in emotional complexity and the importance of emotions. However, as emotional complexity levels and the significance of emotions vary across countries, including more cultural samples would increase our understanding of emotional experience variability. Moreover, although this study assessed overall relationship quality using a single item, relationship quality can be evaluated using multiple methods across diverse dimensions (c.f., Fincham and Rogge 2010). Future studies should consider more comprehensive approaches to capture detailed aspects of relationship quality.

Additional variables could enhance the generalizability of our findings. The study measured emotion regulation in terms of difficulty or ease, but various other aspects of emotion regulation could be considered. For instance, examining specific types of emotions would be valuable, as socially engaging emotions (e.g., guilt) and socially disengaging emotions (e.g., anger) can lead to different interpersonal relationship outcomes in socially disengaged contexts like the COVID-19 pandemic (Kitayama et al. 2006). Therefore, measurements of emotion regulation might need to vary depending on the type of emotion being studied.

While this study assumed NA arose from the COVID-19 situation, it couldn’t distinguish between specific sources of distress. For example, negative emotions might have stemmed from reduced social interaction due to social distancing or from financial strain. Additionally, individual differences in baseline negative affect levels weren’t accounted for, as some individuals naturally experience lower levels of negative emotions. Future research should explore these dynamics in more controlled settings or specific personal circumstances beyond the pandemic context. This would help isolate the precise mechanisms through which different types of negative affect influence relationship quality and emotion regulation across cultures. Future research should examine the interplay between cultural, relational, and emotional factors across different contexts that elicit negative affect, to determine the generalizability of these findings beyond the COVID-19 pandemic context.