Does Social Interaction Improve Emotional Experience Across Different Relationships? A Study Using Experience Sampling in Older Chinese Adults

Abstract

Social interaction is an important way that we as humans connect with others. Socioemotional selectivity theory emphasizes the importance of close relationships, and Chinese culture attaches great importance to family members. As such, this study aimed to examine the differences in the quality of interactions that older Chinese adults have with close partners (e.g., children, friends, and relatives) as well as with other partners (e.g., neighbors, colleagues, and strangers) and to examine the association between interaction quality and emotional experience across these interactions. We collected data from 213 older adults over the course of 14 days. Results indicate that (1) compared to other partners, interactions with close partners are considered to be of higher quality; (2) in interactions with children and relatives, interaction quality is positively associated with positive affect (PA) and negatively associated with negative affect (NA); in interactions with friends and neighbors, interaction quality is only associated with PA; in interactions with colleagues and strangers, interaction quality is not associated with either PA or NA. Overall, interactions with close partners were shown to be considered to be of higher quality, and that the quality of interactions with family members was closely associated with emotional experience.

Keywords

social interaction interaction partners close relationship emotional experience experience sampling methods older adults

Introduction

Social interaction is an important way that people seek psychological communication and obtain social support, both of which play an important role in one’s physical and mental health (Chui et al., 2014). Positive social interaction has been found to be related to positive emotional experience (Schmitt et al., 2007). Maintaining connection with others is also vital for older adults, with interpersonal relationships playing a more important role in their happiness than it does for younger adults (Carstensen et al., 2003). Although one’s social circle shrinks as we grow older, socioemotional selectivity theory suggests that older adults pay more attention to the emotional satisfaction they experience from their close relationships (Carstensen, 1991).

The primary aim of this study was to explore whether the quality of older adults’ social interactions with close relationships was higher than that of other types of relationships. Close relationships, which can include a partner, friends, or children, have long been the focus of research into social interaction as they play a very important role for individuals throughout one’s life (for review, see Ramsey & Gentzler, 2015). One study found that compared to interactions with acquaintances, individuals reported more pleasure, satisfaction, and autonomy when interacting with family and friends (Downie et al., 2008). Socioemotional selectivity theory suggests that for older adults, the gains from interaction with unfamiliar social partners are unlikely to merit the energy consumption used in the interaction (Carstensen, 1991), but that in contrast, interactions with close partners have a greater significance in satisfying older adults’ emotional cravings (Donnelly & MacEntee, 2012). To investigate this deeper, we used the Rochester Interaction Record (RIR) to collect information regarding participants’ social interactions and to explore the differences between the quality of social interactions with close partners (e.g., children, friends, and relatives) and with other relationships (e.g., colleagues, neighbors, and strangers).

The secondary aim of this study was to examine the association between interactions and emotional experience when older adults interact with children, relatives, friends, neighbors, colleagues, and strangers. The association between social interaction and emotional experience has been shown to vary across different social partners according to the unequal importance social partners may have for individuals (Liu et al., 2018). For instance, one study found that being with family members (not including one’s spouse) and friends was related to experiencing more positive moods, but spending time with one’s spouse was not associated with either positive or negative affect (Chui et al., 2014). We believed that the association between interaction quality and emotional experience may vary depending on the relationship one has with the person they are interacting with. Furthermore, Chinese traditional culture emphasizes family connections and believes that the family is the fundamental unit of society (Park & Chesla, 2007). For Chinese people, family relationships may be considered to be more important than other relationships, which could in turn strengthen the association between interaction quality and emotion in interactions with family. Therefore, our study asked participants to report their emotional experiences using the Positive-Negative Affect Scale (PANAS), which allowed us to examine the association between social interaction and emotional experience across different interaction partners.

In summary, the study had two goals. First, we examined whether the perceived quality of social interaction with a close partner is higher than that of interactions with other partners. Second, we examined the association between social interaction and emotional experience when interacting with various types of partners.

Method

Participants

A total of 258 older adults were recruited from the Xi’an urban area in China. Because this study involved different emotion-related variables, depression could have affected the results. Therefore, participants were asked to answer the Geriatric Depression Scale (GDS; Aikman & Oehlert, 2001). One person was found to have severe depression according to the GDS scoring and was therefore removed from our participant pool. After deleting 44 participants who responded without any records of interactions during the time period, a total of 213 (64.80% female participants and 36.20% male participants) participants were included in the present study. Their ages ranged from 53 to 84 years (M = 64.88 years, SD = 6.33). Fifty-nine percent of participants had a college or university education. Every participant signed the informed consent form. After completing all forms and questionnaires, each participant was paid 300 RMB. The protocol was approved by the Ethical Committee of Shaanxi Normal University.

Materials

Rochester Interaction Record (RIR)

Following similar previous studies, we chose to use the RIR as it has been developed to record more comprehensive information regarding interactions (Reis & Wheeler, 1991). Each RIR interaction record includes the gender and age of the partner the participant interacted with and their relationship to the participant. The relationship type includes neighbors, colleagues (for retired older adults, colleagues refer to their former colleagues), friends, relatives, children, and strangers. Participants must also evaluate the quality of each interaction. Scores of all items are averaged to obtain an overall interaction quality score. We used the psych 2.1.3 R package to calculate both between-individual reliability (R_KRN) and within-individual reliability (R_CN) for the seven items of interaction quality. A high R_CRN is required for the scale to be considered reliable, while low to moderate R_CN is acceptable for within-person variability. In the present study, R_KRN was .87 and R_CN was .52.

Emotional experience

Emotional experience was measured using the Positive-Negative Affect Scale (PANAS), which has been shown to be a reliable and valid measure of the constructs it was intended to assess (Watson et al., 1988). The PANAS consists of two 10-item mood scales and can be used to quickly measure both positive affect (PA) and negative affect (NA). A higher PA score coupled with a lower NA score means that the participant feels happier. Participants were asked to choose their instant affect experience using a seven-point Likert scale (1 = no affect at all, 7 = very strong affect). In the present study, R_KRN was .98 and R_CN was .64 for PA and R_KRN was .98 and R_CN was .66 for NA.

Covariates

We used a basic questionnaire to collect relevant demographic information from participants to use as covariates. These covariates included gender (0 = male, 1 = female); marital status (0 = never married or divorced, 1 = currently married); university education (0 = no college education, 1 = college-level education or higher); and economic situation (i.e., How has your economic situation been over the past half year? 1 = not difficult at all, 5 = very difficult).

Procedure

The daily social interactions and emotion experiences of the older adult participants were measured using an event-contingent design (Reis & Wheeler, 1991). At the introduction meeting, participants were informed of the purpose of the study and the patterns of social interaction were explained. After asking participants to complete the demographic questionnaire, we then explained that they needed to use the RIR to record and describe every interaction they had over the following 14 days, and that they must then use the PANAS to evaluate their moods after each interaction. An interaction was described as any encounter with another person which caused a change in the participant’s behavior. Participants were told to use Wenjuanxing (an online questionnaire platform) to complete the records. One detail worth noting is that identities of friends and neighbors or colleagues can overlap. For example, someone can be both a friend and a neighbor. Therefore, we told participants to classify their interaction partners according to their feelings about that person. Specifically, while someone might have an identity as a neighbor, if the participant believed that they were a friend, they should choose “friend” as the relationship descriptor in the RIR.

In total, we obtained 10,193 RIR records, with an average number of 47.85 (SD = 26.65) records per participant over the 14 days, and an average number of 3.42 (SD = 1.90) interactions per day, which is similar to figures gathered in previous RIR diary studies (e.g., Nezlek et al., 2007).

Analysis strategy

The interaction dataset used had a hierarchical structure in which the repeatedly measured interaction and PA and NA (level 1) were nested in individuals (level 2). We chose the multilevel model (MLM) as the more suitable statistical analysis method for this type of dataset. MLM can tolerate missing values, has no strict limit on the time length between measurements, and can simultaneously consider the influences of both level 1 and level 2 factors on dependent variables (Marsh et al., 2006). We developed two models using MLM in Mplus 8.0.

Model 1: Social interaction and interaction partners

Before performing the formal analyses, we needed to establish the null model first. If the intraclass correlations (ICCs) in the null model were large (see Table 1), it signaled that we could not ignore differences between individuals. In the two-level model, the interaction partners were divided into “close partners” and “other partners.” This categorical variable was named partner and was taken as within-individual variables in the repeatedly measured within-person level (level 1), and covariates of participants were considered as between-person variables (level 2). The models were established as follows

{Outcome}_{ij} = β_{0j} + β_{1 j} partner + ε_{ij}

(1)

β_{0j} = γ_{00} + γ_{01} age + γ_{02} gender + γ_{03} marriage + γ_{04} education + γ_{05} economy + μ_{0j}

(2)

β_{1 j} = γ_{10} + μ_{1 j}

(3)

where Outcome_ij indicates the interaction quality or duration time for measured interaction i in individual j. The intercept β_0j was allowed to vary between individuals, where the individual-level intercept was γ₀₀ and the residual was μ_0j. In addition, γ₁₀ represents the individual-level partner’s influence, and μ_1j are errors brought by individual j to the slope.

Table 1.

Means and SDs of quality and time duration of interaction and positive and negative affect.

	Children M (SD)	Relatives M (SD)	Friends M (SD)	Neighbors M (SD)	Colleagues M (SD)	Strangers M (SD)	Total M (SD)	ICC
Interaction quality	4.47 (±0.62)	4.24 (±0.56)	4.18 (±0.54)	3.96 (±0.57)	4.06 (±0.54)	3.77 (±0.60)	4.14 (±0.59)	0.23
Pleasure	4.69 (±1.20)	4.35 (±1.08)	4.44 (±1.03)	4.02 (±1.13)	4.21 (±1.00)	3.84 (±1.25)	4.31 (±1.12)	0.20
Intimacy	4.87 (±1.31)	4.41 (±1.12)	4.10 (±1.10)	3.65 (±1.14)	3.91 (±1.08)	3.09 (±1.14)	4.08 (±1.23)	0.26
Satisfaction	4.79 (±1.20)	4.47 (±4.16)	4.51 (±1.00)	4.13 (±1.02)	4.28 (±0.95)	4.00 (±1.17)	4.40 (±1.07)	0.20
Self-disclosure	4.29 (±1.09)	4.16 (±1.06)	3.94 (±1.06)	3.73 (±1.11)	3.81 (±1.10)	3.49 (±1.16)	3.94 (±1.11)	0.14
Other-disclosure	4.38 (±1.11)	4.37 (±1.07)	4.41 (±0.98)	4.23 (±1.09)	4.30 (±1.04)	4.20 (±1.20)	4.34 (±1.06)	0.16
Initiative	4.18 (±1.04)	3.98 (±1.01)	3.87 (±0.97)	3.90 (±0.96)	3.91 (±0.98)	3.78 (±1.12)	3.93 (±1.00)	0.11
Influence	4.09 (±0.83)	3.98 (±0.78)	3.99 (±0.74)	4.08 (±0.78)	4.01 (±0.78)	4.03 (±0.97)	4.02 (±0.80)	0.08
Time duration (mins)	35.78 (±45.26)	37.50 (±42.65)	43.93 (±47.79)	32.83 (±35.45)	37.55 (±40.96)	32.71 (±34.80)	38.09 (±43.08)	0.21
Positive affect	36.87 (±10.53)	37.23 (±9.95)<	37.35 (±10.93)	36.13 (±11.07)	36.94 (±10.85)	38.42 (±11.57)	37.08 (±10.77)	0.79
Negative affect	23.04 (±9.12)	22.83 (±8.91)	23.50 (±9.14)	23.88 (±9.18)	23.45 (±9.58)	22.81 (±10.37)	23.32 (±9.28)	0.78

Note. Interaction quality is the average score of items of pleasure, intimacy, satisfaction, self-disclosure, other-disclosure, initiative, and influence. ICC: intraclass correlations.

Model 2: Social interaction and emotional experience in different partners

In this model, we wanted to know the association between social interaction and emotional experience in different types of relationships. Therefore, we split the data according to six types of interaction partners, specifically children, relatives, friends, neighbors, colleagues, and strangers. We also established a null model first and established the formal models as follows

{Outcome}_{ij} = β_{0j} + β_{01} quality + β_{2 j} duration + ε_{ij}

(4)

β_{0j} = γ_{00} + γ_{01} age + γ_{02} gender + γ_{03} marriage + γ_{04} education + γ_{05} economy + μ_{0j}

(5)

β_{1 j} = γ_{10} + μ_{1 j}

(6)

β_{2 j} = γ_{20} + μ_{2 j}

(7)

where Outcome in Model 2 is PA or NA, quality is an average score of the interaction quality of the seven items, and duration is duration time for each interaction. Similar to Model 1, the intercept and slopes are both set as random effects. Particularly, quality and duration are centered within a person.

Results

Overview of analysis

Face-to-face interaction was shown to be the primary form of interactions for the older adults in our study, as shown in Table 2. It accounted for 79% (7044/10,193) of the total number of interactions. The chi-square test was conducted, and results showed that the times of interaction with close partners were significantly higher than those with other types of partners, χ² = 1381.83, df = 1, p < .001.

Table 2.

The frequency of interaction depends on interaction partners and way of interaction.

	Close Partners			Other Partners			Total
	Children	Relatives	Friends	Neighbors	Colleagues	Strangers	Total
Frequency	1349	1871	2866	1564	1765	778	10,193
Interaction way
Face to face	800	1180	1716	1473	1254	621	7044
Telephone	340	494	490	63	297	81	1765
Online	209	197	660	28	214	76	1384

Model 1: Social interaction and interaction partners

The results of Model 1 are summarized in Table 3.

Table 3.

The model results of social interaction and interaction partners and ICC for outcomes in Model 1

	Pleasure Est. (S.E.)	Satisfaction Est. (S.E.)	Intimacy Est. (S.E.)
Intercept	4.500***(0.227)	4.642***(0.218)	3.893***(0.282)
Sex	0.101 (0.073)	0.066 (0.070)	0.143 (0.093)
Age	−0.004 (0.005)	−0.002 (0.004)	0.005 (0.006)
Marriage	−0.440 (0.272)	−0.400 (0.261)	−0.765*(0.343)
Education	−0.074 (0.065	−0.080 (0.062)	−0.076 (0.082)
Economy	0.095**(0.036)	0.055 (0.035)	0.073 (0.046)
Partner	0.292***(0.023)	0.268***(0.022)	0.712***(0.053)
	Initiative	Influence	Self-disclosure
	Est. (S.E.)	Est. (S.E.)	Est. (S.E.)
Intercept	3.674***(0.170)	4.326***(0.028)	4.068***(0.194)
Sex	−0.027 (0.055)	−0.055 (0.039)	0.084 (0.062)
Age	0.008*(0.003)	0.001 (0.002)	0.001 (0.004)
Marriage	−0.420*(0.208)	−0.225 (0.148)	−0.667**(0.235)
Education	0.003 (0.049)	−0.039 (0.035)	−0.032 (0.056)
Economy	0.033 (0.027)	−0.004 (0.019)	0.065*(0.031)
Partner	0.192***(0.034)	0.009 (0.028)	0.376***(0.038)
	Other-disclosure	Quality	Duration
	Est. (S.E.)	Est. (S.E.)	Est. (S.E.)
Intercept	4.616***(0.198)	4.198***(0.139)	36.977***(10.109)
Sex	0.122*(0.064)	0.066 (0.041)	5.550 (3.107)
Age	−0.001 (0.004)	0.002 (0.003)	−0.050 (0.194)
Marriage	−0.653**(0.238)	−0.509**(0.152)	3.350 (11.499)
Education	−0.004 (0.057)	−0.050 (0.036)	−4.976 (2.755)
Economy	0.051 (0.031)	0.053**(0.020)	−0.124 (1.532)
Partner	0.062 (0.038)	0.261***(0.020)	−1.803 (1.357)

Note. Est: estimated of unstandardized coefficient; S.E.: standard error; partner = 1 means close partners, partner = 0 means other partners.

* p < .05, ** p < .01, *** p < .001.

The results of Model 1 show that the type of interaction partner was significantly associated with the quality of the social interaction, as shown in Table 3. The pleasure score when interacting with close partners was 0.292 higher than when interacting with other partners; the intimacy score when interacting with close partners was 0.712 higher than it was with other partners; satisfaction, initiative, self-disclosure, and quality scores when interacting with close partners were also higher than those when interacting with other partners. However, influence and other-disclosure scores when interacting with close partners were not significantly higher than those when interacting with other partners. Additionally, the interaction partner was not significantly associated with the time duration of the social interaction.

Model 2: Social interaction and emotional experience across different partners

The null model showed that the ICC was 0.79 for PA and 0.78 for NA. Model 2 results are summarized in Table 4.

Table 4.

The model results of association between quality of social interaction in six types of partners.

	Interaction with Children		Interaction with Relatives
	PA Est (S.E)	NA Est (S.E)	PA Est (S.E)	NA Est (S.E)
Intercepts	32.554^***(6.338)	20.060^***(5.311)	36.379^***(5.526)	17.149***(5.028)
Sex	2.594 (1.695)	−0.425 (1.435)	1.398 (1.450)	0.293 (1.317)
Age	0.081 (0.115)	−0.013 (0.097)	0.020 (0.093)	0.048 (0.084)
Married	−11.672 (7.890)	2.657 (6.653)	−5.154 (5.138)	0.946 (4.663)
Education	0.641 (1.510)	2.209 (1.279)	−1.388 (1.249)	1.044 (1.135)
Economy	1.400 (0.862)	−0.075 (0.730)	0.970 (0.713)	0.030 (0.647)
Quality	0.903**(0.319)	−0.870**(0.316)	0.696*(0.337)	−0.623*(0.280)
Duration	−0.002 (0.003)	0.005 (0.003)	0.008*(0.004)	−0.001 (0.003)
	Interaction with friends		Interaction with neighbors
	PA	NA	PA	NA
	Est (S.E)	Est (S.E)	Est (S.E)	Est (S.E)
Intercepts	33.612***(6.338)	17.221***(4.278)	34.509***(5.526)	17.483***(1.350)
Sex	0.978 (1.473)	−0.362 (1.275)	1.874 (1.559)	0.952 (1.350)
Age	0.012 (0.093)	0.038 (0.081)	−0.033 (0.097)	−0.027 (0.084)
Married	−4.132 (5.213)	0.966 (4.517)	−1.087 (5.445)	2.497 (4.728)
Education	0.102 (1.288)	2.486 (1.115)	−0.547 (1.392)	2.148 (1.204)
Economy	1.362 (0.741)	0.118 (0.642)	0.947 (0.762)	0.236 (0.660)
Quality	0.682**(0.262)	−0.234 (0.210)	0.801*(0.375)	−0.165 (0.280)
Duration	0.002 (0.003)	0.001 (0.002)	0.002 (0.005)	0.003 (0.004)
	Interaction with colleagues		Interaction with strangers
	PA	NA	PA	NA
	Est (S.E)	Est (S.E)	Est (S.E)	Est (S.E)
Intercepts	39.447***(5.713)	18.168**(5.438)	65.364***(9.148)	12.809 (8.930)
Sex	1.968 (1.570)	0.199 (1.491)	1.544 (1.992)	−0.018 (1.956)
Age	−0.015 (0.097)	0.008 (0.092)	−0.007 (0.164)	−0.119 (0.161)
Married	−7.045 (5.357)	0.357 (5.069)	−36.098*(15.893)	16.516 (15.583)
Education	−1.118 (1.391)	1.732 (1.324)	1.022 (1.545)	0.885 (1.518)
Economy	1.244 (0.775)	0.515 (0.736)	1.390 (0.922)	0.284 (0.906)
Quality	0.626 (0.119)	0.232 (0.327)	0.838 (0.490)	−0.104 (0.506)
Duration	−0.008 (0.005)	−0.005 (0.004)	0.013 (0.009)	−0.008 (0.009)

Note. Est: estimated of unstandardized coefficient; S.E.: standard error; PA: positive affect; NA: negative affect.

* p < .05, ** p < .01, *** p < .001.

As Table 4 shows, associations differ between quality of interaction and emotional experience, depending on the particular interaction partner. When interacting with children, the interaction quality score was significantly associated with both PA and NA scores. As the interaction quality score increased by 1 over an individual’s average score, the PA score would increase by 0.903, and the NA score would decrease by 0.870. When interacting with relatives, the interaction quality score was again significantly associated with PA and NA scores. When interacting with friends, the interaction quality score was significantly positively associated with PA scores. As the interaction quality score increased by 1 over an individual’s average score, the PA score would increase by 0.682. A similar association also existed when interacting with neighbors. In contrast, the interaction quality score was not significantly associated with either PA or NA scores when interacting with colleagues or strangers. However, duration time was significantly positively associated with PA score only when interacting with relatives. As duration time increased by 1 minute over an individual’s average time, the PA score would increase by 0.005.

Discussion

First, we found that interactions with close partners were experienced as being of higher quality than interactions with other partners. This is similar to previous findings (Downie et al., 2008). Older adults retain a stable inner circle which they use as social convoys (Antonucci & Akiyama, 1995). In such close, secure relationships, older adults are more proactive and willing to express themselves, showing higher levels of self-disclosure and initiative. Meanwhile, they are able to feel more pleasure, satisfaction, and intimacy from interactions with their close partners. As socioemotional selectivity theory suggests, older adults change their aims in interaction as they age, turning to people and activities that better meet their emotional needs (Carstensen, 1991). Therefore, in interactions with children, relatives, and friends, older adults focus more on the sense of emotional support they receive from the experience.

We found that the association between the quality of social interaction and emotional experience varies depending on the type of partner, which is also consistent with previous findings (Chui et al., 2014). When interacting with children or relatives, the quality of the social interaction is associated with both PA and NA; in interactions with friends or neighbors, interaction quality is only positively associated with PA; interactions with colleagues or strangers is not associated with either PA or NA. When interacting with children or relatives, as the interaction quality increases, PA increases and NA decreases, reflecting that older adults feel happier and experience a closer association between interaction quality and emotional experience in interactions with family members. However, a previous study done in the United States found that interactions with friends had a stronger positive affect than those with family for older adults (Vogel et al., 2017). The different result in our study could imply that culture might play an important role in measuring the quality of personal interactions.

There are some limitations in the present study. First, more than half of the participants in the current study had a university education. This may limit the representativeness of our sample, and thus, we must be cautious in generalizing our conclusions. Second, we did not control some variables that could have possibly affected the research results. For example, if the time between two interactions is too short, emotions stimulated by the first interaction could influence the emotions stimulated by the second. Additionally, we also did not control for participants’ physical functioning level (e.g., the flexibility of legs and vision) nor their proficiency with electronic devices in the current study. With these limitations in mind, this study suggests that the quality of family interaction is a key factor affecting the quality of life of older adults in the context of Chinese culture.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Key R&D Program of China [Grant/Award Number: 2020YFC2003000] and the National Natural Science Foundation of China [Grant/Award Number: 71,771,027].

Ethical approval

The studies involving human participants were reviewed and approved by the Shaanxi Normal University. The participants provided their written informed consent to participate in this study.

Open research statement

As part of IARR’s encouragement of open research practices, the authors have provided the following information: This research was not pre-registered. The data used in the research are not available.

ORCID iDs

Xingyu Zhang

Xiying Li

References

Aikman

G. G.

Oehlert

M. E.

(2001). Geriatric depression scale. Clinical Gerontologist, 22(3-4), 63–70. http://doi.org/10.1300/J018v22n03_07

Antonucci

Akiyama

(1995). Convoys of social relations: Family and friendships within a life span context. In Handbook of aging and the family (pp. 355–371). Greenwood Press/Greenwood Publishing Group.

Carstensen

L. L.

(1991). Selectivity theory: Social activity in life-spancontext. In Schaie

K. W.

Lawton

M. P.

(Eds.), Annual review ofgerontology and geriatrics (11, pp. 105–217). Springer.

Carstensen

L. L.

Fung

H. H.

Charles

S. T.

(2003). Socioemotional selectivity theory and the regulation of emotion in the second half of life. Motivation and Emotion, 27(2), 103–123. http://doi.org/10.1023/a:1024569803230

Chui

Hoppmann

C. A.

Gerstorf

Walker

Luszcz

M. A.

(2014). Social partners and momentary affect in the oldest-old: The presence of others benefits affect depending on who we are and who we are with. Developmental Psychology, 50(3), 728–740. http://doi.org/10.1037/a0033896

Donnelly

L. R.

MacEntee

M. I.

(2012). Social interactions, body image and oral health among institutionalised frail elders: An unexplored relationship. Gerodontology, 29(2), e28–e33. http://doi.org/10.1111/j.1741-2358.2011.00523.x

Downie

Mageau

G. A.

Koestner

(2008). What makes for a pleasant social interaction? Motivational dynamics of interpersonal relations. The Journal of Social Psychology, 148(5), 523–534. http://doi.org/10.3200/SOCP.148.5.523-534

Liu

Xie

Q. W.

Lou

V. W. Q.

(2018). Everyday social interactions and intra-individual variability in affect: A systematic review and meta-analysis of ecological momentary assessment studies. Motivation and Emotion, 43(2), 339–353. http://doi.org/10.1007/s11031-018-9735-x

Marsh

H. W.

Hau

K.-T.

Artelt

Baumert

Peschar

J. L.

(2006). OECD's brief self-report measure of educational psychology's most useful affective constructs: Cross-cultural, psychometric comparisons across 25 countries. International Journal of Testing, 6(4), 311–360. http://doi.org/10.1207/s15327574ijt0604_1

10.

Nezlek

J. B.

Schütz

Sellin

(2007). Self-presentational success in daily social interaction. Self and Identity, 6(4), 361–379. http://doi.org/10.1080/15298860600979997

11.

Park

Chesla

(2007). Revisiting confucianism as a conceptual framework for Asian family study. Journal of Family Nursing, 13(3), 293–311. http://doi.org/10.1177/1074840707304400

12.

Ramsey

M. A.

Gentzler

A. L.

(2015). An upward spiral: Bidirectional associations between positive affect and positive aspects of close relationships across the life span. Developmental Review, 36, 58–104. http://doi.org/10.1016/j.dr.2015.01.003

13.

Reis

H. T.

Wheeler

(1991). Studying social interaction with the rochester interaction record. In Advances in Experimental Social Psychology (24, pp. 269-318). Academic Press.

14.

Schmitt

Kliegel

Shapiro

(2007). Marital interaction in middle and old age: A predictor of marital satisfaction? The International Journal of Aging and Human Development, 65(4), 283–300. http://doi.org/10.2190/AG.65.4.a

15.

Vogel

Ram

Conroy

D. E.

Pincus

A. L.

Gerstorf

(2017). How the social ecology and social situation shape individuals' affect valence and arousal. Emotion, 17(3), 509–527. http://doi.org/10.1037/emo0000244

16.

Watson

Clark

L. A.

Tellegen

(1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070. http://doi.org/10.1037/0022-3514.54.6.1063