More Than Words? The Effect of Line Character Sticker Use on Intimacy in the Mobile Communication Environment

Abstract

This study investigated how messaging app Line’s character sticker use may contribute to the perception of intimate experience and enhance relationship satisfaction in both positive and negative emotion situations. A 2 (situation valence: positive emotion and negative emotion) × 3 (response style: text, sticker, and text and sticker) × 3 (scenario: career, romance, and education) mixed design, with situation valence and response style between-subjects variables and scenario a within-subjects variable, was employed. The results revealed the combination of a text and sticker response to a partner’s disclosure can produce the highest level of intimate experience, followed by text- and sticker-only responses. It further suggests that the cartoon-like Line sticker may better convey positive than negative emotions because the detailed illustration is more similar to real-life nonverbal behavior expressing humor and happiness and that may facilitate imagined closeness between communication partners when moving through space. The hyperpersonal affordance of text-based messages to foster relationship may be further distinguished between the cognitive and the affective levels in the messaging app context.

Keywords

smartphone application (app)emoticon intimacy self-disclosure hyperpersonal

With a 203% year-over-year growth in 2013, the adoption of mobile messaging/communication applications (apps), such as Line, WhatsApp, WeChat, KakaoTalk, and Facebook Messenger, is increasing rapidly (Khalaf, 2014). Like text-based short message service (SMS), which developed as a form of instant messaging (IM; Leung, 2007), messaging apps offer real-time text transmissions, but their users are also able to exchange images and voice and video media. In addition, most of these popular messaging apps provide idiosyncratic character emoticons/stickers that differ from traditional emoticons (Table 1) during chat sessions, allowing for rich emotional communication. Line, for example, is now the world’s number one nongaming app in terms of revenue, and 20% of the company’s revenue source is from stickers (App Annie, 2013). A market report also found that Line users tend to choose stickers instead of text when they need to express themselves, and couples even get over fights more easily when using multiple stickers to apologize (Olson, 2013).

Table 1.

The Evolution of the Emotions.

Emoticons are typographical symbols created to resemble facial expressions and to facilitate the representation of nonverbal cues in the computer-mediated communication (CMC) environment (Walther & D’Addario, 2001). Research on facial expressions in off-line settings has demonstrated how social presence (Short, Williams, & Christie, 1976) affects the extent to which one expresses emotion (Fridlund, 1991). Closely related to social presence is the concept of immediacy (Wiener & Mehrabian, 1968) and intimacy (Argyle & Dean, 1965). The former is a measure of psychological distance, whereas the latter is expressed by verbal and nonverbal behaviors. Nonverbal cues are important components of intimacy (Prager, 1997), as they convey warmth and affection (Floyd & Burgoon, 1999). In much the same way as nonverbal behaviors function to reduce the ambiguity of the intended emotion expression and intensify or lessen the emotion expression in the face-to-face (FtF) setting, emoticons help to clarify, accentuate, or emphasize textual messages during CMC interactions (Derks, Fischer, & Bos, 2008). In the text-based CMC, emoticons serve as nonverbal cues that supplement the meaning of text messages when social context cues are absent (Lo, 2008; Rezabek & Cochenour, 1998).

Line as a messaging app on smart devices also demonstrates the difference between the notion of social presence (Short et al., 1976) and connectedness (Kuwabara, Watanabe, Ohguro, Itoh, & Maeda, 2002). Social presence is the degree of salience of the other person using a communication medium, while connectedness is an emotional involvement that may be evoked by the presence of the other person. IM conveys connectedness through the awareness of online others, even without message exchange, while the exchange of messages through mobile devices may lead to connectedness. The sense of connectedness is a feeling of staying in touch, sharing, belonging, and intimacy (IJsselsteijn, van Baren, & van Lanen, 2003), and mobile phones allow for perpetual contact (Katz & Aakhus, 2002). Given this rationale, although messaging apps are developed based on the idea of IM allowing users to interact instantly with their peers, they are installed on mobile devices as a means to constantly connect with family and friends and to strengthen users’ close social bonds while moving across physical space (Auter, 2007; Wei & Lo, 2006).

On the other hand, however, the development of communication technology may have a negative influence on human relationships. As Turkle (2012) has argued, people unnecessarily use mobile phones to clean up social relationships, and by doing so, they sacrifice their conversations for mere connection. While past research based on the hyperpersonal model (Walther, 1996) has demonstrated that the characteristics of text-based CMC, such as anonymity and the absence of nonverbal cues, facilitate more frequent and intimate self-disclosure, leading to increased intimacy (Jiang, Bazarova, & Hancock, 2011), it is inconclusive whether the perception of intimacy can also be enhanced in the mobile phone use context.

Further, researchers have suggested that facial expressions may even function better than vocal and spatial nonverbal cues in judging positivity because message receivers tend to connect a smile with positivity, which has no equivalent in voice or body language (Wu, Lu, & Tao, 2010). As such, while emoticons are used in different social contexts to symbolize one’s feelings, particularly to express humor (Derks et al., 2008), it is still unclear whether intimate relationships may be developed and perceived by using them in different emotional contexts. Moreover, although emoticons are now commonly recognized and used in CMC, they may even have become ritual over time as they require less effort than composing text messages (Walther & D’Addario, 2001). In fact, recent research investigating private communications data from individuals’ smartphones found that the overall amount of emoticon use in SMS was not very high (Tossell et al., 2012). However, unlike emoticons formed using ordinary characters on computer keyboards, Line stickers are used much more often in mobile communication. These oversized cartoon-like and character-driven emoticons portray both body language and facial reaction. The comprehensive illustration of these character stickers is likely to be more elaborate and expressive than traditional emoticons, thus enhancing users’ socioemotional experience. Therefore, this study explores how Line character stickers may be used as nonverbal cues to contribute to a sense of intimacy and how that may increase relationship satisfaction in the mobile communication context.

Model of Intimacy

Intimacy is the feeling of closeness and affection that involves sharing one’s inner thoughts and feelings with another person, nonverbal engagement, and sexual activity (Berscheid & Reis, 1998). It is a component in different types of relationships, such as marriage, kinship, and friendship (Aries & Johnson, 1983). Intimacy is an experiential outcome of an interpersonal and transactional process that reflects two principal components: self-revealing disclosure and partner responsiveness (Reis & Shaver, 1988). It is a process that can sequentially reveal relevant thoughts, feelings, and behaviors, each of which is affected by antecedent conditions and anticipated consequences (Reis & Patrick, 1996).

A three-tiered definition elucidates intimacy as a process of interaction in which partners (1) self-disclose, (2) feel a positive affect (PA), and (3) attentively listen to and understand each other (Prager, 1997). Andersen, Guerrero, and Jones (2006) extend Prager’s (1997) notion of intimacy and place nonverbal behaviors at the core of intimate interaction. While verbal messages enhance closeness in content or in form, such as the use of inclusive first person pronouns, intimacy is further conveyed through nonverbal behaviors such as displaying warm facial expressions, gazing, and touching (Andersen, 1998). Nonverbal intimacy expression and intimate experience are interrelated, and they ultimately lead to the enhancement of a relationship.

Intimacy is both experienced and expressed in social interactions. It situates in internal processes related to perceptions and emotions. An intimate experience includes three dimensions: PA, shared knowledge and understanding (SKU), and feelings of warmth and closeness (FWC; Andersen, Guerrero, & Jones, 2006). At the perceptional level, knowledge is gained through intimate interaction and stored in cognitive schemas, while the perception of being liked, accepted, and understood is an important part of the intimacy experience. At the emotional level, intimacy is experienced via interactions in close relationships with the presence of positive verbal and nonverbal behavior that reflects and creates feelings of warmth.

Although the past research has revealed different aspects of intimacy, one constant is that communication, both verbal and nonverbal, plays a central role in the intimacy exchange process (Andersen, 1998). Communicators can convey multiple layers of messages through verbal and nonverbal communication. Relational communication reflects identifiable affective qualities of interaction that are embedded in verbal messages and nonverbal behaviors expressed by communication partners (Andersen et al., 2006). The current study approaches the construct of intimacy based on Prager’s (1997) theories that emphasize self-disclosure and partner responsiveness and Andersen et al.’s (2006) extended model of intimate interaction to explore how self-disclosure along with partners’ verbal and nonverbal responses may contribute to the perception of intimacy in the mobile messaging app context.

Self-Disclosure and Supportive Communication

Self-disclosure is a process in which one reveals private or personal feelings, thoughts, or attitudes to another (Derlega, Metts, Petronio, & Margulis, 1993). Intimacy has been used synonymously with self-disclosure, as it is the sense that one can openly and honestly share personal thoughts and feelings with another person that would otherwise not be expressed in other relationships (Mackey, Diemer, & O’Brien, 2000). While self-disclosure may facilitate intimacy, self-disclosure on its own is not sufficient to guarantee intimacy. The link between disclosure and intimacy is subject to how the message receiver makes sense of, interprets, and responds to the disclosure by making attributions about the discloser’s reasons for revealing (Derlega et al., 1993). More intimate and reciprocated self-disclosures generate a more satisfying relational experience (Falk & Wagner, 1985).

Emotional support, which typically occurs in close relationships, refers to the expression of concern, compassion, and comfort for individuals when they experience emotional events (Reis & Collins, 2000). Receiving emotional support is associated with psychological wellness and the success of the relationship (Burleson, 2003; Weber, Johnson, & Corrigan, 2004). Thus, self-disclosure and emotional support have been largely suggested as key predictors of intimacy in both FtF and CMC contexts (Harper, Schaalje, & Sandberg, 2000; Hu, Wood, Smith, & Westbrook, 2004).

Self-disclosure also involves the risk of making oneself vulnerable to another (Derlega et al., 1993). It is a form of risk taking in which the potential rewards of sharing are evaluated against the costs of vulnerability to rejection, ridicule, exploitation, or betrayal (Hacker, 1981; Kelvin, 1977). Negative experience disclosure is more personal and private than positive disclosure (Prager, 1997), and it involves putting aside the masks one wears every day (Rubin, 1983). While revealing negative emotions involves making oneself vulnerable, it is also likely to enhance the sense of intimacy in a relationship. When one’s partner discloses in times of difficulty, providing emotional support, reassurance, and expressions of positive regard through verbal or nonverbal communication is an important aspect of intimate interaction (Prager, 2009).

However, revealing negative events may diminish a sense of self-worth in the eyes of one’s partner, and feeling inferior to one’s partner is associated with less relationship satisfaction and less love for the partner (Gable, Gonzaga, & Strachman, 2006). Sharing positive events, on the contrary, has often been linked to psychological wellness, and sharing good news is even related to affective benefits beyond the event itself (Wang, 2013). Feeling understood, validated, and cared for during the positive event disclosure is strongly and consistently associated with relationship satisfaction, commitment, and love (Gable et al., 2006). The ability to present one’s desired self and has it be perceived and accepted by one’s partner is fundamental to relationship development both online and off-line (Jiang et al., 2011; Lambert et al., 2013). People derive meaning from a message based not only on content but also on situational context (Antaki, Barnes, & Leudar, 2005). Therefore, this study investigates the situation valence, the positive and negative character of emotion (Colombetti, 2005), in the process of intimate interaction through Line chat. Messaging apps provide the possibility of continuous online presence, and sharing good news in real time may particularly facilitate intimate experience:

Hypothesis 1: There will be a significant difference in situation valence (positive vs. negative) such that revealing positive emotions will be more likely to enhance all three dimensions (PA, SKU, and FWC) of the intimate experience.

Emotion communication is the recognition, expression, and sharing of emotions between two or more communication partners (Derks et al., 2008). The quality of intimate interaction is embedded in both verbal messages and nonverbal behaviors (Andersen et al., 2006), and the latter can be meaningful between partners because of their mutual recognition, understanding, and shared experiences (Prager, 1997). The nonverbal primacy model also highlights the tendency for individuals to rely on nonverbal behaviors to interpret emotion support (Jones & Guerrero, 2001). Further, although both verbal and nonverbal communication can be used alone, they are more frequently used in conjunction (Guerrero, Farinelli, & McEwan, 2009).

In the FtF setting, the salience of social presence and the visibility of the communicators influence how emotions are communicated based on the social presence perspective (Short et al., 1976). Although some social context cues may be filtered out in text-based CMC, abundant research has demonstrated that the characteristics and affordance of CMC can even surpass the level of intimacy of parallel FtF interaction, given that enough time is spent communicating (Hian, Chuan, Trevor, & Detenber, 2004). In addition, emoticons can serve as a nonverbal surrogate to express emotion, to strengthen a message, and to express humor in a CMC setting (Derks et al., 2008). Increases in communicated emoticon quantity can lead to significant increases in perceived intimacy through affective technology (Janssen, Ijsselsteijn, & Westerink, 2014).

Although Line stickers may function as oversized emoticons used in CMC, they are character driven and depict more detailed facial expression and body language and may thus provide more social–emotional cues than traditional emoticons. Further, the use of mobile devices characterizes the nature of human mobility—the constant movement through geographical space. The need for connectedness is particularly salient in mobile communication, and mobile phones are used not for interaction but rather for strengthening close relationship (De Gournay, 2002). Line allows users to quickly and instantly connect with their social networks, but users may not be able to spend much time composing messages while moving across space. It appears reasonable to assume that in response to a partner’s self-disclosure, stickers may also facilitate socially oriented communication to enhance intimate relationships; however, typing text is more involving and effortful (Walther & D’Addario, 2001) and may be perceived as more intimate, particularly in time-constrained situations. While people tend to use both verbal and nonverbal expression online and off-line, the combination of text and stickers may convey the current mood of the message sender more clearly. Thus,

Hypothesis 2: There will be a significantly different effect of style (text vs. sticker vs. text and sticker) in response to a partner’s disclosure on the perceived intimate experience (PA, SKU, and FWC).

Hypothesis 2a: Those who were exposed to a response with a combination of text and Line sticker will perceive a higher level of intimate experience (PA, SKU, and FWC), followed by a text- and sticker-only response.

Intimacy is developed through a dynamic process that includes disclosure of relevant personal information to a partner, a partner’s responses to that information, and how that response is interpreted (Reis & Shaver, 1988). In positive event disclosure, for example, affection reaches a peak only when a positive experience is shared and a supportive response is provided (Lambert et al., 2013). In the supportive communication context, verbal (Burleson, 2003) and nonverbal messages (Jones & Guerrero, 2001) differ in quality to convey emotional support. However, there has been little investigation in terms of the effect of forms of supportive communication in different situation valences on intimate experience in mediated communication settings. Hence, we investigate the following research question:

Research Question 1: Does the perceived intimate experience (PA, SKU, and FWC) associated with all three response styles (text vs. sticker vs. text and sticker) differ between positive and negative situations?

Relationship Satisfaction

Relationship satisfaction is the extent to which individuals are satisfied with their relationships, and it is a strong predictor of the length and success of an intimate relationship, both online and off-line (T. L. Anderson & Emmers-Sommer, 2006). Relationship satisfaction is considered both an individual and a dyadic construct and is affected by how an individual perceives his or her partner’s various attitudes, behaviors, and communications (Guerrero, 1996). Self-disclosure, in particular, can enhance relationship satisfaction (Keelan, Dion, & Dion, 1998). One’s perception of closeness, intimacy, or satisfaction can be used to measure the development of a relationship. Intimate experience, that is, the level of affective closeness of intimate behaviors in friendships and romantic relationships (Schaefer & Olson, 1981), is closely related to the extent to which intimacy is perceived in that relationship (Levinger & Senn, 1967). Partners with high relationship satisfaction also reported higher levels of intimacy and commitment (Rusbult & Buunk, 1993). The extent to which affective closeness and intimacy are perceived can also better predict relationship satisfaction both off-line (Deenen, Gijs, & Van Naerssen, 1994) and online (Tong & Walther, 2011). The link between intimacy and relationship satisfaction in an off-line context may also be carried to the messaging app context. However, the relationship between the perception of intimate experience and relationship satisfaction under different types of responses in both positive and negative situations is underexplored. Therefore, this study investigates:

Research Question 2: In both positive and negative situations, in what way do they differ among the three intimate experience (PA, SKU, and FWC) under the three different types of responses (text vs. sticker vs. text and sticker) and relationship satisfaction?

Method

Participants

A recent market report found that Facebook usage among young adults worldwide is dropping as more people aged under 25 are gravitating toward mobile messaging apps (Olson, 2013). Thus, to better approach the population who may be most likely to use Line, survey announcements were made through popular discussion forums and the PTT Bulletin Board System (BBS), the largest BBS in Taiwan. An online survey was hosted for 2 weeks. As a reward, all participants were eligible to participate in a raffle of 50 gift certificates valued at the equivalent of US$3.5.

Participants who visited the online survey site were randomly assigned to one of the six experimental conditions. Each condition was automatically closed after 50 responses were met to ensure equal cell size, yielding a total of 300 respondents (177 males and 123 females). The mean age for the sample was 25.76 years (SD = 7.52), and participants ranged from 15 to 39 years old. Among them, 16 (5.4%) had a high school education, 193 (64.3%) had a college level of education, and 91 (30%) had a graduate level of education. On average, they have had 1 to 1½ years of Line use experience.

Research Design

A 2 (situation valence: positive emotion and negative emotion) × 3 (response style: text, sticker, and text and sticker) × 3 (scenario: career, romance, and education) mixed design, with situation valence and response style between-subjects variables and scenario a within-subjects variable, was employed (Table 2). Three scenarios were chosen as these are likely the life experience that most of the target samples would have been through based on their demographics. Further, within-subject variables were used to examine whether the effect of response styles on intimate experience would differ depending on life scenarios. Experimental assessments of the effectiveness of emotion support include four paradigms: the naturalistic paradigm, the interaction analysis paradigm, the experimental paradigm, and the message perception paradigm (Greene & Burleson, 2003). The first approach usually involves coping with an acute stressor, the second approach focuses on couples in a laboratory setting, and the third approach uses varied procedures to induce stress. This study adopted the message perception paradigm to present participants with sets of supportive messages to allow them to evaluate both the message and the message sender on a variety of dimensions. Extending this rationale, the scenario-based research that usually asks participants to imagine that close friends provide comforting messages was further adopted. The scenario-based research produces stronger effect of person-centered messages on support outcomes than confederate-based studies, as the latter is more artificial due to the many competing stimuli presented in live interaction (High & Dillard, 2012). Participants were asked to rate their intimacy experience and relationship satisfaction based on the situation they were exposed to and the messages they received in this study.

Table 2.

Experimental Design and Estimated Means of Model 0.0.

Situation Valance	Scenario	Response Type	Experimental Group	Variables/Mean

				PA	SKU	FWC
Positive emotion	Career	Text-only	G1	5.36	5.17	5.24
		Sticker-only
		Text + sticker
	Romance	Text-only	G2	5.49	5.38	5.35
		Sticker-only
		Text + sticker
	Education	Text-only	G3	5.42	5.33	5.30
		Sticker-only
		Text + sticker
Negative emotion	Career	Text-only	G4	5.00	4.87	4.94
		Sticker-only
		Text + sticker
	Romance	Text-only	G5	4.83	4.80	4.65
		Sticker-only
		Text + sticker
	Education	Text-only	G6	5.15	5.06	4.88
		Sticker-only
		Text + sticker

Note. PA/SKU/FWC is the average of its corresponding variables in all conditions of response style. PA = positive affect; SKU = shared knowledge and understanding; FWC = feelings of warmth and closeness.

Stimuli and Pilot Study

Two separate pilot tests were conducted. The first checked the stimulus materials to ensure that the situation valence of real-life scenarios were made specific distinct. Text messages were designed based on Burleson (2003) and Gable and Reis’s (2010) concept of highly supportive message quality, providing active constructive messages in positive scenarios and high person-centered messages in negative scenarios. The messages were tested by asking respondents to imagine themselves going through the real-life scenarios (career, romance, and education) and to rate which text messages they perceived as offering the most emotional support. Further, in the FtF context, there are two characteristics of nonverbal intimacy, PA and involvement and positive involvement behaviors, such as gaze, smile, and nod, which are essential to intimacy (Andersen et al., 2006; Prager, 1997). Therefore, Line default stickers that are available to users upon downloading the app were also tested to assure the stickers selected for the current study best depict emotional expressions of positive involvement for each scenario. Finally, the second pilot test was conducted to check the reliability of all constructs used in this study.

Twenty-eight participants (12 males and 16 females), 15 aged from 25 to 34 years old and 13 aged from 15 to 24 years old (M = 27, SD = 5.85), participated in the first pilot study. They were exposed to three paired experimental scenarios with different situation valence (career: layoff/promotion; romance: breakup/in love; and education: passed entrance exam/failed in entrance exam) and were asked to rate their levels of emotion using a 7-point Likert-type scale from very negative to very positive. Three separate independent sample t tests were conducted to test three different scenarios (each with their own pair). The results suggested that each paired scenario was distinct in situation valence, with t(27) = −10.58, p < .001; t(27) = −7.54, p < .001; t(27) = −10.66, p < .001, respectively. In terms of the text and stickers used for emotional support, three text-based responses for each scenario and 38 corresponding default stickers were tested. The top-rated text and its matching sticker were chosen for the formal experiment.

As the current study was conducted in Taiwan and some scale items adapted from previous studies were translated into Chinese, a second run of the pilot study was designed to test the measurement reliability. Respondents were exposed to the scenario chosen from the first run of the pilot study and were asked to rate their perceived intimate experience and relationship satisfaction. A total of 50 respondents (14 males and 36 females) completed the questionnaire, and the mean age for this sample was 25.83 (SD = 8.23). Cronbach’s α for all constructs exceed .70 suggesting that the measurement scales were reliable.

Procedure

Participants who visited the online survey link were directed to a consent form that included a short description of the study, information about confidentiality, and incentives. At the beginning of the questionnaire, a filter question first determined whether the respondents were Line users to ensure that only users would complete the survey. The remaining participants were then exposed to one of the six experimental scenarios (Table 2) with a short text description asking them to imagine the situation and to share with one of the participant’s close tie. For example, the description of the education scenario was: “After all the hard work preparing the most important entrance exam in your life, you passed/failed it and you would like to share this information with one of your close ties (family members, friends, or romantic partners).” Respondents were then directed to three mock-up Line chat screens with randomly displayed sticker, text, and text and sticker responses (Figure 1) and reported their intimate experience and relationship satisfaction each time they were exposed to the mock-up screens. Demographic and other basic information about participants’ Line use were then collected.

Measures

A 7-point Likert-type scale, ranging from 1 (strongly disagree) to 7 (strongly agree), was used to measure all scale items. Participants were asked to rate the extent to which they agreed or disagreed with each statement.

Intimate experience

Andersen et al.’s (2006) three dimensions of intimate experience were adapted in this study. Items were modified with wording changes to refer to the context of the current study. The intimate experience measured in the current study included 3 items of PA, 3 items of SKU, and 4 items of FWC. These were measured and the reliability of each construct was .96, .95, and .96, respectively.

Relationship satisfaction

Hendrick, Dicke, and Hendrick’s (1998) Relationship Assessment Scale (RAS) was adapted with wording changes. A total of 4 items were used, including “I am satisfied with my relationship with him/her” and “I am always thankful to have him/her in my life.” The reliability of this construct was .96.

Assessment of the Measurement

The normality of all variables was checked before the measurement and the model were tested. Structural equation modeling (SEM) with LISREL 8.8 and the model parameters using the robust maximum likelihood method were employed to analyze the data (Jöreskog & Sörbom, 2006). A confirmatory factor analysis was conducted to assess model fitness. Following current conventions (Jackson, Gillaspy, & Purc-Stephenson, 2009; Schreiber, Nora, Stage, Barlow, & King, 2006), comparative fit index (CFI), Tucker–Lewis index (TLI), and root mean square error of approximation (RMSEA) were reported. CFI and TLI values are recommended to be greater than .90, and RMSEA ranging from .05 to .08 is acceptable (R. E. Anderson, Hair, Tatham, & Black, 2006). All the fit statistics demonstrated that the measurement model (χ² = 144.05, df = 73, χ²/df = 1.973, p = .000) was a good fit (CFI = .99; TLI = .99; RMESA = .057, with a 90% confidence interval [CI] = [0.043, 0.071]). Thus, evaluation of construct reliability and convergent and discriminant validities was able to proceed.

Reliability and Validity Tests

The reliability, convergent validity, and discriminant validity of all the constructs used in this study were examined. Reliability and convergent validity were examined based on Cronbach’s α coefficients, factor loadings, composite reliabilities (CRs), and average variance extracted (AVE). The value of Cronbach’s α for each construct surpassed .90, and CR values for each construct also surpassed .90, indicating internal consistency of the measures (R. E. Anderson et al., 2006). Convergent validity was tested by factor loadings and AVE. Factor loadings of most items were close to or above .90, and all the AVEs were above .80, indicating the acceptance of convergent validity, as suggested by Fornell and Larcker (1981). Discriminant validity was tested by comparing the shared variance between the constructs with the AVEs of the individual constructs, which were greater than any squared correlation among constructs, confirming discriminant validity (Fornell & Larcker, 1981; Table 3). Bias related to demographic features such as gender, age, education, and occupation was also assessed. An analysis of variance for each categorical demographic variable indicated that none would affect the research outcomes.

Table 3.

Descriptive Statistics, Reliability, Zero-Order Correlation Coefficients, and Analysis of Construct Reliability and Validity.

Construct	Cronbach’s α	CR	AVE	Mean	SD	Square Root of AVE and Correlations

						PA	SKU	FWC	RS
PA	.960	.960	.900	5.200	1.310	1***^a
SK	.950	.950	.870	5.100	1.290	.98***	1***
						.010^b
						.96–1^c
FWC	.960	.960	.860	5.050	1.310	.97***	.97***	1***
						.010	.010
						.95–.99	.95–.99
RS	.960	.960	.860	5.350	1.160	.93***	.93***	.94***	1***
						.010	.010	.010
						.91–.95	.91–.95	.92–.96

Note. PA = positive affect; SKU = shared knowledge and understanding; FWC = feelings of warmth and closeness; RS = relationship satisfaction; AVE = average variance extracted; CR = composite reliability; SK = shared knowledge.

^aCorrelation coefficient between two constructs. ^b SE between two constructs. ^cConfidence interval.

***p < .001.

Results

Structured Means Analysis

To test the measurement invariance across groups, a mean structure analysis was employed. A χ² difference test was the procedure used to estimate the equality of factor structures. An evaluation of competitive models was recommended to assess the individual influencing factors (Anderson & Gerbing, 1988). The significance of individual parameters was assessed using a series of nested model comparisons. A nonsignificant change in χ² between the two models suggests the acceptance of a more parsimonious, supporting null hypothesis.

This study first assessed whether the effect of response style on intimate experience under the same situation valence would differ depending on three different scenarios. A structural model (Model 0.0) was established to compare with a more constrained model (Figure 2). The results showed no significant change, accepting null hypothesis (χ²/df = 1.76, p < .05; TLI = .98; CFI = .99; RMSEA = .13 with a 90% CI [0.01, 0.21]). That is, the mean differences in intimate experience (PA/SKU/FWC) within three experimental groups (G1, G2, and G3) under positive situation and three experimental groups (G4, G5, and G6) under negative situation were not statistically significant (Table 2). Thus, G1, G2, and G3 were able to be combined into the positive situation, while G4, G5, and G6 were combined into the negative one.

To test the effect of differences in situation valences (positive vs. negative) on all three dimensions of intimate experience (PA, SKU, and FWC), a structural model (Model 1.0) that has no between-group invariance constraints on estimated parameters (Figure 3) was established to compare with a more constrained model. The results showed a significant change in favor of Model 1.0 which is a saturated model (df = 0), rejecting the null hypothesis (Table 4). A series of between-group comparisons to test the difference between means in each group were also conducted. To avoid familywise error—the probability that any one of a set of comparisons or significance tests is a Type I error—a Bonferroni post hoc test was employed (p < .05/3 = .0017). The results showed that the perception of all three dimensions of intimacy experience in the positive situation were higher than in the negative situation (Table 5), supporting Hypothesis 1.

Table 4.

Fit Indices for Invariance Tests: Situation Valence on Intimate Experience.

Model	χ² (df)	χ²/(df)	RMSEA (90% CI)	NFI	TLI	Δχ² (Δdf)	CFI (ΔCFI)	Comparison	Decision
M1.0	0 (0)	—	0	1	1	—	—	—	Accept
M1.1: M11 = M12, M21 = M22, M31 = M32	25.16 (3)	8.39	0.22 ([0.15, 0.31])	0.98	0.93	25.16 (3)	0.98 (—)	M1.0 versus M 1.1	Reject
M1.2: M11 = M12	19.69 (1)	19.693	0.36 ([0.23, 0.50])	0.99	0.83	19.69 (1)	0.99 (—)	M1.0 versus M 1.2	Reject
M1.3: M21 = M22	15.04 (1)	15.037	0.31 ([0.18, 0.46])	0.99	0.88	15.04 (1)	0.99 (—)	M1.0 versus M 1.3	Reject
M1.4: M31 = M32	23.09 (1)	23.085	0.39 ([0.26, 0.53])	0.98	0.8	23.09 (1)	0.98 (—)	M1.0 versus M 1.4	Reject

Note. Positive affect/shared knowledge and understanding/feelings of warmth and closeness is the average of its corresponding variables in all conditions of response style. RMSEA = root mean square error of approximation; NFI = normed fit index; TLI = Tucker–Lewis index; CFI = comparative fit index; CI = confidence interval.

Table 5.

Estimated Means of Model 1.0 and 2.0 Derived From Multigroup SEM.

Estimated means of Model 1.0
Variables	PA	SKU	FWC
Mean (positive)	5.42	5.30	5.30
Mean (negative)	5.00	4.91	4.82
Estimated means of Model 2.0
Variables	T_PA	T_SKU	T_FWC
Mean	5.48	5.38	5.39
Variables	S_PA	S_SKU	S_FWC
Mean	4.26	4.18	4.02
Variables	T&S_PA	T&S_SKU	T&S_FWC
Mean	5.89	5.75	5.77

Note. T = text; S = sticker; T&S = text & sticker; PA = positive affect; SKU = shared knowledge and understanding; FWC = feelings of warmth and closeness; SEM = structural equation modeling.

To test the effect of different response types (text, sticker, and text and sticker) on all three dimensions of intimate experience, a more general structural model (Model 2.0) was established (Figure 4) to compare with a more constrained model. The results showed a significant change (χ² = 169.331, df = 6, p < .05) in favor of Model 2.0, rejecting the null hypothesis (Table 6). Hypothesis 2 was therefore supported. A post hoc comparison using Bonferroni’s procedure (p < .05/9 = .0056) further revealed that the combination of text and sticker in response to a partner’s disclosure facilitates the highest level of intimate experience in all three dimensions, followed by text-only and then sticker-only responses (Table 5). Therefore, Hypothesis 2a was supported.

Table 6.

Fit Indices for Invariance Tests: Response Type on Intimate Experience.

Model	χ² (df)	χ²/(df)	RMSEA [90% CI]	GFI	NFI	TLI	Δχ² (Δdf)	CFI (ΔCFI)	Comparison	Decision
M2.0	167.86 (48)	3.50	0.09 [0.08, 0.11]	.914	.976	.976	—	0.98 (—)	—	Accept
M2.1: Mt1 = Ms1 = Mts1, Mt2 = Ms2 = Mts2, Mt3 = Ms3 = Mts3	432.26 (54)	8.00	0.16 [0.14, 0.17]	.886	.932	.926	264.40 (6)	0.94 (0.04)	M2.0 versus M2.1	Reject
M2.2: Mt1 = Ms1	307.05 (49)	6.27	0.14 [0.12, 0.15]	.906	.952	.945	139.20 (1)	0.96 (0.02)	M2.0 versus M2.2	Reject
M2.3: Ms1 = Mts1	365.17 (49)	7.45	0.15 [0.14, 0.16]	.90	.94	.93	197.32 (1)	0.95 (0.04)	M2.0 versus M2.3	Reject
M2.4: Mt1 = Mts1	236.38 (49)	4.82	0.12 [0.10, 0.13]	.91	.97	.96	68.52 (1)	0.97 (0.01)	M2.0 versus M2.4	Reject
M2.5: Mt2 = Ms2	302.85 (49)	6.18	0.13 ([0.12, 0.15])	.91	.95	.95	134.99 (1)	0.96 (0.02)	M2.0 versus M2.5	Reject
M2.6: Ms2 = Mts2	354.63 (49)	7.24	0.15 [0.13, 0.16]	.90	.94	.93	186.78 (1)	0.95 (0.04)	M2.0 versus M2.6	Reject
M2.7: Mt2 = Mts2	227.23 (49)	4.64	0.11 ([0.10, 0.13])	.91	.97	.96	59.38 (1)	0.97 (0.01)	M2.0 versus M2.7	Reject
M2.8: Mt3 = Ms3	326.086 (49)	6.65	0.14 [0.13, 0.15]	.90	.95	.94	158.23 (1)	0.96 (0.03)	M2.0 versus M2.8	Reject
M2.9: Ms3 = Mts3	373.02 (49 )	7.61	0.15 [0.14, 0.17]	.90	.94	.92	205.16 (1)	0.94 (0.04)	M2.0 versus M2.9	Reject
M2.10: Mt3 = Mts3	226.77 (49)	4.63	0.11 [0.10, 0.13]	.91	.97	.96	58.92 (1)	0.97 (0.01)	M2.0 versus M2.10	Reject

Note. RMSEA = root mean square error of approximation; NFI = normed fit index; TLI = Tucker–Lewis index; CFI = comparative fit index; CI = confidence interval; GFI = goodness-of-fit index.

To answer Research Question 1, a baseline model was established to compare with the nested model (Figure 5), and the results showed a significant difference (χ² = 29.007, df = 9, p < .05) in favor of Model 3.0 (Table 7). A Bonferroni post hoc test (p < .05/9 = .0056) further revealed that the perception of all three dimensions of intimate experience was higher in the positive situation with all three types of response styles than in the negative situation, except that no significantly different SKU was perceived when exposed to a text response in both positive and negative situations (Table 8). An interaction effect between situation valence and response style was also assessed (Model 4.0). None of the interactions was statistically significant, yielding significant main effects on the dependent variables (Table 7).

Table 7.

Fit Indices for Invariance Tests: Interaction Effect and Intimate Experience on Relationship Satisfaction.

Model	χ² (df)	χ²/(df)	RMSEA [90% CI]	GFI	NFI	TLI	Δχ² (Δdf)	CFI (ΔCFI)	Comparison	Decision
M3.0	183.15 (96)	1.91	0.08 [0.06, 0.10]	.91	.97	.98	—	0.99 (—)	—	Accept
Interaction
M4.0: Mts11-Ms11 = Mts12-Ms12, Ms11-Mt11 = Ms12-Mt12; Mts21-Ms21 = Mts22-Ms22, Ms21-Mt21 = Ms22-Mt22; Mts31-Ms31 = Mts32-Ms32, Ms31-Mt31 = Ms32-Mt32	186.80 (102)	1.83	0.08 [0.06, 0.09]	.91	.97	.98	3.65 (6)	0.99 (0.00)	M3.0 versus M4.0	Accept
Simple effect
M3.1: Mt11 = Mt12, … , Mts31 = Mts32	212.16 (105)	2.02	0.08 [0.07, 0.10]	.91	.97	.98	29.01 (9)	0.98 (0.00)	M3.0 versus M3.1	Reject
M3.2: Mt11 = Mt12	192.39 (97)	1.98	0.08 [0.07, 0.10]	.91	.97	.98	9.24 (1)	0.99 (0.00)	M3.0 versus M3.2	Reject
M3.3: Ms11 = Ms12	194.55 (97)	2.01	0.08 [0.07, 0.10]	.91	.97	.98	11.40 (1)	0.99 (0.00)	M3.0 versus M3.3	Reject
M3.4: Mts11 = Mts12	195.50 (97)	2.02	0.08 [0.07, 0.10]	.91	.97	.98	12.35 (1)	0.99 (0.00)	M3.0 versus M3.4	Reject
M3.5: Mt21 = Mt22	189.91 (97)	1.96	0.08 [0.06, 0.10]	.91	.97	.98	6.76 (1)	0.99 (0.00)	M3.0 versus M3.5	Accept
M3.6: Ms21 = Ms22	192.54 (97)	1.98	0.08 [0.0654, 0.0996]	.91	.97	.98	9.39 (1)	0.99 (0.00)	M3.0 versus M3.6	Reject
M3.7: Mts21 = Mts22	193.02 (97)	1.99	0.08 [0.07, 0.10]	.91	.97	.98	9.87 (1)	0.99 (0.00)	M3.0 versus M3.7	Reject
M3.8: Mt31 = Mt32	194.31 (97)	2.00	0.08 [0.07, 0.10]	.91	.97	.98	11.16 (1)	0.99 (0.00)	M3.0 versus M3.8	Reject
M3.9: Ms31 = Ms32	197.54 (97)	2.04	0.08 [0.07, 0.10]	.91	.97	.98	14.39 (1)	0.98 (0.00)	M3.0 versus M3.9	Reject
M3.10: Mts31 = Mts32	196.56 (97)	2.03	0.08 [0.07, 0.10]	.46	.97	.98	13.41 (1)	0.99 (0.00)	M3.0 versus M3.10	Reject

Note. RMSEA = root mean square error of approximation; NFI = normed fit index; TLI = Tucker–Lewis index; CFI = comparative fit index; CI = confidence interval; GFI = goodness-of-fit index.

Table 8.

Estimated Means of Model 3.0 Derived From Multigroup SEM.

Variables	T_PA	S_PA	T&S_PA	T_SKU	S_SKU	T&S_SKU	T_FWC	S_FWC	T&S_FWC
Mean (positive)	5.64	4.53	6.09	5.52	4.43	5.94	5.58	4.33	5.98
Mean (negative)	5.31	3.99	5.69	5.23	3.93	5.57	5.21	3.71	5.55

Note. T = text; S = sticker; T&S = text & sticker; PA = positive affect; SKU = shared knowledge and understanding; FWC = feelings of warmth and closeness; SEM = structural equation modeling.

Path Analysis

To answer Research Question 2, an empirical structural equation model was developed (Figure 4). The analysis of this model produced a χ² of 183.15 (df = 96, χ²/df = 1.91, p < .05). The fit indices exhibited a reasonable good fit to the data (CFI = .99, TLI = .98, RMSEA = .08 with a 90% CI [0.06, 0.10]). This study estimated each path for statistical significance of the path coefficient. As shown in Figure 5, all the proposed paths were supported except the paths from the perception of SKU to relationship satisfaction under the text alone response in the positive situation, the perception of SKU to relationship satisfaction under the sticker alone response in the negative situation, and the perception of SKU to relationship satisfaction under the combination of the text and sticker response in the positive situation.

Discussion

The results of this study revealed that the combination of a text and sticker response to a partner’s disclosure can produce the highest level of intimate experience, followed by text- and sticker-only responses. The findings here illustrate how stickers used in the mobile communication app may serve the same function as the emoticons used in text-based CMC, which can compensate for the facial mimics and gestures that would otherwise be hard to express in writing and that support verbal communication (Derks et al., 2008). This study further revealed that the perception of intimate experience under all three types of responses significantly differs between situation valences. All three types of responses were generally perceived as resulting in a higher level of intimacy in the positive situation than in the negative situation, except that text responses in both the positive and the negative situations contributed to no significantly different effect on the SKU dimension of intimate experience.

In the text-based response condition, it found that the perception of SKU did not differ between situation valences. This finding is consistent across the current study. Except for three paths from SKU to relationship satisfaction, this study found positive relationships between all three dimensions of intimate experience under the three types of responses and relationship satisfaction in both positive and negative situations. While SKU is stored in cognitive schemas at the perceptual level (Andersen et al., 2006), in the text-based CMC environment, message receivers can form perceptions regarding senders’ positive and negative emotions, as the hyperpersonal model suggests (Harris & Paradice, 2007). The extent to which these schemas are accurate and perceived under different message styles in response to a partner’s disclosure differ between situation valences, thus leading to the varying quality of intimate relationships. At the emotional level, intimacy is perceived during social interaction in close relationships in the presence of PA and FWC, and thus, intimate connections via all types of responses may contribute to relationship satisfaction across situation valences.

This study also showed the significant main effects of situation valence and response type on intimate experience. In particular, positive event disclosure can better predict the sense of intimate experience when compared to negative situation disclosure. Researchers have suggested that sharing a positive experience increases one’s PA and life satisfaction both online (Wang, 2013) and off-line (Lambert et al., 2013). Enthusiastic responses not only convey a listener’s understanding and appreciation (Gable & Reis, 2010) but also signify the pride the listener feels about a partner’s good news (Aron, Aron, & Norman 2001). The notion that positive forms of social interaction are associated with elevated PA may also be carried out in the message app use context.

Theoretical Implications

The empirical findings of this study suggest a number of theoretical implications. First, the hyperpersonal model explains how one’s selective self-disclosure and the receiver’s idealization provoke exchanges that are more intimate than in the FtF communication (Walther, 1996). In the text-based CMC interaction, while socioemotional and social identity cues are limited, message receivers tend to engage in an overattribution process leading to intensified impressions of the senders and a higher level of intimacy. The findings here are in line with research in text-based CMC, which states that verbal message content still prevails over emoticons in conveying emotions (Walther & D’Addario, 2001). Text messages still have greater effect on affective interaction in the mobile communication context. It also extends the hyperpersonal model in terms of the effect of response type in creating intimacy under different situation valences in that text messages were found to trigger the same level of intimate experience at the perceptual level (SKU) in both positive and negative situations. It resonates with the schema theory that individuals develop cognitive structures called schemata to enter interactions with some sense of what behaviors are appropriate for given situations and relationships (Andersen, 1998). Texts can only provide directions for receivers to how they should retrieve or construct meaning from their own, previously acquired knowledge (Rumelhart 1980). While intimacy schemas are organized based on the extent their content focuses on the self, the partner, or the relationship (Prager, 1997), text message receivers may attribute the appropriateness of another individual’s intimate response to the relationship based on the knowledge stored in their information packets. In this context, text-only response may achieve the same level of intimacy in both positive and negative situation as communication partners construct and interpret meaning based on their past intimate experience with each other at the cognitive level. At the emotional level, however, text messages facilitate a higher level of intimacy in positive situations compared to negative situations, as intimacy is experienced mainly at an affective state with subjective feeling and attachment. Therefore, the hyperpersonal affordance of text-based messages to foster relationship may be further distinguished between the cognitive and the affective levels.

Second, there is a distinct body of literature examining the disclosure–intimacy link in the FtF and CMC contexts, which has suggested that an online environment facilitates even more self-disclosure, leading to an intensified perception of intimacy (Jiang et al., 2011). While there has been a general consensus that self-disclosure triggers intimate interaction, more attention has been paid to partner’s responses (Reis & Collins, 2000), how those responses are interpreted and attributed in this process, particularly in the online environment (Jiang et al., 2011). In the off-line context, researchers have suggested that sharing positive event makes the experience more salient while a supportive listener can further validate the experience (Lambert et al., 2013). In the mediated environment, individuals tend to use emoticons to display humor or sarcasm (Tossell et al., 2012). Emoticons convey a nuance of playfulness, indicating that the speech act performed may not be taken entirely seriously (Dresner & Herring, 2010). Thus, extending the rationale of emotion expression, this study found in the Line communication context that a partner’s supportive response to positive event disclosure is generally perceived as more intimate than in the negative event situation. Line stickers depicting a cartoon-style sense of humor may function better in positive event disclosure to strengthen a listener’s positive mood when good news is shared and a real-time response is highly anticipated. While intimate experiences at the perceptual level are not manifest in communication, they remain internal processes leading to a lesser effect on relationship satisfaction (Andersen et al., 2006). The perception of SKU in the sticker-only response condition may enhance relationship satisfaction only in the positive and not in the negative situation, as found in this study. These findings may broaden our understanding of supportive responses under different situation valences and how these intimate interactions may sustain and enhance relationships in the mobile communication environment.

Finally, building upon the theoretical framework of intimacy in the off-line context (Andersen et al., 2006; Prager, 1997), this study explored the effect of different forms of response on the perception of intimate experience and how intimate experience may influence relationship satisfaction in both positive and negative emotional situations in the mobile app context. It further extends the notion that intimacy is located at both the perceptual (SKU) and the emotional levels (PA and FWC) to suggest that the experience of intimacy may also transit from off-line to the mediated environment through nonverbal and verbal cues under different emotional valences. As such, while people express emotions such as warmth and affection through nonverbal cues off-line (Floyd & Burgoon, 1999), they use stickers to express emotions online, leading to an intimate experience at the emotional level. While people tend to express sadness when they are alone and show smiley faces more on the social condition (Jakobs, Manstead, Fischer, 2001), the cartoon-like Line sticker may better convey positive than negative emotions because the detailed illustration is more similar to real-life nonverbal behavior expressing humor and happiness and that may facilitate imagined closeness between communication partners when moving through space. While emoticons may not be sufficient enough to solve a problem in a negative situation (Derks et al., 2008), stickers in general produced a less intimate experience in the negative situation compared to the positive situation. Sticker-based SKU at the perceptual level may not enhance relationship satisfaction when sadness is expressed either.

Limitation and Future Directions

Limitations of this study included that it only explored Line users’ perception of intimacy through their partners’ interactions but did not compare the differences between users and nonusers. Line users may be more adapted to the use of stickers as a form of nonverbal expression and thus may perceive a higher level of intimacy from them. Future studies may want to investigate the extent to which emotions can be perceived when nonusers are exposed to Line stickers. Further, future studies may also want to compare the difference in effect between Line stickers and traditional emoticons on emotion expressions. Finally, although the message perception paradigm (Greene & Burleson, 2003) adopted in this experimental design is useful to test the well-developed model of message effectiveness, as researchers are able to manipulate the messages used in the study based on theoretically constructed concepts, it may not guarantee the external validity of the findings.

Footnotes

Acknowledgments

The author would like to thank the editor and two anonymous reviewers for their helpful comments and Miss Elaine Peng for her assistance in the data collection process.

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 study is fully funded by Aim for the Top University Project of Electronic Commerce Research Center, NSYSU.

References

Anderson

J. C.

Gerbing

D. W.

(1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological bulletin, 103, 411.

Andersen

P. A.

(1998). The cognitive valence theory of intimate communication. In Palmer

Barnett

G. A.

(Eds.), Progress in communication sciences (pp. 39–72). Norwood, NJ: Ablex.

Andersen

P. A.

Guerrero

L. K.

Jones

S. M.

(2006). Nonverbal behavior in intimate interactions and intimate relationships. In Manusov

Patterson

M. L.

(Eds.), The Sage handbook of nonverbal communication (pp. 259–277). Thousand Oaks, CA: Sage.

Anderson

R. E.

Hair

J. F.

Tatham

Black

(2006). Multivariate data analysis. Upper Saddle River, NJ: Pearson.

Anderson

T. L.

Emmers-Sommer

T. M.

(2006). Predictors of relationship satisfaction in online romantic relationships. Communication Studies, 57, 153–172.

Antaki

Barnes

Leudar

(2005). Self-disclosure as a situated interactional practice. British Journal of Social Psychology, 44, 181–199.

App Annie. (2013, 3 7). App Annie index: Social networking connects on iOS. Retrieved from http://blog.appannie.com/app-annie-index-social-networking/

Argyle

Dean

(1965). Eye-contact, distance and affiliation. Sociometry, 28, 289–304.

Aries

E. J.

Johnson

F. L.

(1983). Close friendship in adulthood: Conversational content between same-sex friends. Sex Roles, 9, 1183–1196.

10.

Aron

E. N.

Norman

(2001). Self-expansion model of motivation and cognition in close relationships and beyond. In Fletcher

G. J. O.

Clark

M. S.

(Eds.), Blackwell handbook of social psychology: Interpersonal processes (pp. 478–501). Oxford, England: Blackwell.

11.

Auter

P. J.

(2007). Portable social groups: Willingness to communicate, interpersonal communication gratifications, and cell phone use among young adults. International Journal of Mobile Communications, 5, 139–156.

12.

Berscheid

Reis

H. T.

(1998). Attraction and close relationships. In Gilbert

D. T.

Fiske

S. T

Lindzey

(Eds.), The handbook of social psychology (Vol. 2, pp. 193–281). Boston, MA: McGraw-Hill.

13.

Burleson

B. R.

(2003). The experience and effects of emotional support: What the study of cultural and gender differences can tell us about close relationships, emotion, and interpersonal communication. Personal Relationships, 10, 1–23.

14.

Colombetti

(2005). Appraising valence. Journal of Consciousness Studies, 12, 8–10.

15.

De Gournay

(2002). Pretense of intimacy in France. In Katz

J. E.

Aakhus

M. A.

(Eds.), Perpetual contact (pp. 193–206). Cambridge, England: Cambridge University Press.

16.

Deenen

Gijs

Van Naerssen

(1994). Intimacy and sexuality in gay male couples. Archives of Sexual Behavior, 23, 421–431.

17.

Derks

Fischer

A. H.

Bos

A. E.

(2008). The role of emotion in computer-mediated communication: A review. Computers in Human Behavior, 24, 766–785.

18.

Derlega

V. J.

Metts

Petronio

Margulis

S. T.

(1993). Self-disclosure. Newbury Park, CA: Sage.

19.

Dresner

Herring

S. C.

(2010). Functions of the nonverbal in CMC: Emoticons and illocutionary force. Communication theory, 20, 249–268.

20.

Falk

D. R.

Wagner

P. N.

(1985). Intimacy of self-disclosure and response processes as factors affecting the development of interpersonal relationships. The Journal of Social Psychology, 125, 557–570.

21.

Floyd

Burgoon

J. K.

(1999). Reacting to nonverbal expressions of liking: A test of interaction adaptation theory. Communications Monographs, 66, 219–239.

22.

Fornell

Larcker

D. F.

(1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39–50.

23.

Fridlund

A. J.

(1991). Sociality of solitary smiling: Potentiation by an implicit audience. Journal of Personality and Social Psychology, 60, 229.

24.

Gable

S. L.

Gonzaga

G. C.

Strachman

(2006). Will you be there for me when things go right? Supportive responses to positive event disclosures. Journal of Personality and Social Psychology, 91, 904.

25.

Gable

S. L.

Reis

H. T.

(2010). Good news! Capitalizing on positive events in an interpersonal context. Advances in Experimental Social Psychology, 42, 195–257.

26.

Greene

Burleson

(Eds.). (2003). Handbook of communication and social interaction skills. Mahwah, NJ: Lawrence Erlbaum.

27.

Guerrero

L. K.

(1996). Attachment-style differences in intimacy and involvement: A test of the four-category model. Communications Monographs, 63, 269–292.

28.

Guerrero

L. K.

Farinelli

McEwan

(2009). Attachment and relational satisfaction: The mediating effect of emotional communication. Communication Monographs, 76, 487–514.

29.

Hacker

H. M.

(1981). Blabbermouths and clams: Sex differences in self-disclosure in same-sex and cross-sex friendship dyads. Psychology of Women Quarterly, 5, 385–401.

30.

Harper

J. M.

Schaalje

B. G.

Sandberg

J. G.

(2000). Daily hassles, intimacy, and marital quality in later life marriages. American Journal of Family Therapy, 28, 1–18.

31.

Harris

R. B.

Paradice

(2007). An investigation of the computer-mediated communication of emotions. Journal of Applied Sciences Research, 3, 2081–2090.

32.

Hendrick

S. S.

Dicke

Hendrick

(1998). The relationship assessment scale. Journal of Social and Personal Relationships, 15, 137–142.

33.

Hian

L. B.

Chuan

S. L.

Trevor

T. M. K.

Detenber

B. H.

(2004). Getting to know you: Exploring the development of relational intimacy in computer-mediated communication. Journal of Computer-Mediated Communication, 9. http://onlinelibrary.wiley.com/doi/10.1111/j.1083-6101.2004.tb00290.x/full

34.

High

A. C.

Dillard

J. P.

(2012). A review and meta-analysis of person-centered messages and social support outcomes. Communication Studies, 63, 99–118.

35.

Wood

J. F.

Smith

Westbrook

(2004). Friendships through IM: Examining the relationship between instant messaging and intimacy. Journal of Computer-Mediated Communication, 10. http://onlinelibrary.wiley.com/doi/10.1111/j.1083-6101.2004.tb00231.x/full

36.

IJsselsteijn

van Baren

van Lanen

(2003). Staying in touch: Social presence and connectedness through synchronous and asynchronous communication media. Human-Computer Interaction: Theory and Practice (Part II), 2, 924–928.

37.

Jakobs

Manstead

A. S.

Fischer

A. H.

(2001). Social context effects on facial activity in a negative emotional setting. Emotion, 1, 51.

38.

Janssen

J. H.

IJsselsteijn

W. A.

Westerink

J. H.

(2014). How affective technologies can influence intimate interactions and improve social connectedness. International Journal of Human-Computer Studies, 72, 33–43.

39.

Jackson

D. L.

Gillaspy

J. A.

Jr. Purc-Stephenson

(2009). Reporting practices in confirmatory factor analysis: An overview and some recommendations. Psychological Methods, 14, 6.

40.

Jiang

Bazarova

N. N.

Hancock

J. T.

(2011). The disclosure–intimacy link in computer-mediated communication: An attributional extension of the hyperpersonal model. Human Communication Research, 37, 58–77.

41.

Jones

S. M.

Guerrero

L. K.

(2001). The effects of nonverbal immediacy and verbal person centeredness in the emotional support process. Human Communication Research, 27, 567–596.

42.

Jöreskog

Sörbom

(2006). LISREL 8.8 for Windows [Computer software]. Lincolnwood, IL: Scientific Software International.

43.

Katz

J. E.

Aakhus

(2002). Perpetual contact: Mobile communication, private talk, public performance. Cambridge, England: Cambridge University Press.

44.

Keelan

J. P. R.

Dion

K. K.

Dion

K. L.

(1998). Attachment style and relationship satisfaction: Test of a self-disclosure explanation. Canadian Journal of Behavioural Science/Revue Canadienne Des Sciences Du Comportement, 30, 24.

45.

Kelvin

(1977). Predictability, power and vulnerability in interpersonal attraction. In Duck

(Ed.), Theory and practice in interpersonal attraction (pp. 355–378). London, England: Academic Press.

46.

Khalaf

(2014). Mobile use grows 115% in 2013, propelled by messaging apps. Retrieved from http://blog.flurry.com/bid/103601/Mobile-Use-Grows-115-in-2013-Propelled-by-Messaging-Apps

47.

Kuwabara

Watanabe

Ohguro

Itoh

Maeda

(2002). Connectedness oriented communication: Fostering a sense of connectedness to augment social relationships. Paper presented at the Applications and the Internet, 2002 (SAINT 2002). In Proceeding 2002 Symposium on Applications and the Internet, IEEE Press, 186–193.

48.

Lambert

N. M.

Gwinn

A. M.

Baumeister

R. F.

Strachman

Washburn

I. J.

Gable

S. L.

Fincham

F. D.

(2013). A boost of positive affect: The perks of sharing positive experiences. Journal of Social and Personal Relationships, 30, 24–43.

49.

Leung

(2007). Unwillingness-to-communicate and college students’ motives in SMS mobile messaging. Telematics and Informatics, 24, 115–129.

50.

Levinger

Senn

D. J.

(1967). Disclosure of feelings in marriage. Merrill-Palmer Quarterly of Behavior and Development, 13, 237–249.

51.

S. K.

(2008). The nonverbal communication functions of emoticons in computer-mediated communication. CyberPsychology & Behavior, 11, 595–597.

52.

Mackey

R. A.

Diemer

M. A.

O’Brien

B. A.

(2000). Psychological intimacy in the lasting relationships of heterosexual and same-gender couples. Sex Roles, 43, 201–227.

53.

Olson

(2013). Teenagers say goodbye to Facebook and hello to messenger apps. Retrieved from http://www.theguardian.com/technology/2013/nov/10/teenagers-messenger-apps-facebook-exodus

54.

Prager

K. J.

(1997). The psychology of intimacy. New York, NY: Guilford Press.

55.

Prager

K. J.

(2009). Intimacy. In Reis

H. T.

Sprecher

S. K.

(Eds.), Encyclopedia of human relationships (Vol. 1, pp. 919–923). Thousand Oaks, CA: Sage.

56.

Reis

H. T.

Collins

(2000). Measuring relationship properties and interactions relevant to social support. In Cohen

Underwood

L. G.

Gottlieb

B. H.

(Eds.), Social support measurement and intervention: A guide for health and social scientists (pp. 136–192). New York, NY: Oxford University Press.

57.

Reis

H. T.

Patrick

B. C.

(1996). Attachment and intimacy: Component processes. In Higgins

E. T.

Kruglanski

A. W.

(Eds.), Social psychology: Handbook of basic principles (pp. 523–563). New York, NY: Guilford.

58.

Reis

H. T.

Shaver

(1988). Intimacy as an interpersonal process. In Duck

(Ed.), Handbook of personal relationships (pp. 367–389). Chichester, England: John Wiley.

59.

Rezabek

L. L.

Cochenour

J. J.

(1998). Visual cues in computer-mediated communication: Supplementing text with emoticons. Journal of Visual Literacy, 18, 201–215.

60.

Rubin

L. B.

(1983). Intimate strangers: Men and women together. New York: Harper & Row.

61.

Rumelhart

D. E.

(1980). Schemata: The building blocks of cognition. In Spiro

R. J.

Bruce

B. C.

Brewer

W. F.

(Eds.), Theoretical issues in reading comprehension. Hillsdale, NJ: Lawrence Erlbaum.

62.

Rusbult

C. E.

Buunk

B. P.

(1993). Commitment processes in close relationships: An interdependence analysis. Journal of Social and Personal Relationships, 10, 175–204.

63.

Schaefer

M. T.

Olson

D. H.

(1981). Assessing intimacy: The pair inventory*. Journal of Marital and Family Therapy, 7, 47–60.

64.

Schreiber

J. B.

Nora

Stage

F. K.

Barlow

E. A.

King

(2006). Reporting structural equation modeling and confirmatory factor analysis results: A review. The Journal of Educational Research, 99, 323–338.

65.

Short

Williams

Christie

(1976). The social psychology of telecommunications. London, England: John Wiley.

66.

Tong

S. T.

Walther

J. B.

(2011). Relational maintenance and CMC. In Wright

K. B.

Webb

L. M.

(Eds.), Computer-mediated communication in personal relationships (pp. 98–118). New York, NY: Peter Lang.

67.

Tossell

C. C.

Kortum

Shepard

Barg-Walkow

L. H.

Rahmati

Zhong

(2012). A longitudinal study of emoticon use in text messaging from smartphones. Computers in Human Behavior, 28, 659–663.

68.

Turkle

(2012). Alone together: Why we expect more from technology and less from each other. New York, NY: Basic Books.

69.

Walther

J. B.

(1996). Computer-mediated communication impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 23, 3–43.

70.

Walther

J. B.

D’Addario

K. P.

(2001). The impacts of emoticons on message interpretation in computer-mediated communication. Social Science Computer Review, 19, 324–347.

71.

Wang

S. S.

(2013). “I Share, Therefore I Am”: Personality Traits, Life Satisfaction, and Facebook Check-Ins. Cyberpsychology, Behavior, and Social Networking, 16, 870–877.

72.

Weber

Johnson

Corrigan

(2004). Communicating emotional support and its relationship to feelings of being understood, trust, and self-disclosure. Communication Research Reports, 21, 316–323.

73.

Wei

V.-H.

(2006). Staying connected while on the move cell phone use and social connectedness. New Media & Society, 8, 53–72.

74.

Wiener

Mehrabian

(1968). Language within language: Immediacy, a channel in verbal communication. New York, NY: Ardent Media.

75.

L.-L.

H.-P.

Tao

Y.-H.

(2010). The effect of emoticons in simplex and complex task-oriented communication: An empirical study of instant messaging. Computers in Human Behavior, 26, 889–895.