Up Close and Personal on Social Media: When Do Politicians’ Personal Disclosures Enhance Vote Intention?

Abstract

Two experiments investigated how politicians’ personal disclosures on social media might affect individuals’ vote intention. In Study 1 (n = 240), a male politician’s Facebook posts centering on his private life (vs. impersonal posts highlighting public activities) enhanced U.S. participants’ intention to vote for him, mostly by heightening likability. By contrast, a female politician’s personal Facebook posts lowered perceived competence, and thereby, vote intention among those who considered Facebook ill-suited for relational purposes. Using Twitter, Study 2 (n = 258) mostly replicated the findings among South Korean participants, confirming bounded benefits of publicizing politicians’ private persona via social media.

Keywords

privatization of politics social media role congruity theory task-media fit hypothesis expectancy violations theory

With approximately 2.7 billion active social media users worldwide (Chaffey, 2017), politicians have been quick to leverage these nascent platforms to their advantage, disseminating campaign information and establishing bonds with voters (Ross & Bürger, 2014). Among the candidates running for the U.S. Senate in 2012, for example, 97% were using Facebook, as were 90.2% of those running for the House of Representatives (Gulati & Williams, 2015). Similarly, 95.3% of the members of the National Assembly in South Korea were active on at least one social media platform (National Assembly Research Service, 2013).

As compared with traditional mass media, social media are deemed to enable political candidates to “instantly engage voters in a more direct, personal, and potentially interactive manner” (Lee & Oh, 2012, p. 932). To narrow the psychological distance with voters and project a more down-to-earth, personable image, politicians often reveal mundane everyday interests (Jackson & Lilleker, 2011) and disclose intimate personal stories on social media, rather than merely publicizing professional, work-related activities (Marwick & boyd, 2011). In fact, politicians’ personal disclosure on social media can be seen as a part of the increasing trend of personalization of politics, which entails both individualization and privatization. Although the former refers to the shift of focus from political parties and institutions to individual leaders and candidates, the latter points to the growing emphasis on politicians’ private lives (vs. public performance) in political campaigns, media coverage, and voter behavior (Van Aelst, Sheafter, & Stanyer, 2012; Yaniv & Tenenboim-Weinblatt, 2016). Extending this line of inquiry, we examined how politicians’ conscious and voluntary revelation of their private personae via social media affects voters’ impressions of, and subsequently, vote intention for them.

Drawing upon role congruity theory (Eagly & Karau, 2002), we first investigated if personal disclosure confers greater advantages on male politicians than their female counterparts whose communal qualities are often seen at odds with agentic qualities necessary for successful leaders. Second, in light of the task-media fit hypothesis (Mennecke, Valacich, & Wheeler, 2000), we explored how individuals’ normative expectation of the communication channel (i.e., how appropriate it is to fulfill relational goals) moderates the effects of a politician’s personal disclosure on social media. To address these questions, we conducted two experiments, first in the United States (Study 1) and then in South Korea (Study 2), employing two different social media platforms, Facebook and Twitter, respectively.

Self-Disclosure for Relationship Building

According to social penetration theory (Altman & Taylor, 1973), interpersonal relationships develop with the disclosure of personal information. Individuals exchange impersonal, superficial, and nonintimate information at the initial stage of a relationship and forge relational intimacy by gradually revealing personal and intimate information (see Collins & Miller, 1994, for a meta-analysis). Originally proposed to account for interpersonal attraction and relationship development in a dyad, the presumed benefits of personal disclosures mostly stem from reciprocal processes of self-disclosure and, thus, may not apply to politicians’ social media posts. After all, politicians do not use social media for genuine, two-way interactions with citizens, but rather unilaterally broadcast their posts to a large audience (Ross & Bürger, 2014).

However, research on parasocial interaction has shown that audience members develop quasi-intimate relationships with media personalities in the complete absence of interaction (Horton & Wohl, 1956), and such illusory relationships are not qualitatively different from ordinary interpersonal relationships for those engaging in them (Giles, 2002). If so, it stands to reason that politicians’ pseudo-social interactions with voters via social media may nonetheless incur similar consequences as those induced by personal disclosures among peers.

Indeed, recent studies suggest that fairly nominal cues to social interactions are sufficient to induce parasocial relationships with politicians. People were more likely to feel as if they were having a one-to-one conversation with a politician when they read his responses to his followers’ questions on Twitter than his self-initiated, one-sided tweets with identical content, and such illusory feelings of interaction led to more positive candidate evaluations (Lee & Shin, 2012). Even when the topics of the tweets were identical, when a politician’s generic policy statements were couched within personal experiences, they heightened perceived intimacy with the politician among those higher on affiliative tendency (Lee & Oh, 2012). More directly germane to the present study, even though people were aware that politicians’ personal revelations on Twitter were “strategically managed,” they viewed “corporate-speak” or “work topics” as less authentic than “personal, ‘human’ revelations” (Marwick & boyd, 2011, p. 127). Collectively, these findings suggest that politicians’ public display of their private personae on social media may help forge (pseudo-)personal connections with voters and foster positive impressions.

Double Standard for Personal Disclosure?

The aforementioned benefits of personal disclosures, however, may not occur when leadership is at stake, especially for women. According to role congruity theory (Eagly & Karau, 2002), stereotypes can be binding in two different ways: (a) when there is a lack of fit between the stereotypes and traits desired for a role (e.g., leader) and (b) when the individual deviates from a prescriptive stereotype. In this view, women suffer discrimination from leadership positions because they are perceived to be less competent, ambitious, and competitive than men, lacking the traits required for a successful leader (Rudman & Phelan, 2008). If women attempt to overcome this bias by acting agentic, however, they often invite negative evaluations and even socioeconomic sanctions for not being feminine enough (i.e., backlash effect; Rudman & Phelan, 2008). Indeed, women were less liked when they succeeded in a stereotypically male domain (Heilman & Okimoto, 2007) and penalized for seeking power (Okimoto & Brescoll, 2010).

Whereas it is possible that a female politician’s social media posts revealing her private persona may alleviate negative reactions that arise from presumed deficit in communal attributes for female leaders (Heilman & Okimoto, 2007), such posts run the risk of validating traditional gender stereotypes that associate women with the private sphere/home and men with the public sphere/work (Van Zoonen, 2006; Yaniv & Tenenboim-Weinblatt, 2016), thereby reinforcing the belief that women are not well-suited for leadership roles. Considering that female leaders who exhibited a task-oriented style were less liked, but perceived as more effective than relationship-oriented females (Forsyth, Heiney, & Wright, 1997), female politicians’ personal messages may serve as a double-edged sword, eliciting mixed evaluations—likable, but incompetent.

By contrast, male politicians’ intimate personal disclosures are less likely to prompt negative reactions, because they do not shoulder the burden of proving their leadership ability (Johnson, Murphy, Zewdie, & Reichard, 2008; Rudman & Phelan, 2008). Men are generally perceived to possess traits reflecting greater competency and instrumentality than women, and even when no objective sex difference in performance was observed, participants rated men’s performance more highly than women’s (Carli, 1991). Hence, instead of inviting questions about their competence, male politicians’ personal revelations may signal that they are “complete human beings combining caring and working responsibilities” (Van Zoonen, 2006, p. 298). Moreover, when the violation of expectancy involves a positively valued behavior, it engenders even more favorable reactions than when the expectancy is confirmed (Burgoon, 1993). If so, a male politician’s communal side, as inferred from his social media posts, might be taken as a positive violation of expectations and elicit more favorable evaluations than a female politician’s intimate disclosures do.

H1: A politician’s personal disclosures on social media heighten likability as compared with impersonal posts that promote his or her public activities. Such effects are more pronounced for a male than a female politician.

H2: A politician’s personal disclosures (vs. impersonal posts) on social media lower perceived competence for a female politician, but heighten it for a male politician.

Media Appropriateness: What Are Social Media Good for?

Just as people have generalized expectations for how men and women (should) behave, they also have expectations for how certain communication media (should) function. Using different labels, such as media appropriateness (Rice, 1993) and normative images (Perse & Courtright, 1993), communication scholars have long acknowledged that there exists “a broad social consensus . . . about what these media are good for and how they are most appropriately used” (Markus, 1994, p. 523). That is, “just as different people are thought to be more appropriate for fulfilling certain interpersonal communication needs, certain media channels are thought to be satisfying for specific motives” (Westmyer, DiCioccio, & Rubin, 1998, p. 29). Although these views focus on media choice as a direct consequence of such perceptions, perceived media appropriateness for specific communicative goals may also shape the message recipients’ evaluations of the communicator (e.g., a boss who fired her subordinate via text messaging) as well as their reactions (Westmyer et al., 1998).

In a similar vein, the task-media fit hypothesis (Mennecke et al., 2000) postulates that when the objective properties of a communication channel, such as synchronicity and modality, are matched to the requirements of the communication task at hand, it enhances the efficiency of performance. However, the task-media fit may not be determined solely by technological affordances. Instead, users may judge the task-media fit differently, depending on their subjective assessments of the communication channel on various dimensions, such as its capacity to fulfill specific needs, social norms regarding its use, and peer evaluations (Flanagin & Metzger, 2001). The findings that intimate disclosures via status update (vs. direct message) of Facebook reduced perceived appropriateness, and thereby, liking for the discloser (Bazarova, 2012), show that people not only hold normative expectations as to which channel is appropriate for a given communication, but they also respond negatively when such expectations are violated.

Unlike traditional media for which cultural norms about their typical usage are well-established, however, “the normative images of new, widely used, and rapidly changing technologies are evolving quickly” (Flanagin & Metzger, 2001, p. 159). If so, there may be sufficient variance in users’ perceptions of the appropriateness of social media for interpersonal goals, which have entered our media landscape only recently. More important, a politician’s personal posts may evoke different reactions depending on how well they conform to the normative image individual users hold of the medium. Considering that media effects are amplified when the messages resonate well with the user’s existing opinions, values, and norms (Valkenburg & Peter, 2013), politicians’ intimate disclosure on social media may exert more positive influence among those who believe that the given platform is well-suited for interpersonal exchanges.

H3: A politician’s personal disclosures (vs. impersonal posts) on social media heighten (a) likability and (b) perceived competence, especially among those who consider social media more appropriate for fulfilling interpersonal goals.

Candidate sex and perceived appropriateness of social media for relational purpose may also operate jointly to moderate the effects of politicians’ self-disclosure. For instance, participants’ prior beliefs of how fitting it is to exchange personal messages on social media may serve as a precondition for politicians’ personal messages to garner any positive effects, regardless of candidate sex. Due to the paucity of theoretical or empirical work that would inform specific predictions, we proposed the following:

RQ1: Do candidate sex and perceived appropriateness of social media for fulfilling interpersonal goals jointly affect how a politician’s personal disclosures affect (a) likability and (b) perceived competence?

Positive candidate evaluations, in turn, are likely to enhance the intention to vote for the candidate. For example, individuals’ attitude and normative pressure about voting in line with various political groups significantly predicted their intention to vote, but attitude was a much stronger predictor than normative pressure (Hennessy, Delli Carpini, Blank, Winneg, & Jamieson, 2015). Likewise, candidate evaluation played a leading role in deciding how to vote in various political settings, such as U.S. presidential elections (Funk, 1999), U.S. presidential primaries (Campbell, 1983), and U.K. general elections (Stewart & Clarke, 1992). To assess behavioral implications of politicians’ personal disclosures on social media, above and beyond candidate evaluations, we predicted the following:

H4: Both (a) likability and (b) perceived competence of a politician positively predict the participant’s willingness to vote for the politician.

Processing of Personal Messages

Although the current research focused on candidate evaluations and vote intention, individuals’ prior expectations may guide their message processing as well. Specifically, researchers have reported that people find it easier to assimilate expectancy-congruent information to their existing cognitive structure, and thus can process such information more efficiently (Snyder & Uranowitz, 1978). For example, after watching a videotaped vignette of a woman identified either as a waitress or a librarian, participants better recognized characteristics of the woman that were consistent with the occupational information (Cohen, 1981). This is particularly true for well-established social categories, such as gender and race (Fyock & Stangor, 1994). After watching political ads of identical content featuring male and female candidates with varying tones (positive vs. neutral vs. negative ads), the female candidates received more attention than the male candidates when their ads concerned their families in a neutral tone (Hitchon & Chang, 1995). These findings suggest that people preferentially encode the information that supports their existing expectations. If there is a “generalized readiness to process information on the basis of the sex-linked association” (Bem, 1981, p. 345), considering that people generated more communal “prototypes” (i.e., mental representations for how leaders should behave) for female leaders and more agentic prototypes for male leaders (Johnson, Murphy, Zewdie, & Reichard, 2008, p. 40), politicians’ personal disclosures might be better recognized than their impersonal posts when made by female than male politicians.

Likewise, an individual’s normative beliefs about media appropriateness may direct his or her attention differently. In Lee and Shin’s (2014) study, participants recognized the issues mentioned in a politician’s newspaper interview better than those in his tweets, suggesting medium-specific message processing: They might have anticipated to read the politician’s issue stance in a newspaper, while expecting his personal stories on Twitter, and heeded the content that matched such expectations more carefully. Rather than comparing two different media channels, for which people presumably have different normative expectations, we measured how appropriate people think a given social media platform is for fulfilling interpersonal goals, and examined how such belief alters their processing of personal (vs. impersonal) messages conveyed through it.

H5: Participants recognize a politician’s personal social media posts better than impersonal posts, especially (a) when the politician is a woman and (b) when they consider social media more appropriate for fulfilling interpersonal goals.

Study 1

Method

Pretest

Eighty respondents (39 men; age M = 35.6, SD = 9.47) recruited via Amazon Mechanical Turk (MTurk) rated 25 posts on how intimate, personal, private, and in-depth they were (α = .89; Bazarova, 2012; see “Manipulation Check” section). Nine personal posts that scored significantly higher, all ts > 2.57, ps < .02, and nine impersonal posts that scored significantly lower than the neutral point, all ts < −8.21, ps < .001, were selected. Message length was equivalent between the personal and impersonal messages, t(10) = .16, p = .88.

Participants and procedure

A total of 240 U.S. adults (116 men; age M = 34.86, SD = 9.74) were recruited via MTurk. Within each candidate sex, they were randomly assigned to either personal or impersonal message condition. To ensure familiarity with the medium, only current Facebook users were allowed to participate.

Participants were informed that the study was on politicians’ social media use and then saw a mock-up Facebook page (see Online Appendix A1), which displayed 11 posts including two filler messages (see Online Appendix B). Personal messages concerned the politician’s interactions with friends and family or mundane everyday interests (e.g., “Happy 25th Anniversary! Thank you, Pat, for staying by my side for all these years. #MarriageBliss #FamilyLove,” “My daughter convinced me to switch to a treadmill desk at my office. I’m going to live forever”), whereas impersonal posts featured the politician’s public activities (e.g., “Today I visited the CDC to see how America is tackling adolescent obesity,” “Preventing child abuse is everyone’s responsibility. Visit http://GIO.FD/SlnHr1 and pledge to help stop abuse!”). The candidates’ pictures were blurred to prevent potential biases, but allowed for gender identification.

Measures

For likability and competence, participants indicated how well a given adjective described the politician whose Facebook page they had viewed (1 = very poorly, 7 = very well; Funk, 1999). Responses to “friendly,” “likable,” and “attractive” were averaged for likability (α = .82, M = 4.86, SD = 1.12), and those to “competent,” “intelligent,” and “has leadership abilities” were averaged for competence (α = .88, M = 4.91, SD = 1.07). Vote intention was measured by two items: “I would like this politician to run in the next election” and “If I had the opportunity, I would vote for this politician in the next election” (1 = strongly disagree, 7 = strongly agree; r = .87, M = 4.06, SD = 1.31). For message recognition, participants were shown 11 topics and indicated whether the politician had mentioned each topic or not (correct =1, incorrect = 0, M = 7.96, SD = 2.21). Media appropriateness was measured using six items capturing how appropriate the participants perceived Facebook to be for fulfilling interpersonal goals: for example, “keeping in touch with friends and family,” “socializing with people who have interests that are similar to mine,” and “making new acquaintances” (Hughes, Rowe, Batey, & Lee, 2012; Ku, Chu, & Tseng, 2013; 1 = strongly disagree, 7 = strongly agree; α = .81, M = 5.48, SD = 0.93).

Results

Manipulation check

Participants evaluated the Facebook posts they had read on the following bipolar scales (α = .83, M = 3.81, SD = 1.47): “non-intimate (1)–intimate (7),” “impersonal–personal,” “public–private,” and “superficial–in-depth” (Bazarova, 2012). Personal posts (M = 4.65, SD = 1.14) were rated significantly higher than impersonal ones (M = 2.96, SD = 1.28), t(238) = 10.85, p < .001. Both personal and impersonal messages were rated significantly different from the scale midpoint (4) in the intended direction, one-sample t(119) = 6.30, p < .001; t(119) = −8.91, p < .001, respectively. When asked to identify the candidate’s sex at the end of the questionnaire, two participants gave the incorrect answer.

Hypothesis tests

All hypotheses and research questions, except H5 on message recognition, point to the conditional indirect effects of message type on vote intention. Thus, we used structural equation modeling (SEM) with maximum likelihood estimation to test them simultaneously. We constructed a path model with the following specifications: (a) likability was regressed on message type (impersonal vs. personal), candidate sex (female vs. male), media appropriateness (MA; perceived appropriateness of Facebook for relational purposes), and their two-way and three-way interactions; (b) competence was regressed on message type, candidate sex, MA, message type × candidate sex, message type × MA (the candidate sex × MA interaction and the three-way interaction were not significant; see below for details); (c) seven exogenous variables were allowed to be correlated with each other; (d) the residuals of likability and competence were allowed to be correlated; (e) vote intention was regressed on likability and competence, but not on the exogenous variables.¹ The results indicated a good model fit to the data (see Table 1): χ²(9) = 19.16, p = .02; comparative fit index (CFI) = .972; root mean square error of approximation (RMSEA) = .069, 90% confidence interval (CI) = [.024, .111], the p value for the close-fit test (RMSEA = .05) = .21; standardized root mean square residual (SRMR) = .027.

Table 1.

SEM of Message Type Effects on Vote Intention (Study 1).

	Likability	Perceived competence	Vote intention
Message type: Personal (vs. Impersonal)	.28 (.18)	−.19 (.18)
Candidate sex: Male (vs. Female)	−.34 (.18)^†	−.14 (.18)
MA	.46 (.13)^***	.05 (.11)
Message Type × Candidate Sex	.68 (.25)^**	.51 (.26)^*
Message Type × MA	−.03 (.18)	.47 (.14)^***
Candidate Sex × MA	−.39 (.14)^**
Message Type × Candidate Sex × MA	.48 (.19)^**
Likability			.32 (.10)^**
Perceived competence			.34 (.10)^**
R ²	.22^***	.14^***	.26^***

Note. n = 240. Cell entries are unstandardized coefficients with standard errors in parentheses from SEM with maximum likelihood estimation. The residual correlation between likability and competence was .75, p < .001. MA was mean-centered. SEM = structural equation modeling; MA = media appropriateness.

†

p < .10. *p < .05. **p < .01. ***p < .001.

H1 predicted that a politician’s personal Facebook posts would increase likability, especially for a male politician, whereas H2 predicted that they would lower the perceived competence for a female politician but heighten it for a male politician. H3a and H3b predicted that personal posts would enhance likability and perceived competence, especially for those high on MA. RQ1a and RQ1b addressed if the two moderators (i.e., candidate sex, MA) would jointly alter the effects of personal posts.

There was a significant interaction between message type and candidate sex on likability, b = .68, 95% CI [0.18, 1.18], p = .01. Personal posts, as compared with impersonal posts, increased likability for the male candidate, b = .96 [0.61, 1.32], p < .001, but not for the female candidate, b = .28 [−0.08, 0.63], p = .12, supporting H1. Moreover, there was a significant three-way interaction (RQ1a), b = .48 [0.12, 0.85], p = .01. As shown in Figure 1a to Figure 1c, the interaction between candidate sex and message type was not significant for those with low MA (M − SD), b = .23 [−0.38, 0.84], p = .45, but significant for those with moderate (M), b = .68 [0.18, 1.18], p = .01, and high (M + SD) MA, b = 1.14 [0.53, 1.74], p < .001.

Figure 1.

Interaction effects (Study 1).

A significant interaction effect between message type and candidate sex emerged for perceived competence as well, b = .51, 95% CI [0.01, 1.01], p = .047. Although the effect for the male politician, b = .32 [−0.04, 0.67], p = .08, and that for the female politician, b = −.19 [−0.55, 0.16], p = .28, were in opposite and expected directions, neither was statistically significant. Therefore, H2 was not supported.

To test the interaction between message type and MA, we effect-coded candidate sex (−1 vs. 1) to obtain an average effect. The two-way interaction was not significant for likability, b = .22, 95% CI [−0.06, 0.49], p = .13, failing to support H3a. By contrast, there was a significant interaction for competence (see Figure 1d), b = .47 [0.19, 0.75], p < .001, supporting H3b. For those low on MA, a politician’s personal posts lowered the perceived competence, b = −.38 [−0.74, −0.02], p = .04; for those with moderate MA, message type made no difference, b = .06 [−0.19, 0.31], p = .64; and for those high on MA, personal posts enhanced the perceived competence: b = .50 [0.14, 0.86], p = .01. For competence, we found no significant interaction between candidate sex and MA, b = −.07 [−0.34, 0.20], p = .62, or the three-way interaction (RQ1b), b = .30 [−0.25, 0.86], p = .29.

H4a and H4b predicted that likability and perceived competence would positively predict vote intention. Both hypotheses were supported: b = .32, 95% CI [0.12, 0.51], p = .001, for liking; b = .34 [0.13, 0.54], p = .001, for perceived competence.

To probe the conditional indirect effects of message type on vote intention, we calculated its indirect effects via each mediator at each level of the moderators using bootstrapping (see Table 2). For the male candidate, personal messages increased vote intention by heightening likability, but their effect on perceived competence was limited to those with high MA. By contrast, the female politician’s personal disclosures decreased vote intention by lowering perceived competence among those with low MA.

Table 2.

Conditional Indirect Effects of Message Type on Vote Intention (Study 1).

	Low MA	Moderate MA	High MA
Message type → likability → vote intention
Female candidate	.096 [−0.037, 0.321]	.088 [−0.003, 0.251]	.081 [−0.051, 0.292]
Male candidate	.170 [0.009, 0.443]	.306 [0.106, 0.582]	.443 [0.158, 0.800]
Message type → perceived competence → vote intention
Female candidate	−.214 [−0.448, −0.067]	−.066 [−0.230, 0.025]	.083 [−0.052, 0.275]
Male candidate	−.042 [−0.255, 0.103]	.107 [−0.008, 0.278]	.255 [0.096, 0.497]

Note. n = 240. Cell entries are unstandardized coefficients with 95% bias-corrected bootstrap confidence intervals in brackets (5,000 resamples) for indirect effects of message type on vote intention for each mediator at each candidate sex and each level of MA for interpersonal goals (low [M – SD], moderate [M], high [M + SD]). MA = media appropriateness.

To test if individuals would recognize personal messages better than impersonal messages when the candidate is a woman (H5a) and when they are high on MA (H5b), we ran an ordinary least squares (OLS) regression with message type, candidate sex, MA, message type × candidate sex, and message type × MA as predictors, F(5, 234) = 4.09, p = .001, R² = .08. A significant interaction emerged between message type and candidate sex (see Figure 1e), b = −1.12, 95% CI [−2.22, −0.03], p = .04. As predicted, the female politician’s personal posts were better recognized than impersonal ones, b = .85 [0.08, 1.62], p = .03, but no such effect was found for the male candidate, b = −.28 [−1.05, 0.50], p = .48. Although the message type by MA interaction was not statistically significant, b = .52, 95% CI [−0.07, 1.12], p = .09, personal messages were better recognized than impersonal ones only by those with high MA, b = .78 [0.001, 1.55], p = .05. No significant difference was observed among those with low MA, b = −.20 [−0.99, 0.59], p = .61, or moderate MA, b = .29 [−0.26, 0.83], p = .30 (see Figure 1f). Thus, H5b was supported.²

Discussion

Overall, the results suggest bounded benefits of politicians’ personal disclosure on Facebook. After reading a male politician’s personal stories, as opposed to those highlighting his public activities, participants liked him more and became more willing to vote for him. For those who considered Facebook appropriate for relational purposes, a male politician’s personal disclosure even made him look more competent. By contrast, sharing her private life backfired for the female candidate, as those who considered Facebook unsuitable for interpersonal exchanges rated her as less competent and became less willing to vote for her than when she promoted her public activities. Findings on message recognition suggest that people tend to process the posts congruent with their expectations more easily, but differential encoding does not account for why politicians’ private disclosures yielded different outcomes for male and female politicians and those holding different normative images about the medium.

Study 2

Study 1 demonstrated how politicians’ personal disclosures on Facebook alter individuals’ reactions. However, the extent to which politicians use their private lives for publicity purposes and the public’s perceptions of the significance of such information vary across cultures (Stanyer & Wring, 2004). Similarly, the degree to which sex roles are emphasized in the public’s evaluation of male and female leaders may also be culture-bound (Sczesny, Bosak, Neff, & Schyns, 2004). If so, the interaction between candidate sex and message type observed in Study 1 may dissipate or take different forms in a different country. For instance, in a society where traditional gender-role orientation is more strictly held and the division between work and home as gendered domains is more impermeable, a male politician’s posts on his family and relationships may be received less favorably than those emphasizing his public performance.

Likewise, not all social media platforms are created equal. For example, politicians believe that Facebook attracts more diverse audiences than Twitter which appeals to “political junkies” (Ross & Bürger, 2014, p. 55) and they tend to emphasize personal aspects more in the “semi-private sphere” of Facebook than in the more “public arena” of Twitter (Enli & Skogerbø, 2013, p. 763). If so, the effects of Facebook posts found in Study 1 may not emerge with Twitter.

To address these issues, we replicated Study 1 in South Korea, using Twitter as a platform. In addition, we measured masculinity and femininity as well as political orientation as individual differences to explore (a) if more masculine or feminine individuals respond differently to male and female politician’s personal disclosures and (b) if liberals and conservatives differ in the extent to which they penalize female politicians’ intimate disclosures while welcoming those of male candidates.

Method

Pretest

Ninety-three respondents (44 men; age M = 36.92, SD = 10.38) were recruited via an online survey company Macromill Embrain in South Korea. They rated 25 posts on the same message intimacy scale used in Study 1 (α = .87). Nine personal posts rated significantly higher than the scale midpoint, all ts > 3.58, ps < .01, and nine impersonal posts rated significantly lower than the scale midpoint, all ts < −4.63, ps < .001, were selected. Message length was held equivalent between the two message conditions, t(10) = .03, p = .98.

Participants and procedure

Participants were 258 current Twitter users in South Korea (131 men; age M = 36.31, SD = 9.95) recruited from Macromill Embrain. The procedure for Study 2 was identical to Study 1, except that participants viewed a mock-up Twitter page (see Online Appendix A2) and answered questions regarding their political ideology and gender-role orientation. To achieve comparability between the two studies, Korean messages were crafted to address similar topics to those used in Study 1 with minor adjustments (see Online Appendix B). For instance, “I send my deepest sympathies to the Charleston shooting victims. We stand with you” was modified to “I send my deepest sympathies to the families of the victims of the Suwon construction site collapse. May the victims rest in peace.” Personal messages included, “Today is our 25th wedding anniversary. Thank you for staying by my side for all these years. #Marriage #Anniversary # FamilyLove” and “I recently started jogging, and tried on the sneakers that my eldest recommended. I felt as light as a feather! Maybe I should consider running a half-marathon this autumn?!” Impersonal posts included, “According to a recent report by the Ministry of Health and Welfare, adolescent obesity is a serious problem. I’ll continue to work on tackling this issue” and “The best way to prevent child abuse is to be engaged. Protect children from the shadows of abuse! http://GIO.FD/SlnHr1.”

Measures

Measures for the following variables were identical to Study 1: likability (α = .88, M = 3.59, SD = 1.28), perceived competence (α = .90, M = 3.48, SD = 1.17), vote intention (r = .81, M = 3.68, SD = 1.25), message recognition (M = 6.25, SD = 2.01), and appropriateness of Twitter for interpersonal goals (α = .85, M = 4.44, SD = 1.07).

Political ideology was assessed using a single item (1 = very liberal, 7 = very conservative, M = 3.91, SD = 1.27). Gender-role orientation was measured by the short-form Bem Sex-Role Inventory (Colley, Mulhern, Maltby, & Wood, 2009). Participants reported how well a given adjective described themselves (1 = very poorly, 7 = very well). Responses to 10 items, such as “dominant,” “independent,” and “assertive,” were averaged for masculinity (α = .85, M = 4.27, SD = 0.82). Femininity score was computed by averaging responses to 10 items, such as “gentle,” “affectionate,” and “sympathetic” (α = .89, M = 4.73, SD = 0.88).

Results

Manipulation check

Using the same items as Study 1 (α = .75, M = 4.10, SD = 1.28), personal tweets (M = 4.78, SD = 1.08) were rated as more personal, intimate, private, and in-depth than impersonal tweets (M = 3.41, SD = 1.08), t(256) = 10.14, p < .001. Personal tweets were scored higher, one-sample t(129) = 8.20, p < .001, whereas impersonal tweets were scored lower than the neutral point (4), t(127) = −6.14, p < .001. When asked to identify the candidate’s sex, only nine participants answered incorrectly.

Hypothesis tests

We used SEM with the following specifications: (a) likability and competence were regressed on message type, candidate sex, media appropriateness (MA; perceived Twitter appropriateness for relational goals), message type × candidate sex, message type × MA, femininity, masculinity, and political ideology (the interaction between candidate sex and MA, and the three-way interaction were not significant; see below); (b) correlations among the exogenous variables were allowed; (c) the residual correlation between likability and competence was estimated; and (d) vote intention was regressed only on likability and competence. The model fitted the data well (see Table 3): χ²(8) = 6.21, p = .62; CFI = 1.000; RMSEA = .000, 90% CI [0.000, 0.061], the p value for the close-fit test (RMSEA = .05) = .90; SRMR = .010.

Table 3.

SEM of Message Type Effects on Vote Intention (Study 2).

	Likability	Perceived competence	Vote intention
Message type: Personal (vs. Impersonal)	.27 (.19)	−.35 (.18)^†
Candidate sex: Male (vs. Female)	−.38 (.19)^*	−.47 (.18)^*
MA	.39 (.10)^***	.29 (.09)^**
Femininity	.07 (.09)	.10 (.08)
Masculinity	−.02 (.09)	.05 (.09)
Political ideology	.06 (.05)	.07 (.05)
Message Type × Candidate Sex	.57 (.27)^*	.79 (.26)^**
Message Type × MA	.24 (.13)^†	.21 (.12)^†
Likability			.41 (.07)^***
Perceived competence			.37 (.08)^***
R ²	.28^***	.21^***	.54^***

Notes. n = 258. Cell entries are unstandardized coefficients with standard errors in parentheses from SEM with maximum likelihood estimation. The residual correlation between likability and competence was .80, p < .001. MA was mean-centered. MA = media appropriateness; SEM = structural equation modeling.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Supporting H1, a significant message type by candidate sex interaction emerged for likability (see Figure 2a), b = .57, 95% CI = [0.04, 1.09], p = .04. A male politician’s personal disclosures on Twitter heightened likability, b = .84 [0.46, 1.21], p < .001, whereas those of a female politician had no significant effect, b = .27 [−0.10, 0.64], p = .15. The two-way interaction was also significant for competence (see Figure 2c), b = .79 [0.28, 1.29], p = .002, supporting H2. Personal tweets tended to decrease perceived competence of a female candidate, b = −.35 [−0.70, 0.01], p = .06, but made a male candidate look more competent, b = .44 [0.08, 0.80], p = .02.

Figure 2.

Interaction effects (Study 2).

A marginally significant interaction between message type and MA emerged for likability, b = .24, 95% CI [−0.01, 0.49], p = .06, as well as competence, b = .21 [−0.03, 0.45], p = .09. To probe this two-way interaction, we effect-coded candidate sex. For those with low MA (M – SD), personal tweets did not affect likability, b = .30 [−0.08, 0.67], p = .12, but they heightened likability for those with moderate (M) and high (M + SD) MA, b = .55 [0.29, 0.81], p < .001; b = .80 [0.43, 1.18], p < .001, respectively (see Figure 2b). Therefore, H3a was supported. As for competence, personal tweets had no effect on competence regardless of MA, failing to support H3b, b = −.18 [−0.54, 0.18], p = .34; b = .04 [−0.21, 0.30], p = .73; b = .26 [−0.10, 0.62], p = .15, for low, moderate, and high MAs. The three-way interaction was neither significant for likability (RQ1a), b = .22 [−0.28, 0.72], p = .39, nor for competence (RQ1b), b = .08 [−0.40, 0.56], p = .74.

Consistent with H4a and H4b, both likability and perceived competence positively predicted vote intention: b = .41, 95% CI [0.27, 0.55], p < .001; b = .37 [0.21, 0.53], p < .001.

The conditional indirect effects of message type on vote intention (see Table 4) mostly replicated those from Study 1. The male candidate’s personal tweets enhanced the participants’ intention to vote for him by improving perceived competence (except for those with low MA) as well as likability. For the female politician, personal tweets increased vote intention by boosting likability, but only among those with high MA. By contrast, her personal tweets decreased vote intention by lowering perceived competence among those with low and moderate MA.

Table 4.

Conditional Indirect Effects of Message Type on Vote Intention (Study 2).

	Low MA	Moderate MA	High MA
Message type → likability → vote intention
Female candidate	.006 [−0.182, 0.194]	.111 [−0.035, 0.294]	.215 [0.014, 0.478]
Male candidate	.239 [0.043, 0.480]	.344 [0.170, 0.570]	.448 [0.228, 0.747]
Message type → perceived competence → vote intention
Female candidate	−.210 [−0.426, −0.058]	−.129 [−0.306, −0.005]	−.047 [−0.261, 0.131]
Male candidate	.082 [−0.074, 0.285]	.163 [0.037, 0.360]	.244 [0.071, 0.511]

Note. n = 258. Cell entries are unstandardized coefficients with 95% bias-corrected bootstrap confidence intervals in brackets (5,000 resamples) for indirect effects of message type on vote intention for each mediator at each candidate sex and each level of MA for interpersonal goals (low [M – SD], moderate [M], high [M + SD]). MA = media appropriateness.

For message recognition, an OLS regression analysis showed no interaction between message type and candidate sex, b = −.04, 95% CI [−0.99, 0.90], p = .93, or MA, b = .001 [−0.45, 0.45], p = .995, so we removed these interaction terms from the final model, F(6, 251) = 4.93, p < .001, R² = .11. Participants recognized personal tweets better than impersonal ones, b = .84 [0.37, 1.32], p < .001, and the male politician’s tweets better than the female politician’s, b = .57 [0.10, 1.04], p = .02. MA was positively associated with recognition, b = .35 [0.11, 0.59], p = .004. Gender-role orientation and political ideology were unrelated to recognition, ps > .18.

Finally, we tested if the effects of message type and its interactions with candidate sex and/or MA were further moderated by femininity, masculinity, or political ideology. We found no significant interactions, all ps > .12.

Discussion

Study 2 replicated most of the key findings of Study 1. The beneficial effects of personal disclosures on candidate evaluation and vote intention were more pronounced for a male than a female politician and among those who considered Twitter better suited for relational goals. Unlike Study 1, however, message type and candidate sex affected recognition independently, ruling out the possibility that differential information encoding might explain the conditional effects of personal messages on candidate evaluation and vote intention. Finally, the observed treatment effects were not qualified by gender-role orientation or political ideology, demonstrating their robustness.

General Discussion

Privatization of politics, or “the politicization of private persona” (Langer, 2010), is often characterized by an increased media focus on politicians’ personal lives and personal qualities and “the rising importance of the politician as ‘ordinary’ person” (Van Aelst et al., 2012, p. 206). Two experiments examined (a) how politicians’ voluntary revelations of their private lives via social media affect individuals’ evaluations of and intention to vote for them, and (b) how the politician’s sex and the message recipients’ normative image of the communication channel moderate such effects, using two different platforms in the United States and in South Korea. Results were largely consistent with role congruity theory, as female politicians’ personal disclosures lowered the perceived competence, and thereby, vote intention, whereas male politicians’ competence evaluation was not affected (Study 1) or even enhanced (Study 2) by their posts on private life. Apparently, male politicians’ private posts were taken as positive expectancy violations and increased liking, and subsequently, vote intention. The task–media fit hypothesis was also supported as those who perceived social media as appropriate tools for relationship building and maintenance were more appreciative of politicians’ public display of private persona on those venues.

Theoretical Implications

Comparing a political candidate’s campaign Facebook page and the official campaign website, Dunn and Nisbett (2014) found that participants not only indicated greater emotional intimacy with the politician, but also evaluated the quality of messages more positively after viewing the Facebook page. Similarly, other researchers have documented heightened social presence and stronger (illusion of) intimacy in response to politicians’ tweets, as compared with their traditional media appearance, albeit in conjunction with the participants’ predispositions (Lee, 2013; Lee & Shin, 2014). Although acknowledging social media’s potential to emulate a more up-close and personal encounter with public figures, we focused on the variability in users’ perceptions of media appropriateness to account for the differing effects of politicians’ personal disclosures via social media. As predicted, politicians’ intimate self-disclosure garnered positive outcomes mostly for male politicians, especially when it resonated with the normative images people held about the platform.

First, personal disclosure conferred greater benefits on the male than the female candidate in terms of candidate evaluation and vote intention, validating the speculation that “convergence of personal and political life into a hybrid political persona” works out better for men than for women (Van Zoonen, 2006, p. 299). As such, female politicians’ efforts to protect the traditional boundaries between the private and the public (Van Zoonen, 2006) may indeed be a wise strategy. However, if the capacity to offer a “human” persona is crucial for political success as a way of gaining media coverage and appealing to those less interested in politics (Langer, 2010), female leaders are inherently disadvantaged as they feel forced to refrain from using their personal life for publicity and remain strictly within the public domain.

In fact, researchers have shown how gender stereotypes persist in the media coverage of election campaigns, with more frequent mentions of children and marital status when reporting on female than male candidates (e.g., Banwart, Bystrom, & Robertson, 2003; Bystrom, Robertson, & Banwart, 2001; Kittilson & Fridkin, 2008), thereby highlighting “the images of mother and wife” that carry “less authority and competence in the public arena” (Banwart et al., 2003, p. 671). Our results confirm that female candidates’ personal disclosures indeed prompt questions about their competence whereas male candidates’ do not. Although social media have the potential to enable political candidates to counter biased media portrayals by allowing them to deliver their own messages directly to the voters, the added control does not seem to mitigate gender-typed reactions.

Two possibilities were examined to further probe why participants responded more favorably to male than female politicians’ personal disclosures. First, the male candidates’ personal disclosure might have demanded greater attention due to its deviation from expectancy, thereby exerting greater influence on subsequent judgments. Contrarily, personal messages were better recognized than impersonal messages (a) only when delivered by the female candidate (Study 1) or (b) regardless of candidate sex (Study 2). Moreover, personal messages were not recognized any better when delivered by the male than the female politicians in both studies. Second, personal and impersonal messages might have been perceived differently, depending on the communicator’s sex. That is, people may hold differing thresholds of what counts as a “personal” disclosure for men and women, and thus perceive men’s private revelation as more intimate, personal, private, and in-depth, than women’s. However, in both studies, we found no significant interaction effect between message type and candidate sex on perceived message personalness (see “Manipulation Check” sections), both ps > .37. Collectively, these results indicate that it is not the greater amount of attention or different interpretations of the personal messages that elicited more favorable reactions to the male than female candidates’ intimate disclosures. Rather, participants seemed to make different inferences about the communicator from equally personal messages in light of deep-rooted sex roles, if not consciously so.

Our results also suggest that the task-media “fit,” which supposedly affects the communication outcome (Mennecke et al., 2000), is defined not only by technological affordances but also by users’ subjective assessment of the media—how fitting a given communication channel is for a particular task. Thus far, media appropriateness has been mostly conceptualized as an antecedent to media choice or media use from the communicator’s perspective (Perse & Courtright, 1993; Rice, 1993), but we demonstrated that media appropriateness, as judged by the message recipient, moderates communication effects as well. Notably, the violation of normative expectancy hurts only female, not male politicians; female leaders were more likely to be penalized for their “inappropriate” communicative behaviors than their male counterparts.

Theoretically, one might argue that those who consider social media to be more suitable for relationship building and maintenance would respond more negatively to politicians’ carefully crafted, strategic self-disclosures, which fall short of genuine, reciprocal interaction. However, just as people develop imaginary parasocial relationships with media personalities from a distance, people do not seem to distinguish social interactions among peers from pseudo-social interactions enacted by public figures on social media platforms and respond similarly.

Limitations and Future Directions

Although the current research sheds light on the bounded benefits of politicians’ personal self-disclosure via social media, some limitations deserve note. First, because we used fictitious politicians, it remains unknown if the present findings would apply to real politicians, about whom people have fixed impressions and attitudes. Not only would it be more difficult to alter voters’ existing evaluations of political leaders than to create impressions about non-existing figures, but voters may also hold specific expectations about known politicians and interpret their social media posts in light of such expectancies, which may suppress or amplify the effect of candidate sex observed herein.

Similarly, female candidates’ personal disclosures may be more favorably received if their competence is unambiguously established a priori. Unlike their male counterparts, female leaders’ competence is often subject to additional scrutiny unless their specific skills at the task are clearly demonstrated (Shackelford, Wood, & Worchel, 1996). Given that female leaders need to exhibit both communal and agentic behaviors to be perceived as effective (Johnson et al., 2008), female politicians’ personal disclosure may induce positive reactions, if and only if their agentic ability is proven.

Based on actual politicians’ social media posts and personal blogs, we created a mixture of personal messages to avoid the fixed message effect, most of which concerned family life and leisure time activities. Although this choice conforms to the common operationalization of privatization, which entails family life, past life and upbringing, leisure time, and love life (Langer, 2010; Van Aelst et al., 2012), these messages highlight the candidate’s communal side, a traditionally feminine domain. Also, we adopted similar messages in both studies to ensure message equivalence so that we could rule out alternative explanations based on content variations in case divergent findings were to emerge in Study 2. As it turned out, the findings from Study 1 were mostly replicated, adding confidence to our tentative conclusions, but it remains unclear if other kinds of personal information would elicit similar responses, and if not, why. Therefore, to enhance both internal and external validity of the current findings, future research will need to identify conceptual dimensions on which personal and impersonal messages are judged differently (e.g., perceived relevance and significance) and examine how such judgments relate to candidate evaluations.

To separate out the effects of message type, while avoiding conflation with any other attributes, we removed other information normally available on social media pages, such as profile information, the number of friends/followers, and the visitors’ comments. Such exclusions could have not only inflated the artificiality of the study stimuli, but also precluded the opportunity to examine important features that might affect individuals’ reactions to politicians’ intimate disclosures on social media. Specifically, in the absence of clear instruction, participants in the personal message condition might have believed that it was the candidate’s private Facebook page, whereas those in the impersonal message condition thought it was his or her official page. If so, the positive effects of personal messages might be due, in part, to perceived legitimacy of personal disclosures in the candidate’s private space, which may not replicate when the messages are conveyed via his or her official page. Given that matching message content to the message recipients’ normative expectations about the channel induces more positive responses, future research should explore what dispositional and contextual factors may define individuals’ prescriptive perceptions of a communication channel and the very nature of communication performed through it, and subsequently, shape various outcomes.

Conclusion

As one of the newest additions to our media repertoire, social media were once thought to be a medium through which one could create and maintain interpersonal relationships (boyd & Ellison, 2007). Realizing their potential to reach mass audiences, politicians worldwide have embraced various social media to promote their policy agenda and create (pseudo-)personal bonds with their electorate, often revealing mundane and private details that emphasize their ordinariness. Although effective for some communicators and with some audiences, personal disclosures are, by no means, a panacea for political hopefuls who strive to win the voters’ hearts. All in all, people bring different expectations about the communicator as well as the medium to the virtual conversation with politicians, and such expectations guide their reactions to identical messages.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) declared receipt of the following financial support for the research, authorship, and/or publication of this article: this work was supported by the National Research Foundation of Korea Grant funded by the Korean Government (NRF-2015S1A5A2A01012690).

Supplemental Material

The online appendices are available online at .

Notes

Author Biographies

Eun-Ju Lee (PhD, Stanford University) is a professor in the Department of Communication at Seoul National University. She examines social influence and social cognition in computer-mediated contexts.

Soo Youn Oh (MA, Seoul National University) is a doctoral candidate in the Department of Communication at Stanford University. She studies social dynamics of computer-mediated communication.

Jihye Lee (MA, Seoul National University) is a doctoral candidate in the Department of Communication at Stanford University. She studies the economics of news markets and information inequality.

Hyun Suk Kim (PhD, University of Pennsylvania) is an assistant professor in the School of Communication at the Ohio State University. His research centers on message effects in viral and persuasive comunication.

References

Altman

Taylor

D. A.

(1973). Social penetration: The development of interpersonal relationships. New York, NY: Holt, Rinehart & Winston.

Banwart

M. C.

Bystrom

D. G.

Robertson

(2003). From the primary to the general election: A comparative analysis of candidate media coverage in mixed-gender 2000 races for governor and U.S. Senate. American Behavioral Scientist, 46, 658-676. doi:10.1177/0002764202238491

Bazarova

N. N.

(2012). Public intimacy: Disclosure interpretation and social judgments on Facebook. Journal of Communication, 62, 815-832. doi:10.1111/j.1460-2466.2012.01664.x

Bem

S. L.

(1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354-364. doi:10.1037/0033-295x.88.4.354

boyd

Ellison

N. B.

(2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13, 210-230. doi:10.1111/j.1083-6101.2007.00393.x

Burgoon

J. K.

(1993). Interpersonal expectations, expectancy violations, and emotional communication. Journal of Language and Social Psychology, 12, 30-48. doi:10.1177/0261927x93121003

Bystrom

D. G.

Robertson

T. A.

Banwart

M. C.

(2001). Framing the fight: An analysis of media coverage of female and male candidates in primary races for governor and U.S. Senate in 2000. American Behavioral Scientist, 44, 1999-2013. doi:10.1177/00027640121958456

Campbell

J. E.

(1983). Candidate image evaluations: Influence and rationalization in presidential primaries. American Politics Quarterly, 11, 293-313. doi:10.1177/004478083011003002

Carli

L. L.

(1991). Gender, status, and influence. In Lawler

Markovski

Ridgeway

Walker

(Eds.), Advances in group processes (Vol. 8, pp. 89-113). Greenwich, CT: JAI Press.

10.

Chaffey

(2017, April 27). Global social media research summary, 2017. Retrieved from http://www.smartinsights.com/social-media-marketing/social-media-strategy/new-global-social-media-research/

11.

Cohen

C. E.

(1981). Person categories and social perception: Testing some boundaries of the processing effect of prior knowledge. Journal of Personality and Social Psychology, 40, 441-452. doi:10.1037/0022-3514.40.3.441

12.

Colley

Mulhern

Maltby

Wood

A. M.

(2009). The short form BSRI: Instrumentality, expressiveness and gender associations among a United Kingdom sample. Personality and Individual Differences, 46, 384-387. doi:10.1016/j.paid.2008.11.005

13.

Collins

N. L.

Miller

L. C.

(1994). Self-disclosure and liking: A meta-analytic review. Psychological Bulletin, 116, 457-475. doi:10.1037/0033-2909.116.3.457

14.

Dunn

S. G. S.

Nisbett

G. S.

(2014). Parasocial interactions online: Candidate intimacy in webpages and Facebook. The Journal of Social Media in Society, 3, 26-41.

15.

Eagly

A. H.

Karau

S. J.

(2002). Role congruity theory of prejudice toward female leaders. Psychological Review, 109, 573-598. doi:10.1037/0033-295X.109.3.573

16.

Enli

G. S.

Skogerbø

(2013). Personalized campaigns in party-centered politics: Twitter and Facebook as arenas for political communication. Information, Communication & Society, 16, 757-774. doi:10.1080/1369118X.2013.782330

17.

Flanagin

A. J.

Metzger

M. J.

(2001). Internet use in the contemporary media environment. Human Communication Research, 27, 153-181. doi:10.1111/j.1468-2958.2001.tb00779.x

18.

Forsyth

D. R.

Heiney

M. M.

Wright

S. S.

(1997). Biases in appraisals of women leaders. Group Dynamics: Theory, Research, and Practice, 1, 98-103. doi:10.1037/1089-2699.1.1.98

19.

Funk

C. L.

(1999). Bringing the candidate into models of candidate evaluation. The Journal of Politics, 61, 700-720. doi:10.2307/2647824

20.

Fyock

Stangor

(1994). The role of memory biases in stereotype maintenance. The British Journal of Social Psychology, 33, 331-343. doi:10.1111/j.2044-8309.1994.tb01029.x

21.

Giles

D. C.

(2002). Parasocial interaction: A review of the literature and a model for future research. Media Psychology, 4, 279-305. doi:10.1207/s1532785xmep0403_04

22.

Gulati

G. J.

Williams

C. B.

(2015). Congressional campaigns’ motivations for social media adoption. In Farrar-Myers

Vaughn

(Eds.), Controlling the message: New media in American political campaigns (pp. 32-52). New York: New York University Press.

23.

Heilman

M. E.

Okimoto

T. G.

(2007). Why are women penalized for success at male tasks? The implied communality deficit. The Journal of Applied Psychology, 92, 81-92. doi:10.1037/0021-9010.92.1.81

24.

Hennessy

Delli Carpini

M. X.

Blank

M. B.

Winneg

Jamieson

K. H.

(2015). Using psychological theory to predict voting intentions. Journal of Community Psychology, 43, 466-483. doi:10.1002/jcop.21697

25.

Hitchon

J. C.

Chang

(1995). Effects of gender schematic processing on the reception of political commercials for men and women candidate. Communication Research, 22, 430-458. doi:10.1177/009365095022004003

26.

Horton

Wohl

R. R.

(1956). Mass communication and para-social interaction; Observations on intimacy at a distance. Psychiatry, 19, 215-229. doi:10.1080/00332747.1956.11023049

27.

Hughes

D. J.

Rowe

Batey

Lee

(2012). A tale of two sites: Twitter vs. Facebook and the personality predictors of social media usage. Computers in Human Behavior, 28, 561-569. doi:10.1016/j.chb.2011.11.001

28.

Jackson

Lilleker

(2011). Microblogging, constituency service and impression management: UK MPs and the use of Twitter. The Journal of Legislative Studies, 17, 86-105. doi:10.1080/13572334.2011.545181

29.

Johnson

S. K.

Murphy

S. E.

Zewdie

Reichard

R. J.

(2008). The strong, sensitive type: Effects of gender stereotypes and leadership prototypes on the evaluation of male and female leaders. Organizational Behavior and Human Decision Processes, 106, 39-60. doi:10.1016/j.obhdp.2007.12.002

30.

Kittilson

M. C.

Fridkin

(2008). Gender, candidate portrayals and election campaigns: A comparative perspective. Politics & Gender, 4, 371-392. doi:10.1017/S1743923X08000330

31.

Y.-C.

Chu

T.-H.

Tseng

C.-H.

(2013). Gratifications for using CMC technologies: A comparison among SNS, IM, and e-mail. Computers in Human Behavior, 29, 226-234. doi:10.1016/j.chb.2012.08.009

32.

Langer

A. I.

(2010). The politicization of private persona: Exceptional leaders or the new rule? The case of the United Kingdom and the blair effect. The International Journal of Press/politics, 15, 60-76. doi:10.1177/1940161209351003

33.

Lee

E.-J.

(2013). Effectiveness of politicians’ soft campaign on Twitter versus TV: Cognitive and experiential routes. Journal of Communication, 63, 953-974. doi:10.1111/jcom.12049

34.

Lee

E.-J.

S. Y.

(2012). To personalize or depersonalize? When and how politicians’ personalized tweets affect the public’s reactions. Journal of Communication, 62, 932-949. doi:10.1111/j.1460-2466.2012.01681.x

35.

Lee

E.-J.

Shin

S. Y.

(2012). Are they talking to me? Cognitive and affective effects of interactivity in politicians’ Twitter communication. Cyberpsychology, Behavior, and Social Networking, 15, 515-520. doi:10.1089/cyber.2012.0228

36.

Lee

E.-J.

Shin

S. Y.

(2014). When the medium is the message: How transportability moderates the effects of politicians’ Twitter communication. Communication Research, 41, 1088-1110. doi:10.1177/0093650212466407

37.

Markus

M. L.

(1994). Electronic mail as the medium of managerial choice. Organization Science, 5, 502-527. doi:10.1287/orsc.5.4.502

38.

Marwick

A. E.

boyd

(2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society, 13, 114-133. doi:10.1177/1461444810365313

39.

Mennecke

B. E.

Valacich

J. S.

Wheeler

B. C.

(2000). The effects of media and task on user performance: A test of the task-media fit hypothesis. Group Decision and Negotiation, 9, 507-529. doi:10.1023/A:1008770106779

40.

National Assembly Research Service. (2013). Characteristics and states of the use of Internet and SNS by the members of the 19th national assembly (NARS Report No. 28). Retrieved from https://www.nars.go.kr/eng/brdList.do?cmsCd=CM0139&sgrp=S02

41.

Okimoto

T. G.

Brescoll

V. L.

(2010). The price of power: Power seeking and backlash against female politicians. Personality and Social Psychology Bulletin, 36, 923-936. doi:10.1177/0146167210371949

42.

Perse

E. M.

Courtright

J. A.

(1993). Normative images of communication media mass and interpersonal channels in the new media environment: Mass and interpersonal channels in the new media environment. Human Communication Research, 19, 485-503. doi:10.1111/j.1468-2958.1993.tb00310.x

43.

Rice

R. E.

(1993). Media appropriateness. Human Communication Research, 19, 451-484. doi:10.1111/j.1468-2958.1993.tb00309.x

44.

Ross

Bürger

(2014). Face to face(book): Social media, political campaigning and the unbearable lightness of being there. Political Science, 66, 46-62. doi:10.1177/0032318714534106

45.

Rudman

L. A.

(1998). Self-promotion as a risk factor for women: The costs and benefits of counterstereotypical impression management. Journal of Personality and Social Psychology, 74, 629-645. doi:10.1037/0022-3514.74.3.629

46.

Rudman

L. A.

Phelan

J. E.

(2008). Backlash effects for disconfirming gender stereotypes in organizations. Research in Organizational Behavior, 28, 61-79. doi:10.1016/j.riob.2008.04.003

47.

Sczesny

Bosak

Neff

Schyns

(2004). Gender stereotypes and the attribution of leadership traits: A cross-cultural comparison. Sex Roles, 51, 631-645. doi:10.1007/s11199-004-0715-0

48.

Shackelford

Wood

Worchel

(1996). Behavioral styles and the influence of women in mixed-sex groups. Social Psychology Quarterly, 59, 284-293. doi:10.2307/2787024

49.

Snyder

Uranowitz

S. W.

(1978). Reconstructing the past: Some cognitive consequences of person perception. Journal of Personality and Social Psychology, 36, 941-950. doi:10.1037/0022-3514.36.9.941

50.

Stanyer

Wring

(2004). Public images, private lives: An introduction. Parliamentary Affairs, 57, 1-8. doi:10.1093/pa/gsh001

51.

Stewart

M. C.

Clarke

H. D.

(1992). The (un)importance of party leaders: Leader images and party choice in the 1987 British election. The Journal of Politics, 54, 447-470. doi:10.2307/2132034

52.

Valkenburg

P. M.

Peter

(2013). The differential susceptibility to media effects model. Journal of Communication, 63, 221-243. doi:10.1111/jcom.12024

53.

Van Aelst

Sheafer

Stanyer

. (2012). The personalization of mediated political communication: A review of concepts, operationalizations and key findings. Journalism, 13, 203-220. doi:10.1177/1464884911427802

54.

van Zoonen

. (2006). The personal, the political and the popular: A woman’s guide to a celebrity politics. European Journal of Cultural Studies, 9, 287-301. doi:10.1177/1367549406066074

55.

Westmyer

S. A.

DiCioccio

R. L.

Rubin

R. B.

(1998). Appropriateness and effectiveness of communication channels in competent interpersonal communication. Journal of Communication, 48, 27-48. doi:10.1111/j.1460-2466.1998.tb02758.x

56.

Yaniv

N. W.

Tenenboim-Weinblatt

(2016). Politically relevant intimacy: A conceptual and empirical investigation. International Journal of Communication, 10, 5186-5205.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.28 MB