خودارائه گری دانش آموزان در فیس بوک: بررسی عوامل شخصیتی و خودتفسیری

Abstract The present research seeks to extend existing theory on self-disclosure to the online arena in higher educational institutions and contribute to the knowledge base and understanding about the use of a popular social networking site (SNS), Facebook, by college students. We conducted a non-experimental study to investigate how university students (N = 463) use Facebook, and examined the roles that personality and culture play in disclosure of information in online SNS-based environments. Results showed that individuals do disclose differently online vs. in-person, and that both culture and personality matter. Specifically, it was found that collectivistic individuals low on extraversion and interacting in an online environment disclosed the least honest and the most audience-relevant information, as compared to others. Exploratory analyses also indicate that students use sites such as Facebook primarily to maintain existing personal relationships and selectively used privacy settings to control their self-presentation on SNSs. The findings of this study offer insight into understanding college students’ self-disclosure on SNS, add to the literature on personality and self-disclosure, and shape future directions for research and practice on online self-presentation.

Introduction The large and burgeoning influence of social networking sites (SNSs) such as Facebook and Twitter has brought significant changes to information distribution and the cultural norms of social relationship. These technologies provide new means for individuals to present themselves, access and broadcast information, articulate their social networks, and establish and maintain connections with others. In higher educational institutions, on one side, the explosion of social media provides students with unprecedented learning and networking opportunities (Chen & Bryer, 2012). On the other side, educators expressed concerns about the openness over negative consequences of students’ self-disclosure on SNSs (Bryer & Chen, 2010). According to recent media reports, individuals’ posts on SNSs were used by potential employers in assessing their employment qualifications and individuals were penalized or criminally charged on the basis of their text or photo posts (Christofides et al., 2009 and Peluchette and Karl, 2008). Understanding the way individuals behave and disclose information on such sites is a potentially valuable source of information for practitioners and researchers. Our goal in this paper is to investigate how young adults use SNSs, and factors that might impact their self-presentation on those sites, such as general disclosiveness, cultural heritage and personality. We examine these factors using a non-experimental survey-based method. We hope to provide insights to help educators understand students’ online behaviors and potentially facilitate these behaviors to best utilize such networked learning opportunities. From a broader psychological perspective, we hope to shed light on the individual difference variables that determine how individuals disclose information about and present themselves in online SNS environments such as Facebook.

. Results All analyses were conducted using SPSS 20. SPSS Frequencies was used to analyze information regarding SNS usage habits. Tests of study hypotheses were conducted using SPSS GLM Repeated-Measures, with a mixed-subjects design, and utilizing mixed-model Multivariate Analysis of Covariance (mixed-model MANCOVA). Modality (online disclosure vs. in-person disclosure) was included as a categorical repeated-measures independent variable, extraversion and idiocentrism were included as continuous between-subjects independent variables, respondent race and sex were included as covariates, and the six measured aspects of self-disclosure (intent, amount, positivity, honesty, control, relevance) were included as dependent variables. One potential issue in our dataset was that respondent age varied widely, between 18 and 51. Although age has been found to predict frequency of Facebook use but not self-disclosure, friendship networks, or profile privacy protections (Hew, 2011), we nevertheless examined whether the large variability in respondent age may have affected results. To examine this possibility, we ran a Multivariate Analysis of Variance (MANOVA) with age as the predictor and all twelve self-disclosure variables as criteria. The multivariate main effect of age was not statistically significant (F12,454) = 1.313, p = .208, η2 = .034). Thus, this latter variable was not considered for further analyses. Alpha was set at α = .05 for tests of main effects and two-way interactions, and to α = .10 for tests of three-way interactions. As noted by Liakhovitski, Stone-Romero, and Jaccard (2008), increasing the alpha level is an acceptable strategy for increasing the power to detect the presence of higher-order interactions, because such interactions are smaller by nature, and thus require more statistical power to detect (c.f., Liakhovitski et al., 2008). Data were screened for multivariate outliers prior to hypotheses testing, in order to help prevent possible Type I and/or Type II errors (c.f., Fidell & Tabachnick, 2003). Four observations were discarded from analyses on grounds of being multivariate outliers, resulting in a final dataset of N = 463 respondents. In order to aid interpretability, plots of statistically significant interactions were created using dichotomized transformations of the continuous independent variables (individualism and extraversion), at one standard deviation below and above the mean, respectively. Thus, although hypotheses tests were run using the full range of these variables, only the observations on the lower (one SD below the mean) and higher ends (one SD above the mean) of individualism and/or extraversion are depicted. Means, standard deviations, and intercorrelations between all study variables are listed in Table 1. As shown in Table 1, the various in-person vs. online aspects of self-disclosure are more highly correlated with their mirror aspects (e.g., correlation between amount of self-disclosure in-person vs. amount of self-disclosure online: r = .40) than with other aspects (e.g., correlation between amount of self-disclosure in-person vs. honesty of self-disclosure online: r = .20); none of the mirror dimensions are so highly correlated as to be indistinguishable from each other (i.e., the highest such correlation is between honesty of self-disclosure in-person vs. honesty of self-disclosure online, at r = .59). Thus, the various aspects of self-disclosure are generally weakly to moderately correlated with each other, but generally highly correlated to their mirror aspects. This pattern of intercorrelations supports the notion that these aspects of self-disclosure are independent dimensions ( Wheeless, 1978, Wheeless and Grotz, 1976 and Wheeless and Grotz, 1977), and that in-person vs. online self-disclosure are different but related constructs. 6.1. SNS usage habits Exploratory analyses regarding SNS usage habits showed that almost 99% of the respondents reported that they used SNSs. The majority (n = 338) reported that they only used one or two SNSs and less than a quarter of the respondents (n = 109) used three or four SNSs. Few respondents used more than four SNSs and the maximum number of SNSs students used was seven. The most preferred favorite SNS was Facebook (n = 422). Twenty-seven respondents used Twitter, four used LinkedIn and three used MySpace. Most respondents (n = 446) maintained one single profile on their favorite SNS. Most participants (n = 313) were active on their most preferred SNS as they reported that they visited the site either very frequently or frequently. Most of our participants set up some privacy settings so that their full profile or photos were not released to the public; n = 413 respondents selectively restricted access to their full profile, n = 410 selectively restricted access to photos, and n = 399 selectively restricted access to their posts/updates on their wall. Fig. 1 indicates that the goal for using online social networks among our participants was mostly staying in touch with friends and classmates. Few respondents used the sites for making new friends or for professional or academic purposes. The importance ratings of motivational goals for using SNS usage. Fig. 1. The importance ratings of motivational goals for using SNS usage. Figure options 6.2. Hypotheses tests Hypothesis 1 tested the main effect of modality (in-person self-disclosure vs. online self-disclosure). Hypothesis 2 tested the 2-way interaction between modality and self-construal. Hypothesis 3 tested the 2-way interaction between modality and extraversion, and Hypothesis 4 tested the 3-way interaction between modality, self-construal, and extraversion. The multivariate repeated-measures main effect of modality was not statistically significant (F (6, 444) = 2.02, p > .05, η2 = .027), thereby yielding no support for Hypothesis 1. The multivariate mixed-subjects interaction between modality and idiocentrism was not statistically significant (F (6, 444) = 1.86, p > .05, η2 = .024), thereby yielding no support for Hypothesis 2. The multivariate mixed-subjects interaction between modality and extraversion was statistically significant (F (6, 444) = 2.33, p < .05, η2 = .03). Analysis of the univariate effects indicated a statistically significant interaction for relevance of self-disclosure (F (1, 449) = 4.416, p < .05, η2 = .01), and a marginally significant interaction for amount of self-disclosure (F (1, 449) = 2.925, p < .10, η2 = .006). Fig. 2 depicts this interaction for relevance of self-disclosure. As shown, individuals scoring one standard deviation below the mean on extraversion disclosed just as audience-relevant information in-person (M = 3.30; SD = .98) as individuals scoring one standard deviation above the mean on extraversion (M = 3.28; SD = .6). However, contrary to the hypothesized directions, low-extraversion individuals disclosed less audience-relevant information online (M = 3.18; SD = .96) than highly-extraverted individuals (M = 3.4; SD = .80). Fig. 3 depicts the interaction for amount of self-disclosure. As expected, the least amount of information shared was by low-extraversion individuals interacting online (M = 2.8; SD = 1.09). These results yield mixed support for Hypothesis 3. Interactive relation between modality and extraversion on relevance of ... Fig. 2. Interactive relation between modality and extraversion on relevance of self-disclosure. Figure options Interactive relation between modality and extraversion on amount of ... Fig. 3. Interactive relation between modality and extraversion on amount of self-disclosure. Figure options The above results are qualified, however, by the existence of a higher-order, 3-way multivariate interaction between modality, idiocentrism, and extraversion (F (6, 444) = 2.27, p < .05, η2 = .029). Analysis of the univariate effects indicated a statistically significant interaction for relevance of self-disclosure (F (1, 449) = 5.471, p < .05, η2 = .012), and a statistically significant interaction for honesty of self-disclosure (F (1, 449) = 2.881, p < .10, η2 = .006). Fig. 4a and Fig. 4b depict this 3-way interaction for relevance of self-disclosure. As expected, individuals low on both extraversion and idiocentrism (allocentrics) disclosed more audience-relevant information online (M = 3.44; SD = 1.16) than individuals low on idiocentrism but high on extraversion (M = 3.06; SD = .95); however, the former also disclosed the most audience-relevant information in-person (M = 3.56; SD = 1.16 for allocentric non-extraverts; M = 3.28; SD = .76 for allocentric extraverts). For idiocentrics, those low on extraversion disclosed less audience-relevant information online (M = 3; SD = .79) than those high on extraversion (M = 3.59; SD = .66); the same pattern was detected when in-person disclosure was considered, albeit to a lesser degree (M = 3.12; SD = .85 for idiocentric non-extraverts; M = 3.28; SD = .56 for idiocentric extraverts). Interactive relation between modality and extraversion for allocentrists on ... Fig. 4a. Interactive relation between modality and extraversion for allocentrists on relevance of self-disclosure. Figure options Interactive relation between modality and extraversion for idiocentrists on ... Fig. 4b. Interactive relation between modality and extraversion for idiocentrists on relevance of self-disclosure. Figure options Fig. 5a and Fig. 5b depict this 3-way interaction for honesty of self-disclosure. As expected, individuals low on both extraversion and idiocentrism (collectivists) disclosed the least honest information online (M = 4.07; SD = 1.1 for allocentric non-extraverts; M = 4.78; SD = .78 for allocentric extraverts); the former also disclosed less honest information in-person, albeit to a lesser degree (M = 4.37; SD = 1.2 for allocentric non-extraverts; M = 4.85; SD = .75 for allocentric extraverts). For idiocentrists, those low on extraversion also disclosed less honest information online (M = 4.46; SD = 1.06) than those high on extraversion (M = 4.67; SD = .66); the same pattern was detected when in-person disclosure was considered, albeit to a lesser degree (M = 4.51; SD = 1.07 for idiocentric non-extraverts; M = 4.83; SD = .77 for idiocentric extraverts). Interactive relation between modality and extraversion for allocentrists on ... Fig. 5a. Interactive relation between modality and extraversion for allocentrists on honesty of self-disclosure. Figure options Interactive relation between modality and extraversion for idiocentrists on ... Fig. 5b. Interactive relation between modality and extraversion for idiocentrists on honesty of self-disclosure. Figure options Thus, the least honest information disclosed was as hypothesized, by allocentric individuals low on extraversion, and interacting in an online environment. As expected also, allocentrists low on extraversion disclosed more audience relevant information online than allocentrists high on extraversion. Overall, the pattern of results across these two dependent measures generally supports the hypothesized direction for allocentric non-extraverts, thereby supporting Hypothesis 4.