SEARCH

SEARCH BY CITATION

Keywords:

  • Personal Journal Blogs;
  • Self-Disclosure;
  • Online Disinhibition;
  • Anonymity;
  • Content Analysis

Abstract

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

The connections between anonymity and self-disclosure online have received research attention, but the results have been inconclusive with regard to self-disclosure in blogs. This quantitative content analysis of 154 personal journal blogs tested some assumptions of the online disinhibition effect in order to examine the effect of types of anonymity on the amount, breadth, and depth of self-disclosure in blog entries. Results showed that participants disclosed more information in their blog entries when they were more visually identified (sharing a picture of themselves), contrary to the assumptions of the online disinhibition effect. Overall, a trend emerged where visual anonymity led to less disclosiveness, and discursive anonymity (sharing one's real name) led to less disclosiveness for particular types of bloggers.

It is well established that self-disclosure drives relationships (e.g., Altman & Taylor, 1973; Derlega, Metts, Petronio, & Margulis, 1993), and close relationships are characterized by disclosure marked by increasing breadth and depth (Altman & Taylor, 1973; Wheeless, 1978). However, as more relationships are being formed and maintained online, our traditional theories of self-disclosure in face-to-face (FtF) relationships must be tested in the electronic environment because we cannot assume that our conventional understandings of relationships apply equally in different contexts. The connection between identities and self-disclosure is one such area that deserves a closer look. Unlike in FtF relationships, where disclosure of personal information occurs in established relationships between interactants who know one another well, a different pattern has developed on the Internet.

Specifically, the anonymity afforded by the Internet has been found to result in decreasing inhibitions and increasing self-disclosures, a condition known as the online disinhibition effect (Suler, 2004). Bargh, McKenna, and Fitzsimons (2002) argued that online anonymity can produce results similar to the “stranger on a train” phenomenon, wherein people share intimate self-disclosures with strangers they may never see again (Rubin, 1975). Under conditions of anonymity, people self-disclose more intimate information because there are few if any related risks and constraints (Bargh et al., 2002). The prevailing research in computer-mediated communication (CMC) supports this positive relationship between anonymity and self-disclosure (e.g., Bailenson, Yee, Merget, & Schroeder, 2006; Chiou, 2006, 2007; Joinson, 2001; McKenna & Bargh, 2000). However, these predictable effects of anonymity on self-disclosure may not hold in the blogosphere, where people write about their inner selves through personal journal blogs (Blood, 2002). Unlike traditional journals, which are often kept hidden under lock and key, bloggers often post their innermost thoughts for any Internet user to read (Bortree, 2005; Viegas, 2005). Frye and Dornisch (2010) found that Internet users considered public blogs to be the least private medium when compared to other forms of Internet communication. Additionally, research suggests that blogs are often used to maintain relationships (Hollenbaugh, 2011; Stefanone & Jang, 2007). These highly personalized postings, displayed in a public setting, are often accompanied by identifying information, such as one's name, age, location, place of employment, and pictures (see Qian & Scott, 2007; Viegas, 2005). These findings challenge existing theories, such as the online disinhibition effect (Suler, 2004), and suggest that research findings must be tested in the blogging context to determine their applicability.

In recent years, mainstream media have profiled stories of bloggers who suffer the professional and social consequences of self-disclosure in such a public forum. Bloggers who are caught revealing personal, potentially damaging information on their blogs are likely to be those who also disclose their true identities, sacrificing the anonymity characteristic of CMC. If these bloggers were not identifiable, they would not be able to be held accountable for their blog content. Indeed, 87 of 207 bloggers surveyed expressed concern about the potential ramifications of self-disclosure in their blogs (Qian & Scott, 2007). However, of those bloggers who were concerned, only 42% actually self-censored themselves when posting content in their blog entries. Although the concern for privacy seems to be there, bloggers do not appear to be taking actions to protect their private self-disclosures. Therefore, the relationship between anonymity and disclosiveness in blogs appears tenuous and merits further investigation.

Prevailing literature on self-disclosure in blogs has been problematic for two reasons. First, many studies measure self-disclosure by examining the identifying characteristics revealed in profiles, such as name, age, and interests (e.g., Boyle & Johnson, 2010; Mesch & Beker, 2010; Nosko, Wood, & Molema, 2010). Although this method of measurement is convenient, it represents only a small part of disclosure when compared to the amount of information shared in the content of blog entries. Second, other studies have relied on self-report assessments to measure self-disclosure (e.g., Frye & Dornisch, 2010; Mesch & Beker, 2010; Qian & Scott, 2007; Viegas, 2005). These more global measures may or may not reflect the depth and breadth of actual disclosure. This study attempts to compensate for some of these problems by examining both profile and blog entry information through quantitative content analysis.

The purpose of this study is to examine the applicability of existing research findings in the blogosphere by testing the effects of discursive and visual anonymity on bloggers' amount, breadth, and depth of self-disclosure, in accordance with the online disinhibition effect. Toward this end, literature is reviewed that discussed the unique characteristics of blogs, the prevalence of self-disclosure in blogs, and the connection between anonymity and online self-disclosure. Control variables are introduced to increase accuracy of the study's findings. Following this literature review, a content analysis is presented that tests the impacts of anonymity on self-disclosure in blog entries.

Literature Review

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

Blogs Described

Blogs are websites that are frequently updated and organized in reverse chronological order (Herring, 2004). The majority of blogs maintained on the Web are personal journal blogs (Herring, Scheidt, Wright, & Bonus, 2005; Viegas, 2005), or blogs composed of short posts concerning the blogger's life and internal self (Blood, 2002). When compared to social network sites such as Facebook, blogs provide a more isolated space with less interactivity that may reduce users' awareness of their audiences. Thus, blog posts may be more disclosive than those on social network sites. Personal journal blogs are more likely to be characterized by self-disclosure than filter blogs, which are devoted to external content like political news (Herring & Paolillo, 2006). Higher amount of disclosure in personal journal blogs is likely due to bloggers' motivations for maintaining these blogs. People are motivated to maintain personal journal blogs to archive and organize their thoughts, to help others, for social connection, to get feedback (Hollenbaugh, 2011), to express creativity, and to entertain others (Lenhart & Fox, 2006).

Self-Disclosure in Blogs

Self-disclosure, defined as the revelation of personal information (Derlega et al., 1993), is a multifaceted concept that includes dimensions such as amount, breadth, and depth (Altman & Taylor, 1973; Wheeless, 1978). Amount of self-disclosure is the volume of self-disclosure in blog entries, whereas breadth is the variety of topics disclosed and depth refers to the level of intimacy of information disclosed (Altman & Taylor, 1973).

Given the freedom to construct blogs around anything they wish, bloggers more often than not simply talk about themselves. People often describe their daily lives in their blogs (Bortree, 2005; Trammell & Keshelashvili, 2005). However, researchers have also confirmed the existence of large amounts of intimate disclosure on blogs. For example, in a study by Viegas (2005), one-fourth of the respondents reported they posted very personal information on their blogs fairly often, whereas less than 20% claimed they never posted intimate information on their blogs. The teenage girls participating in Bortree's (2005) ethnography disclosed very personal information about their despairs and frustrations, as well as other feelings. These respondents said they would not openly disclose these things FtF (Bortree, 2005). It is obvious that personal experiences dominate the content of most blogs. What is less clear is how bloggers' level of anonymity affects their postings. Those potential effects are explored below.

Anonymity and Online Self-Disclosure

Attributes of the Internet provide more opportunities for anonymity, which may allow for more intimate disclosure than is typical in FtF communication. Suler (2004) hypothesized that there is an online disinhibition effect that explains this phenomenon. Factors impacting disinhibition include dissociative anonymity, invisibility, asynchronicity, the creation of a conversational partner in one's mind, the creation of an online reality, and the equalization of status (Suler, 2004). In reference to anonymity specifically, Suler (2004) posited that “When people have the opportunity to separate their actions online from their in-person lifestyle and identity, they feel less vulnerable about self-disclosing and acting out” (p. 322). When anonymous, people are not held accountable in the offline world for their online communication. Therefore, they are able to disclose more information, which is often more intimate than they may disclose FtF (Suler, 2004).

Anonymity is defined as “the degree to which a communicator perceives the message source is unknown and unspecified” (Anonymous, 1998, p. 387). However, anonymity is not an either/or construct; it exists on a continuum (Anonymous, 1998). Two types of anonymity are especially relevant to CMC research. Discursive anonymity (what Suler [2004] labeled dissociative anonymity) occurs when a message cannot be linked to a specific source (Scott, 2004). Bloggers manage their degree of discursive anonymity through the use of identity management cues, or information identified on bloggers' profiles that disclose a true offline identity (Viegas, 2005). For example, a person's name, age, location, schools attended, and employer would be considered identity management cues. Disclosing these identity management cues reduces the amount of discursive anonymity bloggers maintain. Visual anonymity (what Suler [2004] calls invisibility) refers to the extent to which people can see and/or hear the blogger (Scott, 2004). Therefore, any pictures or videos of the blogger can disclose information about his/her offline identity.

The online disinhibition effect establishes a link between anonymity and online self-disclosure, such that under conditions of anonymity, online communicators will be more likely to disclose more information that is more personal than one would expect in FtF contexts. This link has been tested in several studies of online communication with varying results.

For example, a recent study explored the effects of visual and discursive anonymity on self-disclosure in blogs (Qian & Scott, 2007). In an online survey of over 200 bloggers, Qian and Scott (2007) found no significant effects of anonymity on self-disclosure in blog entries. However, when they excluded those bloggers who revealed more identity management cues than just their names (18.8% of their sample), they discovered predictable effects of discursive anonymity. Specifically, bloggers revealed more private information under conditions of partial or full discursive anonymity (Qian & Scott, 2007). These findings are not conclusive for bloggers in general, namely because almost one-fifth of their sample did not adhere to this self-disclosure norm. Additionally, Qian and Scott (2007) relied on self-report data, which may provide less valid conclusions than experimental or content analysis methods (Babbie, 2004; Kerlinger & Lee, 2000). The impact of anonymity on self-disclosure in blogs is still questionable and deserves a more detailed look.

CMC research has largely supported the predictions and explanations mentioned above (e.g., Bailenson et al., 2006; Chiou, 2006, 2007; Joinson, 2001). Although much of this research tested the effects of visual anonymity, it is assumed that the effects of visual anonymity will also explain the impact of other cues contributing to anonymity, namely identity management cues such as name. Given the tenets of the online disinhibition effect and existing research on anonymity and disclosure, it should be expected that when bloggers disclose more identity management cues and share pictures of themselves, they will disclose less intimate information covering less topics, or aspects of their lives. However, this link between anonymity and self-disclosure has not been studied in blog entries specifically, and it is questionable whether such a relationship exists in this context. Therefore, several research questions are posed to explore the effects of anonymity on self-disclosure in blog entries:

RQ1: How does visual anonymity affect the amount, breadth, and depth of self-disclosure in blog entries?

RQ2: How does discursive anonymity affect the amount, breadth, and depth of self-disclosure in blog entries?

RQ3: Are there interaction effects of visual and discursive anonymity on amount, breadth, and depth of self-disclosure in blog entries?

Control Variables

Given the existing research examining the impact of other variables on self-disclosure both offline and online, sex and age were added as control variables in this study. Age is one individual characteristic that likely impacts disclosure dimensions in blogs. In traditional FtF literature, younger people are often identified as more disclosive than older people (e.g., Knapp, Ellis, & Williams, 1980; Sinha, 1972). Some research in CMC also shows a trend whereby younger people are more disclosive online than older people (e.g., Lenhart & Fox, 2006; Ma & Leung, 2006; Valkenburg & Peter, 2007). One would expect this trend to hold true in the blogosphere, as well.

Gender is studied with regard to disclosure more often than any other individual characteristic (Dindia, 2002). Sex differences in disclosure have emerged in both FtF communication and CMC. Studies have consistently shown that in FtF communication, women tend to disclose more information (e.g., Buhrke & Fuqua, 1987; Dolgin & Minowa, 1997; Jourard, 1971) that is more intimate (e.g., Davidson & Duberman, 1982; Morton, 1978) than men. Although certain variables, such as the target's sex (e.g., Hacker, 1981) and the method of measurement (e.g., Dindia & Allen, 1992) may moderate sex differences, these differences persist. This trend appears to apply in CMC, too (e.g., Nosko et al., 2010; Peter, Valkenburg, & Schouten, 2005; Punyanunt-Carter, 2006).

However, when considering blogs specifically, findings are mixed. Child (2007) explored the privacy rules and boundaries that adolescents enact as they blog and found that females and males were equally likely to disclose information in their blogs (Child, 2007). On the other hand, females were significantly more concerned with who would co-own their disclosures when blogging than males (Child, 2007). Although there is no significant difference between men and women in disclosing identity management cues (Huffaker & Calvert, 2005), teenage girls in Bortree's (2005) preliminary analyses did seem to disclose more intimate information in their blogs than teenage boys. Additionally, Lenhart and Fox (2006) found that females were more likely than males to be inspired to post to their blogs by a personal experience, which may suggest that they would disclose more in these posts. More research is needed to tease out sex differences in blogging; however, it is does seem that sex does play a role in disclosure dimensions in blogs.

Considering prior findings on the impact of age and sex, the following hypothesis was posed:

H: Control variables (age and sex) will affect amount, breadth, and depth of self-disclosure in blog entries.

Methods

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

A quantitative content analysis of personal journal blog profiles and entries was employed to test the effects of visual and discursive anonymity on amount, breadth, and depth of self-disclosure in blog entries. Previous research (e.g., Altman & Taylor, 1973; Harper & Harper, 2006; Jourard, 1971; Jourard & Lasakow, 1958; Tidwell & Walther, 2002) was consulted to assemble the coding structure. Although prior research utilizing self-reports has been very informative, content analysis is the most valid way to examine mediated messages, which are archived and publicly available (Riffe, Lacy, & Fico, 2005). The participants, procedures, and data analyses are described in the following sections.

Participants

Participants were English-speaking bloggers who responded to a call for research participation. The bloggers in this study maintained public blogs accessible to anyone. To participate, bloggers must have been at least 18 years old, they must have maintained personal journal blogs (which were defined in the call for participation as blogs that are composed of short posts concerning the blogger's life and internal self), and bloggers must have posted to their blogs at least once a month. Following data collection, any bloggers who reported they blogged less than once per month were removed from further analyses.

Participants were recruited for this study in multiple ways. Although sampling was not random, a number of recruitment strategies were used to try to reach the widest, most diverse audience possible given certain practical constraints. Featured bloggers on Blogger.com and Xanga.com were contacted and asked to post a call for research participation containing information about the study on their blogs. Ostensibly, these are the blogs most likely to have high traffic; therefore, it was expected that posting a call for participants would be most effective on these popular blogs. Additionally, bloggers' participation was elicited through e-mails or comments on their blogs found through the random feature on Livejournal.com and Blogger.com, through the most recently updated blogs on Wordpress.com, or through these bloggers' lists of frequented blogs. Finally, announcements were posted on various discussion boards and listservs that bloggers may read. These recruitment strategies are similar to those used in other blogging studies (e.g., Kaye, 2005; Qian & Scott, 2007; Viegas, 2005). In total, this study was announced to 297 individual bloggers, eight discussion boards, and six listservs.

A total of 243 public bloggers completed the survey. However, after excluding participants who provided an invalid blog URL, 157 participants remained. Three additional participants were excluded because upon closer examination, their blog entries were not able to be coded. Therefore, the final sample consisted of 154 bloggers. The sample was primarily composed of women (n = 113, 73.4%; male n = 41, 26.6%), which is characteristic of the larger population of personal journal bloggers (see Herring et al., 2005). Participants' ages ranged from 18 to 70 years (M = 32.22, SD = 12.19). These bloggers maintained anywhere from one to seven blogs (M = 1.77, SD = 1.09), posting an average of 16.61 times in a given month (SD = 13.74).

The most represented ethnicity in the sample was Caucasian (n = 124, 80.5%), with other represented ethnicities including Asian (n = 19, 12.3%), Hispanic (n = 5, 3.2%), African American (n = 4, 2.6%), Middle Eastern (n = 3, 1.9%), Native American/Alaska Native (n = 3, 1.9%), Pacific Islander (n = 2, 1.3%), and African (n = 1, .6%). Most participants lived in the United States at the time of data collection (n = 130, 84.4%), but other bloggers resided in Canada (n = 7, 4.5%), the United Kingdom (n = 7, 4.5%), and Germany (n = 3, 1.9%). Denmark, France, the Netherlands, Singapore, South Korea, and Switzerland were each represented by one participant (.6%) in the sample.

Procedures

Upon approval by the Human Subjects Review Board, an online survey was constructed for this study using SurveyMonkey.com. After providing informed consent, participants supplied demographic and descriptive information, including their age, sex, ethnicity, country of residence, frequency of blog posting, and number of blogs maintained. (Additional self-report data gathered in the survey was used for analyses elsewhere.) Participants were also asked to enter the web address for their blogs.

Researchers visited each of the web addresses to retrieve the bloggers' profiles and archived blog posts. To ensure that the survey did not impact the collected data, the five most recent blog posts prior to the date the survey was distributed were collected. In a few cases there was more than one entry per day. Subsequently, the five most recent posts in those cases were not representative of five different days. Any text, pictures, hyperlinks, or videos posted in the bloggers' profiles and five blog entries were copied into a Microsoft Word document and saved for further analysis. Blogs with less than five entries prior to the survey distribution date or with invalid web addresses were excluded from the study. Each blogger was given an identification number, which allowed the researchers to link the survey data with the content analysis data.

For the bloggers' profiles, the profile (n = 154) was the unit of analysis. For the blog entries, each entry (n = 770) was the unit of analysis. Profile information was coded to yield the visual and discursive anonymity levels of each blogger, whereas blog entries were coded for amount, breadth, and depth of self-disclosure.

Coders recorded the presence/absence of 34 categories in the bloggers' profiles. (See Appendices A and B for the codebook and coding sheet.) Categories of interest in this study included first name, last name, nickname/pseudonym, picture of blogger, picture of more than one person, and other pictures. Coders then examined the self-disclosures in the five blog entries preceding the distribution of the survey from each blog. Self-disclosure was defined as personal information revealed about one's self, inner states, feelings, or opinions. Each self-disclosure was coded for its topic and level of intimacy. Topics, taken from Taylor and Altman (1966) and Tidwell and Walther's (2002) work, included biographic information, sex, school and work, current events, physical appearance and condition, hobbies and interests, money and property, and relationships with others. Level of intimacy (i.e., depth of self-disclosure) was coded as a 1 (peripheral information), 2 (intermediate information), or 3 (core information), in accordance with Altman and Taylor's (1973) explanation of these three levels.

Peripheral information included general information about the topic as it related to the self. This information was superficial disclosure that is largely descriptive or explanatory. Intermediate information included evaluative statements, such as those pertaining to one's attitudes, beliefs, and opinions about specific issues. Core information represented the most private and risky information, such as one's personal needs, fears, inadequacies, and other undesirable characteristics. Additionally, beliefs about one's self-identity that were related to these more private areas were included in this category.

Topics and level of intimacy were identified for each coded instance of self-disclosure. Altman and Taylor (1973) suggested controlling for breadth frequency, or the number of disclosures within a particular breadth category, when the primary research interest is studying breadth and depth of self-disclosure. This practice yields a more accurate mean depth score (Altman & Taylor, 1973). Therefore, each topic/intimacy combination was only recorded once per blog entry (the unit of analysis). For example, once a level-1 disclosure about school and work was made, it was not coded again in that blog entry unless school and work appeared at another level of intimacy in order to control for the bloggers' breadth frequency.

Two coders were assigned to the blogger profiles, and two coders analyzed the blog entries. The researchers conducted several training sessions, where differences among coders were resolved and the codebook was refined. Intercoder reliability was continually assessed throughout the process until acceptable levels were met for all categories on 20% of the data (n = 32 blog profiles, n = 160 blog entries). Categorical variables (e.g., profile information and categories of self-disclosure) were assessed using Cohen's kappa, and reliabilities ranged from .81 to 1.00. According to Neuendorf (2002), kappas greater than or equal to .75 are considered excellent reliability levels. Continuous variables (e.g., depth of information) were also found to be reliable, as Krippendorf's alpha ranged from .79 to .86. (See Appendix Appendix for reliabilities for all coded variables.) Neuendorf (2002) reported that alphas must be greater than or equal to .80 to be considered acceptable. The second level of depth was on the verge of being acceptable at .79, and thus was included in this study. For the 20% of the data that was coded jointly, disagreements were resolved between coders and the final decisions were recorded in the data set. The remaining data (n = 125 blogger profiles, n = 625 blog entries) were split amongst the coders and coded independently.

Computing study variables

Visual anonymity was assessed by analyzing the coded information in bloggers' profiles and assigning them to one of two categories: picture of people (n = 84, 54.5%; one or more people) and no picture of people (n = 70, 45.5%; picture of things or no picture). Discursive anonymity was also assessed by assigning bloggers to categories based on information in their profiles. Two categories were created for discursive anonymity: real names (n = 84, 54.5%; first, last, or first and last names) and no real names (n = 70, 45.5%; pseudonym or no name provided). These methods are similar to research conducted by Qian and Scott (2007).

Amount, breadth, and depth of self-disclosure were measured by analyzing the blog entries (five per blogger). Altman and Taylor's (1973) work was consulted in determining how to calculate scores for self-disclosure dimensions. Amount of self-disclosure was computed by summing the total coded instances for each blogger. Because there were three levels in each of the eight categories of self-disclosure, the maximum number of disclosures that could be coded was 120 when all five blog entries were combined. The actual scores ranged from 1 to 44. (See Appendix Appendix for the means and standard deviations of all continuous variables in the study).

Each blogger was also assigned a score for breadth of self-disclosure by computing the number of topics out of the possible eight that were disclosed across their five blog entries. Breadth scores ranged from 1 to 8. Self-disclosure within a topic can vary in degree of intimacy. Therefore, depth of self-disclosure was computed for each of the eight categories by averaging the scores, with scores ranging from 1 (little intimacy) to 3 (more intimacy).

Results

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

Anonymity and Self-Disclosure

Hierarchical multiple regressions were used to test Research Questions 1 and 2, as well as the hypothesis. This statistical test assumes that cases are independent, the data are normally distributed, the distribution of predicted scores are normal across any given predictive score, and the relationships are linear in nature (Spicer, 2005). Evaluation of multicollinearity diagnostics (VIF and tolerance scores) showed that predictor variables were not overly correlated in the regressions. For each of the analyses, age and sex were entered on the first block of predictors as control variables, whereas visual and discursive anonymity (each dummy coded 0 and 1) were entered on the second step. Amount, breadth, and seven of the eight categories of depth served as the dependent variable in each of nine regression analyses. Because only nine respondents (5.8%) disclosed about sex, the depth of that category was not tested due to insufficient power. Results of these analyses are presented in Appendix Appendix.

Research Question 1 was posed to test the effects of visual anonymity on amount, breadth, and depth of self-disclosure in participants' blog entries. Examination of the standardized betas showed that visual anonymity significantly contributed to the amount of self-disclosure, above and beyond the effects of age, sex, and discursive anonymity (see Appendix Appendix for all standardized betas). Participants who were more visually identified, including pictures of people in their profiles, disclosed a larger amount of private information in their blog entries. Visual anonymity did not have a significant impact on any other dependent variable.

Research Question 2 addressed the potential effects of discursive anonymity on amount, breadth, and depth of self-disclosure in blog entries. As evident in Table 3 (see Appendix Appendix), discursive anonymity did not predict any of the dependent variables beyond age, sex, and visual anonymity. Whether bloggers used their real names or not did not impact the amount, breadth, or depth of their self-disclosures in their blog entries.

Separate factorial design analyses of variance (ANOVAs) were conducted to address Research Question 3, which asked whether there were interaction effects between visual and discursive anonymity on the amount, breadth, and depth of self-disclosure. The 2 (visual anonymity) X 2 (discursive anonymity) factorial ANOVA showed a significant interaction effect for depth of self-disclosure about current events, F(1, 54) = 4.37, p < .05. Specifically, bloggers with less discursive anonymity (using their real names) who maintained more visual anonymity (using pictures without people) disclosed about current events with more depth than others. No other interaction effects were found in the data.

Impact of Control Variables

The hypothesis predicted that age and sex would affect amount, breadth, and depth of self-disclosure in blog entries. These control variables, entered on the first step of the regression analyses, together explained 6% (p < .01) of the amount of self-disclosure and 12.7% (p < .001) of the breadth of self-disclosure in blog entries. When using the conservative Bonferroni method to control for the increased risk of Type I error (α = .05/9 = .006), the models were still significant for amount and breadth of self-disclosure. However, this model was not a significant predictor of depth of self-disclosure using the adjusted alpha level (adjusted R2 = .04, p = .03). Examination of the standardized betas in the regression analyses also provided support for this hypothesis (see Appendix Appendix). In particular, age was a significant predictor of amount and breadth of self-disclosure. In other words, younger participants were more likely to disclose a larger amount of information on a variety of topics than older participants.

Sex also predicted amount and breadth of self-disclosure, as well as depth of self-disclosure about biographic topics. Women were more likely to disclose a larger amount of information about a variety of topics than men. Additionally, they disclosed more intimately about biographic information than men. However, the effects of sex on depth of disclosure about biographic information should be interpreted with reservation, as the total regression model was not significant for that outcome variable.

Additional Analyses

Considering the exploratory nature of this study and the strong impact of age and sex on self-disclosure, a series of factorial design ANOVAs were conducted on subgroups of the sample to further investigate the data. It is possible that the relationships between anonymity and self-disclosure differ for men and women or for different age groups. Additionally, because this study is one of the few to explore anonymity and self-disclosure in blogs, it seemed appropriate to partition out subgroups of the participants and more closely examine anonymity's effects on self-disclosure dimensions. There were not enough participants in each group for depth of self-disclosure in the seven breadth categories to conduct ANOVAs. Therefore, only amount and breadth of self-disclosure were examined in the post hoc analyses.

The assumptions of ANOVA are normality of scores in the population, homogeneity of variances, and independent samples (Maxwell & Delaney, 2004). However, Maxwell and Delaney (2004) explain that ANOVA's F-test is robust against violations of assumptions. Although the samples were small for these post hoc tests, factorial design ANOVA is still the appropriate statistic to run. The tests were not sufficiently powered, increasing the changes for Type II error, but because significant differences did emerge, these differences are addressed in the following sections.

Age

Participants were placed into three groups according to age: young (18–24 years; n = 51, 33.1%), middle (25–35 years; n = 50, 32.5%), and older (36–70 years; n = 51, 33.1%). A series of 2 (visual anonymity) × 2 (discursive anonymity) factorial design ANOVAs were conducted on these groups separately. For each age group, two ANOVAs were run, one with amount of self-disclosure as the dependent variable and another with breadth of self-disclosure as the dependent variable. For younger participants, there were significant main effects of visual anonymity [F(1, 47) = 4.52, p < .05] and discursive anonymity [F(1, 47) = 5.79, p < .05] on breadth of self-disclosure in blog entries. Because both measures of anonymity were dichotomous, the mean scores were evaluated to interpret significant main effects. Younger participants who displayed pictures of people (M = 5.20, SD = 1.30) and used fake or no names (M = 5.38, SD = 1.40) had more breadth in their blog entries than those who used no picture or pictures of things (M = 4.76, SD = 1.41) and their real names (M = 4.77, SD = 1.28). In other words, younger participants had the most breadth in their self-disclosures when they were visually identified but discursively anonymous. However, these findings should be interpreted with care because when using the Bonferroni method (α = .05/2 = .025), there was not a significant difference between those visually anonymous and those visually identified (p = .04).

The factorial design ANOVA showed a significant main effect of discursive anonymity on amount of self-disclosure for participants in the middle category (ages 25–35 years), F(1, 46) = 7.47, p < .01. This finding remained significant when correcting for increased risk of Type I error (α = .05/2 = .025). Specifically, these participants disclosed more when using fake or no names (M = 15.30, SD = 7.33) than when using their real names (M = 10.26, SD = 4.65). Essentially, participants ages 25 to 35 were most disclosive under conditions of discursive anonymity. There were no significant effects of visual or discursive anonymity among the older category of participants (ages 36–70 years).

Sex

A series of 2 (visual anonymity) × 2 (discursive anonymity) factorial design ANOVAs were conducted on men (n = 41, 26.6%) and women (n = 113, 73.4%) separately to test the simple main effects of visual and discursive anonymity on amount and breadth of self-disclosure. For men, there was a significant main effect of visual anonymity on amount of self-disclosure, F(1, 37) = 7.99, p < .01. Men self-disclosed more when they were visually identified, using pictures of people in their blog profiles (M = 14.70, SD = 6.11) rather than no picture or pictures of things (M = 9.56, SD = 5.91). The factorial design ANOVAs showed different effects for women than for men. There was a simple main effect of discursive anonymity on amount [F(1, 109) = 4.84, p < .05] and breadth [F(1, 109) = 5.19, p < .05] of self-disclosure. Women who used fake or no names in their blog profiles disclosed a larger amount of information (M = 17.26, SD = 6.99) on a wider variety of topics (M = 4.90, SD = 1.49) than women who used their real names (M = 14.44, SD = 7.55; M = 4.35, SD = 1.39, respectively). In this way, being discursively anonymous led to more amount and breadth of self-disclosure for women. However, once again the reader should be cautious in interpreting these findings, as when using the Bonferroni method of correction (α = .05/2 = .025), only the difference in discursive anonymity and breadth of self-disclosure was significant.

Discussion

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

The purpose of this study was to test the assumptions of the online disinhibition effect in the blogging context by employing a quantitative content analysis. Visual anonymity was a significant predictor of a smaller amount of self-disclosure, above and beyond age, sex, and discursive anonymity. Participants disclosed the most amount of information when they were more visually identified through pictures of people in their blog profiles. Analyses exposed an interaction effect between visual and discursive anonymity on depth of self-disclosure about current events. Specifically, less discursive anonymity, along with more visual anonymity resulted in more intimate self-disclosure about current events.

The control variables – age and sex – were influential in the analyses, such that younger participants and women disclosed more amount and breadth than older participants and men. Women were also more intimate when disclosing biographic information in their blog entries. This warranted post hoc analyses, which showed several significant relationships when analyzing age and sex groups separately.

Implications of the Current Study

In general, a trend emerged in this study. Visual anonymity tended to have an inverse effect on self-disclosure dimensions. In other words, participants who were more visually identified were likely to be more self-disclosive. Across the board, participants who used pictures of people in their profiles disclosed a larger amount of information. Young participants, ages 18 to 24 years, who used pictures of people disclosed with more breadth, and men who were visually identified in their profiles disclosed a larger amount of information. Post hoc analyses showed some evidence of an opposite effect for discursive anonymity. Young participants and women disclosed more breadth of topics when they were more discursively anonymous, and participants in the middle age category and women had a greater amount of self-disclosure while discursively anonymous. This trend in the data suggests that perhaps visual cues were perceived by bloggers as less identifying than discursive cues, such as one's real name.

Results of this study partially support Suler's (2004) online disinhibition effect. Namely, bloggers were more disclosive under conditions of discursive, or dissociative, anonymity. These findings corroborate Qian and Scott's (2007) study findings. On the other hand, visual anonymity (invisibility) had the opposite effect in this study. Bloggers who were more visually anonymous were actually less disclosive. Perhaps bloggers see pictures of themselves and others as simply another means of disclosing information like that disclosed in blog entries, rather than as cues that identify someone's offline identity. In other words, visual cues may go beyond simply identifying someone to instead constituting an important component of self-disclosure overall. The tenets of the online disinhibition effect should be further scrutinized to determine whether visual anonymity is really a factor leading to online disinhibition, or if it is instead an outcome of this condition.

Not surprisingly, the control variables had predictable effects on self-disclosure in blog entries. In this study, age was inversely related to amount and breadth of self-disclosure. Additionally, women disclosed with more amount and breadth, as well as more depth for biographic information, than men. Existing research supports these findings. This study adds to the growing body of research that illustrates the importance of controlling for such demographic characteristics. Additionally, the rich findings that resulted from the post hoc analyses suggest that researchers should look into differences among groups of people with more depth. Perhaps our existing theories only hold true for certain types of people. Finding nonsignificant results when collapsing these categories might confound the very significant effects for particular demographic groups.

This study improved upon existing research by examining information in both blog profiles and blog entries through quantitative content analysis, rather than relying on self-report data or focusing primarily on blog profiles. Additionally, self-disclosure dimensions were operationalized more comprehensively than in previous research. However, some limitations do exist.

Limitations and Future Research Directions

The implications of this study should be understood with some reservation due to the study's limitations. First, the computation of depth of self-disclosure may have been flawed, as only one significant relationship emerged from this variable. Depth should theoretically vary along with breadth and amount of self-disclosure; therefore, it seems odd that this variable was not predicted by the independent variables. The measure of depth, which was computed for each category of self-disclosure, points toward a need to examine the relationships among amount, breadth, and depth of self-disclosure with more care. Different categories of information contain an inherent level of depth (Altman & Taylor, 1973), which led to the decision to score depth for each category rather than creating a combined score. For example, disclosing information about sex is more personal by virtue of the topic than disclosing about current events. The nuances of these data should be evaluated further to better discover the relationships between amount, breadth, and depth of self-disclosure.

Second, visual anonymity was measured through pictures available in blog profiles. However, most bloggers (almost 88% in this sample) post at least some pictures of people in their blog entries. Although the researchers decided to follow conventions and focus on the profile for information on visual anonymity because it is a more stable feature of the blog, it is possible that participants' conceptualizations of their visual anonymity were not dependent just on their profiles but on their blogs as a whole, including pictures posted in entries. This complicates the measurement of visual anonymity, but also lends credence to the idea that pictures may be seen as another type of disclosure rather than an identity management cue. Due to the growing simplicity and popularity of sharing pictures in blogs, this will no doubt be a rich area for future research.

The third limitation concerns the generalizability of the findings given the data collection methods. The sampling was convenient, so participants may not have represented the population as a whole. Also, there is the possibility that significant life events, such as the loss of a loved one, or even shared social events, such as Valentine's Day, that occurred during the data collection period may have skewed the data. Therefore, future researchers may consider sampling blogs entries from various time frames to help address this potential effect. Additionally, the small sample sizes used in the post hoc factorial design ANOVAs resulted in insufficiently powered tests. Although the findings from these analyses emerged despite the low power, they must be interpreted carefully given the small samples.

The findings of this study suggest a number of areas for future research. As facial recognition technologies become more integrated into everyday use, the relationship between visual anonymity and self-disclosure should be further tested. In December of 2010, Facebook began using facial recognition software, much like that used by photo software Picasa and iPhoto, to suggest people to “tag” in photos uploaded on Facebook (Isaac, 2010). Now when users upload photos, Facebook suggests who should be tagged in these photos based on facial features. It is possible that the way people think about pictures and their offline identities will undergo change as the technology continues to advance.

image

Figure 1. Interaction effect of visual and discursive anonymity on depth of self-disclosure about current events.

Download figure to PowerPoint

Researchers should also consider expanding the model tested in the present study to include people's predispositions. Suler (2004) explained that individual differences, such as one's needs or personality traits, also likely play a role in how disclosive they are likely to be online. More research is needed to tease out the relationships among individual differences, anonymity, and other characteristics of the Internet that may impact online self-disclosure.

Self-disclosure is a multifaceted construct, and this study points to the need for continued efforts to develop measures that accurately reflect the complexities of breadth and depth that occur in observed self-disclosure. Especially given the rapid changes in technology and the decreasing anonymity of the various Internet formats, this continues to be an important area of study in order to understand how existing models of relationship development and intimacy through self-disclosure may also be changing. The use of quantitative content analysis offers a meaningful way in which to explore these complexities.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

The authors wish to thank Dr. Amber Ferris, Frances LeHoty, Alyssa Pearson, and Carly Engrish-Desphy for their help in this project.

References

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies
  • Altman, I., & Taylor, D. A. (1973). Social penetration: The development of interpersonal relationships. New York: Holt, Rinehart & Winston.
  • Anonymous. (1998). To reveal or not to reveal: A theoretical model of anonymous communication. Communication Theory, 4, 381407.
  • Babbie, E. (2004). The practice of social research (10th ed.). Belmont, CA: Wadsworth.
  • Bailenson, J. N., Yee, N., Merget, D., & Schroeder, R. (2006). The effect of behavioral realism and form realism of real-time avatar faces on verbal disclosure, nonverbal disclosure, emotion recognition, and copresence in dyadic interaction. Presence, 15, 359372.
  • Bargh, J. A., McKenna, K. Y. A., & Fitzsimons, G. M. (2002). Can you see the real me? Activation and expression of the “true self” on the Internet. Journal of Social Issues, 58, 3348.
  • Blood, R. (2002). The weblog handbook: Practical advice on creating and maintaining your blog. Cambridge, MA: Perseus.
  • Bortree, D. S. (2005). Presentation of self on the web: An ethnographic study of teenage girls' weblogs. Education, Communication & Information, 5, 2539.
  • Boyle, K., & Johnson, T. J. (2010). MySpace is your space? Examining self-presentation of MySpace users. Computers in Human Behavior, 26, 13921399. doi:10.1016/j.chb.2010.04.015
  • Buhrke, R., & Fuqua, D. (1987). Sex differences in same- and cross-sex supportive relationships. Sex Roles, 17, 339352.
  • Child, J. T. (2007). The development and test of a measure of young adult blogging behaviors, communication, and privacy management. Unpublished doctoral dissertation. North Dakota State University of Agriculture and Applied Sciences, Fargo, ND.
  • Chiou, W.-B. (2006). Adolescents' sexual self-disclosure on the Internet: Deindividuation and impression management. Adolescence, 41, 547561.
  • Chiou, W.-B. (2007). Adolescents' reply intent for sexual disclosure in cyberspace: Gender differences and effects of anonymity and topic intimacy. CyberPsychology & Behavior, 10, 725728. doi:10.1089/cpb.2007.9961
  • Davidson, L., & Duberman, L. (1982). Friendship: Communication and interactional patterns in same-sex dyads. Sex Roles, 8, 809822.
  • Derlega, V. J., Metts, S., Petronio, S., & Margulis, S. T. (1993). Self-disclosure. Newbury Park,CA: Sage.
  • Dindia, K. (2002). Self-disclosure research: Knowledge through meta-analysis. In M. Allen, R. W. Preiss, B. M. Gayle, & N. Burrell (Eds.), Interpersonal communication research: Advances through meta-analysis (pp. 169185). Mahwah, NJ: Erlbaum.
  • Dindia, K., & Allen, M. (1992). Sex-differences in self-disclosure: A meta-analysis. Psychological Bulletin, 112, 106124.
  • Dolgin, K. G., & Minowa, N. (1997). Gender differences in self-presentation: A comparison of the roles of flatteringness and intimacy in self-disclosure to friends. Sex Roles, 36, 371380.
  • Frye, N. E., & Dornisch, M. M. (2010). When is trust not enough? The role of perceived privacy of communication tools in comfort with self-disclosure. Computers in Human Behavior, 26, 11201127. doi:10.1016/j.chb.2010.03.016
  • 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, 385401.
  • Harper, B. V., & Harper, E. J. (2006). Understanding student self-disclosure typology through blogging. The Qualitative Report, 11, 251261.
  • Herring, S. C. (2004). Slouching toward the ordinary: Current trends in computer-mediated communication. New Media & Society, 6, 2636.
  • Herring, S. C., & Paolillo, J. C. (2006). Gender and genre variation in weblogs. Journal of Sociolinguistics, 10, 439459.
  • Herring, S. C., Scheidt, L. A., Wright, E., & Bonus, S. (2005). Weblogs as a bridging genre. Information Technology & People, 18, 142171.
  • Hollenbaugh, E. E. (2011). Motives for maintaining personal journal blogs. Cyberpsychology, Behavior, & Social Networking 14, 1320. doi:10.1089/cyber.2009.0403
  • Huffaker, D. A., & Calvert, S. L. (2005). Gender, identity, and language use in teenage blogs. Journal of Computer-Mediated Communication, 10(2), article 1. Retrieved January 2, 2007, from http://jcmc.indiana.edu/vol10/issue2/huffaker.html
  • Isaac, M. (2010, December 15). Facebook to add facial recognition software to photo tagging. In Forbes online, retrieved March 22, 2011, from http://blogs.forbes.com/mikeisaac/2010/12/15/facebook-to-add-facial-recognition-software-to-photo-tagging/?boxes=Homepagechannels
  • Joinson, A. N. (2001). Self-disclosure in computer-mediated communication: The role of self-awareness and visual anonymity. European Journal of Social Psychology, 31, 177192.
  • Jourard, S. M. (1971). The transparent self (rev. ed.). New York: Van Nostrand Reinhold.
  • Jourard, S. M., & Lasakow, P. (1958). Some factors in self-disclosure. The Journal of Abnormal and Social Psychology, 56, 9198. doi:10.1037/h0043357
  • Kaye, B. K. (2005). It's a blog, blog, blog, blog world. Atlantic Journal of Communication, 13, 7395.
  • Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral research (4th ed.). Belmont, CA: Wadsworth.
  • Knapp, M. L., Ellis, D. G., & Williams, B. A. (1980). Perceptions of communication behavior associated with relationship terms. Communication Monographs, 47, 262278.
  • Lenhart, A., & Fox, S. (2006). Bloggers: A portrait of the Internet's new storytellers. Pew Internet & American Life Project. Retrieved January 18, 2012, from http://www.pewinternet. org/∼/media/Files/Reports/2006/PIP%20Bloggers%20Report%20July%2019%202006.pdf.pdf
  • Ma, M. L., & Leung, L. (2006). Unwillingness-to-communicate, perceptions of the Internet and self-disclosure in ICQ. Telematics and Informatics, 23, 2237. doi:10.1016/j.tele.2005.01.001
  • Maxwell, S. E., & Delaney, H. D. (2004). Designing experiments and analyzing data: A model comparison perspective (2nd ed.). Mahwah, NJ: Earlbaum.
  • McKenna, K. Y. A., & Bargh, J. A. (2000). Plan 9 from cyberspace: The implications of the Internet for personality and social psychology. Personality and Social Psychology Review, 4, 5775.
  • Mesch, G. S., & Beker, G. (2010). Are norms of disclosure of online and offline personal information associated with the disclosure of personal information online? Human Communication Research, 36, 570592. doi:10.1111/j.1468-2958.2010.01389.x
  • Morton, T. L. (1978). Intimacy and reciprocity of exchange: A comparison of spouses and strangers. Journal of Personality and Social Psychology, 36, 7281.
  • Neuendorf, K. A. (2002). The content analysis guidebook. Thousand Oaks, CA: Sage.
  • Nosko, A., Wood, E., & Molema, S. (2010). All about me: Disclosure in online social networking profiles: The case of Facebook. Computers in Human Behavior, 26, 406418. doi:10.1016/j.chb.2009.11.012
  • Peter, J., Valkenburg, P. M., & Schouten, A. P. (2005). Developing a model of adolescent friendship formation on the Internet. CyberPsychology & Behavior, 8, 423430.
  • Punyanunt-Carter, N. M. (2006). An analysis of college students' self-disclosure behaviors on the Internet. College Student Journal, 40, 329331.
  • Qian, H. & Scott, C. R. (2007). Anonymity and self-disclosure on weblogs. Journal of Computer-Mediated Communication, 12(4), article 14. Retrieved March 25, 2009, from http://jcmc.indiana.edu/vol12/issue4/qian.html
  • Riffe, D., Lacy, S., & Fico, F. G. (2005). Analyzing media messages: Using quantitative content analysis in research (2nd ed.). Mahwah, NJ: Erlbaum.
  • Rubin, Z. (1975). Disclosing oneself to a stranger: Reciprocity and its limits. Journal of Experimental Social Psychology, 11, 233260.
  • Scott, C. R. (2004). Benefits and drawbacks of anonymous online communication: Legal challenges and communicative recommendations. In S. Drucker (Ed.), Free speech yearbook 41 (pp. 127141). Washington, DC: National Communication Association.
  • Sinha, V. (1972). Age differences in self-disclosure. Developmental Psychology, 7, 257258.
  • Spicer, J. (2005). Making sense of multivariate data analysis. Thousand Oaks, CA: Sage.
  • Stefanone, M. A., & Jang, C.-Y. (2007). Writing for friends and family: The interpersonal nature of blogs. Journal of Computer-Mediated Communication, 13, 123140.
  • Suler, J. (2004). The online disinhibition effect. CyberPsychology & Behavior, 7, 321326.
  • Taylor, D. A., & Altman, I. (1966). Intimacy-scaled stimuli for use in studies of interpersonal relations. Psychological Reports, 19, 729730.
  • Tidwell, L. C., & Walther, J. B. (2002). Computer-mediated communication effects on disclosure, impressions, and interpersonal evaluations: Getting to know one another a bit at a time. Human Communication Research, 28, 317348.
  • Trammel, K. D., & Keshelashvili, A. (2005). Examining the new influencers: A self-presentation study of A-list blogs. Journalism & Mass Communication Quarterly, 82, 968982.
  • Valkenburg, P. M., & Peter, J. (2007). Preadolescents' and adolescents' online communication and their closeness to friends. Developmental Psychology, 43, 267277. doi:10.1037/0012-1649.43.2.267
  • Viegas, F. B. (2005). Bloggers' expectations of privacy and accountability: An initial survey. Journal of Computer-Mediated Communication, 10(3), article 12. Retrieved January 2, 2007, from http://jcmc.indiana.edu/vol10/issue3/viegas.html
  • Wheeless, L. R. (1978). A follow-up study of the relationships among trust, disclosure, and interpersonal solidarity. Human Communication Research, 4, 143157.

Appendix A

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

Codebook

For each thematic utterance identified as self-disclosure (i.e., the highlighted portions in the blog entries), please assign it to a topic category, described below. (Use the provided handout to define these categories.) NOTE: If you cannot decide which category to assign an utterance, please make a note on the blog entry and hand-write an alternative category in the margins.

Additionally, you should rate the level of intimacy of that disclosure, or how personal it is, by circling the corresponding number beside the appropriate topic. More intimate disclosures are assumed to be riskier and are more likely to be disclosed only to someone very close. Rate each disclosure as a level 1, 2, or 3.

Thematic Utterances are coded by TOPIC and then within the TOPIC by LEVEL. Descriptions of the topics and levels, including examples, are below.

LEVELS

  1. PERIPHERAL: general information about the topic as it relates to self; common, more visible things; accessible to others; superficial; no risk; most people would likely disclose to a stranger; personal information that is descriptive and explanatory; socially desirable

    Example: “I am dating someone.” “I need a job to pay for school.”

  2. INTERMEDIATE: attitudes, beliefs, and opinions about specific issues; semi-private; minimal risk; includes orientation toward authority figures; most people would likely disclose to a casual friend, co-worker, or short-term dating partner, (evaluative intimacy) – judge phenomena (Monsour 1992). Like or dislike of object, person, event (I agree, I think, I hate, I like).

    Example: “I really feel good about this person, and I'm happy with him/her.”

  3. CORE: personal beliefs, needs, fears, inadequacies, weaknesses, vulnerabilities, socially undesirable characteristics, values, self-concept, and self-esteem; beliefs about self-identity; private/personal; risky; most people would likely disclose only to a best friend or close romantic partner

    Example: “This better be the one, because I feel like I am getting too old to meet anyone new, and I'm not confident that anyone else would be attracted to me.” (I feel, I need, I want, I regret, I worry, I care, I am am/not satisfied …)

TOPICS

Biographic Information Information essentially about one's self-concept, personality, and internal states NOT directly related to other categories. Includes “demographic” information about family composition and relationship status, although NOT the quality or characteristics of those relationships in particular. Includes own religion and views about one's own religion. Includes both current and historical events that one is personally relating to the self or to one's own experience.
Sex Utterances about one's sexual performance or sexual history; facts about present sex life.
School and Work Disclosures pertaining to school and work, including attitudes toward school and work, emotions/states relating to education and career, and career goals.
Current Events Discussing opinions, attitudes, and beliefs about current and/or historical events that one is NOT personally relating to one's own experience. Includes views on government, trends, specific events in entertainment/sports, religion (unless talking about one's own religion personally), etc.
Physical Appearance and Condition (Body) Description and/or evaluation of one's own physical appearance, health, attractiveness, attempts at changing appearance, and practices related to the body (e.g., running, lifting weights). Also includes attitudes and/or beliefs about the body abstractly.
Hobbies and Interests Revealing one's own hobbies and interests, including pastimes (like playing or watching sports), favorites, tastes in music, movies, and books. Includes disclosures about pets, as well. Something that is engaged in on a regular basis.
Money and Property Sharing one's own financial worth, attitudes toward money and material property, plans or goals related to money, and the worth of a dollar.
Relationships with Others Disclosures about a real, specific relationship with another person, whether romantic, familial, workplace related, or platonic friendships. Includes information about the quality of a relationship, such as strengths and weaknesses and the feelings one has for another. Also includes how one feels about the actions of friends, family members, co-workers, and romantic partners. Does NOT include revealing the existence of such relationships (which are Biographic), but instead information about the relationships themselves.

Appendix C

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies

Tables

Table 1. Frequencies and Reliability Coefficients for Coded Variables
 n(%)Cohen's kappa
  1. Note. Frequencies for breadth and depth categories reflect the number of participants who were coded as having at least one disclosure in the respective categories.

Profile information  
  First name82 (53.2).867
  Last name22 (14.3)1.000
  Nickname/pseudonym47 (30.5).932
  Picture(s) of blogger – face in full display47 (30.5).920
  Picture(s) of blogger – face obscured24 (15.6)1.000
  Picture(s) of more than one person13 (8.4).840
  Picture(s) of animal, object, or place10 (6.5)1.000
  Comics (in still form)2 (1.3)1.000
  Illustrations/art/clip art34 (22.1)1.000
Breadth categories  
  Biographic Information149 (96.8).920
  Sex9 (5.8).854
  School and Work100 (64.9).924
  Current Events58 (37.7).832
  Physical Appearance and Condition (Body)59 (38.3).930
  Hobbies and Interests135 (87.7).812
  Money and Property106 (68.8).868
  Relationships w/ Others59 (38.3).874
  Krippendorf's alpha (ratio)
Levels of Depth  
  Depth 1149 (96.8).8153
  Depth 2153 (99.4).7933
  Depth 369 (44.8).8557
Table 2. Means and Standard Deviations for Dependent Variables
 MSD
Amount of self-disclosure14.827.30
Breadth of self-disclosure4.381.53
Depth of self-disclosure  
  Biographic1.61.28
  Sex1.40.34
  School and work1.43.37
  Current events1.54.41
  Physical appearance1.61.43
  Hobbies and interests1.58.31
  Money and property1.47.40
  Relationships with others1.84.33
Table 3. Individual Betas in Regressing Disclosure Dimensions on Control Variables, Visual Anonymity, and Discursive Anonymity
 AmountBreadthDepth
BiographicSchool & WorkCurrent EventsPhysical AppearanceHobbies & InterestsMoney & PropertyRelationships with Others
  1. a

    p < .05.

  2. b

    p < .01.

  3. c

    p < .001.

  4. Note. All betas are standardized betas on the last step of the regression analyses.

Age−.17a−.28c−.13−.06.21.01.03.13−.06
Sex.21b.24b.19a−.06−.12.18.01−.01.05
Visual Anonymity−.16a−.14−.090.12.05−.05−.01.16
Discursive Anonymity.15.14.04−.11−.14.14−.01.02.04
Adjusted R2 (total model).09b.15c.03−.02.01−.01−.03−.02−.04

Biographies

  1. Top of page
  2. Abstract
  3. Literature Review
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix A
  10. Appendix B
  11. Appendix C
  12. Biographies
  • Erin E. Hollenbaugh (Ph.D., Kent State University, 2008) is an Assistant Professor in the School of Communication Studies at Kent State University at Stark. Her research interests lie in the blurring intersections between interpersonal and media communication, specifically new media.

    Address: 6000 Frank Ave. NW, North Canton, OH 44720, USA.

  • Marcia K. Everett (Ph.D., Kent State University, 2005) is a Professor in the Communication Arts Department at Malone University. Her teaching and research interests include social support, new media, and gender issues in interpersonal relationships.

    Address: 2600 Cleveland Ave. NW, Canton, OH 44709, USA.