Abstract
Abstract
A review of blog literature suggests that blogs may have many social and emotional benefits. In particular, there is the suggestion that blogging may have benefits in the forms of coping, insight and growth, catharsis, as a free space to communicate, through feedback, through recognition, and in social support. However, while there have been numerous studies on the potential motivations for and effects of blogging, little quantitative research has been conducted on defining the manner in which one experiences blogging. Thus rather than focusing on motivation for blogging, the present study examined the way in which bloggers perceived and experienced their activity, that is, their blogging style. A 16-item Personal Blogging Style Scale (PBSS) was created after a factor analysis of 25 sample items. Participants were 182 experienced bloggers. The PBSS classifies user experiences by four independent and unique subscales: Therapeutic Blogging, Self-Censoring Blogging, Connected Blogging, and Substitution Blogging. Reliability and validity evidence were presented for the subscales of the measure.
Introduction
Blogs, or online diaries, are websites containing an archived series of reverse chronological items of content posted by the author. 2 Blogs exist on several different types of Web sites (social networking sites, blog-specific Web sites, corporate Web sites, etc.), and accordingly, blog styles can vary widely from being impersonal and professional to being informal and chatty. 3 There has been research on the antecedents of blogging2,4,5 and its effects,6,7 but very little quantitative research on different blogging styles. However, the manner in which bloggers interpret their experience of blogging might differ from their original intentions, as evidenced, for example, by the high rate of discontinued blogs, 8 suggesting the experience did not lead to the anticipated rewards. Additionally, the way a blog is experienced may influence outcomes for the author, for example, their level of social integration, friendship development, and psychological well-being.6,7 Thus the ability to describe a blogger's personal style may be of benefit in providing a link or mediator between an individual's original motivations for blogging and subsequent psychosocial outcomes.
Blog usage generally fits well within a uses and gratifications framework, 9 whereby Internet use is postulated to satisfy needs. Motivations described in the literature include information gathering and dissemination, keeping up with trends, passing time, self-promotion, professional advancement, connecting with others, and expressing emotions.2,4,5 Blogs have been described as helpful for learning, organizing ideas, and making sense of experiences.10–12 By providing a space for venting emotions, blogs can be cathartic.12–14 They also seem to serve a social purpose. Through open dialogue using commenting and track-back features, blogs open themselves up to dimensions of social support, friendship, and positive interaction,13,15 and blogging appears to improve an individual's sense of satisfaction with their friendships and their sense of closeness with peers.6,7
Despite most blog literature describing mainly positive interactions and experiences in the blogosphere, 16 there are opportunities for an individual to experience increased feelings of isolation and alienation through personal attack, or indeed from lack of acknowledgment. An author may receive twenty positive comments but focus on a few negative ones. In this way, motivations and benefits of blogging are insufficient to describe the blogging experience, as they do not capture all the dimensions of the personal experience. Bloggers' thinking styles may play a large role in the way they experience blogging, and any resulting positive or negative consequences. There is thus a need to examine how different bloggers interpret the experience of blogging.
The aim of the present study was to develop a measure of the ways in which individuals experience and interpret their blogs—a Personal Blogging Style Scale (PBSS), focusing mainly on the social/communication and emotional aspects of the medium. A list of 25 potential items (see Table 1) was created from some of the experiential aspects reported in the literature. For example, discussion of the therapeutic aspects of blogging in Nardi et al.'s 12 qualitative study of bloggers' experiences led to the development of items such as “When I'm upset, blogging helps ease my distress.” As well, the writings of Doctorow 11 and Blood 10 on blogger cognitions provided ideas for generating items such as “I think about how my readers will react to a blog entry while I'm writing it” and “My blog has helped me to organize my thoughts.”
Boldface type indicates the item exceeds the factor loading criterion.
Participant responses to the items were factor analyzed, scales developed, reliabilities assessed, and relationships with other scales analyzed to establish validity. Construct validity was assessed through testing of predictions that (a) the tendency to use more socially oriented styles of blogging would correlate with measures of friendship quality and perceived social support, and (b) the tendency to use blogging styles characterized by efforts to use the blog to vent, cope with, and gain support for personal problems would correlate with the more general use of coping styles such as venting, seeking emotional and instrumental support, and active coping. The rationale behind these predicted relationships was that individuals would bring to their blogs a style of interaction that paralleled their offline persona. Additionally, relationships between blogging style, blog activity and readership, and psychological distress were explored, as were relationships between blogging style and a range of other coping strategies for which specific hypotheses could not be derived from past research or theory.
Methodology
Participants
A total of 182 bloggers (87 males, 95 females; age range 18–64 years, M = 29.08, SD = 9.06) recruited from
Participants' mean number of post-initial blogs posted over a 1-week period (see Measures) was 5.02 (SD = 9.98). Over the same week, the mean number of comments received was 82.66 (SD = 195.30), the mean number of subscribers was 88.13 (SD = 160.82), and the mean number of blog readers (provided by
Sampling method
Measures
Blogging style
Participants rated their agreement with the 25 PBSS items (expressed as statements) on 5-point Likert scales ranging from 1 = “completely disagree” to 5 = “completely agree.”
Blog activity and readership
Blogs of the recruited participants were examined by the first researcher, and information on the number of blog entries posted, comments received, subscribers, and readers was collected over a 1-week period. These measures were positively skewed and were transformed using log transformations to meet the assumption of a normal distribution.
Social interaction
The Social Provisions Scale (SPS; 24 items) assessed perception of social support in six domains (α = 0.92). 17 Participants rate statements about their social support on a 4-point scale ranging from 1 = “strongly disagree” to 4 = “strongly agree.” Participant ratings of satisfaction with number and closeness of their online and offline friends were also assessed using 5-point Likert-type scales ranging from 1 = “completely dissatisfied” to 5 = “completely satisfied.”
Coping
The 28-item Brief COPE 18 assessed participants' preferred coping styles. Ten of the 14 COPE subscales were relevant to the current study because they tapped emotional and social aspects of coping. These subscales, with their median alpha reliabilities, were: Active Coping (0.68), Planning (0.73), Positive Reframing (0.64), Acceptance (0.57), Emotional Support (0.71), Instrumental Support (0.64), Self-distraction (0.71), Denial (0.54), Venting (0.50), and Self-blame (0.69). 17 Participants responded to statements on a 4-point scale ranging from 1 = “haven't been doing this at all” to 4 = “have been doing this a lot.”
Affect
The Depression, Anxiety and Stress Scale-21 (DASS-21) assessed psychological affect/distress (depression a = 0.88, anxiety a = 0.82, stress a = 0.90).19,20 Participants responded to statements about their recent mood on a 4-point scale ranging from 0 = “did not apply to me at all” to 3 = “applied to me very much or most of the time.”
Results
Exploratory factor analysis
The 25 items of the PBSS were subjected to a principal components analysis (PCA) using SPSS statistical software (V16; SPSS, Inc., Chicago, IL). Prior to this, the suitability of the data for factor analysis was assessed. Inspection of the correlation matrix revealed the presence of many coefficients of 0.3 and above. The Kaiser-Meyer-Olkin value was 0.842, exceeding the recommended value of 0.6,21,22 and the Bartlett's test of sphericity 23 reached significance (χ2 = 1774.98, df = 300, p < 0.001), supporting the factorability of the correlation matrix.
PCA revealed the presence of seven components with eigenvalues exceeding 1, explaining respectively 27.1%, 10.3%, 7.8%, 6.1%, 5.0%, 4.6%, and 4.0% of the variance. Inspection of the scree plot using Cattell's 24 scree test in combination with parallel analysis indicated a four-component solution.
To aid the interpretation of these components, varimax rotation was performed. The solution explained 51.4% of the variance (component variance shown in Table 1). As the sample size was <200, the factor loading criterion was increased to 0.50 for each item. Using this criterion, the rotated solution revealed the presence of a simple structure, 25 with a number of strong loadings and no cross loadings (see Table 1).
On the basis of item content, the factors were named Factor 1: Therapeutic Blogging; Factor 2: Self-Censoring Blogging; Factor 3: Connected Blogging; Factor 4: Substitution Blogging.
Creation of four subscales and confirmatory analysis
Factor 1 contained eight items, while factors 2, 3, and 4 each contained four items. It was decided to use the four highest loading items on Factor 1 and discard the remaining four items in order to limit each factor scale to four items. This decision was supported by the high intercorrelation between the full eight-item factor and the reduced four-item factor (r = 0.97, p < 0.01). Additionally five items from the original set were not included in the final scale as they did not load at 0.50 or above on any factor.
A second PCA was conducted using the 16 retained items (4 items × 4 factors) to confirm a four-factor structure. The Kaiser-Meyer-Olkin value was 0.779, exceeding the recommended value of 0.6,21,22 and the Bartlett's test of sphericity 23 reached significance (χ2 = 965.64, df = 120, p < 0.001), supporting the factorability of the correlation matrix.
PCA revealed the presence of five components with eigenvalues exceeding 1, explaining 26.1%, 15.1%, 10.8%, 8.0%, and 6.6% of the variance respectively, but parallel analysis and inspection of the scree plot using Cattell's 24 scree test indicated a four-component solution. The varimax rotated solution explained 60.1% of the variance. Using the criterion of 0.5 as the minimum value for acceptable factor loadings, the rotated solution revealed the presence of a simple structure. 25 The factor analysis confirmed the original four factors. Four subscales were created from these items.
Subscales of the PBSS
Scores for the subscales were calculated by adding relevant item ratings. Correlations between the subscales are shown in Table 2, as are Cronbach's alphas (diagonal). To test for construct validity of the 16-item PBSS, correlations between the subscales and measures of blog activity and readership, social provisions, satisfaction with friendship, coping styles, emotional distress, age, and gender were examined (Table 3).
p < 0.001; **p < 0.01; *p < 0.05.
p < 0.001; **p < 0.01; *p < 0.05.
Boldface type indicates a statistically significant correlation.
Discussion
The current study involved the development of a scale that assessed four distinct blogging styles. Each of these styles is described in turn. The Therapeutic Blogging subscale assessed a style that is open and expressive, focused on affective exchange and symptom relief, and more directed to the concerns of the blogger than the blog readers. Individuals scoring high on this subscale were less satisfied with their friendships; scored higher on depression, anxiety and stress; and endorsed numerous coping mechanisms employed to deal with stressors, including venting, positive reframing, seeking emotional and instrumental support, and acceptance. The profile of therapeutic bloggers suggests individuals who are emotionally distressed, in need of social support, have a strong need for personal comfort from others, but who are able to use a range of coping styles involving social connection to relieve their distress. This fits well with the idea of blogging providing a space for emotional outlet, to connect with others, and to seek support.12–14
High scorers on Connected Blogging tended to be less stressed and depressed, more satisfied with friendships, in receipt of more comments from others, and have more subscribers and readers to their blogs. Connected Blogging was also positively related to most of the social provisions scales and negatively associated with the maladaptive coping technique of denial, suggesting that these bloggers utilize their blogs to connect and communicate with others rather than solve emotional problems. Their highly interactive blogs appear to enhance friendship networks in a range of ways, as suggested in previous research,6,7,13,15 beyond just seeking assistance for problems and stresses as in the case of the therapeutic bloggers.
The Self-Censoring style also blogs to communicate, but those scoring high on this subscale focus on positive portrayal and self-presentation, 14 rather than a more open style of communication usually favored among friendship groups. Items in the Self-Censoring Blogging subscale on face value suggest the restriction of emotional expression and caution in sharing personal information. This was supported by correlations demonstrating fewer blog posts, lower negative affect, and less emotional and instrumental support seeking. Self-Censoring bloggers were also less likely to share feelings by venting, or using other coping mechanisms such as acceptance, self-blame, planning, and positive reframing. Thus the Self-Censoring style appears focused on substantial positive identity construction and impression management in line with observations by Nardi et al. 12 and Jacobs 14 that many bloggers utilize significant self-censorship and self-presentation strategies when authoring their blogs.
Items comprising the Substitution Blogging subscale suggest a strong focus on interaction with others through the blog. Those high on this subscale appear to use their blogs to substitute for, rather than enhance, face-to-face friendships and social networks. This type of blogging may be characteristic of authors using the Internet to overcome face-to-face loneliness or social anxiety. 26 Indeed, the correlations suggest that substitution bloggers were relatively dissatisfied with the number of and closeness with face-to-face friends, yet despite these face-to-face deficits, tended to have higher numbers of blog subscribers and readers. Thus while substitution bloggers may be dissatisfied with their offline relationships, their focus on feedback from others and readership appears to be successful, as those scoring high on this blogging style report a higher number subscribers and comments from their readership. There were no significant correlations of the style with mood or coping strategies, indicating that substitution bloggers were no more or less likely than others to be stressed or use particular styles of coping.
The four blogging styles fit well with the initial hypotheses. As predicted, the more socially oriented connected style of blogging correlated with measures of friendship quality and perceived social support. Substitution bloggers appeared to be successfully utilizing their blog to compensate for face-to-face deficits. Likewise, the more cathartic therapeutic style correlated with active coping strategies to deal with higher levels of emotional distress. No predictions were made regarding self-censorship, but this style appears to fit well with the self-presentation focused blogger described by Nardi et al. 12 and Jacobs, 14 or alternatively with the “lurking” bloggers described by IP and Wagner 27 who are more focused on reading other blogs than posting their own.
In the current research, females were more likely to use a therapeutic style of blogging, and males more likely to be self-censoring or substitution-style bloggers. While some research has suggested that sex has little effect on the content of blogs,28,29 other studies show women to be more focused on writing about personal life 30 and socialising 31 than men, 32 including greater use of emoticons33,34 and diary-type blogs. 35 Past research has also suggested that personal journal blogs can be therapeutic in reducing emotional distress and increasing social provisions.6,7 Thus the relationship between gender and Therapeutic Blogging style may be a consequence of more interpersonal content and motivation. Males were more likely to use blogging as substitution for face-to-face friendship, which may be evident of a problem-focused method of coping with social deficits where male bloggers concentrate on creating positive social and professional online connections.
Older bloggers were more likely to blog in a connected style. Research on older bloggers is still somewhat unclear, perhaps because most blogs are created by young adults. 29 Schler et al. 30 found that older bloggers were more likely to post blogs topical in content and containing hyperlinks, while in partial contrast, Fullwood et al. 3 found older bloggers more likely to post negatively toned blogs. In the present study, older bloggers were more inclined to use their blogs to communicate with others, and experience their blogs as topical extensions of their normal interests and conversations.
This initial study into blogging styles illustrated that there are indeed different styles with which individuals interpret their blogging experience, and these different dimensions can overlap for individual bloggers, as seen by intercorrelations between styles. Just as blogs can change in content and style from post to post, 36 it makes sense that blog authors can experience their blog in different ways at different times depending on their mood, commentary, feedback, and other relevant factors. Still, the simple structures of each factor and moderate to high internal reliabilities suggest that the subscales are suitable to be used individually or in combination to assess the different ways in which bloggers may experience their blogs.
Conclusion
The primary and secondary factor analysis resulted in four independent subscales for the PBSS that describe relatively unique styles of blogging. The blogging scales demonstrated high face validity as evidenced by the relationships between their items, good construct validity as seen in their correlations with relevant constructs, and adequate reliability. Bloggers can be profiled on the four subscales to assess the relatively strength of each style. The present study is limited by its focus on social and emotional experiences, and its small sample size. Future research could examine further validity evidence for the PBSS in a larger sample.
Footnotes
Disclosure Statement
No competing financial interests exist.
