Reality Television as a Model for Online Behavior: Blogging, Photo, and Video Sharing



This research explores traditional mass media as an antecedent to nondirected self-disclosure online. New Internet-based tools allow users to communicate with global audiences, and to make intimate personal information available to this audience. At the same time, a culture that rewards the public performance of private thoughts and emotions is increasingly evident in “reality” television (RTV) programming. This study used survey data to examine RTV consumption, authoritarianism, and users' offline social context as potential antecedents for nondirected self-disclosure via blogs, online photo sharing, and online video sharing. RTV consumption correlated with blogging and video sharing, but not photo sharing. Social support network size was a significant correlate of photo sharing, indicating that photo sharing may be a more relational activity.

Cultivating Celebrity: How Reality Television Influences Blogging, Photo, and Video Sharing

Despite annual predictions that it is just a passing fad, reality television programming continues to gain prominence in U.S. network and cable television lineups. Reality television (RTV) which purports to focus on the unscripted interaction of amateur actors (or “real” people) is lauded by network executives for its lucrative ability to attract young adult viewers. “Fifty-six network reality series will air this season, up from 51 last year, and pilot orders are up 50” while cable networks such as MTV and Bravo use the genre exclusively to fill their lineup of original content (Levin, 2007). Other economic factors, such as the relatively low upfront development costs also promote adoption of such shows: even shows that rate poorly and “burn out” quickly tend to be enormously profitable (Calvert, 2000; Levin, 2007).

In the tradition of social cognitive theory, our goal is to explore the relationship between RTV consumption and social behavior manifested online. We suggest that a media diet heavy in RTV functions to model nondirected self-disclosure (NDSD). Some self-disclosure portrayed in RTV is consistent with traditional norms; RTV characters in engage in personal relationships and self-disclose to those other characters. Disclosure during interpersonal communication is traditionally characterized by personalism, as it is directed at specific others. However, self-disclosure portrayed in RTV is typically nondirected, as participants regularly share thoughts and feelings in a variety of ways to groups of recipients. For example, often “confessional” style formats are used so that viewers can learn about characters' inner thoughts and feelings. Within competitive shows like American Idol, interviews and spotlight segments are the mechanisms through which nondirected self-disclosure is enacted, albeit more subtly than confessionals. In both of these cases, participants impersonally disclose personal thoughts and feelings to abstracted viewing audiences. We suggest that this impersonal NDSD which is broadly portrayed via RTV is analogous to behavior exhibited online by users of communication tools such as blogs, social networking sites, and video and photo sharing sites. We propose to explore the relationship between traditional mass media consumption in the form of RTV and the use of three types of online tools: web logs (blogs), photo sharing sites, and video sharing sites, all social technologies that contribute to the “Web 2.0.”

Web 2.0 (O'Reilly, 2005) refers to a changing orientation between online content producers, consumers, and web technologies. While the utility of this particular term remains questionable (e.g., Berners-Lee, 2006) it does serve to highlight the increasing prevalence of personal, user-focused content online. Recent research indicates that many concepts behind Web 2.0 are more than just marketing hype, and that younger people in particular are increasingly engaged with these social technologies. Over half of all Internet-using teens are “content creators” who create websites or blogs, share original media like photos and videos, or remix content into new creations (Lenhart & Madden, 2005). Social networking sites like MySpace ( and Facebook ( are becoming cornerstones of this informational space, with many recent surveys finding that 95% or more of college students maintain site profiles (e.g., PACS survey, 2007).

We see substantial congruence between Web 2.0's culture of personal self-disclosure and the “reality culture” that has come to dominate some segments of the television market. Recent research on blogging, for example, finds that disclosures via personal-journal style blogs are often nondirected in nature (Stefanone & Jang, 2008). The following review discusses the normative values of RTV, variables we believe affect the social cognitive process, and poses hypotheses about the intersection of “reality” media culture and new media behavior.

Television and Reality Culture

The relationship between the content of mass media and cultural attitudes is among the most examined issues in mass communication research. Previous studies have explored the relationship between mass media and attitudes towards violence (Dominick, 1984), sex (McGee & Frueh, 1980), and smoking (Shanahan, Scheufele, Yang & Hizi, 2004; Wakefield, Flay, Nichter & Giovino, 2003) among many additional topics. One trend observed in the last two decades is the relative increase of reality-framed television programming. As the term is used here, the defining characteristic of RTV is that ordinary people (not professional actors) serve as the main characters (Reiss & Wiltz, 2004), and includes programs such as Survivor, Big Brother, and Temptation Island. RTV makes the personal thoughts, behaviors, and interactions of the characters the main focus of audience attention. Bente and Feist (2000) refer to this genre as affect TV, which presents viewers with “the most private stories of nonprominent people to a mass audience, crossing traditional borders of privacy and intimacy” (p. 114). Calvert (2000) refers to this realignment of the private and the public as “mediated voyeurism,” and suggests that this is becoming endemic to culture at large. This culture of mediated voyeurism may influence those who are most involved in it, and specific personality traits like authoritarianism may exacerbate these effects.

Mediated Voyeurism

As early as 1922, Lippman suggested that we live in a mediated “pseudoenvironment,” while a host of later cultural scholars have debated the nature of the “spectacle” and the “simulation” of the media-saturated culture. The present focus of mass media research is social constructivism which recognizes that media exert influence on the way audiences structure social reality (cf. Shanahan & Morgan, 1999). This influence is moderated by discourse about media, as peer groups negotiate meanings among themselves. Although theoretical approaches may vary, it is reasonable to hypothesize that heavy consumers of specific media genres, such as RTV, may be influenced by its messages (Shrum, Wyer, & O’Guinn, 1998).

One of RTV's most consistent messages regards NDSD, where personal revelations are not targeted toward specific, individual others, but rather directed at broader audiences. As the personal thoughts of the characters are not (yet) directly accessible to the viewing audience, the narrative structure of many RTV shows requires the characters to sacrifice traditional privacy by disclosing thoughts and feelings to a broad viewing audience. The high level of NDSD is presented as necessary and fundamentally normal. Ultimately, RTV participants are rewarded for this behavior with financial gain in the form of prize money, and they often retain their celebrity status even after these programs run their course. While previous discussions of the genre have included shows such as COPS where some participants (the suspects) are ostensibly unwilling, this discussion is limited to socially rewarding RTV with participants who willingly disclose their private selves.

Modeling Disclosive Behavior

A variety of media effects models have been proposed, including cultivation theory and social cognitive theory. Cultivation theory (Gerbner, Gross, Morgan, & Signorielli, 1982; Shanahan & Morgan, 1999) suggests that mass media consumption influences individuals' conceptions of social reality. More specifically, the theory posits that heavy consumers of media align their beliefs, attitudes, and behaviors with those observed in the media. Cultivation studies have found that as audiences are exposed to versions of televised reality they begin to cultivate a shared idea of reality. However, cultivation research tends to disregard individual differences and treats viewers as one homogeneous group (Potter, 1993). To account for these distinctions, the current project examines media use and behavior from the social cognitive perspective.

Social cognitive theory (formerly social learning theory; Bandura, 1986) attempts to explain how and why people acquire and maintain certain behavioral patterns. Human functioning is explained as the product of the dynamic interaction of personal, behavioral, and environmental influences. Personal influences include cognitive, affective, and biological factors. Environmental factors include social context and the informational environment. Finally, Bandura includes behavior as a component of function because individuals can reflect on the effects of their own behavior. This tripartite construct is thus dynamic and highly contextual.

Social cognitive theory uses the term modeling to characterize the process through which an individual observes others, interprets the observed behavior, and adjusts their own behavior in response. Such observational learning may be the intended outcome of a given behavioral process, such as teaching a child to feed itself. However, Bandura (1986) notes that modeling may occur in many other contexts, indeed wherever an individual is able to observe others' behavior. The development of television is viewed by Bandura as an especially important source of behavior models, enabling people to “transcend the bounds of their immediate social life” (p. 55). In comparison to the quantity of information about the world available in daily life, the amount of environmental information provided via media is vast. To the extent that one's images of reality are mediated and vicarious rather than directly experiential and experimental, the greater the impact of the media. Bandura is careful when he explains that modeling is a more complex process than simple mimicry or imitation, and identifies several specific functions of the process.

The observational learning process requires a model, a learnable attitude or behavior, and a conducive personal/behavioral/environmental context. A recent study by Ferris, Smith, Greenberg and Smith (2007) argues that RTV presents behavioral models. Ferris et al. examined the relationship between content of reality dating shows, viewers' attitudes, and behaviors, and found that exposure to RTV was associated with viewer adoption of the dominant themes portrayed. In the present study, the characters in RTV programming also serve as models, but the Web 2.0 environment provides the context for enacting observed behavior. The wide adoption of Internet platforms such as social network sites and media-sharing sites among the largest demographic of RTV viewers—young adults (Hill, 2005)—suggest significant potential for interrelated media behavior.

Consistent with Rosen (2004), we argue that RTV promotes a culture which equates public visibility with personal success. Specifically, heavy viewers of socially rewarding RTV should be more likely to believe that behavior exhibited on these shows is acceptable and normal. These viewers may then model this behavior with the tools that are available to them, such as social web technologies. We examine the role that a personality trait—authoritarianism—plays in the context of RTV consumption and subsequent online behavior. The following discussion explains why this trait is of particular interest in to context of social cognitive theory and RTV consumption.


Previous research has found links between personality traits and media use, such as aggressiveness with violent media use (Slater, Henry, Swaim & Anderson, 2003) and openness to experience and television consumption (Finn, 1997). RTV may communicate a distinct set of values that heavy consumers emulate, and such emulation may be enhanced by authoritarian (Altemeyer, 1988; 1996) personality characteristics.

The study of authoritarianism is rooted in social psychology and political communication (Eagly & Chaiken, 1993) and has been understood as both a personality trait and an attitude (Christie, 1991). Mass media research tends to frame television consumption as the cause of emergent authoritarian attitudes in viewers (Oliver & Armstrong, 1995; Shanahan, 1998). Little research involving these variables and mass media, however, operationalizes authoritarianism as a trait which influences behavior.

Altemeyer (1988) suggests authoritarianism consists of three “attitudinal clusters” (p.2): submission, where people tend to follow authority rule or follow others who are perceived to be legitimate power sources in a society; aggression, where general aggression is perceived to be sanctioned by established authorities; and conventionalism, characterized by a high degree of adherence to social rules and norms. The conventionalism dimension of the authoritarian construct would seem to resonate most strongly with social cognition and cultivation theories, as cultural norms and values may be fostered by heavy media use.

Altemeyer (1996) suggests that people high in authoritarianism exhibit tendencies toward conformity to group norms where being “normal” means modeling behavior and attitudes perceived to be socially accepted and desirable. People high in authoritarianism tend to respect and defer to those in traditional roles of power, such as political and religious leaders, police, and parents. Thus, they tend to see little social value in “deviant” attitudes and behaviors. Interestingly, authoritarianism appears to operate independently of both age and gender, as no significant interactions have been found in decades of research. As research has indicated that television can become an influential social “other” to heavy viewers (Brown, Halpern, & L’Engle, 2005), RTV may exert a normalizing influence. It is likely that heavy consumers of a given television genre, such as socially rewarding RTV, may believe that the genre's values are conventional and normal. Individuals with highly authoritarian personalities may be increasingly likely to bring their attitudes and behaviors in line with this frame of conventionality.

In sum, television content presents norms and behaviors which may differ from those of the audience. In the specific context of RTV, these norms reflect the relaxation of personal boundaries of privacy and emphasize the benefits of public visibility. As individuals find themselves competing for a finite resource–the attention of others in a mediated environment–they may look for behavioral models that are successful in this social sphere. From a normative perspective, NDSD is a hallmark of RTV content, and within the genre of socially rewarding RTV participants are rewarded with attention (and often celebrity status) and cash prizes. Consistent with Calvert (2000), we suggest this RTV genre demonstrates a value system which equates visibility and celebrity status with social prestige and personal value.

New media technologies allow people to participate in an encompassing and active global media ecology which seemingly allows average people to achieve, or at least emulate, this celebrity. Webpages, podcasting, and webcasting provide access that can rival the reach and influence of major media organizations. The following section discusses these new technologies and presents our hypotheses about their relationship to changing notions of privacy and disclosure.

New Media and Self-Disclosure

Research on blogs suggests that hallmark features include format and content analogous to traditional journals. Schiano, Nardi, Gumbrecht, and Swartz (2004) found that bloggers use this technology for personal expression and expect interaction with their audience through a variety of communication channels. Bloggers leverage the interactive nature of the Internet to disseminate traditionally personal, private information. The public sharing of photos, via specialized platforms like Flickr ( or more general social platforms like Facebook, is an increasingly common online behavior. Likewise, posting personal videos, often shot using inexpensive digital cameras or mobile devices, is gaining in popularity as evidenced by the popularity of YouTube ( and the integration of video sharing capabilities into social networking sites. Although motivations and contexts for these behaviors may vary, media sharing presents strong elements of NDSD when access to the media is not restricted to a particular set of others. This section discusses each of these activities and presents specific hypotheses for three online behaviors: blogging, photo sharing, and video sharing.


Blogs are a reconfiguration of existing web-based tools: frequently updated web pages that generally have current and archived text-based posts. These tools enable anyone with access to a computer and the Internet to create and maintain a blog because little technical knowledge (i.e., HTML) is required. Although blogging tools afford different levels of privacy, ranging from password protected sites to publicly listed and accessible sites, the overwhelming majority of users do not restrict access to content (Lenhart & Fox, 2006). Stefanone and Jang (2007) suggest that some bloggers use this medium to maintain strong tie relationships, although disclosure via blogs is still nondirected in nature. In social network analysis, the most frequently measured aspect of relationships is tie strength, which operationalizes intensity as an indication of how close people are to each other (McCarty, Bernard, Killworth, Shelly, & Johnsen, 1997). Strong tie contacts are characterized by frequent, reciprocal communication and usually a long, stable history of interaction and typically constitute relationships with family and close friends. This group of contacts is also characterized as a social support network (SSN). On the other hand, weak ties are characterized by infrequent communication, low reciprocity, and a lack of emotional closeness (Granovetter, 1973; 1982). As social support network size increases so does the likelihood that blogs are used to maintain existing relationships (Stefanone & Jang, 2007), although blog posts were overwhelmingly public in nature as bloggers tended not to restrict access. Thus, accounting for the social context of people's technology use adds explanatory power and should be included when examining online behavior.

We contend that RTV exhibits NDSD as a core theme or value. In accordance with social cognitive theory, we suggest that people who observe NDSD presented as a positive behavior may model this activity, provided an appropriate context. Personal blogs may provide one such context. Thus,

H1: Heavy RTV viewing has a positive relationship with the likelihood to maintain personal-journal style blogs.

Also recall that authoritarianism is associated with conventionalism, the tendency to model behavior deemed to be appropriate and normal. As NDSD is perhaps more conventional within the discourses of RTV than in the broader culture, more authoritarian viewers may be more apt to model this apparently normalized behavior. It is likely that the interaction between RTV consumption and this personality trait will result in stronger modeling tendencies in the context of NDSD. Thus,

H2: The combination of heavy RTV viewing and authoritarianism has a positive relationship with the likelihood to maintain personal-journal style blogs.

Social cognitive theory also suggests that the environment of the learning process, such as the social context, may influence behavioral outcomes. Watching RTV with friends may introduce a ritualistic or normative aspect that is not necessarily present when watching television alone. To the extent that watching RTV cultivates a set of beliefs and attitudes about self-disclosure and celebrity, it is likely these norms will be further reinforced if participants view RTV with friends, as these social networks function to sustain and promote these norms in a variety of other social contexts. Thus,

H3: Watching RTV with friends has a stronger positive relationship with the likelihood to maintain blogs, opposed to watching RTV alone.

Video sharing.

Video has only recently become a highly distributed and accessible medium. Video production required expensive cameras and recording media (film, VHS), distribution required bulky and fragile physical media (cartridges, tapes), and playback was only possible on further specialized equipment (projectors, televisions, and VCRs). Advances in digital imaging technologies have rapidly changed video into a practical and normalized communication medium.

YouTube allows users to easily share and discuss video content uploaded from webcams, mobile devices, and other sources. YouTube currently ranks among the top four most-visited sites on the web (Alexa traffic rankings,, October 2008). Young adults are voracious consumers of online video, with three-quarters of the age 18–29 demographic reporting that they download or stream video (Madden, 2007). Of these young video viewers, over half watch videos with others or recommend links to others, while nearly a quarter of this demographic rates videos or post comments about them. Further, fully 20 percent of young adult video viewers report that they have uploaded their own videos (Madden, 2007). An explosion of competitors have emerged in an attempt to chip away at YouTube's popularity, and popular social networking sites such as MySpace and Facebook have since integrated video-sharing.

Video sharing would seem to be the behavior most in line with the values of RTV. The YouTube site, for example, contains prominent visual cues that indicate the popularity or fame of participants. Users rate the popularity of videos and an explicit tally shows the number of times the video has been viewed. While blogging is behavior symptomatic of NDSD modeled via RTV, sharing videos adds the “celebrity” aspects of socially rewarding RTV. Thus,

H4: Heavy RTV viewing has a positive relationship with sharing videos online.

H5: The combination of heavy RTV viewing and authoritarianism has a positive relationship with sharing videos online.

In keeping with our previous hypothesis about the social context of RTV consumption, it is likely that watching RTV with friends functions to reinforce the norms and values modeled in the programming. Thus,

H6: Watching RTV with friends has a stronger positive relationship with sharing videos online, opposed to watching RTV alone.

Photo sharing.

Digital devices have radically transformed the social landscape of photography. Digital cameras began to outsell film cameras in 2004 (Musgrove, 2006), while camera-enabled mobile phones began to outsell digital cameras just a year later (Sharma, 2005). The profusion of digital cameras and camera devices suggest that digital photography, like other new media practices, is an increasingly banalized activity (Graham, 2004) that may play a subtle but important role in social relationships. A host of web services such as Flickr and Snapfish ( have emerged to support the storage, organization, and sharing of digital photos, while general social networking sites such as Facebook and MySpace include photo-sharing as a key functionality.

Unlike blogging or video sharing, photography has a long and social history. As Americans have grown increasingly mobile in recent decades, photos have helped keep distant family members and friends in close emotional proximity. In many cases the digitalization of photography simply made these sharing processes faster and more convenient. The web has also enabled relatively new forms of interaction through photos. Miller and Edwards (2007) note that two relatively distinct modes of photo sharing can be observed online–both the traditional sharing of photos with an existing social network of friends and family, and an emergent form of public sharing with strangers and online acquaintances. These two groups are perhaps better understood as representing ends of a spectrum of sharing behaviors, as the boundaries of intimacy are increasingly blurred by technological affordances.

Consistent with the hypotheses above addressing blogging and video sharing, it is likely that heavy RTV viewers are more likely to share photos online. Further, the interaction of RTV viewing and authoritarianism should result in heightened levels of photo sharing online. Thus,

H7: Heavy RTV viewers are more likely to share photos online.

H8: The interaction of heavy RTV viewing and authoritarianism has a positive relationship with photo sharing online.

Following our previous hypotheses regarding social context, we predict a similar association for photo sharing:

H9: Watching RTV with friends has a stronger positive relationship with photo sharing online, opposed to watching RTV alone.

However, as noted earlier a primary goal of digital photo sharing, like analog photo sharing, may be the development and maintenance of interpersonal relationships. This is especially true of photo sharing on social network sites like Facebook. Competing hypotheses about photo-sharing behavior emerge from this theoretical framework. On the one hand, photo sharing may be linked to traditional media use in the form of RTV. Photo-sharing may also hinge upon an individual's personal social context, consistent with Stefanone & Jang's (2007) results suggesting people with large SSNs appropriate new online communication tools to support existing relationships. Thus,

H10: SSN size has a positive relationship with photo sharing frequency.

Finally, prior research suggests men and women communicate in different styles and for different social objectives (Eakins & Eakins, 1978; Tannen, 1990). Gender differences in terms of self-disclosure have also long been observed (Dindia & Allen, 1992; Morgan, 1976). Females were found to have higher tendency to disclose their emotions than males, particularly to their spouses and lovers (Snell, Miller & Belk, 1988). Dindia and Allen's (1992) meta-analysis concluded that females talked about themselves to their close social networks more than males. There is evidence these differences manifest themselves in cyberspace as well (Gefen & Ridings, 2005; Barrett & Lally, 1999; Hamburger & Ben-Artzi, 2000). These differences suggest that gender should be taken into consideration when studying Internet-based communication tools like blogs, as well as photo and video sharing sites. Given the gender differences in terms of self-disclosure and communication motives, it is expected that females are more likely to share photos online opposed to males. Thus,

H11: Females share photos online more often than males.


Pilot Study

Because some of the dependent variables used in this study were single-item measures, a pilot study was conducted to test their validity in terms of the accuracy of respondent recall. Two variables were selected to test in the pilot study: the size of participants' mediated networks and the number of photographs shared. These items were chosen because objective measurement of these variables is easily accessible by simply viewing online profiles. One hundred students volunteered from an undergraduate class in the U.S. and were given research credit for participating in this pilot study. They were asked to first recall the size of their online networks (on Facebook, for example) and the number of photos of themselves they share on these sites, and report those numbers. Upon collecting these responses, participants were then required to log on to their networking site profile page and record the actual numbers of friends and photos shared.

Fifty-three participants were female. The majority of respondents were Caucasian, and the entire group averaged 19.5 years (SD = 1.87) of age. For the recall data, participants reported an average of 259 network contacts (SD = 202) and an average of 84.4 photos shared (SD = 73.8). Pearson correlation coefficients between the recall and actual data were .64 and .61 for network size and number of photos shared, respectively.

To determine if systematic differences existed in recall based on age and gender, and to test how well recall data predicted actual data, two separate regression models were calculated. The model predicting network size was significant (F, 3, 100 = 33.15, p < .001); the recall variable was a significant predictor (β = .61, p < .001) of actual network size. Similarly, the regression model for number of photos shared was significant (F, 3, 100 = 29.05, p < .001), and the recall data was a significant predictor (β = .59, p < .001). These results suggest that young people are reliably able to recall these specific characteristics of their online behavior.

Main Study

A total of 452 online surveys were completed by a sample of university students enrolled in undergraduate communication courses. T-tests were used to ensure these participants did not systematically differ from those in the pilot study. All participation was voluntary and the University Institutional Review Board for Human Subjects approved all materials. Approximately 58% of the sample was female; the average age of participants was 20.3 years (SD = 2.7). The majority of participants identified their ethnic background as Caucasian (approximately 62%). About 16% were Asian, 6% were African-American, and 3% were Hispanic. The rest (about 13%) identified with a variety of other ethnicities.


Social support network size

Because people have the capacity to accurately identify people with whom they have frequent interaction (Freeman, Romney, & Freeman, 1987; Marsden, 1990), SSN size was measured using a single item that explained the specific characteristics of these affiliations in detail. While responses to this question are not perfect, they do represent people's typical “interpersonal environment” (McPherson, Smith-Lovin, & Brashears, 2006, p. 355). Specifically, the question stated “A strong tie is defined as a person you have known for a long time, have frequent communication with, and positive feelings for. These relationships may include your immediate family members, as well as close friends. How many strong ties do you have?” Several responses were outliers and replaced with the mean score as a conservative correction to the distribution abnormality. Responses ranged from 0 to 50 with a mean of 9.57 (SD = 6.9), which is consistent with extant sociological research measuring SSN size.


This variable was measured by using a set of four questions from the larger 34-item scale developed by Altemeyer (1996), where responses ranged from 1 = strongly disagree to 7 = strongly agree. Because the full scale has been deemed unidimensional such that all items load onto one factor and all items have been demonstrated to equally predict all three characteristics of authoritarianism including conventionalism (Altemeyer, 1996), a subset of questions was chosen to help limit respondent fatigue. Cronbach's alpha for this scale was .78 (M = 2.8; SD = 1.2).

Media use measures

Reality television viewing.

RTV consumption was measured by asking a series of questions. First, respondents were asked how many days during a typical week they watch television. Next, they were asked to indicate for a typical day how many hours they watched news, fiction, nonfiction/educational, and RTV programming. These items are consistent with Salomon and Cohen (1978) who argue this approach is appropriate when measuring time spent viewing television. Overall, participants reported watching about 32 hours of television per week (SD = 28.6). RTV consumption items were prompted with examples of socially rewarding programs like the Real World and American Idol. These variables were used to determine how many hours per week respondents watch different content categories of television. RTV programming was heavily skewed to the right, with a mean of 6.3 hours per week (SD = 9.6). This variable was log transformed to normalize the distribution for further analysis. After the transformation, the mean was .88 (SD = .39).

A list of 36 popular RTV shows was created based on Nielsen ratings for the past 5 years (Appendix), and participants were asked to indicate how frequently they watched each of these shows, how frequently they watched these specific shows with friends, and on average how many friends were present during viewing. This list represents a comprehensive collection of the most popular RTV programming. Again, the sample of shows was restricted to “socially rewarding” content in the sense that participants on these shows are remunerated with positive outcomes in the form of financial gains and attention, and included shows like The Bachelor, Big Brother, Laguna Beach, Making the Band, Real World Denver, and Temptation Island, among many others. Using a scale ranging from 1 (never) to 7 (always), respondents reported their viewing frequency of for each of these popular shows. The distribution for responses regarding the frequency at which shows were watched with friends was skewed to the right (M = 5.6; SD = 6.6). Here, on average respondents reported watching between 5 and 6 shows with friends (after a logarithmic transformation to normalize the data, M = .72, SD = .39). Finally, on average respondents reported watching RTV shows with 2.1 friends (SD = 1.5). The product of these variables was used to create a variable reflecting how often participants watched RTV with friends (M = 14.49; SD = 24.06). Because this variable was skewed to the right, it was log transformed (M = 0.72; SD = 0.39).

Blogging, photo and video sharing.

Participants were asked to indicate whether or not they maintained a personal-journal style blog. Overall, 36 participants indicated they did. Respondents rated the frequency of posting using the following scale: 1 = once a month or less, 2 = a few times per month, 3 = once per week, 4 = several times per week, 5 = once a day, and 6 = many times per day. On average respondents indicated they post to their blog about once a week (M = 3.2, SD = 1.1), suggesting they are generally active in the maintenance of their blogs. Participants were also asked to indicate the number of photos of themselves they have shared via popular online services like Facebook, Flickr, and Photobucket. The distribution of responses to this question was heavily skewed right with a mean of 133 (SD = 185.7) and mode of 0. This variable was log transformed to normalize the distribution for further analysis. Upon transformation, the variable was normally distributed (M = 1.69; SD = .77). Finally, participants were asked to indicate the number of videos they have created and posted online to video sharing sites like YouTube. One hundred participants indicated they had posted videos online. The distribution of responses to this question were also heavily skewed (M = .8; SD = 2.2; Mode = 0).


Table 1, below, summarizes the relationships between variables used in the analyses. Age was negatively correlated with sharing photos online. SSN size had a positive relationship with sharing photos online and watching RTV with friends. On the other hand, authoritarianism had a strong negative correlation with sharing photos online. A strong relationship is evident between general television viewing and RTV viewing, as expected. Finally, RTV viewing in hours was positively correlated with watching RTV with friends.

Table 1.  Zero-order Correlations Between Variables Used
Scale ItemAgeStrong TiesAuthGeneral TVRTV hrsRTV frndPhotoVideo
  1. Note:*p < .01,

  2. **p < .001.

Strong Ties 0.030.05–0.06.12*.18*0.13
Authoritarianism  0.023–0.06–0.02–.22**0.1
General TV   0.78**.15*– .03.093
RTV viewing, Hrs.    .19*–0.02–0.12*
RTV viewing, frnds     0.070.12*
Photo sharing      0.01
Video Sharing       

Recall that hypotheses 1 through 3 posit positive relationships between RTV viewing and the likelihood to maintain personal-journal style blogs. Due to the infrequency of maintaining personal blogs, the first set of analyses uses analysis of variance (ANOVA) to test the hypotheses by comparing bloggers to nonbloggers. First, the log transformation of overall television viewing in hours was used as the between group factor to test whether general television consumption had a relationship to the likelihood of blogging (M = 1.34, SD = .46). Results show that general television viewing did not have a significant relationship with the tendency to blog, F(1,22) = .99, ns. Next, the log transformation of RTV viewing in hours was used as the between groups factor via median split. Heavy viewers of RTV were significantly more likely to maintain a blog, F(1,22) = 2.73, p < .001, support for hypothesis 1. When the analysis was limited to those participants who were high in authoritarianism, there were no significant differences regarding likelihood to maintain a blog. Surprisingly, when the interaction between RTV viewing and authoritarianism was analyzed results indicated that although heavy RTV viewing and authoritarian participants were not more likely to blog, they were significantly less likely to share photos online, F(1,19) = 1.82, p < .05. Hypothesis 2 which proposed that the interaction of RTV viewing and authoritarianism, was not supported. Finally, participants who watched RTV with friends were more likely to blog, F(1,22),= 2.01, p < .01, support for hypothesis 3.

Logistic regression was used to test hypotheses 4 through 6 which address the relationship between RTV viewing and video sharing. Although 100 participants reported sharing videos online, they tended to do so infrequently, so the posting videos variable was dichotomized and coded so that 0 = no and 1 = yes.

Logistic regression analyses were used to test the hypothesized relationships between RTV viewing and the likelihood to share videos. These models are presented in table 2. Model 1 includes RTV viewing as an independent variable, while Model 2 differs in that it includes a general measure of television viewing instead of RTV. The intent is to delineate the relative contributions of both kinds of television viewing regarding online video sharing while avoiding using both independent variables in a single model due to their correlation. The χ2value for model 1 was 16.70 (p < .01), indicating that the model performed well. Additionally, the Cox and Snell R2was 6.5 percent. Model 2 was not significant (χ2 = 12.4).

Table 2.  Results Summarizing Logistic Regression Model Predicting Video Sharing
  1. Note: *p < .01,**p < .001; Model 1 χ2 = 16.7*; Model 2 χ2 = 12.4, ns; CI = Confidence Interval; Female = 1, Male = 2. Interaction in model 1 is between RTV viewing and authoritarianism; Interaction in model 2 is between General TV viewing and authoritarianism.

age0., [0.91, 1.13]–, [0.87, 1.12]
gender0.76**0.335.290.47, [0.25, 0.89]0.69*0.344.051.99, [1.01, 3.88]
watch RTV w/friends0.29*0.099.551.34, [1.11, 1.61]0.09*0.044.471.08, [1.01, 1.16]
RTV viewing, hours0.350.410.721.42, [0.63, 3.17]
General TV Viewing, hours0.650.432.371.92, [0.84, 4.43]
authoritarianism–, [0.77, 1.28]–, [0.76, 1.25]
interaction0.010.1601.01, [0.73, 1.38], [0.61, 1.22]

Hypothesis 4 suggested that heavy RTV viewing would be associated with sharing videos online, but was not supported (B = .35, ns). Further, the interaction between hours spent watching RTV and authoritarianism was not a significant predictor of sharing videos online. Hypothesis 5 was not supported. However, participants who reported watching RTV with friends were almost 1.5 times as likely to post videos online (B = .29, p < .001, Model 1), opposed to those who did not report watching with friends. Hypothesis 6 was supported. It is noteworthy that gender and watching RTV with friends were significant in both models.

Given the popularity of sharing photos online, OLS hierarchical regression was used to test hypotheses 7 through 11 and to control for a set of demographic variables including age and gender. Hierarchical regression was used to test these hypotheses because in this kind of model one can determine the relative contribution of each block of variables beyond the contribution of earlier blocks. Table 3, below, summarizes the results from the analyses testing the relationship between RTV consumption and photo sharing online. All the variables in the model accounted for 15.6% of the variance in predicting the extent participants shared photos online (F,5,399 = 14.42, p < .001). Age, gender, and SSN size were entered into the first block which explained 10.9 percent of the total variance. Age, gender, and SSN size all had significant relationships with photo sharing online. Younger participants, females, and those with larger SSNs tended to share more photos online. Further, those who watched RTV with friends also shared more photos online. These results support hypotheses 7, 9, and 10.

Table 3.  Standardized Beta Coefficients and Adjusted R2 for Regression Model Predicting Photo Sharing Online
 Before entry βFinal β
  1. Note: Model Summary: F(5,399) = 14.42**, Adj. R2 = 15.6; *p < .01,**p < .001; Female = 1, Male = 2.

SSN size0.179**0.150**
Incremental R = 10.9%  
RTV hours0.0250.019
RTV viewing with friends0.124*0.129*
Incremental R = 4.5%  
Interaction term (Auth & RTV hrs)n/a–0.063
Incremental R = .2%  

Hypothesis 8 predicted that the interaction between hours spent watching RTV and authoritarianism would be a significant predictor of photo sharing, and was not supported (β = −.063, ns), although authoritarianism (β = −.181, p < .001) and RTV viewing with friends (β = .129, p < .01) were significant predictors as separate independent variables. Surprisingly, authoritarianism exhibited a significant relationship with the dependent variable, albeit in the opposite direction expected.


This research examined viewer consumption of socially rewarding RTV and found support for the hypothesis that there is a relationship between RTV consumption and behavior online. This finding is consistent with previous research of other television genres (Rössler & Brosius, 2001) and suggests a novel framework for understanding some aspects of social behavior online. Specifically, RTV presents NDSD as a conventional, positive behavior, and can be empirically linked to specific types of online behavior. The three behaviors examined in this study–blogging, photo sharing, and video sharing–exhibit distinct relationships with RTV consumption, authoritarianism, and in some cases the social context of traditional media consumption.

The writing of personal, journal-style blogs is an active endeavor which requires the author to reflect on his or her life and how to frame that experience in text. The decision to disclose personal information to a public forum may be linked with a broad range of cultural phenomena including celebrity, consumerism, and structures of self-actualization. Cultural products like RTV appear to share many characteristics and values with journal-style blogs. The relationship between authoritarian personality and quantity of RTV consumed per week, and the decision to maintain a journal-style blog was examined.

We identified a significant association between the quantity of RTV consumed and the decision to maintain a blog. This relationship was consistent with our expectations. RTV presents the disclosure of personal information as a normal and prosocial behavior. Characters in reality television are expected to make their motivations and personal thoughts known to the audience, and this disclosure may be highly structured within the program (isolated “confessional” disclosures) or less structured, in the form of explicit dialog and interaction. Heavy viewers of RTV become familiar with the genre's previously novel convention of ongoing, long-term personal disclosures to a mediated audience. Personal blogging is also a disclosive, episodic, and potentially abstracted media behavior which correlates with RTV viewership.

Video sharing most closely mimics the “celebrity” aspects of RTV, as it allows an individual to broadcast videos of intimate thoughts and behavior. The popularity of sites like YouTube is a testament to the fact that many people are willing to indulge the desire to broadcast themselves. Evidence from the current research supports the notion that heavy viewers of RTV would be more likely to participate in video sharing; both the time spent watching RTV programs and the number of different RTV programs watched were significantly associated with video sharing. Sharing videos allows individuals to actively participate in the voyeuristic and celebrity aspects of the culture.

There is further evidence that RTV influences video sharing behavior in the sense that those who watch RTV with friends are more likely to share videos. It may be the case that RTV is further normalized by the proximity of “real” others, and such RTV consumers may be subject to increased pressure to model prominent behaviors such as NDSD online.

The posting of photos represents a different level of personal disclosure, and perhaps an entirely different dimension. Where a blogger has a great deal of control over the symbolic content of and can shape a text to meet their communicative needs with relative ease, photos present iconic information about an individual. While a photographer can manipulate the context and presentation of a photo to a great extent (to say nothing of the potential of digital editing), a photo still presents something essential about its subject. Whereas a blog need only represent thoughts, photos represent bodies which may be a more intimate type of disclosure in the online environment.

Photo posting is a common behavior among the sampled population. Two demographic factors were significant predictors of photo posting. Age was a significant negative predictor of photo posting even given the limited variation in age among our undergraduate student sample. It seems that youth culture is embracing this behavior. Further, females were more likely to post photos than males. The female image maintains strong connotations in our culture, and this finding may indicate that young females are following broader cultural prescripts about image use.

SSN size was also a significant predictor of photo posting behavior. This points to a highly social function of this behavior as those who are socially active may post photos of themselves and their friends to show support and solidarity. Posting photos can contribute to relationship maintenance and may also promote reciprocal behavior. The high numbers of photos shared by some participants (many hundreds, in some cases) points towards the value that they attribute to sharing personal images.

Unlike the other two disclosure behaviors measured in this study, blogging and video posting, we found no association between RTV consumption and photo sharing. Our data indicate that photo sharing may be a more fundamentally social and relational activity than the other two behaviors. The context of photo posting–young, female individuals with large social support networks–drives photo posting to a greater extent than external media influences.

The authoritarian personality trait interacted with the other independent variables in unexpected ways. Although we anticipated that an authoritarian outlook would serve to promote the reproduction of media behaviors that were presented as normative, we did not find this to be the case with RTV viewing and the sharing of personal media online. The only significant interaction of authoritarianism and RTV viewing was found with regard to photo sharing, and this was in the opposite direction as predicted–photo sharers were found to be less authoritarian than those who do not share photos online. This may follow from the fact that high authoritarians may have more conventional understandings of personal privacy, and may be reluctant to make photos available for public perusal, even when exposed to RTV's messages regarding exposure.

Limitations of this study include the small proportion of participants who reported maintaining a personal-journal style blog. Although significant differences were found, our results would be strengthened by a larger sample of active bloggers. The small sample of bloggers in this study limited the analyses to relatively simple means comparisons. Also, in this study maintaining personal-journal style blogs was operationalized as a form of NDSD. We acknowledge that bloggers target specific audiences typically comprised of people they have meaningful relationships with. However, when bloggers post content their behavior is consistent with NDSD because they are targeting groups of others, as opposed to targeting messages at specific others. Further, these blogs are overwhelmingly public forums for personal self-disclosure as they rarely restrict access.

Similarly, participants in this study tended not to post videos of themselves online. The video sharing variable was dichotomized for use in logistic regression analyses which limits our understanding of this behavior. While an understanding of the content characteristics of videos shared online would add clarity to the results, this undertaking was beyond the scope of the current research but would be a productive area for future scholarship.

The current research also relies on a broad characterization of socially rewarding RTV content. While the list of shows selected for this study was comprehensive and spanned several years of programming, formal and detailed content analyses was not done on this corpus of material. However, the list of RTV shows used all exhibited specific characteristics consistent with our operationalization of programming that models nondirected disclosure and a culture of celebrity. Future research would benefit from a more comprehensive and systematic analysis of behaviors modeled in this genre of television.

Additionally, the data do not allow for statistical control over how much time participants' spend with their friends, generally. There may be a connection between time spent engaging in social activities with friends offline and the dependent variables used in this study.

In the future, we hope to address some of these limitations and explore further some of the findings. For example, the authoritarian scale used in this study, while backed by decades of significant previous research, proved to be less reliable than in previous studies. Many of the items address traditional concepts that may not reflect conventional attitudes in this population. Attitudes toward homosexuality, for instance, have changed dramatically in recent decades, especially among younger generations. Traditional negative attitudes towards homosexuality may no longer be conventional among college-age populations. We hope to revisit and revise the authoritarianism scale to more accurately gauge conventionalist traits in this population. Further, it will be useful to control for enacted privacy behaviors–e.g. restricting access to personal content, efforts to maintain anonymity or pseudonymity–that may result from privacy concerns.

Among the most promising findings in this study were the strong associations between gender and certain online disclosive behaviors including video and photo sharing. Contrary to earlier theories about the decreasing influence of gender in Internet-mediated environments, gender is a powerful predictor of some online behaviors. Finally, there are intuitive links between the culture of mediated voyeurism discussed here and celebrity culture (e.g. McCutcheon, Lange, & Houran, 2002). Video sharing sites such as YouTube, and live video streaming sites such as Justin.TV ( increasingly allow anyone access to “stardom” in ways similar to that of traditional celebrities.

About the Author

Michael A. Stefanone is an assistant professor in the Department of Communication at the State University of New York at Buffalo. His research focuses on the intersection of people, organizations, and technology. His current research explores the relationship between traditional mass media and new media use and how people's social context influence technology adoption and use.

Address: 359Baldy Hall, University at Buffalo, Buffalo, NY 14260-1060, USA

Derek Lackaff is a doctoral candidate in the Department of Communication at the State University of New York at Buffalo. His research explores the impacts of technological mediation on communication processes and social structure. His present research focus is the effects of communication technology use on social support structures during a life transition.

Address: 359 Baldy Hall, University at Buffalo, Buffalo, NY 14260-1060, USA

Appendix A

List of Popular Reality Television Series

  • American Family

  • Airline

  • Amazing Race

  • America's Got Talent

  • American Idol

  • The Apprentice

  • The Bachelor

  • Big Brother

  • The Biggest Loser

  • The Contender

  • Dancing with the Stars

  • Makeover: Home

  • Fear Factor

  • Gold Rush

  • Hell's Kitchen

  • I Love New York

  • Intervention

  • Laguna Beach

  • Making the Band

  • MTV Cribs

  • The Mole

  • Project Runway

  • Nashville Star

  • Paradise Hotel

  • Project Greenlight

  • Queer Eye for the Straigh Guy

  • Real World Denver

  • Real World Challenge

  • Road Rules

  • Rock Star

  • The Simple Life

  • Starting Over

  • The Surreal Life

  • Survivor

  • Temptation Island

  • Top Chef

  • Top Model