Volume 21 Numbers 1 & 2, 2011
Journalists, social media, and the use of humor on Twitter
Avery E. Holton
Seth C. Lewis
In much of the developed world, news consumption across traditional platforms such as print newspapers and broadcast television has been on the decline for years (Chyi, 2009), forcing news organizations to rethink the way news is organized and distributed in an increasingly hyper-competitive digital environment. In this search for better ways to gain and maintain consistent followers, legacy news organizations generally have struggled to keep pace with changes in technology, consumer preferences, and cultural habits (e.g., see Chyi, 2009; Mindich, 2005; Napoli, 2010). This comes at a time when the internet has given rise to novel applications for social networking and information sharing (Harrison, 2009; Palloff & Pratt, 2009; Shifman & Blondheim, 2010; Thelwall, Buckley, & Paltoglou, 2010). The explosive growth and success of social networking sites (SNS) such as Facebook and Twitter, which together claim more than 700 million users worldwide (Facebook, 2011; Smith, 2011), suggests that news media organizations could learn something from these efforts—from the nature of these social media—as they work to cultivate a following online.
Indeed, journalists have taken notice. Many have made social media a normal part of their practice (Hermida, 2010a; Thurman & Hermida, 2010), using social tools to facilitate the process of gathering, filtering, and distributing news, and in driving traffic to their parent news sites (Hermans, Vergeer, & Pleijter, 2009; Lariscy et al., 2009). For online users, too, social media have made it increasingly likely that everyday people can work alongside journalists, as it were, in contributing to the flow of news and information (Braun & Gillespie, 2011). Yet, questions remain about the extent to which journalists, in fact, are rethinking their relationship with active users (Singer et al., 2011), and the degree to which the nature of news presentation has ch anged in the social media context. With so much information available through so many outlets, today’s news and information consumers may indeed want more than just news (Boczkowski, 2010; Purcell et al., 2010); they may want, as Hermida (2010a) has suggested, to feel part of the news and information ecology, having a more personal connection with news messages and those journalists behind them. At least one study has illustrated this point. In an analysis of the most prominent journalists on Twitter, Lasorsa and colleagues (2011) found that journalism practices on Twitter are shifting to include more non-traditional elements, including the sharing of opinion and personal information and more active engagement with fellow users, pointing to an emerging news process in which journalists and publics are more interconnected in real time.
This study explores that question of connectivity. Academic literature has highlighted the motives and uses of SNS such as Twitter (Chen, 2010; Coursaris, Yun, & Sung, 2010; Java, Song, Finin, & Tseng, 2007), pointing to the psychological desire for connection. Yet, the literature has not fully clarified just how such a connection is fulfilled—nor, in particular, how that connection might occur in the context of journalism as journalists modify their traditional norms and practices to develop a stronger relationship with online audiences (Lasorsa, Lewis, and Holton, 2011). This paper argues that a potential point of connection is the use of humor, which is an important facet of online communication (Shifman and Blondheim, 2010), and which in recent years has played a key role in gen erating large followings for fake news shows on television. Programs such as The Daily Show with Jon Stewart and The Colbert Report have attracted viewers by weaving humor and satire into general news and information (Fox, Koloen, & Sahin, 2007). Fox et al. (2007) found that news messages conveyed using humor on The Daily Show were better received by viewers than similar news messages on traditional news broadcasts. If humor can help facilitate news and provide a sense of connectedness, it is reasonable to consider that journalists might benefit from its incorporation.
This paper examines the extent to which journalists use humor on Twitter, so as to assess how this use of humor is (1) associated with other forms of Twitter activity generally and (2) related to indicators of connectivity in particular. Twitter is important because of its public nature, such that journalists on Twitter are acting under the aegis of their news organization’s brand when they post information and discuss with fellow users, and also because it represents a crucial space for experimentation, innovation and participation in journalism (Hermida, 2010b). The question becomes: In their use of Twitter, to what extent are journalists using humor, a non-traditional news element that might allow them to better connect with online audiences?
More than 90% of Americans now get their daily news from more than one source, using an average of four to six platforms (Purcell et al., 2010). Increasingly, those platforms are accessed through lightweight mobile devices such as tablet computers and cell phones, making the incorporation of digital news a pivotal component of today’s more successful news organizations. Many of those organizations have welcomed social media as portals of news gathering, processing, and distributing information, developing presences on Facebook and Twitter that allow them to stay plugged in to the changing needs of consumers—consumers who now demand around-the-clock access to news (Hermida, 2010b). In this context, the emerging question is to what degree journalists alter traditional styles of newswork (or not) to meet the changing dynamics of technological platforms and user expectations associated with them (Lasorsa et al., 2011; cf. Singer 2005).
The recent explosive growth of Twitter—a micro-blogging service that allows for users to instantly share fragments of news and information with other users—has prompted a wealth of scholarly research, much of which has focused on the why of user engagement (e.g., see boyd, Golder, & Lotan, 2010; Chen, 2010; Coursaris et al., 2010; Johnson & Yang, 2009). Research suggests that Twitter use is closely associated with a desire to connect with others, but as yet scholars have not thoroughly explored how key news producers might be altering their routines—which historically have emphasized dispassionate distance from the audience (Singer et al., 2011)—to meet such needs. Meanwhile, humor has been employed to great effect on TV news programs for some years now, but its use as a tool for connection on SNS has yet to be examined. This study proceeds by discussing the rise of Twitter as a reporting tool and how journalists might be employing a non-traditional practice—the use of humor—to suit the Twitter environment.
Twitter as a reporting tool
Since launching publicly in 2006, Twitter has grown into a global news-and-information resource used across myriad platforms by more than 200 million people (Ionescu, 2010; Smith, 2011). Twitter is a public site that invites users to post messages known as “tweets.” Messages of 140 characters or less are sent out on the user's personal site and to the user’s list of “followers,” who are users who have chosen to receive those messages. Users have the capability to publicly reply to or direct messages toward other users (i.e., @username), to “retweet” (or re-post) messages posted by others, and to observe or engage in trending topics (i.e., #trendtopic). Twitter users may also use multiple forms of media content, including text, links, images, and videos. All of these characteristics help users stay connected, share information, extend knowledge, create and disseminate content, an d collect sources that benefit their interests (Kaplan & Haenlein, 2010; Skoler, 2009).
Despite restrictions enforced by some news organizations, journalists are increasingly using Twitter to break news, share information, search out sources, engage with other users, promote their own work, and comment on the work of others (Ahmad, 2010; Farhi, 2009; Hermida, 2010a, 2010b). Given the continued rise in activity on Twitter, which drew more than 4,000 messages per second in the hours following the announcement of Osama bin Laden’s death in May 2011 (Kanalley, 2011), many news outlets have asked their reporters to use Twitter as a means of connecting people to their coverage (Gleason, 2010; Tenore, 2007). Journalists appear to be heeding the call, as their use of Twitter has grown substantially (Baym, 2010; Hermida, 2010b ; Thurman & Hermida, 2010). Meanwhile, news audiences have likewise adopted Twitter, whether for following news organizations (e.g., The New York Times, with 3 million followers) or individual journalists (e.g., The Times’ Nick Kristof, with 1 million followers).
The relative ease of creating and sharing content through Twitter allows individuals to observe and interact with information like never before (Stassen, 2010). Users are increasingly turning to SNS such as Twitter for reliable information (Jansen, Zhang, Sobel, & Chowdury, 2009), to break news or share breaking news posted by other users, and to offer a new form of live coverage, or “eyewitnessing” (Kawk, Lee, Park, & Moon, 2010). Twitter is blurring the lines between journalists and citizens (Bruns, 2008; Purcel et al., 2010; cf., Rosen, 2006, 2008). Such a distortion of the traditional producer-consumer relationship is worth exploring, and could have a significant impact on journalists as they try to stay connected with the wants and needs of the public.
Already, Twitter contributes to what Hermida (2010b) describes as “ambient journalism.” Thanks in large part to digital and mobile technologies and outlets, journalism is no longer constrained by time, textual limitations, or traditional editorial processes, but rather is omnipresent and without bounds. Journalists and non-journalists alike now have the opportunity to contribute to the news creation and dissemination process 24 hours a day, engaging in an increasingly fluid and ambient news process. Yet, as reporters work to incorporate Twitter into their daily routines, they are also adding to the evolution of journalism routines, which may be fueled, at least in part, by connections and relationships between journalists and the individuals who choose to follow them on SNS such as Twitter (Skoler, 2009). Thus, it is important to consider elements beyond news and information sharing that may link journal ists to the public. Daily chatter and conversation play a large role in Twitter use (Java et al., 2007), affecting user expectations and possibly changing the way journalists approach the medium.
Evolving journalistic practices
For journalism and its professional moorings, the incorporation of Twitter (Lasorsa et al., 2011) and other forms of engagement with users (Braun & Gillespie, 2011) points to a larger evolution occurring as the field learns to adjust not only to a digital environment but also to a truly networked one (Singer, 2010). This negotiation of networked modalities online raises both philosophical and practical questions for newswork (Deuze and Marjoribanks, 2009; Lewis, Kaufhold, and Lasorsa, 2010), and plays out both in the way that journalists adopt new technologies such as blogs (Hermida, 2009; Phillips et al., 2009; Robinson, 2006; Steenson, 2010) and in how they think about the implicat ions of digital production for traditional norms of professional authority and control (Mitchelstein and Boczkowski, 2009; Singer et al., 2011).
As journalists increasingly turn to Twitter, they also help construct norms of its application in the profession. For example, the Rocky Mountain News posted coverage of a child’s funeral to Twitter, drawing criticism that forced both the publication and the industry to consider acceptable and unacceptable uses of Twitter (Arceneaux & Weiss, 2010). Octavia Nasr, a reporter for CNN, lost her job after posting a message to Twitter indicating her sadness over the death of a Shiite cleric (Ingram, 2010). Twitter also has ignited debate over whether journalists should offer opinions and other non-traditional information via the medium (Ingram, 2011; Poniewozik, 2010). A May 2011 report from the American Society of Newspaper Editors suggested that journalists should approach Twitter with traditional ethics in mind, avoiding opinion or other bits of information that don’t provide news or context (Hohmann, 2011).
Professional norms of journalism make it clear that journalists are expected to provide timely, unbiased information while keeping their opinions to themselves (Kovach & Rosenstiel, 2001; Kovach & Rosenstiel, 2010; Wolfsfeld, 2004). The pursuit of truth and objectivity remains a cornerstone of journalistic integrity (Schudson and Anderson, 2008), even in light of the challenges presented by digital media (Robinson, 2006). Thus, it becomes important to understand if and how journalists are deviating from these norms of neutral, staid, and buttoned-up approaches to storytelling in their use of humor via social media spaces such as Twitter. In the most comprehensive study of “j-tweeters” to date, Lasorsa et al. (2011) found that journalists express opinions more freely on Twitter, write more often about their lives and their jobs, engage in direct conversations with users, and share user-generated content—all which stand in contrast to traditional notions of news content as free of opinion or personal information, let alone such engagement with audiences and their content. While the study did not directly answer why journalists are modifying traditional norms, it did confirm the evolving nature of journalism practice on SNS. This study attempts to build on that work by singling out a key component of the changing nature of journalism on Twitter and examining its possible use as a new way to connect with news consumers.
Connecting through humor
Social networking sites provide a number of online and offline functions for users, including information sharing, emotional support, and social connectivity (Baltaretu & Balaban, 2010; Chen, 2010; Joinson, 2008). SNS such as Facebook and Twitter help people stay connected, develop identities, find content, investigate others, search for new connections, and update the statuses of their own lives (Joinson, 2008). By offering new ways for people to connect without restrictions of time and place, SNS provide a social environment where members of multiple societies can converge, connect, and build relationships like never before (Stafford & Stafford, 1998; Stafford, Stafford, & Schkade, 2004). Chen (2010) argued that SNS such as Twitter provi de para-social structures for people, offering them a place to belong, to share, and to connect—the last of which is of particular interest to this paper.
In order to achieve and maintain connections in a digital environment, at least one researcher has indicated the need to consider the role of sentiment. In a study of more than 34 million messages posted to Twitter, Thelwall et al. (2010) noted an attachment of sentiment to general news events. These attachments—typically positive or negative in nature—could play a role in what news and information people seek, whom they seek it from, and how likely they are to return to that source. The public, especially young adults, tend to seek out news and information through sources that weave in entertainment, often in the form of humor (Feldman, 2007). Younger people, who represent the news consumers of today and tomorrow, also seek out satire, irony, parody, and comedy—all constructs of humor—for a number of reasons, including social acceptance, relaxation, and connectivity (Calavi ta, 2004). Collectively, humor fulfills a social need to connect by helping convey emotions and knowledge and seal bonds between people (Martin, 2007). However, the role of humor as a connective agent through Twitter, let alone other SNS, has yet to be explored, possibly because of its abstract nature. Because humor means so many different things to so may different people, it can be both difficult to define and yet easily comprehended and appreciated from the vantage point of those using it within a common environment (Chapman & Foot, 1995; Fry, 2010).
Humor has long had a place in journalism, traditionally showing up in less-structured formats such as newspaper and magazine columns and television commentary (Weaver & Wilhoit, 1996). More recently, fake news programs such as Saturday Night Live and The Daily Show with Jon Stewart have engaged the public, weaving humor with traditional news offerings to draw millions of regular viewers (Farhi, 2010; Feldman, 2007; Fox et al., 2007). While these programs are often criticized for confusing fact with humor, at least one study showed The Daily Show provides just as much informative news as televised network news broadcasts, and may in fact be better received by those intimidated or turned off by more serious programs (Fox et al., 2007). Feldman (2007) connec ted this finding more narrowly to news, arguing that a substantial number of people now get their daily news from comedy programs. However, no studies to date have examined how journalists might be incorporating humor into their existing practices on emerging platforms like Twitter. Given the ability of humor to connect individuals across societies (Forester, 2004; Sultanoff, 2003; Welch, Maiuri, & Poleshchuk, 2010), it is important to note how humor functions in the communication process, especially considering the evolving communication patterns brought on by technology.
Humor is often shared among groups of people to help them connect and relax (Morreall, 1983). Research has examined such benefits as applied against stress and loneliness (Booth-Butterfield, Booth-Butterfield, & Wanzer, 2007), in advertising and corporate communication (Weinberger & Gulas, 1992), in healthcare settings (Wrench & Booth-Butterfield, 2003), and in the political arena (Baumgartner, 2007; Fox et al., 2007). From a communications approach, messages of humor often require individuals to have at least some prior knowledge of the subject matter being discussed are more readily accepted by those with shared knowledge and shared emotional constructs who can collectively decipher their meanings (Chiaro, 1992; Oring, 1992). Taken out of context or without some sort of prior knowledge, humor can be perceived as irrelevant, hurtful, or tactless (Carrell, 1992; Duncan, 1982; Hackman & Barthel-Hackman, 1993). As such, humor can either unify or divide, helping construct and deconstruct identities (Baym, 1995), an important consideration as media organizations continue to try to maintain and build connections with news consumers.
Applied to communication settings where connectivity is key—such as in the sustainability of journalism platforms through continued relationships with news and information consumers—humor allows individuals to connect with other individuals, improving group cohesion and allowing people to decide for themselves whether or not the humor of another matches up with their own (Graham, Papa, & Brooks, 1992; Gruner, 1985). Humor, as Meyer (2000) observed, allows an individual to move closer to an audience by providing a bridge between the two. More broadly, humorous people or those who employ humor strategically, often represent the most popular or most endeared individuals in society (Wanzer, Booth-Butterfield, & Booth-Butterfield, 1996).
Humorous messages are often better received than more serious ones (Gruner, 1970; Markiewicz, 1974), making the argument for news augmented by humor (e.g., The Daily Show) one worth considering. Messages reinforced with humor also have the capacity to enhance or persuade attitudes (Markiewicz, 1974), making them powerful tools in the consideration of messages framed or presented by the media, which has long been noted for setting agendas and guiding public perception (McCombs & Shaw, 1972). Along with influencing the salience of a news topic, journalists also have the potential to affect how the people understand or consider a subject (Scheufele, 1999; Scheufele & Tewksbury, 2007), making the element of humor important to consider. Yet, humor constructed through digitiz ed communication has not been adequately explored (Shifman & Blondheim, 2010).
The purpose of this study is to assess the extent to which journalists use humor on Twitter, so as to assess how this use of humor is (1) associated with other forms of Twitter activity generally and (2) related to indicators of connectivity in particular. Previous research (in particular Lasorsa et al., 2011) has identified that journalists, in some sense, have adapted their behaviors to suit Twitter’s cultural milieu—e.g., of sharing opinions, talking about personal/job details, and linking to others’ content—while at the same time “normalizing” the new medium to fit professional norms, much as Singer (2005) found in her study of early blogging by journalists. A particularly interesting finding of both Singer (2005) and Lasorsa et al. (2011) is that journalists from more prestigious or elite media organi zations, which already have large audiences, are less likely to engage in non-traditional behaviors, whether on blogs or microblogs. They’re more apt to stick to business as usual, whereas journalists from less well-known news organizations—who perhaps must work harder to get attention—are relatively more inclined to take advantage of the dialogue and engagement with users afforded by digital media. Thus, it would be useful to understand how elements of Twitter use—not only the content of tweets, but also the relative intensity of Twitter use—and the prominence of the “j-tweeters” themselves might be associated with humor and its potential for attracting and maintaining a following. Against the backdrop of these considerations, this study poses the following research questions:
It is impossible to obtain a comprehensive list of all journalists using Twitter, let alone a list of English-language journalists who have signed up for the service. Although a random sample would have been ideal, the researchers obtained a suitable substitute: a list of the 500-most followed journalists on Twitter, made available by Muck Rack (see muckrack.com), a website that aggregates tweets of professional journalists, mostly from the United States. Muck Rack’s database includes information on thousands of journalists on Twitter, organized by organizational affiliation and beat expertise (Seward, 2009). The database allows for journalists to be sorted according to number of followers. This sample of the 500 most-followed journalists, while not representative of all journalists on Twitter, nevertheless represents perhaps the best compilation available.
The sample derived from Muck Rack included each journalist's organizational affiliation and number of followers as of September 2009. At that time, George Stephanopoulos of ABC News was atop the list with 1,224,118 followers, while Andy Newman of the New York Times was No. 500 with 690 followers. Since some journalists closed or changed their Twitter accounts after September 2009, and others did not use Twitter during the time frame of this study, the actual number of journalists studied was 430.
To gain a content sample, the first 10 tweets the journalists posted each day were coded, starting at 12 a.m. and ending at 11:59 p.m., for two weeks, starting on Oct. 5, 2009 and ending on Oct. 18. The tweet was the unit of analysis. If a journalist posted fewer than 10 tweets on a given day, those that they did post were coded. Only the first 10 tweets per day were coded, primarily for logistical reasons (e.g., to keep particularly active journalists from overly influencing the results). Ultimately, 22,248 tweets were coded.
Using the primary purposes and dominant content of journalists on Twitter as identified by Lasorsa et al. (2011), each tweet was first analyzed for its primary function—to convey information, to seek information, or to offer opinion.
Those that primarily conveyed information were coded for the presence of absence of job talking, personalizing, and lifecasting. Job talking included information about a journalist’s job, but excluded self-promotion. Personalizing included messages about or pertaining to the journalist’s life. This differed from lifecasting, which included mundane information about a journalist’s everyday life, such where she was having lunch or what she was doing after work.
Tweets seeking information directly asked for responses about certain topics. For instance, a journalist working on a story about antique cars might have sent a message out saying, “Anyone with a ‘59 Mustang? Tell me about it.” Additionally, each tweet was coded for the presence of discussing and retweeting, both unique activities encouraged on Twitter. Discussing is similar to “in reply to” and usually includes a public reply directed at another user (ex: “text @otheruser”). Retweeting is similar to discussing, but does not necessarily imply a conversation. Rather, retweeting is a message that contains the original message of another user, typically preceded by a comment on that tweet (e.g., “Cool topic RT @username Dug into holistic medicine today”). Because tweets often contain fragments of news and information, they sometimes point other users to more information using hyperlinks. Coders noted the presence or absence of linking.
Tweets that were judged as primarily conveying opinion were labeled as cases of “major opining.” For example, the following tweet would be regarded as major opining: “Microsoft buys Skype? Just great. Another useful service doomed to failure at the hands of Microsoft.” In addition, tweets that did not primarily opine but instead primarily conveyed information were coded in terms of whether they nonetheless contained an element of opinion. Such tweets were labeled as “minor opining.” Here’s an example: “Covering Microsoft’s announcement on the Skype deal. Hope Steve Ballmer says something newsworthy this time.”
Finally, all tweets were coded for evidence of humor. Humor included any attempt by the journalist to be funny, regardless of the purpose of the tweet. To assess the use of humor in journalists’ Twitter postings, coders were asked to assess the following: “Regardless of what else the journalist is doing (seeking information, stating an opinion, or conveying information), is the journalist trying to be funny? Yes or no.” For this and the other variables described below, the number of tweets coded as “yes” for a particular journalist were summed and divided by the total number of tweets that same journalist posted during the period in order to produce a proportion figure (e.g., Journalist A expressed humor in 15% of her tweets).
Beyond coding for the above forms of Twitter activity (see RQ2), coders also gathered information on measures of engagement and prominence on Twitter—elements that might indicate some evidence of connectivity with followers (see RQ3). By visiting the Twitter profile page for each journalist (e.g., twitter.com/Pogue for The New York Times’ David Pogue), coders recorded the following information: the number of followers; the number of other Twitters the journalist was following; the number of times the journalist had been listed on a Twitter list organized by other users (usually around a theme); the journalist’s total number of days on Twitter; the journalist’s total tweets; the journalist’s total number of tweets over the 2-week period (i.e., all tweets posted during the coding per iod, and not just those actually coded; see above for this distinction). Finally, with this information, a score of relative productivity for each journalist was calculated by taking her total tweets divided by total number of days on Twitter.
To gauge whether journalists working for different news media might differ in their microblogging activities, journalists affiliated with national newspapers, the news divisions of the major television broadcasters, and the cable television news channels (47.6 percent) were roughly grouped together as “elite” news media, and they were compared to those working for the other news outlets (e.g., as in Singer, 2005). In this analysis, the dichotomous variable was labeled prestige, where 1 = a more elite news organization and 0 = a less elite news organization.
Coders were students in an undergraduate mass communication course who received credit for their work. These 60 coders engaged in extensive practice sessions involving all coders coding the same tweets both as class exercises and homework assignments, which then were discussed in class. In addition, six graduate students conducted supplemental coding to address gaps left by the undergraduate coders. Intercoder reliability was determined by selecting a subset of 488 pairs of tweets that were independently coded by two different coders. Cohen’s kappa was used to estimate intercoder reliability (Cohen, 1968). This statistic is a more conservative measure of reliability than some other measures, including percent of agreement, because it does not give credit for chance agreement. Thus, a kappa value of .80 represents very high intercoder reliability and a value of .60 represents acceptable intercoder reliability (Viera and Garrett, 2005). The reliability estimates for the variables used in the analyses are: humor, .63; discussing, .86; linking, .66; job talking, .71; lifecasting, .79; prestige, .82; minor opining, .67; major opining, .77; personalizing, .62; retweeting, .80; and seeking information, .85.
Because of the heavy skewness and kurtosis of many of the variables—particularly metrics such as number of followers, which can range from a few thousand for some journalists to several million for celebrity journalists—all of the interval-level measures analyzed here were first recoded as five-part binned variables (i.e., equally sized groupings of Lowest, Lower, Medium, Higher, Highest). These new ordinal variables were examined using Pearson correlations.
The journalists included sent an average of 5.6 tweets per day. That number varied widely, with some sending no tweets at all and one sending an average of almost 60 a day. Two-thirds (67%) of the journalists were male. Twenty-seven percent worked for national newspapers, 21.8% for local newspapers, 15.2% for magazines, 10.8% for national television broadcast networks, 9.8% for cable news networks, 9.4% for online news sites, 2.2% for radio stations, 2% for wire services, and 1.8% for other news outlets. A descriptive profile of the journalists in the sample is displayed in Table 1.
The results show the extreme variation in the extent to which these most-followed journalists used Twitter—with some not posting anything during our two-week coding span, while one journalist notched up to 810 tweets. There is wide variance, too, in the number of followers that a given journalist has, and in the number of people that journalist, in turn, is following. Additionally, the number of times that a journalist on Twitter is included on a “list” is an indication of the extent to which she is known by other Twitter users and categorized as such—and, again, this measure also shows a wide range of scores. Finally, the change over time in the number of followers from September 2009 to March 2010—from a mean of 32,260 to a mean of 52,133 followers for these journalists, or a 62% increase—can be taken as one more indicator of Twitter’s explosive growth in recent times.
RQ1 sought to address a baseline question: To what degree are journalists attempting to be funny in their tweets? Overall, 22.5% of all tweets were coded as seeking to convey humor. However, the extent to which individual journalists used humor varied dramatically: the median journalist used humor about 17% of the time; some 78 journalists (18.1% of the sample) used none in the course of the coding period; and only 10 journalists convey humored in more than 80% of their tweets.
Table 2 put this use of humor in perspective by comparing it to the frequency with which journalists engaged in other types of Twitter activity. From this it is apparent that humor was one of the more frequent activities overall. While it is unsurprising that nearly half of all tweets included a link—as linking is commonplace on Twitter—it is noteworthy that humor featured so prominently among other forms of use.
RQ2 attempted to take this a step further by exploring how the use humor might be associated with other types of Twitter use—namely, those activities outlined in Table 2. This is important for understanding how humor is related (or not) with other forms of engagement. The results of a zero-order Pearson correlation analysis are in Table 3.
Humor was found to be significantly associated with all of these common forms of Twitter activity, with the exception of talking about one’s job. However, it’s important to note the relative strength of these associations: for example, both major (r = .311, p < .001) and minor (r = .313, p < .001) types of opinion are the most closely related to humor, followed by lifecasting (r = .282, p < .001) and personalizing (r = .264, p < .001). These four variables, the strongest in connection with humor, suggest that the use of humor may be particularly related to connecting on a personal level: sharing one’s viewpoint, talking about outside-of-work life details, and altogether focusing on topics that are personal in nature. The significance of linking (r = .219, p < .001) might point to the frequency with which Twitter users generally point to humorous video clips and other such things online. A number of studies have shown that humor is most of ten used when discussing others, so perhaps these links concern other people (Martin, 2007). Notably, the significance of seeking information (r = .216, p < .001), retweeting (r = .136, p < .01), and discussing (r = .128, p < .01) may suggest another form of connectivity through humor. All three variables represent virtual conversations or exchanges, which might be best accomplished through personalized elements such as humor (Baym, 1995; Martin, 2007).
RQ3 sought to explore the extent to which this use of humor is associated with measures of engagement and prominence on Twitter—in essence, to go beyond the cluster of activities outlined in Table 3 and actually measure the extent to which humor use is related to factors such as a journalist’s history on Twitter and his potential reach with audiences (i.e., number of followers). A series of zero-order and partial correlations were calculated (see Table 4). Humor use was positively correlated with total tweets, or having a high volume of tweets overall (r = .195, p < .001); with tweeting more during the two-week coding period in particular (r = .132, p < .05); and with productivity, or a measure of journalists’ aggregate tweets divided by the number of days they’ve been on Twitter (r = .183, p < .001). However, in the strongest finding of this group, humor was negatively correlated with the prestige of a given journalist’s news organization (r = -.259, p < .001); this indicates that journalists at the elite news outlets (e.g., the broadcast news networks) attempted to be funny less frequently than their less prestigious counterparts in the sample.
Looking beyond humor use, it’s clear that the number of followers a journalist tends to garner is strongly associated both with consistent activity (see total tweets, tweets over the 2-week period, and productivity) as well as the immediate branding offered by working for a major national news organization (r = .272, p < .001). This built-in prominence, in particular, makes it possible for journalists who do rather little with their Twitter profiles to gather massive follower counts simply by virtue of their celebrity; this is especially true in the case of TV journalists working for networks, such as CNN, that make a point of highlighting their staffers’ social media presence.
Because of the potential for this element of prestige to obscure the real associations here between a type of activity (humor) and indicators of engagement, a partial correlation was run to control for this effect. The results in Table 4 show that rather little changed: The use of humor remained unrelated to the number of followers, followees, and “listed” mentions that a journalist attained, and has an insufficient connection with the overall number of days that have elapsed since a journalist signed up for Twitter. Likewise, the association between humor and measures of activity—total tweets (r = .177, p < .01), tweets over the 2-week period (r = .130, p < .05), and productivity (r = .162, p < .01)—remained significant and positive. These findings seem to suggest a few key points: (1) that just “being funny,” by itself, has no particular connection with one having more (or fe wer) followers; (2) that being an early adopter of Twitter also has little association with making humorous use of the microblogging medium; but that (3) being an active member of the Twitter community is related to the extent to which a journalist is likely to use humor. This final point would indicate that, to the extent journalists immerse themselves in the culture of Twitter, they are more likely to step outside their traditional, serious persona and adopt some of the interpersonal humor and flavor of social media.
While some research has indicated the possibility of humor to promote news consumption, those studies have primarily focused on traditional media formats such as television. This study contributes to the literature by noting that, as journalists grow accustomed to conducting newswork in social media spaces, they may indeed be refining their professionalism to make room for emerging forms of engagement with users—a kind of “ethic of participation” (Lewis, 2010; see also Lewis, in press) that both encourages and accommodates dialogical connection with audiences. Given Twitter’s increasing popularity as a news-sharing platform, along with its importance as a platform that may give shape to future forms of social networking and “ambient” news delivery (Hermida, 2010b), how it gets used as a reporting device is important to examine. As the results of this study show, many of the most-followed journalists on Twitter are making humor a regular part of their tweeting, representing a further evolution in the negotiation of norms around news production in digital spaces (Lasorsa et al., 2011; Singer, 2005).
When it comes to social connections, people often use humor to let one another know they’re welcome—that they belong (Booth-Butterfield et al., 2007; Forester, 2004; Graham et al., 1992). Considering that research has shown people seek to gain and maintain connections through social networking sites (SNS), the importance of humor as a connective device should be considered more fully both by media researchers and practitioners. While this study did not examine the effects of Twitter on user interests, nor did it measure what impact the use of humor might have on those users, the results do indicate that journalists, at least in this social space, are tweaking professional norms by incorporating humor. That alone is a significant finding, and one that deserves further research to uncover the qualitative elements behind this development. As legacy news org anizations search for ways to maintain their footing in an unsteady environment, a little humor may go a long way in helping them better connect with current and potential followers. While television programs like The Daily Show successfully mingle news messages with humor to entice viewers, longstanding journalistic institutions haven’t been so quick to follow along, even and perhaps especially in the case of SNS use by their journalists. A number of major news organizations have restricted journalists’ use of Twitter or provided stringent guidelines for engaging the public (e.g., see Grove, 2010; Lavrusik, 2010), and the American Society of News Editors (ASNE), one of the leading trade associations, has discouraged using Twitter for breaking news or sharing opinions (Hohmann, 2011). Yet, these same restrictions may be working against the very goals of media outlets—to gain and maintain followers.
Bureaucratic hurdles notwithstanding, a number of journalists appear to be incorporating non-traditional practices in their everyday routines on SNS such as Twitter, as this study demonstrates. Nearly a quarter of all tweets coded indicated that journalists were “trying to be funny”—making humor one of the most common forms of Twitter use among the journalists in this sample. Correlation results suggest that, while humor is associated with nearly all forms of Twitter activity (i.e., various kinds of conveying information, seeking information, or sharing opinion), it is most strongly connected with major and minor forms of opinion, lifecasting and personalizing—elements that speak to the journalist’s “self” being especially present in the sharing of behind-the-scenes life details and personal musings. In this sense, humor may be an additional pathway through which journalists seek to connect with users on a more personal level. F urthermore, humor was found to be significantly associated with seeking information, discussing and retweeting on Twitter, reinforcing the literature on the importance of humor as a form of interpersonal connection (Martin, 2007). This might suggest that at least some journalists are embracing the “social” nature of Twitter’s environment, taking advantage of the opportunity to connect with and gain insight from the “people formerly known as the audience” (Rosen, 2006) in a more individualized and responsive fashion.
This study went on to explore the degree to which this use of humor is associated with measures of engagement and prominence on Twitter, which together can speak to evidence of connection with users. Simply “being funny” was found to have no association with the relative number of followers (or mentions on lists) that a journalist might collect, nor was it related to the number of days that a journalist had been on the service. But there was a rather significant positive correlation between a journalist’s level of activity—e.g., total tweets, tweets during the two-week coding period, etc.—and the use of humor in her tweeting. This would suggest that, to the extent that journalists become familiar with the Twitter platform—not merely signing up for the service as an early adopter, but actually using it regularly—they are more likely to become part of the fabric of Twitter and pattern their practices around its social mili eu: informal, conversational, opinionated—and, yes, often witty and satirical too.
Finally, this paper put these findings in the context of the relative prestige of one’s news organization: Having an affiliation with Fox News versus one with a local newspaper may have a bearing not only on the number of followers a journalist gathers on Twitter, but also be associated with how that journalist engages with the service. Prestige was negatively correlated with humor use, indicating that journalists from less elite institutions—such as local newspapers—are more likely to try to be funny in the course of their Twitter use. There are several possible explanations for this finding. Perhaps journalists from elite news media are more tightly bound by company restrictions, as in the case of those employed by Reuters (Grove, 2010) and The Washington Post (Lavrusik, 2010). Or perhaps, as the May 2011 ASNE report recommended (Hohmann, 2011), they see Twitter as more of a news-and-information platform and less of a social arena—in part because of their positioning at the pinnacle of journalism’s profession. Or, as previous findings have suggested (Lasorsa et al., 2011), it could be that journalists from more prestigious outlets are more comfortable with “business as usual” in professional norms, while their less prestigious counterparts feel that they must be more active and interesting on Twitter to catch and maintain attention. Such innovations to attract an audience might include humor.
Entering the Twitter arena—one about which journalists have been warned, or only recently have begun exploring in large numbers—naturally includes an element of risk, as evidenced in accounts of reporters facing organizational punishment or professional shame for thoughts shared on Twitter (see examples in Hohmann, 2011). And yet, journalists willing to take such risk might be pioneering new means of connecting with users in ways that either haven’t been possible or haven’t been encouraged via traditional norms and media platforms. Because its social graph and dialogue are generally public, Twitter promotes information dissemination across weak ties (Baym, 2010), providing chance encounters with news that might guide unlikely users to “find” news organizations—and ultimately become regular followers of the kind needed to sustain journalism in the long run. However, news organiza tions must first be willing to allow their journalists to venture onto Twitter without the fear of penalty (Ingram, 2011). As McGuire (2010) and Poniewozik (2010) recently argued, restricting journalists from freedoms such as opinion and humor—freedoms afforded to everyday social situations—could deter people from turning to those journalists and the outlets they represent. While norms of objectivity inevitably collide with the sharing of opinion or humor, news consumers increasingly appear to want journalists, above all, to be transparent—to provide more information about themselves, personally, and where they’re coming from in approaching a news item (Ingram, 2010; McGuire, 2010). To the degree that journalists use a space like Twitter to reveal more about themselves in the name of transparency, it might point to a reformulation of journalistic processes and journalistic authority beginning to play out in digital media at large (Karlsson, 2011).
This study drew a sample of the 500 most-followed journalists on Twitter (as of September 2009). While their large number of followers certainly should not be ignored, followers do not necessarily represent influence or popularity (Cha, Haddadi, Benevenuto, & Gummadi, 2010). As such, the results of this study should not be generalized beyond the sample population. What is humorous for one journalists or news organization may not be so for another. Because humor often requires prior context or knowledge to comprehend, journalists must be cautious when using it to engage audiences. Furthermore, this study examined journalists’ use of Twitter during a particular two-week period of late 2009; in a quickly evolving realm such as SNS, continuously updated and replicated forms of research are needed to provide a more complete picture of social media activity and content. This study also emphasized one possible way journalists might be conne cting with their followers, but as Lasorsa and his colleagues (2011) have indicated, there are myriad other ways in which journalists may be changing their routines in social media spaces—and thus many possible approaches for identifying and measuring such changes. Notably, humor proved somewhat difficult to code as a variable, as evidenced by a relatively low but still acceptable Cohen’s kappa score of .63 for intercoder reliability. Explaining humor can be difficult for a number of reasons. Most notably, humor is subjective and frequently defined, especially in the case of this study, by the receiver (coder). While overt messages of humor might be easy to pick up on, separate receivers may take more subtle comments differently. Prior knowledge of a subject can certainly help a receiver understand the intent of humor, but without it a receiver is left to judge that intent based solely on the message. Future research may s eek to take previous messages into account. Such research might also explore ways to delineate a more exact definition of humor. As a final limitation, incorporations of humor into social interactions tend to favor younger audiences (Calavita, 2004; Martin, 2007). Examining the nature of age—both that of the journalist as well as her followers—and other demographic considerations (e.g., years on the job) might be interesting points of entry for future research on news production via SNS.
Despite these limitations, the findings of this study offer a useful step forward, for journalism studies and news organizations alike, as the professional field wrestles with the philosophical and practical challenges of user participation generally and social media engagement in particular.
Braun, J., & Gillespie, T. (2011). Hosting the public discourse, hosting the public: When online news and social media converge. Journalism Practice, published via iFirst. doi:10.1080/17512786.2011.557560
Bruns, A. (2008). The active audience: Transforming journalism from gatekeeping to gatewatching. In C. Patterson & D. Domingo (Eds.), Making News Online: The Ethnography of New Media Production. New York: Peter Lang.
Carrell, A. (1992). The need to incorporate audience and situation into a theory of humor. Paper presented at the Speech Communication Association annual convention.
Cha, M., Haddadi, H., Benevenuto, F., & Gummadi, K. P. (2010). Measuring user influence on Twitter: The million follower fallacy. Paper presented at the The 4th International Association for the Advancement of Artificial Intelligence Conference on Weblogs and Social Media, Washington, DC.
Chyi, H. I. (2009). Information surplus in the digital age: Impact and implications. In Z. Papacharissi (Ed.), Journalism and citizenship: New agendas (pp. 91-107). New York: Lawrence Erlbaum/Taylor and Francis.
Coursaris, C. K., Yun, Y., & Sung, J. (2010, June 13-15). Twitter uses vs. quitters: A uses and gratifications and diffusion of innovations approach in understanding the role of mobility in microblogging. Paper presented at the Ninth International Conference on Mobile Business, Athens, Greece.
Facebook. (2011). Facebook press room. Retrieved April 20, 2011: http://www.facebook.com/press/info.php?statistics
Farhi, P. (2010). Just who does Jon Stewart think he is? Washington Post. Retrieved from http://www.washingtonpost.com/wpdyndyn/content/article/ 2010/10/24/AR2010102402645.html
Fox, J. R., Koloen, G., & Sahin, V. (2007). No joke: A comparison of substance in The Daily Show with Jon Stewart and broadcast network television coverage of the 2004 presidential election campaign. Journal of Broadcasting & Electronic Media, 51(2), 213-227.
Grove, J. V. (2010). Reuters to journalists: Don't break news on Twitter. Mashable. Retrieved from http://mashable.com/2010/03/11/reuters-social-media-policy/
Hackman, M. Z., & Barthel-Hackman, T. A. (1993). Communication apprehension, willingness to communicate, and sense of humor: United States and New Zealand perspectives. Communication Quarterly, 41, 282-291.
Hohmann, J. (2011). 10 best practices for social media: Helpful guidelines for news organizations. Report for the American Society of News Editors (ASNE). Available from http://asne.org/article_view/articleid/ 1800/asne-issues-guide-to-10-best-practices-for-social-media.aspx
Ingram, M. (2010). Twitter forces media to confront the myth of objectivity. Retrieved from http://gigaom.com/2010/07/08/ twitter-forces-media-to-confront-the-myth-of-objectivity
Ingram, M. (2011). News editors still don't want journalists to be human. Retrieved from http://gigaom.com/2011/05/12/ news-editors-still-dont-want-journalists-to-be-human/
Ionescu, D. (2010). Twitter use explodes, hits 50 million tweets per day. PC World. Retrieved from http://www.pcworld.com/article/ 190026twitter_use_explodes_ hits_50_million_tweets_per_day.html
Java, A., Song, X., Finin, T., & Tseng, B. (2007). Why we twitter: Understanding microblogging usage and communities. Paper presented at the International Conference on Knowledge Discovery and Data Mining, San Jose, CA.
Johnson, P. R., & Yang, S.-U. (2009). Uses and gratifications of Twitter: An examination of user motives and satisfaction of Twitter use. Paper presented at the Association for Education in Journalism and Mass Communication, Boston, MA.
Joinson, A. N. (2008, April 5-10, 2008). 'Looking at', 'looking up' or 'keeping up with' people? Motivates and uses of Facebook. Paper presented at the The 26th annual SIGCHI Conference on Human Factors in Computing Systems, Florence, Italy.
Kanalley, C. (2011). Twitter reactions to Osama bin Laden's death. The Huffington Post. Retrieved May 2, 2011 from http://www.huffingtonpost.com/2011/05/02/ osama-bin-laden-death-tweets_n_856119.html#s272181&title=Keith_Urbahn_Rumsfeld
Kawk, H., Lee, C., Park, H., & Moon, S. (2010, April 26-30, 2010). What is Twitter, a social network or a news media? Paper presented at the 19th International Conference on the World Wide Web, Raleigh, NC.
Lasorsa, D. L., Lewis, S. C., & Holton, A. E. (2011). Normalizing twitter: Journalism practice in an emerging communication space. Journalism Studies, published via iFirst. doi:10.1080/1461670X.2011.571825
Lavrusik, V. (2010). Washington Post tells journalists not to engage on Twitter. Retrieved from http://mashable.com/2010/10/20/ washington-post-tells-journalists-not-to-engage-on-twitter/
Lewis, S. C. (2010). Journalism innovation and the ethic of participation: The case of the Knight Foundation and its News Challenge (Doctoral dissertation). Available from Dissertation and Theses database. (UMI No. 3429005)
Lewis, S. C., Kaufhold, K., & Lasorsa, D. L. (2010). Thinking about citizen journalism: The philosophical and practical challenges of user-generated content for community newspapers. Journalism Practice, 4(2), 163-179. doi:10.1080/14616700903156919
McGuire, T. (2010). Twitter is not killing journalism, journalists are killing journalism. McGuire on Media. Retrieved from http://cronkite.asu.edu/mcguireblog/?p=158
Poniewozik, J. (2010). CNN, Twitter and why hiding journalists' opinions is (still) a bad idea. Time Magazine. Retrieved from http://tunedin.blogs.time.com/2010/07/08/ cnn-twitter-and-why-hiding-journalists-opinions-is-still-a-bad-idea/
Purcell, K., Rainie, L., Mitchell, A., Rosenstiel, T., & Olmstead, K. (2010) Understanding the participatory news consumer. Pew Internet and American Life Project. Retrieved from http://www.pewinternet.org/reports/2010/online-news.aspx?r=1
Rosen, J. (2006). The people formerly known as the audience. PressThink. Retrieved from http://archive.pressthink.org/2006/06/27/ppl_frmr.html
Rosen, J. (2008). A most useful definition of citizen journalism. PressThink. Retrieved from http://archive.pressthink.org/2008/07/14/a_most_useful_d.html
Schudson, M., & Anderson, C. (2008). Objectivity, professionalism, and truth seeking in journalism. In K. Wahl-Jorgensen & T. Hanitzsch (Eds.), Handbook of journalism studies. (pp. 88-101). New York: Routledge.
Seward, Z. M. (2009). Muck Rack, home of 140-character dispatches from the field, gets useful. Nieman Journalism Lab, Harvard University. Retrieved from http://www.niemanlab.org/2009/05/muck-rack-home-of-140 character-dispatches-from the-field-gets-useful/
Singer, J. B., Domingo, D., Heinonen, A., Hermida, A., Paulussen, S., Quandt, T., Reich, Z., & Vujnovic, M. (2011). Participatory journalism: Guarding open gates at online newspapers. Malden, MA: Wiley-Blackwell.
Skoler, M. (2009). Why the news media became irrelevant and how social media can help. Nieman Reports. Retrieved from http://www.nieman.harvard.edu/ reportsitem.aspx?id=101897
Smith, C. (2011). Twitter user growth statistics show stunning growth. The Huffington Post. Retrieved April 20, 2011 from: http://www.huffingtonpost.com/2011/03/14/twitter-user-statistics_n_835581.html
Stafford, T. F., & Stafford, M. R. (1998). Uses and gratifications of the World Wide Web: A preliminary study. Paper presented at the 1998 American Academy of Advertising Conference, Washington State University.
Tenore, M. J. (2007). Experimenting with Twitter: How newsrooms are using it to reach more users. Poynter. Retrieved from http://www.poynter.org/latest-news/top-stories/84153/ experimenting-with-twitter-how-newsrooms-are-using-it-to-reach-more-users/
Thurman, N. J. & Hermida, A. (2010). Gotcha: How newsroom norms are shaping participatory journalism online. In: Web Journalism: A New Form of Citizenship. Sussex Academic Press, Eastbourne, pp. 46-62.
Wanzer, M., Booth-Butterfield, M., & Booth-Butterfield, S. (1996). Are funny people more popular: The relationship of humor orientation, loneliness, and social attraction. Communication Quarterly, 44, 42-52.
Welch, K. J., Maiuri, T., & Poleshchuk, A. (2010). Building international partnerships through an online humor exchange. Progress in community health partnerships. Research, Education, and Action, 265-269.
Wrench, J., & Booth-Butterfield, M. (2003). Increasing patient satisfaction and compliance: An examination of physician humor orientation, compliance-gaining strategies, and perceived credibility. Communication Quarterly, 51, 482-503.
 In its top 500, Muck Rack included some “journalists” who may not follow typical avenues of information delivery. For example, at least one of the journalists included is known to mix satire with news. However, removing this individual from the data set had virtually no bearing on the results. Moreover, because studies have shown that people pay more attention to news and information when delivered with humor (Fox et al., 2007), it seemed appropriate to keep the sample intact as-is.
 As evidenced by the low but acceptable reliability score, humor proved to be a difficult variable for coders to agree upon. Because humor often requires prior knowledge of a subject or, at the very least, knowledge of intent, coding can become an issue of perception. In other words, what was intended to be humorous to one coder may not have been so for another. However, other messages were overt in their humor such as this one by New York Times technology columnist David Pogue: “Husband: ‘Honey, what setting do I use on the washer?’ Wife: ‘Depends. What’s it say on the shirt?’ Husband: ‘University of Oklahoma!’ (see https://twitter.com/#!/Pogue/status/4700562797)
 For an explanation about Twitter lists, see https://support.twitter.com/entries/76460-how-to-use-twitter-lists.
Copyright 2011 Communication Institute for Online Scholarship, Inc.
This file may not be publicly distributed or reproduced without written permission of