Abstract
Television (TV) audiences are increasingly using portable communication technologies to multitask, look up information online, check social network sites, and comment on the programs being watched. Although multitasking can distract audiences away from the TV content, the use of a second screen in a manner that complements the mass communication content is a unique phenomenon that may lead to positive outcomes. This study, based on survey data collected from a national stratified random sample (N = 1417), supports a theoretical model linking frequency of complementary simultaneous media use to engagement, which in turn mediates incidental learning. Findings may be useful for mass communication scholars and practitioners seeking to understand the effects of dual electronic media use.
Keywords
Introduction
The proliferation of Internet-enabled mobile communication technologies, such as laptops, tablets, and smartphones, has given rise to the so-called second screen phenomenon: using another electronic device while watching television (TV) (Nielsen, 2013a). Although audiences have frequently multitasked with the TV on in the background, the simultaneous consumption of content on another electronic screen is a new development worthy of exploration by mass communication scholars and practitioners (Hassoun, 2014; Phalen and Ducey, 2012). Frequently, consumers use multiscreens in a manner unrelated to the content on TV. For example, consumers check e-mail, use social media, and browse the Internet as forms of second screen activities, according to studies conducted by Google (2012) and Nielsen (2013a). However, consumers increasingly also are using a second device in a manner that complements what they are watching on TV, particularly during major live events (Sasseen et al., 2013). A survey by the Nielsen ratings company (2013a) found nearly half of second screen users (49%) look up information related to the show they are watching, at least some of the time, and approximately one-fifth of tablet and smartphone owners have read conversations about a TV program on social network sites simultaneous to the broadcast.
Complementary simultaneous media use (CSMU) is defined as cooperating a computer, tablet, or mobile device while watching TV, where the same topic or content is the user’s focus on both screens at the same time. Exploring the possible outcomes of this trend may provide new perspectives on the ongoing academic questions about the impact of new technologies on audience behaviors. Are new technologies deterministic of human behavior? Does technology contribute to societal fragmentation and isolation? Or do these new technologies afford new uses and opportunities for obtaining knowledge and participating in the public sphere?
The Nielsen TV ratings company now tracks ‘social TV’ conversations among US audiences. Nielsen has found a reciprocal causal relationship between TV ratings for a program and the volume of Twitter conversation around the program – in other words, the social media buzz often generates more viewers, and vice versa (Nielsen, 2013b). Big media events, such as sports and awards shows, attract the largest volume of second screen interactions, while dramatic series and reality TV shows have a smaller but steadier following on Twitter. The 2014 FIFA World Cup playoff rounds and the final were among the top 10 sports events of the year that simultaneously attracted a Twitter audience in the United States (Nielsen, 2014). Although much of the second screen interactions are around popular culture programming, televised events of the 2012 US presidential election drew a social TV audience. The Pew Research Center reported that 11% of the people watched the second US presidential debate between Barack Obama and Mitt Romney dual screened in a complementary way with a computer or mobile device, and 27% watched election night results in a similar manner (Sasseen et al., 2013).
Although much has been written about the negative potential of multitasking with technology to cognitively overwhelm, distract (Hallowell, 2006; Marois and Ivanoff, 2005), or fragment society (Carr, 2010; Putnam, 2000; Turkle, 2011), CSMU may have some positive outcomes, depending on several situational factors and psychological motivations. Jenkins (2006) presented a convergence culture paradigm in which new media would not displace old media, but instead would interact with traditional media in complex ways. Consumers would drive this shift by actively participating with televised content, forming knowledge communities around specific shows, and socially connecting with one another. As social media use has grown, TV has become an ‘intrinsic part of “new” media’ (Gray and Lotz, 2012: 3). This shift in how consumers relate and interact with televised media presents challenges and opportunities for media elites, who are not accustom to sharing the role of media producer and knowledge broker.
The purpose of this study is to present and test an interdisciplinary theoretical model that links CSMU to engagement with, rather than distraction from, televised content. Drawing from education, cognitive psychology, and mass communication frameworks, the authors propose that consumers will perceive a sense of increased knowledge gain from digital sources, as an outcome of simultaneous engagement with old and new media. Knowledge gain is operationalized here as the consumers’ perception that learning has taken place through their activities online. The type of information learned varies from news and events to facts regarding the TV show being watched and opinions of other people. Learning may happen intentionally, through the specific act of looking up information or incidentally by stumbling upon social media posts generated by other audience members.
Using national survey data gathered in the United States (N = 1417), the authors test a theoretical model (as shown in Figure 1) that the frequency of CSMU leads to engagement with the televised content, which then increases the chance of incidental or informal learning through online sources or social networks. For the theoretical model presented here, what is learned is less important than the consumers’ perception that learning has occurred. This model will assist mass media scholars who are just beginning to research the possible effects of the convergence of consumers’ online behaviors associated with the televised content they are simultaneously watching. Findings also add a new dimension to previous TV studies that have sought to characterize active audiences.
A theoretical model linking frequency of complementary simultaneous media usage, engagement, and incidental learning.
Review of relevant literature and theory construction
Audience studies
Uses and gratifications (U and G) theory has long been used to analyze audience behaviors and motivations for choosing certain types of media over another (Kippax and Murray, 1980; Rubin, 1983). U and G theory presumes an active audience, who seeks specific genres of programming to fulfill specific extrinsic needs. Many scholars, however, have argued that TV viewing is a more passive, and therefore inferior, activity than consuming other media, such as cinema or newspapers that require more concentration. Ellis (1993), for example, negatively compared TV viewing as a passing glance to movie watching, which is a more intense experience. Others have suggested that traditional TV viewers are more passive than the active, goal-seeking consumers that U and G theory presumes (LaRose and Eastin, 2004; Ruggiero, 2000). This approach can be traced back to McLuhan’s (1994) assertion that the medium is the message, but McLuhan also has been criticized as assigning too much importance to the media platform and not enough emphasis on how the audience processes the message. Williams (1974) argued against technological deterministic approaches, saying that society and social interactions determine the uses of technology, not the technology itself.
This argument is in line with earlier British Cultural Studies from 1960s to 1980s that analyzed consumers as active participants who shape the meaning of mass-produced products or media. Hall (1980), for example, defined audience activity as the cyclical process of encoding and decoding mass media messages based on one’s individual frames and experiences. Fisk (1989) likewise suggested that popular culture is the result of audiences actively assigning and sharing meaning to texts produced by culture industries. Jenkins’ (1992) studies of fan communities, who coproduce and remake televised content by creating stories, art, and songs, predate the diffusion of Internet technologies. Some scholars have criticized the active audience theory as being overly optimistic about consumers’ abilities to shape content and skewed to the comfortable middle class who has access and leisure time to explore such pursuits (Morris, 1988; Seaman, 1992). However, the widespread adoption of digital and social platforms, spread through the use of mobile phones, has exponentially increased the ability of audiences to simultaneously produce and consume media. In industrialized countries, the barriers to entry have decreased, allowing access to audiences from a range of demographics and incomes. Bruns (2008) has characterized contemporary audience as ‘produsers’ – who act as both producers and consumers. Gray and Lotz (2012) noted that audience studies should now turn toward defining how consumers and the industry are adapting to their changing roles. Useful frameworks for the present phenomenon under study include learning theories, which address the informal educational outcomes of computer-mediated technologies, and the conceptual constructs of media interactivity and engagement.
CSMU: Definitions and behaviors
Although little academic research has been conducted regarding the effects of multiscreen viewing behaviors, Phalen and Ducey (2012) noted that media viewing styles could be categorized by intentionality as well as activity level. Habitual–ritualistic viewers tend to be more medium oriented, while intentional viewers focus more on finding the content and seeking the appropriate device. Passive viewers might watch whatever is most convenient. Active viewing could be defined along a continuum of behaviors, from seeking out additional information online to accessing related content, or socially conversing with others regarding the program (Costello and Moore, 2007). The widespread adoption of the Internet brought praise from some authors who saw the potential of more user-friendly Web 2.0 technologies to allow the audience to take a more physically active role in creating content and engaging with others about the issues of the day (Rosen, 2006; Shirky, 2010). Social networks have provided the added opportunity for individuals to obtain a sense of shared experience with others regardless of geographical boundaries or location (Rainie and Wellman, 2012). Shirky (2010) envisioned that consumers would use the cognitive surplus acquired through generations of passive TV watching to actively post memes, blogs, and social messages online. Examples of the behavior Shirky described were rampant during the 2012 presidential debates as consumers created Internet memes or parody Twitter accounts almost instantly after unintended blunders or gaffes left the candidates’ mouths. Other users then engaged in that activity by sharing and commenting on the user-generated media that other consumers had created.
Some scholars have argued that the most high-intensity media users will not trade one medium for another but will instead use a number of communication technology platforms and traditional media to a greater extent than the general population (Cooper and Tang, 2009; Enoch and Johnson, 2010; Jenkins, 2006). Jenkins noted that many fans form knowledge communities around shows of mutual interests and, by doing so, assert power over media producers by sharing spoilers, for example. He also suggested that this shift in how consumers relate to and participate with media would occur first through popular culture and then spill over to political culture. Therefore, in the realm of simultaneous media use, the most intentional and active users could be defined as those who purposively seek information or social interactions on another device in a complementary manner to the content being watched, as opposed to true multitaskers who may be habitually inclined to use another device in a way that is diversionary and unrelated to the TV content. CSMU may also lead to interaction with the content or other people, increasing the likelihood of greater engagement.
Defining dual-screen interaction
The concepts of interactivity and engagement are changing as communication technologies evolve, creating opportunities unimagined by prior scholars or practitioners. Interactivity is defined here as using another screen in a complementary way to the content being watched on TV either for intrinsic goals (e.g. being informed, affiliation, and sense of community) or extrinsic goals (e.g. social status, recognition, and success). Examples of dual-screen interactivity include the following: (1) searching for related information or products online, (2) connecting with others watching the same program through social networks, and (3) creating user-generated content in the form of blog posts or memes.
Scholars have defined interactivity as process, system feature, or perception (Stromer-Galley, 2004). The interactive process emphasizes human-to-human interaction, the exchange of information, and reciprocity of conversations in which both parties have interchangeable roles (Pavlik, 1998; Stromer-Galley, 2004). Scholars use system feature to describe the capabilities for technological tools to mediate human-to-human interactions, which Bucy (2004) identified as a key function of new media technology. Conceptualizing interactivity as perception focuses on the users’ experience of feeling engaged and mutually understood by others during the interaction (Burgoon et al., 2000).
Jenkins (2006) differentiated interactivity from participation. He described interactivity as a feature of a new technology, shaped by the designer, which allows a user to converse with and offer feedback to the producer. Participation is shaped by the consumer and may or may not include the producer. Thus, the social TV second screen interactions analyzed here are more similar to Jenkin’s definition of participation. Cable companies and networks have attempted to inject interactive system features into TV programming long before the digital age. In the late 1970s, for example, Warner Cable launched QUBE, an interactive TV system that allowed users to vote, shop, play games, and take classes through the TV (Parsons, 2008). QUBE’s failure to take hold was partly due to the company’s financial difficulties, but as Arceneaux (2013) observed, the audience might not have been ready for or even wanted the active involvement with the TV at that time.
Jenkins (2006) did not view media convergence as resulting in one box that would serve multiple functions. Instead, he described a culture in which interactive system features would create a process for the public to participate in TV shows. But the bulk of the participation Jenkins described, such as voting for American Idol contestants through calling or texting, occurred asynchronously, before or after the airing of the show – not during. The growth of second screen activity alongside a TV program is a system feature, or technological affordance of information communication technologies (ICTs), which allows fans to participate synchronously in a show as it airs, thereby providing greater opportunities for both the process and the perception of interactivity to occur. For the purpose of this study, interactivity is investigated from the consumer’s perception. Thus, this perspective is more in line with Jenkins’ definition of participatory culture than interactivity as a feature of new media technology to provide feedback.
Conditions of engagement
For the purpose of this study, engagement is defined as a cognitive state of the TV viewer marked by high levels of involvement and absorption with the content of the TV programming. Attention is another attribute that must be considered when assessing the degree of learning that may occur. The most relevant research is from cognitive psychology regarding attention as a physiological and cognitive activity of the brain, which may lead to a core flow state, during which people become highly engrossed in an activity (Csikszentmihalyi, 1990). Adding another competing screen, even regarding the same topic, can create inattention to the TV, however. Thus, broadcasters no longer can assume attention, and as Russell (2010) argued, they must now earn engagement.
Although engagement has been difficult for scholars to measure (Phalen and Ducey, 2012), interactivity, amount of time spent, intensity of use, and sense of connectedness, all characteristics of experiences, may lead to heightened media engagement (Evans, 2008). The possibility then exists that those who intentionally and actively dual screen televised programs, for social or informational purposes related to what they are watching, have heightened attention and a greater chance of becoming engaged with the content and overall TV viewing experience (Costello and Moore, 2007; Phalen and Ducey, 2012). Therefore, the first hypothesis is proposed:
Incidental learning as an outcome of engagement
Contemporary education theory suggests that learning, whether formal or informal, occurs most effectively when the learner is actively coproducing content, socially involved in the process, or seeking information to satisfy curiosities (Dabbagh and Kitsantas, 2012; McLoughlin and Lee, 2007; Sefton-Green, 2004). Education researchers are finding ICTs to be useful in mediating learning experiences, particularly among younger people (Castells, 2009; Ito et al., 2008). Play, interactivity, and social engagement are seen as important constructs in creating an environment where learning takes place. Gee (2003), for example, used computer games to illustrate how technology can foster self-directed learning, role playing, and social stimulation.
Formal learning is distinguished from informal learning based on structure, environmental factors, and the presence of measurable outcomes, such as grades. Formal learning takes place in schools, courses, and classrooms, and informal learning is often initiated by the learner and happens through observation, having social conversations, or as a result of being stimulated by an interest (Cross, 2007). Dabbagh and Kitsantas (2012) noted that educators are increasingly using social media to enhance student learning in both formal and informal contexts.
Prior mass communication research has shown that different learning outcomes are related to specific media platforms. Newspapers were once associated with active use and greater political knowledge, while TV was seen as a more passive medium that did not contribute as much to learning about politics (Becker and Dunwoody, 1982; Chaffee, Zhao, and Leshner, 1994). Tewksbury et al. (2001) looked at incidental, or nonpurposive, learning online and found frequency of Internet use to be positively associated with incidental exposure to news online. Younger people and those who had the goal of getting news were the most likely to experience incidental learning online. Factors positively influencing incidental learning included the frequency of use, repetition, and goal direction (Tewksbury et al., 2001). The Pew Research Journalism Project made a similar finding with regard to Facebook use. Nearly half of Facebook users consume news on the social network site, but the majority of those (78%) do so incidentally (Mitchell et al., 2013). The time spent on Facebook also was positively related to the amount of news consumed.
Because CSMU is likely to increase cognitive and social engagement with the content being watched and because the use of a second screen implies active media consumption and creation, the likelihood of incidental learning should increase with level of engagement. Therefore, the following hypothesis is proposed:
Interactive complementary screen use alone does not necessarily imply greater attention. When engagement occurs, however, the likelihood of more focused attention increases, as does the possibility of incidental learning. The authors theorize that engagement is critical to incidental learning (see Figure 1). A third hypothesis is posited:
That is, engagement fully mediates the relationship between CSMU and incidental learning.
Methods
A national telephone survey was conducted in July–August, 2012, utilizing a stratified random sample. The goal of the sampling strategy was to ensure that the sample was statistically representative of all adults living in the United States. In telephone sampling, participating respondents tend to be older and more female, when compared to the population. Further, younger people tend not to have landlines, so the sampling strategy sought to sample cell-only households, as well as households with landlines. Researchers used random digit dialing to generate a probability sample of all households in the United States with telephones (over 97%). The sample was stratified by age and gender, using national data from the 2010 US Census (U.S. Census Bureau, 2011). Age was divided into four strata (18–35, 36–47, 48–66, and 67 or older). This stratification reflected the commonly used age parameters for the Silent generation, Baby Boomers, Gen X, and Millennials. The sample was further stratified by gender for each age strata, ensuring that men and women were represented proportionally in each age-group, based on population parameters. In addition to landline calls, the sample was supplemented with cell phone interviews. Interviews were conducted in both English and Spanish. However, most respondents that self-identified as Hispanic or Latino (11.2% of the sample) preferred to conduct the interview in English, and only 1.5% of the total interviews were conducted in Spanish. On average, interviews lasted 20 min.
Respondents were qualified as follows: (1) 18 years old or older; (2) lived in the United States for the last 12 months as a citizen, permanent resident, or visitor; (3) personal access to the Internet by computer, smartphone, and/or tablet; and (4) accessed the Internet at least once in a typical day. A total of 53,153 phone numbers were called. Of those, 26,363 were invalid due to disconnects, no answer after multiple attempts, business or government numbers, language barriers, or non-qualifying respondents in the household. Of the valid sample of 26,790, 19,610 were not successfully contacted, yielding a noncontact rate of 73.2%. Of the valid sample, 5773 refused, yielding a refusal rate of 21.5%. Of the valid sample, 1417 completed the interview, yielding a response rate of 5.3%. The response rate is lower than ideal, in part because of the length of the interviews conducted.
CSMU was measured with two items. All respondents were asked this qualifier question, ‘How often do you use your smartphone, computer tablet, or your laptop computer while watching something on your TV screen?’ Those who replied ‘never’ or ‘almost never’ were disqualified, yielding a subsample of simultaneous media users. Those indicating that they engaged in simultaneous media use ‘sometimes’ or ‘often’ were then asked, ‘How often do you connect the content of what’s on TV with what you are doing – at the same time – on your smartphone, tablet, or laptop?’ Usable responses included ‘sometimes’, ‘often’, and ‘always’. Those indicating that they ‘never’ or ‘almost never’ engaged in CSMU were not asked the engagement items. Respondents in this latter group were classified as ‘multitasking’, meaning that TV served largely as background to the focal activities conducted on the second screen. Theoretically, the second screen would detract from engagement with TV content. This elimination process yielded a subsample of 338 respondents (23.9% of the overall sample) who engage in CSMU while watching TV.
Three items measured engagement with TV content during CSMU. The three items were introduced to respondents as follows, ‘Please think about the times when you are watching something on TV while using a smartphone, tablet, or laptop
The question about synchronous dual-screen use did not specify the particular Internet platforms that respondents accessed during CSMU. Therefore, the measure of incidental learning was aggregated for respondents reporting on their use of social networks, e-commerce, content communities, or search engines. The computer-assisted telephone interviewing software inserted the name of the particular Web site (e.g., Facebook, Amazon.com, New York Times, and Google search engine) into each of the following probes, based on the respondents’ identification of their favorite Web site.
Before the following questions were asked, respondents were instructed, ‘When answering the following, think about the things you
H1 and H2 were tested using the Pearson product–moment correlation coefficient. One-tailed tests were used since the hypotheses specify the direction of the posited relationships. H3 was tested using the partial correlation coefficient. The value of α was set at 0.05 for all tests.
Findings
The average age of respondents was 46.7 years (median = 45.0). Average income was US$70,406 (median = US$60,000). Regarding gender, 52.2% of respondents were women. Regarding ethnicity, 76.4% were White, 11.2% were Hispanic or Latino, 7.1% were African-American, and 2.5% were Asian-American. Regarding education, 21.0% had a high school diploma or less, 31.0% had some college or technical training, 29.0% had graduated from a 4-year college, and 19.0% reported some graduate studies or an advanced degree. On average, respondents reported spending about 3.4 h on the Internet in a typical day (median 2.0 h).
H1 stated that CSMU is positively related to engagement with TV content. This hypothesis was confirmed, r(338) = 0.13, p = 0.01. Therefore, a positive correlation exists between frequency of CSMU and engagement with TV content.
H2 stated that engagement with TV content is positively related to incidental learning through Internet sources. This hypothesis was confirmed, r (338) = 0.10, p < 0.05. The respondents who were most engaged with TV content also reported the highest levels of self-perceived incidental learning online.
H3 stated that CSMU is not related to incidental learning through Internet sources, after controlling for engagement with TV content. This hypothesis was confirmed, partial r (337) = 0.00, p > 0.99. Thus, a direct correlation does not exist between CSMU and incidental learning online without engagement acting as a mediator.
Discussion
This study provides some insight into the process whereby Americans use their laptops, smartphones, and tablets to complement information provided on the TV screen. CSMU is a new phenomenon, technologically dependent on consumer access to the Internet on portable devices while watching TV. CSMU may resemble social conversation as audiences communicate with each other through social networks or other digital means. Or, CSMU may be a solo activity that happens when a consumer decides to look up information related to the show being watched. Based on our probability sampling of US audiences, nearly 24% of American adults who use the Internet at least once a day reported complementary simultaneous media usage. Thus, our sample represents the earliest adopters of this type of participatory audience behaviors.
Findings of this study extend the concept of an active audience to those who participate with televised content by simultaneously using new media technologies. Although this activity occurs on a continuum, as described by Costello and Moore (2007), and could result in distraction, our findings show increased second screen activity positively correlates with greater engagement with the televised content. In the model we have presented, an active audience member is one who is intentionally viewing a TV show and accessing online resources related to that program simultaneously. As the frequency of this activity increases, so does engagement with the TV content. That engagement is necessary for consumers to perceive that their online activities are helping them gain knowledge.
This theoretical model shown in Figure 1, supported by our findings, meshes well with the theory of flow (Csikszentmihalyi, 1990). Passive consumption of TV content can be transformed into a highly interactive, engaging activity through complementary simultaneous media usage. The user can enter into a flow state where challenges of the activity are balanced with the abilities of the user to achieve those activities. In this flow channel, typified by high levels of engagement, the user learns new information that he or she wasn’t looking for. Such incidental learning, albeit covering a wide range of topics, potentially promises great benefits to society through better informed citizens.
As noted in the literature review, incidental learning is an important outcome of media exposure. Specifically, frequency of CSMU increases the likelihood that audience members will become more engaged in the content on the TV screen. Such engagement, in turn, wholly mediates the incidental learning that occurs simultaneously over the Internet, through social media or other sources, while watching TV.
Since the turn of the 21st century, scholars have debated the detrimental versus positive effects of digital media use. One school of thought takes technological deterministic view that time spent online will fragment society, thereby decreasing political and civic engagement (Carr, 2010; Putnam, 2000; Turkle, 2011). Other scholars see the potential of ICTs to engender participation through online networks, particularly among younger people and those of lower socioeconomic status who have become disillusioned with traditional means of civic involvement (Barber, 2001; Delli Carpini, 2000; Rainie and Wellman, 2012). A meta-analysis of research relating to the Internet use and civic engagement (Boulianne, 2009) suggested that the impact was small but positive, particularly among online news seekers. Jenkins (2006) noted that a shift toward a more participatory culture will take time to occur and will begin with consumer interaction through popular culture before spilling into political culture. The US presidential election in 2012 is an example of a major political event that generated temporary knowledge communities around the election as it was covered on TV. Although the more mundane programs (reality TV and serial dramas) generate the bulk of CSMU activity (Nielsen, 2014), this behavior could continue to grow around political events in democratic countries.
The present study takes a variety of digital media activities into account that users may intentionally perform, for intrinsic or extrinsic goals, while watching TV. Users may be searching for information, purchasing products, socializing with others watching the same program, or creating and posting content in response to the mass media content. Interactivity may result in all the above activities as process and perception. But even an active and interactive audience does not ensure focused attention or engagement. For engagement to occur, attention must be focused in a complementary manner to both screens.
Contrary to the belief that all multitasking is distracting and inefficient, using another screen for ancillary purposes may actually increase understanding of the televised content, particularly among those who are seeking information for socio-cognitive motivations. Because of this finding, we would logically expect broadcasters to create these opportunities for consumers of both news and entertainment programming. But as Jenkins (2006) noted, the corporate ownership and interests of media industries are more powerful than the desires and expectations of individual consumers. The gatekeeping of the 2012 Olympics by National Broadcasting Company, for example, conflicted with the expectations of the social media audience who wanted instant televised access to the events so they could converse with others (Nee, 2015). Can TV content providers orchestrate such engagement? Or must such engagement originate organically among audience members? Clearly, further research is needed. Nevertheless, if enough people are intrinsically or extrinsically motivated to engage in CSMU, for the reasons shown in this study, this practice ultimately may lead to a more participatory, informed electorate – a condition necessary and vital for a functioning democracy.
Limitations and further research
The present study depends heavily on a cross-sectional survey of US adults interviewed over the phone. All the limitations of survey research in general and telephone interviewing in particular are applicable to this study. This study also provides only a partial test of the conceptual model in Figure 1. On a theoretical level, incidental learning can come from both TV content and also the Internet. One limitation of the present study is that incidental learning was measured only for the Internet portion of the complementary simultaneous media usage. Future research should measure incidental learning from the TV portion of exposure. Likewise, engagement was measured only for TV use and not use of the other screen. Bifurcating incidental learning outcomes, under conditions of simultaneous use regarding the same topic or content, will be difficult. Experimental designs, where certain content is specific to TV programming and other content is specific to the other screen, could be utilized to measure incidental learning from each screen separately. Such a design could measure interaction effects, whereby an additional experimental condition would include content that is specifically included as TV content and as content on the other screen.
The measure of TV engagement, while face valid and reliable, does not fully address the process whereby audiences engage the content simultaneously through two media platforms. Flow theory (Csikszentmihalyi, 1990) provides a useful theoretical framework for expanding the item set used to measure engagement with both platforms in CSMU. Flow also provides a robust theoretical framework for understanding the process whereby incidental learning occurs. Finally, flow theory allows research on CSMU to tie findings to the burgeoning body of prior studies that utilize flow as an intervening variable that moderates the relationship between simple exposure to media content (including complementary simultaneous use) and desired outcomes, such as incidental learning.
The present study utilized an indirect measure of incidental learning: the perception among audience members that they had gained incidental knowledge. Future audience research studies should link indirect measures of incidental learning to direct, objective measures of knowledge gain.
Footnotes
Acknowledgement
The authors gratefully acknowledge the School of Journalism & Media Studies, San Diego State University, which provided funding for this study.