Positive reading, positive watching: : the underlying mechanisms of perceived valence effects in social TV usage

Positive reading, positive watching: The underlying mechanisms of perceived valence effects in social TV usage

Meng Yang 11354186

Research Master’s Thesis

Graduate school of Communication

Master’s programme Communication Science Thesis Supervisor: Hilde Voorveld

Abstract

Social TV, the simultaneous consumption of TV programs on the first screen and online interactive contents on the second screen, has received more and more attention in the entertainment communication field. Second screen content is believed to have an influence on people’s consumption and reactions to the TV viewing. The study aimed to clarify how the valence people perceive from second screen content affect their evaluations of and commitment to the TV program through two underlying mechanisms (i.e., hedonic enjoyment and eudemonic appraisal). The quasi-experiment (N = 132) showed that perceived valence has a positive influence on people’s evaluations of the TV program via enhancing people’s hedonic enjoyment. Moreover, by recording eye-tracking data it was also found that people’s gaze durations of the second screen content has a positive impact on the valence thy

perceive, whereas the gaze duration of the TV program has positive impact on both TV program evaluations and commitment.

Introduction

Along with the proliferation of various digital devices, people’s media consumption behaviors have changed drastically. Recent research has shown that media multitasking, which refers to the consumption of multiple media contents simultaneously, has become prevalent worldwide (TC Lin, 2018). In recent years, one relatively new form of media multitasking, multi-screening, has also captured widespread attention within the communication field. Defined as “the simultaneous use of multiple screens” (Segijn,

Voorveld, Vandeberg & Smit, 2017, pp.295), multi-screening is believed to have influences on various domains, such as advertising effects, political attitudes, and entertainment experiences (e.g., Van Cauwenberge, Anna, Gabi Schaap, and Rob van Roy, 2014; Raney & Ji, 2017).

In the branch of entertainment communication, multi-screening has been specified as “social TV”, which refers to using interactive online platforms while TV viewing (Mukherjee & Jansen, 2014). Researchers initially focused on clarifying how and why people use social TV. That is, the motivations of social TV usage and predicting individual characteristics.

According to their studies, it could be summarized that people use social TV out of four main motives: communicating about the impressions, curiosity about others’ opinions program recommendation, and information seeking (Krämer, Winter, Benninghoff, & Gallus, 2015; Mukherjee, & Jansen, 2014). Later on, inspired by the first three motivations, scholars gradually moved their focus to the media effects and emphasized the interactive nature of social TV. They took closer looks at how concurrent online interactivity affect people’s overall TV viewing experiences. Many studies exploring the impacts of instant messaging, online commenting, and social media browsing on people’s viewing enjoyment have been done lately (e.g., Ceron & Splendore, 2018; Oviedo, Tornquist, Cameron, & Chiappe, 2015). In the meanwhile, information seeking motived activities, often considered to be less

interactive, are neglected by most researchers.

On the other hand, in the entertainment industry, the information seeking motivation was noticed and fulfilled by providing TV viewers with relevant information on the second screen.

For example, Proulx & Shepatin (2012) introduced in their book that an American TV series, the Kennedys, served text contents about the actual events depicted in the TV series in its mobile app (p.95). Likewise, relevant news reports and in-depth reviews are also provided in the official websites of most TV programs (e.g., Official Website of Modern Family, Game of Thrones, Doctor Foster). This form of information, compared with online comments on social media and online communities that are commonly central in academic studies, is usually more serious, and connects the TV programs to real social phenomena or historical events. Its authenticity may have a different influence on people, and thereby change how they process the TV program. However, such kind of second screen content received little attention in the academic field, and it remains unknown how it may affect people’s overall social TV viewing experience. Thus, this study will take a closer look at this informational type of second screen content and clarify how it affects people during their social TV usage. To better understand how second screen content influence people's overall social TV viewing experience, scholars then took a further step looking at multiple features of second screen contents. Among which, emotion, is asserted as a prominent element which

influences the whole process of social TV usage in various aspects in many studies (e.g., Rubenking, 2017; Krämer, Benninghoff, & Gallus. 2018). Raney & Ji (2017) claimed that the emotionality of the contents on both screens, plays a prominent role during people’s overall TV viewing. More specifically, they stated that one’s emotional reactions to the messages on the second screen could influence how he/she interpret and react to the content on the first screen. Correspondingly, empirical studies focusing on the effects of the valence (i.e., whether the message is positive, negative or neutral) of first and second screen messages have also been done lately. Krämer, Benninghoff, & Gallus (2018) tested how different valence of social media comments affect people’s evaluations of the first screen contents, while Rubenking (2017) found that compared with emotional content, neutral content on the first screen is more likely to trigger people's second screen usage. However, it should be noticed that these studies merely focused on the emotions in the media contents, and the emotion people perceive from media has never been studied before. Furthermore, perceived

emotion has never been discussed in relation to the informational second screen content in the existing studies either. According to cognitive emotion theories, emotions are the results of people’s appraisal processes. Also, the specific emotion one perceives from certain messages could be affected by his/her values and goals. In other words, people may perceive various emotions from the same message due to individual differences (Wirth & Schramm, 2005). Therefore, to fill this gap, the current study also aims to clarify how the emotion people perceive from the second screen content influence their evaluations of the first screen content. Thus, the following research question is formulated:

RQ1: How does the valence people perceive from the second screen influence their evaluations of the TV program on the first screen?

Theoretical Background Perceived valence and social TV usage

Discussions about “active audience” have been last for a long time in entertainment communication and can be traced back to 1970s when Katz, Blumler, and Guervich

proposed uses and gratifications theory. The theory contended that people play active roles during media consumption, and they select and use media to fulfill their gratifications. Thanks to the development of the Internet and digital technologies, this postulation has come into reality and reached a peak (Sundar & Limperos, 2013). TV viewing, known as one of the most common entertainment activities, has also changed drastically recently. Instead of consuming TV content passively, TV viewers are now able to participate actively by conducting relevant activities on a second screen out of their multifaceted motivations. TV programs on the first screen, in the meanwhile, are also believed to benefit from these second-screen activities and get higher evaluations and ratings as well (Lin, Chen, & Sung, 2018).

One underlying reason is that, during the consumption and exchange of program-related information on the second screen, people’ emotion investment on the TV program increases (Rubenking, 2017). Processing emotions from media messages do not differ from our non-media daily emotion elicitations. We firstly perceive emotion from an unconscious appraisal

process, during which (both media and non-media) messages are interpreted and evaluated based on the norms they contained, their novelty and attainability. We then form a series of response patterns consciously, including specific emotions. This process is also known as emotion induction (Wirth & Schramm, 2005). The dimensional theory of emotion has conceptualized emotion as the combination of two underlying dimensions: valence and arousal (Lang, 1995). Valence refers to an affective response and is considered to be a continuum ranges from positive (or pleasant) to negative (or unpleasant). Arousal

distinguishes emotion from extreme calmness to activation. In most existing studies, valence has been testified as an element which has an impact on people’s process of media

messages, yet arousal is more likely to be found as the effect of media consumption (Zillman, 1973). The current study, therefore, also focuses on the valence dimension and aims to clarify how it affect one’s social TV viewing.

Effects of valence on program evaluations

Social TV usage is modeled as a dual process in which first and second screen usage influence each other invariably (Raney & Ji, 2017). The model, named as the model of socially shared TV viewing experience, consisted of 6 processes: motivation, selection, reception, reaction, appraisal, and effects. Within each process, first and second screen usage interact with each other. Consuming specific information on the second screen occurs during the reception process. In the article, the authors focused on people’s consumption of others’ online comments to the TV programs and explained how they perceive and evaluate these comments. During the reception process, people process, interpret and make

meanings of the content on both screens. This process can be influenced by both individual characteristics and the content itself. In turn, in the following reaction process, people respond both emotionally and cognitively to the information they have consumed (Raney & Ji, 2017). That is, their specific emotional reactions are formed, including whether they think the message is positive, neutral or negative (i.e., the valence they processed from the message).

Such emotion reaction during the TV program viewing is also mentioned in the study of Coulter (1998). The author named it as “program induced affect” and claimed it to be a temporary feeling. This temporary feeling is also believed to have an impact on people’s program liking, which represents people’s overall evaluations of the TV program. Therefore, it can be assumed that the valence one perceives from the information on the second screen may influence one’s evaluations of the TV program. The hypothesis is thereby proposed: H1: The valence people perceive while reading information on the second screen will affect their evaluations of the TV program.

The direction of the effect was not clarified in the hypothesis because the multiple underlying mechanisms of this relationship are assumed to have opposite effects on people’s TV program evaluations. Therefore, it is difficult to anticipate the effect direction explicitly. The underlying mechanisms will be explained elaborately in the following part. Underlying mechanisms: Hedonic enjoyment and eudemonic appraisal

As discussed in the former section, specific valence is formed when people consume information on the second screen during the reception process. This process is emphasized in the model of socially shared TV viewing experience as the main difference from Vorderer, Klimmt, & Ritterfeld’s (2004) entertainment model. That is, unlike in entertainment model where people are only assumed to enjoy the media consumption process hedonically, Raney & Ji (2017) stated that people could also appreciate it eudemonically. Moreover, whether they enjoy it hedonically or appreciate it eudemonically is profoundly influenced by their emotion reactions to the information they processed from the content on both screens. Hedonic enjoyment and eudemonic appraisal, therefore, are considered as two underlying mechanisms of the effects of perceived valence in the current study and will be explained in this section.

In the psychology field, people’s subjective well-being has also been categorized into two types: hedonism and eudemonism. The former one reflects the importance of happiness and pleasure of well-being, whereas the latter one distinguishes well-being from happiness and emphasizes the fulfillment and realization of one’s true nature (Ryan & Deci, 2001).

Well-being was initially equated with hedonic pleasure for a long time in psychology, and psychologists asserted the principal target of well-being to be maximizing happiness. Accordingly, the presence of positive mood and the absence of negative mood were taken as two essential indicators when measuring well-being. Later on, the dominance of

hedonism was questioned by more and more scholars. They argued that well-being should be distinct from happiness. According to the self-determination proposed by Ryan & Deci (2000), autonomy, competence, and relatedness should be three vital components of a person’s well-being. Unlike hedonism, negative affects may also lead to higher well-being. Similarly, in the communication field, entertainment experience was initially believed to be purely about hedonic enjoyment. People’s entertainment media selection was initially

considered to be solely out of pleasure-seeking motivations, and scholars also noticed that such pleasure-seeking gratifications could not explain people’s consumption and enjoyment of numerous kinds of entertainment materials. For instance, tear-jerkers like tragic films, sad TV series are not entertaining, yet people may also enjoy them (Oliver & Raney, 2011). To distinguish this different kind of entertainment experience from traditional pleasure-seeking enjoyment (i.e., hedonic enjoyment), Oliver (2008) named it as eudemonic appraisal. Raney & Ji (2017) further explained that hedonic enjoyment could be “characterized by a distinct increase in positive affect, especially pleasure and delight” (pp.3). Eudemonic appraisal, on the other hand, is more related to people’s higher order needs, like seeking the meaning, truth, and purposes of life. These higher order needs are believed to be able to facilitate one’s self-autonomy in the long run (Smallenbroek, Zelenski, & Whelan, 2017). Furthermore, compared with hedonic enjoyment which is merely associated with positive, pleasure

valence, eudemonic appraisal is usually more associated with negative valence. It is explained by Oliver (2008) that after the negative valence is elicited and experienced by viewers, corresponding mood repair and other tangible benefits appear correspondingly, and through this process, people perceive the eudemonic appraisal.

Evaluation and enjoyment are both frequently mentioned in multi-screening studies. These two concepts have also been tested to be positively related to each other in some of the studies. For instance, in the study of Chinchanachokchai, Duff, & Sar (2015), it is found that those who enjoyed multitasking more gave higher evaluations to the advertisements they consumed. However, in these studies, evaluations refer to people’s attitudes toward the media content (e.g., TV programs or advertisements), while the enjoyment (also named explicitly as task enjoyment in some studies) refers to people’s overall feelings about the multi-screening behavior. It should be noticed that, the hedonic enjoyment discussed in the current study is different from the enjoyment (or task enjoyment) in the previous studies. Specifically, hedonic enjoyment in this study refers to the pleasure-seeking-related enjoyment one pursues from the media content. And same as the former studies, TV program evaluations in this study also refers to people's attitudes toward the TV program they consume on the first screen.

In the present study, TV program evaluation, taken as a result of program consumption, is assumed to be closely related to people’s hedonic enjoyment and eudemonic appraisal. We should notice that during the social TV usage, TV is the first screen because all the second screen activities are based on and related to the TV program. Second screen, from where people obtain extra information about the TV program, should be considered as a facilitator to enhance one’s TV program liking (Winter, Krämer, Benninghoff & Gallus, 2018). Therefore, the hypotheses are formulated as follows:

H2: People who perceive more positive valence from the second screen will pursue higher hedonic enjoyment and will, therefore, give higher TV program evaluations, than people who perceive less positive valence.

H3: People who perceive more negative valence from the second screen will pursue higher eudemonic appraisal and will, therefore, give higher TV program evaluations, than people who perceive less negative valence.

Unlike TV program evaluations, TV program commitment is claimed to be a long-term attitudinal disposition. In marketing, committed consumers are assumed to have higher resistance to attitude changes and keep their preferences to the products they are committed to. Similarly, TV program commitment is defined as “a viewer’s long-term attitudinal disposition toward a program, often reflecting an emotional or psychological attachment to the program” (Lin, Sung, & Chen, 2016, p. 172). Viewers with high TV program commitment may be more willing to keep watching the programs as long as possible, and even turn the program viewing into a ritual (Lin, Chen, & Sung, 2018).

Commitment is also claimed to be one of the eudemonia’s long-term trait abilities (Huta & Waterman, 2014). It is further explained by the authors that compared to hedonia which only pursue the pleasure of the moment; eudemonia requires perseverant growth and excellence. When it is specified in entertainment experience, by analogy, we could also assume that viewers who process eudemonic appraisal from TV program may be more willing to continue watching it in order to extend their self-growth and truth-searching experience. In

comparison, those who process hedonic enjoyment may only concentrate on the pleasure affect they sense at the moment and consider less about whether to continue watching the program in the future. Therefore, hypotheses are formulated as follows:

H4: People who perceive more negative valence from the second screen will pursue higher eudemonic appraisal and, therefore, are more likely to be committed to the TV program, than people who perceive less negative valence.

Attention division during Social TV usage

According to the threaded cognition theory proposed by Salvucci & Taatgen (2011), during multitasking, people’s cognition system attempts to attribute different cognitive resources and to satisfy various goals simultaneously. However, when several tasks require the same cognitive resources, some tasks may not be accomplished due to the lack of resources. Social TV usage, similar to other multi-screening behaviors, also entails people to divide their attention when both screens require same cognitive resources (i.e., visual

make recording one’s viewing behavior objectively possible. Actually, such devices have been frequently used in marketing and advertising in recent years (e.g., Boerman, Van Reijmersdal, & Neijens, 2015; Segijn, Voorveld, Vandeberg, & Smit, 2017). To better understand people’s social TV usage, eye-tracking device is also used in the current study. Besides, it would also be interesting to explore if how people divide their attention would affect their reactions to the social TV usage, including the valence they perceive from the second screen, their evaluations of and the commitment to the first screen content. To this end, an extra research question is formulated as follows:

RQ2: Does people’s attention division during social TV usage affect their reactions to social TV usage?

Fig 1. Conceptual Model: effects of perceived valence on TV program evaluations and commitment via hedonic enjoyment and eudemonic appraisal.

Method Sample

A total of 132 people participated in the experiment (Mage = 21.15, SDage = 3.86, 73% female). They were recruited through an online subject pool at the University of Amsterdam

Hedonic

Enjoyment

Valence

TV program

Evaluations

H1

H2

H3

Eudemonic

Appraisal

TV Program

Commitment

H4

RQ2

Attention

Division

(70% communication science students, 25% students in other majors, 5% not enrolled as students). The experiment consisted of two parts: social TV consumption and filling out the questionnaire, and the total duration was approximately 30 minutes per person. During the social TV consumption part, eye-tracker data was logged for every participant. Each participant was paid five euros or research credits as an incentive for their participation. Design and procedure

A quasi-experiment was employed in this study. All independent variables were

measured. The experiment was conducted in a lab which was designed to simulate a living room in order to create a true-to-life environment. One valid concern of lab experiment design is whether it truly reflects people’s everyday multitasking behaviors. That is, whether the ecological validity of the experiment is satisfying or not (Segijn, Xiong, & Duff, 2018). According to former studies, designing the lab to look like a living room is believed to be able to enhance the ecological validity (e.g., Segijn, Voorveld, & Smit, 2016). Before the start of the experiment, participants were first asked to read and sign the informed consent form. The instructions and eye-tracking glasses were then provided, and the experiment started when the participant finished reading the instructions, and the eye-tracking glasses was calibrated. Participants were informed the study was about what people look at during multi-screening, and they were allowed to use the tablet to browse an online magazine while watching TV. Eye-tracking glasses were used in this study to measure participants’ actual multi-screening behavior (see Appendix). Changing TV channels or browsing other websites were forbidden during the experiment. When the social TV consumption is finished,

participants took off the eye-tracking glasses and were required to fill out a questionnaire on a computer in another cubical room. The questionnaire was about their feelings, attitudes, and evaluations of the content on both screens (i.e., TV and tablet).

Stimuli

Television. The TV program was played on a SHARP LC-40LE540E LCD color

Television. The editorial content consisted of a 4.5-minute TV program clip and a 2.5-minute commercial break.To avoid the possible bias resulting from familiarities of the media content,

the TV program and TV commercials were selected from an English-speaking country outside Europe. Both the selected TV program and commercials were originally produced and broadcasted in Australia in the past three years. The TV program clip was chosen from episode 8, season 3 of an Australian sit-com, Upper Middle Bogan. The chosen part of this episode included a Christmas tree decoration and a conversation between a father and his son. Sit-com (i.e., Upper Middle Bogan) is selected as the stimuli in the present study for the following reasons: first, compare to other genres of TV programs, sit-coms require less knowledge of the previous storylines and, therefore, are more accessible for the audience to be involved. Second, sit-coms usually do not contain high-arousal plots and are less likely to cause any physical or mental un-comfortabilities. More specifically, the program Upper Middle Bogan was chosen because it is widely popular and evaluated to be in good quality in Australia. Besides, multiple in-depth reviews are also provided on its official website, among which the corresponding second screen content can be selected as well.

Tablet. Every participant was given a Samsung Galaxy Tab 3 10.1 P5210 Wi-Fi White tablet to browse the digital magazine during the experiment. People’s everyday

multi-screening behavior is believed to be triggered by the emergence and prevalence of portable media devices, which usually refer to smartphones and tablets (Zhong, 2013). That is, people can conduct other tasks on portable devices such as smartphones or tablets while consuming the traditional single-purpose media device like TV. In the present study,

considering text information is used as the second screen content, a larger screen could be more suitable and comfortable for participants to read. On the other hand, it is also more accurate to track participants’ eye-movements on a bigger screen with eye-tracking glasses. Therefore, the tablet is chosen over the smartphone. A 12-page digital magazine was specially designed for the study, and it consisted of 4 articles and some local brand ads released in the past three years. All the articles were selected from the official website of ABC Australia, the broadcaster of the TV program Upper Middle Bogan. The target article was an in-depth review of Upper Middle Bogan, ans was placed as the first article of the magazine; the second article was about mental health, the third one was Australian local

food recommendation, and the last one was about global environment. People’s familiarity and interests, as well as the entertainment value of the content per se, have influences on one’s multitasking processing (Segijn, Xiong, & Duff, 2018). Undergraduate students, which is the majority of participants in the current study, are believed to be highly involved in all the topics selected for the magazine in this study. Thus, participants were expected to switch their attention between the two screens during the multi-screening. Participants could scroll through the magazine freely during the social TV consumption.

The valence manipulation of the in-depth review about Upper Middle Bogan article was pretested in an online questionnaire (N = 26, M = 3.61, SD = 1.11). The pretest was to ensure people perceive various valences from reading the article. In the pretest, participants were asked to read the in-depth review and report their perceived valences.

Variables

The questionnaire included questions about attitudes, evaluations, and feelings. The questions were displayed in the following order to minimalize the possible

cross-contamination among the variables: hedonic enjoyment, eudemonic appraisal, TV program commitment, TV program evaluation, perceived valence, and some demographic questions. Independent variables

Gaze duration. Gaze duration of both TV program and the target article were measured with eye-tracker glasses. The eye tracker glasses (SMI Eye Tracking Glasses 2 Wireless) used in the study was able to log data related to both first and second screens. All the eye-tracking videos were decoded in Observer XT 13. Gaze duration of the clip of the TV program and the in-depth review was measured.

Perceived valence. Perceived valence was measured with a 10-point bipolar item measure (i.e., “how positive or negative would you rate the content of this article”) ranging from 1 = “totally negative” to 10 = “totally positive” (M = 7.20, SD = 4.71) adjusted from the measure developed by Hendriks, de Bruijn, & van den Putte (2012). The continuous scale is chosen over categorical grouping out of the following reasons. To begin with, median splits may cause the loss of individual variability. That is, classifying an individual in the “high

group” even if he/she is only slightly above the median or in the “low group” just because he/she is slightly lower than the median. Such classification is claimed to be problematic considering it increases the possibility of type 2 error. Moreover, some scholars also

declared that median splits could also increase the likelihood of type 1 error. Specifically, it is argued that in some cases significant relationships can only be found when the variable is dichotomized rather than continuous (Iacobucci, Posavac, Kardes, Schneider, & Popovich, 2015). Therefore, to enhance the accuracy of the data analysis, the continuous scale is used in this study instead of classifying participants into different groups by dichotomizing

perceived valence.

Dependent variables

TV program evaluation. TV program evaluation was measured with a 3-item 7-point semantic differential scale (M =4.52, SD = 1.25, Cronbach’s alpha = 0.93) modified for the purpose of this study adapted from the scale developed by Chang & Thorson (2014). The bipolar ends included “likable/ not very likable”; “interesting/ not interesting”; “good/ bad”; and “appealing/ not appealing” (Chang and Thorson 2004).

TV program commitment. TV program commitment was assessed by using a 7-item scale (M = 2.00, SD = 1.04, Cronbach’s alpha = 0.95) developed by Lin, Sung, & Chan (2016) (e.g., “I want my relationship with this TV program to last a long time”, “I am

committed to maintaining my relationship with this TV series”). Response categories ranging from “strongly disagree” to 7 = “strongly agree”, indicated participants’ degree of agreements to the given statements.

Mediators

Hedonic enjoyment. Hedonic enjoyment was measured with three items (M = 4.80, SD = 1.45, Cronbach’s alpha = 0.96) on a 7-point scale ranging from 1 = “strongly disagree” to 7 = “strongly agree”. The items were “The program was entertaining”; “I had a good time watching the program”; “I enjoyed watching the program” (Tukachinsky & Eyal, 2018). Eudemonic appraisal. Eudemonic enjoyment was measured with five items (M = 2.62, SD = 1.16, Cronbach’s alpha =0.90) on a 7-point scale ranging from 1 = “strongly disagree”

to 7 = “strongly agree”. The items were “The program made me focus on meaningful human conditions”; “The program made me more reflective”; “The program made me think”; “The program was moving”; and “The program challenged the way I see things” (Tukachinsky & Eyal, 2018).

Demographic variables

Demographics including the gender (27% male, 73% female), education level (72.4% primary/secondary school; 21.6% bachelor’s degree, 3.7% master’s degree, 2.2% others), and self-reported English proficiency (M = 6.02, SD = 0.85) were asked in the end of the questionnaire.

Results Control variables

The control variables were gender, education levels, and self-reported English proficiency. Linear regression was conducted to test if gender and self-reported English proficiency have influences on both TV program evaluations and TV program commitment (see Table 1). One-way ANOVA was conducted to test if education level was related to TV program evaluation (F (2, 130) = 1.52, p= 1.22) and commitment (F (2, 130) = 2.28, p=0.11). No consistent effects on the two dependent variables were found. Therefore, the variables do not need to be controlled during the hypothesis testing.

Dependent Variable Independent variable b t-value p-value

TV program Evaluations Gender .03 .12 .90 Self-reported English proficiency .25 1.81 .07 TV program Commitment Gender .23 1.14 .26 Self-reported English proficiency -.22 -2.06 .05

Table 1. linear regression between control variables including gender and self-reported English proficiency and dependent variables including TV program evaluations and commitment.

The influence of perceived valence on TV program evaluations

Hypothesis 1 stated that the valence viewers perceived from reading information on the second screen could affect their evaluations of the TV program on the first screen. The linear regression was conducted to test the hypothesis, and results showed a significant causal relationship between perceived valence and TV program evaluation. Participants who perceived more positive valence gave higher evaluations than those who perceived less positive valence (b = 0.05, beta = 0.17, p < 0.05). Hypothesis 1 is thus supported. Mediation effects via hedonic enjoyment and eudemonic appraisal

Hypothesis 2 states that viewers who perceived more positive valence, compared to those who perceived less positive valence, pursue higher hedonic enjoyment, and thus give higher evaluations to the TV program. Hypothesis 3 states that viewers who perceive more negative valence, compared to those who perceived less negative valence, pursue higher eudemonic appraisal, and may, therefore, give higher evaluations to the TV program. The parallel mediation effects were tested with PROCESS model 4 of Hayes (Hayes, 2013). PROCESS was designed based on ordinary least squares (OLS) or logistic regression, by using which researchers can easily estimate both direct and indirect effects in various mediation models, including bootstrap confidence intervals (Hayes, 2013). This study used PROCESS with 1,000 bootstrap samples to estimate the lower and upper confidence intervals (LLCI, ULCIs). Results of the model showed a significant indirect effect of perceived valence to TV program evaluations (indirect effect = 0.15, boot SE = 0.07, 95%LLCIULCI = [0.00, 0.27]). Participants perceived more positive valence pursue higher hedonic enjoyment (b = 0.06, p = 0.03), and gave higher evaluations to the TV program (b = 0.64, p < 0.001). Therefore, hypothesis 2 is supported (see Fig 2a). No significant effect of valence on eudemonic appraisal was found, and eudemonic appraisal was also tested to

have insignificant effects on TV program evaluation. Therefore, hypothesis 3 is rejected (see Fig 2a).

Hypothesis 4 states that viewers who perceived more negative valence, compared to those who perceived less negative valence, pursue higher eudemonic appraisal, and may have higher commitments to the TV program. The results of the PROCESS model showed an insignificant indirect effect of perceived valence to TV program evaluation. Therefore, hypothesis 4 is rejected (see Fig 2b). However, it should be noticed that the effect of

perceived valence on eudemonic appraisal was tested to be insignificant, yet it was showed that viewers who pursue higher eudemonic appraisal did have higher commitments to the TV program (b = 0.37, p < 0.001). a). b).

Perceived

Valence

Hedonic

Enjoyment

TV program

Evaluations

Eudemonic

Appraisal

0.06

*

0.64**

_*

0.04

0.03

0.05

*

Perceived

Valence

Eudemonic

Appraisal

TV program

Commitment

0.04

_0.37**

*

-0.00

Attention

Division

Figure 2. indirect effects of perceived valence on TV program evaluations and TV program commitment via hedonic enjoyment and eudemonic appraisal (b-values provided).

Note: *** p< 0.001, ** p < 0.01, * p< 0.05. Attention division

The eye-tracker data were analyzed in order to obtain objective data of people’s gaze duration on both screens during their social TV usage. It is showed by the result that during the 4.5-minute-long TV program clip was playing, people spent around 30 seconds more on TV (M = 1.87. SD = 0.83) than on the target article (M = 1.34. SD = 0.90).

A linear regression was conducted to test if the gaze duration of the second screen article has an influence on one’s perceived valence, and a significant causal relationship was found (b= 1.26, beta= 0.23, p= 0.01)1. Participants who spent more time reading the target article perceived more positive valence.

It is also interesting to explore if people’s first screen gaze duration could affect their evaluations of and commitment to the TV program. Linear regression was conducted, and results showed that both TV program evaluations (b= 0.29. beta= 0.19, p= 0.04) and commitment (b= 0.35, beta= 0.28, p= 0.002) could be influenced by the TV program gaze duration.

Discussion

The purpose of this study was to investigate how the valence people perceive from the second screen affect their evaluations of and commitment to the TV program on the first screen during social TV usage. Unlike most existing studies, this study focused on less interactive informational second screen content. The results indicated that perceived valence has a positive influence on people’s evaluations of the TV program they consume on the first screen. More specifically, hedonic enjoyment appeared to be the underlying mechanism of such effect. People who perceive more positive valence from the content on the second screen, compared to those who perceive less positive valence, pursue higher hedonic

1 _{The model was also tested with PROCESS Model 81 in order to include the attention division in the model as}

well, yet the results were very similar to the results of Model 4. Therefore, results of Model 81 were not reported.

enjoyment and thus give higher evaluations to the TV program on the first screen. On the other hand, although the mediation effect of eudemonic appraisal on TV program evaluation was tested to be insignificant, eudemonic appraisal was found to have a significant influence on one’s commitment to the TV program. That is, those who pursue higher eudemonic appraisal may have higher commitments to the TV program.

The findings of the present study provide four valuable contributions to social TV literature. First, consistent with previous research, the results of this study supported the importance of valence in media entertainment in general. Not only positive valence per se is believed to be gratifying during media consumption, but it is also one of the rewarding feelings which have a short-term contribution to personal well-being (Bartsch, 2012). Furthermore, positive valence has been tested to be more likely to improve one’s general enjoyment of multi-screening behavior and thus, give better evaluations of the media content as well (Rubenking, 2017). The findings of this study also corroborate Raney & Ji’s (2017) postulation that multi-screening is a dual-process which people’s first and second screen media consumptions interact with each other. More importantly, the researchers only emphasized and discussed interactive second screen media activities such as social media usage and instant online messaging, whereas the current study suggested that less

interactive informational second screen content also has an impact on people’s first screen media consumption experience. More specifically, the present study further provides insights into how first and second screen activities are generated by elaborating hedonic enjoyment being the underlying mechanism.

Second, although scholars in both communication and psychology fields argued that hedonic enjoyment/hedonism has been overemphasized and call for more attention to eudemonic appraisal/eudemonism (Oliver & Raney, 2011; Ryan & Deci,2001), the pivotal role hedonic enjoyment plays during people’s media consumption should not be neglected (Oliver & Bartsch, 2010; Nabi & Krcmar, 2004). In agreements with this argument, the current study also provides strong evidence for it. It is suggested that not only do people

pursue higher hedonic enjoyment when they perceive positive valence during social TV consumption, they also give higher evaluations to the media content afterward.

Unlike evaluations of the TV program, which is stated to be an instant liking, TV program commitment is a long-term attitudinal perception. In entertainment industries, commitment is usually connected with audiences’ loyalty and has considered as one of the most prominent factors of program producing (Lin, Sung, & Chen, 2016). On the contrary, it received much less attention in the academic field. The present study is innovative as it added TV program commitment in a media effect model, and furthermore, indicated that one’s commitment to a TV program could be affected by how much eudemonic appraisal he/she pursued from it. Such finding also offers a valuable practical implication for broadcasters and producers in the entertainment industries. That is, in the long run, viewers are more likely to be committed to the TV programs which enable them to see the truth and meaning of life, even if they may not obtain gratifications immediately.

Fourth, the present study has also contributed to entertainment communication by using objective eye-tracking data to operationalize viewers’ social TV usage. Eye-tracking data has been widely used in the marketing and advertising fields to measure people’s visual behaviors during multi-screening (e.g., Boerman, Van Reijmersdal, & Neijens, 2015; Segijn, Voorveld, & Smit, 2016), whereas in most entertainment studies, people’s social TV usage was often measured with self-reported data, which is argued by many researchers to be subjective and less accurate (e.g., Oviedo, Tornquist, Cameron, & Chiappe, 2015; Harboe et al, 2008). The eye-tracking data of this study showed that more attention is paid on the first screen than the second screen during social TV usage, which corroborates Valkenburg, Peter, & Walther’s (2016) proposition that media modality can be one of the main predictors of media use and effects. In their study, the authors claimed that media modality (e.g., text, auditory, and visual), as one type of media properties, could be a predictor of media effects. Many empirical media comparison studies also implied that information delivered through different modalities have various effects. For example, information delivered through textual modality encouraged higher reading and imagination skills than the identical information

delivered through audiovisual modality (Valkenburg, Peter, & Walther, 2016). The present study provides evidence that screens (i.e., TV) with more modalities is more likely to attract viewers’ attention. This finding expands the existing knowledge by implying that although media with more modalities encourage less reading and imagination skills, people may still pay more attention to it.

Limitations and future directions

Notwithstanding the valuable contributions the current study have made, it also has several limitations. First, an experiment was conducted in a lab with a living room setting in the present study. Such design, compared to most traditional experiments, is more true-to-life and thus improves the ecological validity. However, there are still possibilities that people may act differently in their daily lives (Segijn, Xiong, & Duff, 2018). Observation, on the other hand, enables researchers to know more about daily social TV consumption behaviors, and can further enhance the ecological validity of the research. Therefore, to know more details about people’s social TV usage, observational studies could be conducted in the future. Second, the current study only tested the hypotheses by using one genre of TV program as a stimulus in the experiment. It is argued by some scholars that people’s reactions vary from different genres of TV programs (e.g., Mikos, 2016; Nabi & Krcmar, 2004). More specifically, people may pursue less eudemonic appraisal from sit-coms compared to drama or sci-fi series. It can also be the explanation for the insignificant mediation effect of

eudemonic appraisal in the present study. Thus, to arrive at a more thorough understanding of such effects, it is valuable to distinguish among different genres of TV programs in future studies.

Third, the effects of more emotions on social TV usage could be tested in future studies. Since the present study has proved the crucial role valence plays in social TV usage, future studies could first examine the valence effect as a good start point, and more steps could be taken afterward. Previous studies have found that one may feel multiple emotions

simultaneously when consuming specific media content. For example, some people stated that they feel both happy and sad when watching the movie Life is beautiful (Larsen, Norris,

McGraw, Hawkley, & Cacioppo, 2009). Moreover, multiple emotion scales have been

developed, with which much more than one dimension of emotions could be measured (e.g., PANAS model, SAM model). Therefore, in future studies, researchers could take a further look at different emotion dimensions (including valence, arousal or even dominance), and to see if they have any influence on people’s social TV usage.

Practical Implications

The present study provides several valuable practical implications for program producers in the entertainment industry. First, it is showed that how people react to the media content depends on how much attention they pay to it. Therefore, to enhance audiences’ evaluations of and commitment to the TV program, producers should first ensure that people have paid enough attention to the TV program on the first screen during their social TV usage.

Furthermore, as long as the second screen content does not contain more modalities than the first screen content, according to the implication of the current study, people are not likely to be distracted by the second screen contents from the TV program. Therefore, providing textual or visual second screen content could be a good choice.

Moreover, the study has also proved that the valence people perceived from the second screen have a positive influence on their evaluations of the TV program via enhancing the hedonic enjoyment. Program producers could thus consider offering relevant, entertaining second screen content to the audience. For instance, providing fun related reports, facts on online websites, fan pages, or apps. Second, the result regard to eudemonic appraisal and TV program commitment is also enlightening. For most TV programs which have more than one seasons, audiences’ commitment to the program is one of the most crucial factors producers need to consider. The current study implies that, seeking the truth and meaning of life, and obtain self-growth and autonomy can, in the long run, enhance audiences’

Reference

Bailey, E. J., & Ivory, J. D. (2018). The moods meaningful media create: Effects of hedonic and eudaimonic television clips on viewers’ affective states and subsequent program selection. Psychology of Popular Media Culture, 7(2), 130.doi:10.1037/ppm0000122 Bartsch, A. (2012). Emotional gratification in entertainment experience. Why viewers of

movies and television series find it rewarding to experience emotions. Media Psychology, 15(3), 267-302. doi:10.1080/15213269.2012.693811

Boerman, S. C., Van Reijmersdal, E. A., & Neijens, P. C. (2015). Using eye tracking to understand the effects of brand placement disclosure types in television programs. Journal of Advertising, 44(3), 196-207. doi:10.1080/00913367.2014.967423

Ceron, A., & Splendore, S. (2018). From contents to comments: Social TV and perceived pluralism in political talk shows. New Media & Society, 20(2), 659-675.

doi:10.1177%2F1461444816668187

Coulter, K. S. (1998). The effects of affective responses to media context on advertising evaluations. Journal of Advertising, 27(4), 41-51. doi:

10.1080/00913367.1998.10673568

Chang, Y., & Thorson, E. (2004). Television and web advertising synergies. Journal of Advertising, 33(2), 75-84. doi:10.1080/00913367.2004.10639161

Chinchanachokchai, S., Duff, B. R., & Sar, S. (2015). The effect of multitasking on time perception, enjoyment, and ad evaluation. Computers in Human Behavior, 45, 185-191. doi:10.1016/j.chb.2014.11.087

Choi, I., Koo, M., & Choi, J. A. (2007). Individual differences in analytic versus holistic thinking. Personality and Social Psychology Bulletin, 33(5), 691-705. doi: 10. 1177%2F0146167206298568

Hendriks, H., de Bruijn, G. J., & van den Putte, B. (2012). Talking about alcohol consumption: Health campaigns, conversational valence, and binge drinking intentions. British journal of health psychology, 17(4), 843-853. doi:10.1111/j.2044-8287.2012.02080.x

Harboe, G., Metcalf, C. J., Bentley, F., Tullio, J., Massey, N., & Romano, G. (2008, April). Ambient social tv: drawing people into a shared experience. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 1-10). ACM. doi: 10.1145/1357054.1357056

Huta, V., & Waterman, A. S. (2014). Eudaimonia and its distinction from hedonia: Developing a classification and terminology for understanding conceptual and operational definitions. Journal of Happiness Studies, 15(6), 1425-1456. doi: 10.1007/s10902-013-9485-0

Iacobucci, D., Posavac, S. S., Kardes, F. R., Schneider, M. J., & Popovich, D. L. (2015). Toward a more nuanced understanding of the statistical properties of a median split. Journal of Consumer Psychology, 25(4), 652-665. doi:

10.1016/j.jcps.2014.12.002

audience factors. Media Psychology, 10(3), 364-384. doi:10.1080.15213260701532948

Krämer, N. C., Winter, S., Benninghoff, B., & Gallus, C. (2015). How “social” is Social TV? The influence of social motives and expected outcomes on the usage of Social TV applications. Computers in Human Behavior, 51, 255-262.

doi:10.1016/j.chb.2015.05.005

Larsen, J. T., Norris, C. J., McGraw, A. P., Hawkley, L. C., & Cacioppo, J. T. (2009). The evaluative space grid: A single-item measure of positivity and negativity. Cognition and Emotion, 23(3), 453-480. doi:10.1080/02699930801994054

Lim, J. S., Hwang, Y., Kim, S., & Biocca, F. A. (2015). How social media engagement leads to sports channel loyalty: Mediating roles of social presence and channel

commitment. Computers in Human Behavior, 46, 158–167. doi:10.1016/j.chb.2015.01.013

Lin, J. S., Sung, Y., & Chen, K. J. (2016). Social television: Examining the antecedents and consequences of connected TV viewing. Computers in Human Behavior, 58, 171– 178. doi:10.1016/j.chb.2015.12.025

Lin, J. S., Chen, K. J., & Sung, Y. (2018). Understanding the Nature, Uses, and

Gratifications of Social Television: Implications for Developing Viewer Engagement and Network Loyalty. Journal of Broadcasting & Electronic Media, 62(1), 1-20. doi: 10.1080/08838151.2017.1402904

Mikos, L. (2016). Digital media platforms and the use of TV content: Binge watching and video-on-demand in Germany. Media and Communication, 4(3), 154-161. doi: 10.17645/mac.v4i3.542

Mukherjee, P., & Jansen, B. J. (2014). Social TV and the social soundtrack: significance of second screen interaction during television viewing. In International Conference on Social Computing, Behavioral-Cultural Modeling, and Prediction (pp. 317-324). Springer, Cham.

Nabi, R. L., & Krcmar, M. (2004). Conceptualizing media enjoyment as attitude: Implications for mass media effects research. Communication Theory, 14, 288–310. doi:

10.1111/j.1468- 2885.2004.tb00316.x

Nielsen, (2016, January 20). Nielsen to measure total program-related activity across Twitter and Facebook. Nielsen Center. Retrieved from

http://sites.nielsen.com/newscenter/nielsen- to-measure-total-program-related-activity-across-twitter-and-facebook/

Oliver, M. B. (2008). Tender affective states as predictors of entertainment preference. Journal of Communication, 58, 40–61. doi: 10.1111/j.1460-2466.2007.00373.x Oliver, M. B., & Bartsch, A. (2010). Appreciation as audience response: Exploring

entertainment gratifications beyond hedonism. Human Communication Research, 36, 53–81. doi: 10 .1111/j.1468-2958.2009.01368.x

Oliver, M.B. & Raney, A. (2011). Entertainment as pleasurable and meaningful: Identifying hedonic and eudaimonic motivations for entertainment consumption. Journal of

Oviedo, V., Tornquist, M., Cameron, T., & Chiappe, D. (2015). Effects of media multi-tasking with Facebook on the enjoyment and encoding of TV episodes. Computers in Human Behavior, 51, 407-417. doi: 10.1016/j.chb.2015.05.022

Proulx, M., & Shepatin, S. (2012). Enhancing TV with Synched Content Experiences In Proulx, M., & Shepatin, S. Social TV: how marketers can reach and engage

audiences by connecting television to the web, social media, and mobile. (pp.95-96). John Wiley & Sons.

Raney, A. A., & Ji, Q. (2017). Entertaining each other? Modeling the socially shared

television viewing experience. Human Communication Research, 43(4), 424-435. doi: 10.1111/hcre.12121

Rubenking, B. (2017). Boring is bad: Effects of emotional content and multitasking on enjoyment and memory. Computers in Human Behavior, 72, 488-495. doi: 10.1016/j.chb.2017.03.015

Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A review of research on hedonic and eudaimonic well-being. Annual review of psychology, 52(1), 141-166. doi: 10.1146/annurev.psych.52.1.141

Segijn, C. M. (2016). Second screen advertising: A typology of multiscreening. In Advertising in new formats and media: Current research and implications for marketers (pp. 77-96). Emerald Group Publishing Limited.

Segijn, C. M., Voorveld, H. A., & Smit, E. G. (2016). The underlying mechanisms of multiscreening effects. Journal of Advertising, 45(4), 391-402. doi:

10.1080/00913367.2016.1172386

Segijn, C. M., Voorveld, H. A., Vandeberg, L., & Smit, E. G. (2017). The battle of the screens: Unraveling attention allocation and memory effects when multiscreening. Human Communication Research, 43(2), 295-314. doi: 10.1111/hcre.12106

Segijn, C. M., Xiong, S., & Duff, B. R. (2018). Manipulating and Measuring Media

Multitasking: Implications of Previous Research and Guidelines for Future Research. Communication Methods and Measures, 1-19. doi: 10.1080/19312458.2018.1555797

Smallenbroek, O., Zelenski, J. M., & Whelan, D. C. (2017). Authenticity as a eudaimonic construct: The relationships among authenticity, values, and valence. The Journal of Positive Psychology, 12(2), 197-209. doi: 10.1080/17439760.2016.1187198

Spiller, S. A., Fitzsimons, G. J., Lynch Jr, J. G., & McClelland, G. H. (2013). Spotlights, floodlights, and the magic number zero: Simple effects tests in moderated regression. Journal of marketing research, 50(2), 277-288. doi: 10.1509/jmr.12.0420

Sundar, S. S., & Limperos, A. M. (2013). Uses and grats 2.0: New gratifications for new media. Journal of Broadcasting & Electronic Media, 57(4), 504-525. doi:

10.1080/08838151.2013.845827

TC Lin, T. (2018). Multiscreen Social TV System: A Mixed Method Understanding of Users’ Attitudes and Adoption Intention. International Journal of Human–Computer

Interaction,35, 99-108. doi: 10.1080/10447318.2018.1436115

Valkenburg, P. M., Peter, J., & Walther, J. B. (2016). Media effects: Theory and research. Annual review of psychology, 67, 315-338. doi: 10.1146/annurev-psych-122414-033608

Van Cauwenberge, A., Schaap, G., & Van Roy, R. (2014). “TV no longer commands our full attention”: Effects of second-screen viewing and task relevance on cognitive load and learning from news. Computers in Human Behavior, 38, 100-109. doi:

10.1016/j.chb.2014.05.021

Vorderer, P., Klimmt, C., & Ritterfeld, U. (2004). Enjoyment: At the heart of media entertainment. Communication theory, 14(4), 388-408. doi: 10.1111/j.1468-2885.2004.tb00321.x

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief

measures of positive and negative affect: the PANAS scales. Journal of personality and social psychology, 54(6), 1063-1070. Retrieved from:

https://www.researchgate.net/publication/235726247_Development_and_Validation_ of_Brief_Measures_of_Positive_and_Negative_Affect_The_PANAS_Scales

Winter, S., Krämer, N. C., Benninghoff, B., & Gallus, C. (2018). Shared Entertainment, Shared Opinions: The Influence of Social TV Comments on the Evaluation of Talent Shows. Journal of Broadcasting & Electronic Media, 62(1), 21-37. doi:

10.1080/08838151.2017.1402903

Wirth, W., & Schramm, H. (2005). Media and emotions. Communication research trends, 24(3), 1-43. Retrieved from:

https://www.researchgate.net/profile/Holger_Schramm/publication/233387230_Media _and_Emotion/links/5817411008aeffbed6c30bd1/Media-and-Emotion.pdf

Zhong, B. (2013). From smartphones to iPad: Power users’ disposition toward mobile media devices. Computers in human behavior, 29(4), 1742-1748. doi:

Appendix

Pic 1

Pic 2

Pic 1 & 2: Participants wearing eye-tracker glasses, using tablets while watching TV in a lab designed and decorated as a living room.