Watching Videos Alone but Happy Together: The Effect of
Danmaku on Viewer’s Video Watching Experience
Master’s Thesis Yiyao Liu (11786507) Graduate School of Communication Mastertrack Entertainment Communication
Supervisor: Hande Sungur Word Count: 6,688
Abstract
Danmaku commentary is a novel comment technology which overlays timeline-synchronized user comments directly on the video and creates a unique co-viewing
experience. The present study conducted an online experiment to investigate the effect of the Danmaku density (i.e., the amount of available comments presented on a given time, high vs. low density) on enjoyment of video-watching experience. Social presence was investigated as a mediator between the Danmaku density and enjoyment. Viewer’s previous Danmaku watching experience was examined as a moderator between the Danmaku density and social presence. 112 native Mandarin speakers aged from 18 to 30 years old were recruited to take part in this study. The results showed that the high density Danmaku increased the enjoyment of watching videos, and this was mediated by social presence. When participants were
presented with a high density Danmaku video, they perceived higher degree of social presence and consequently found watching the video more enjoyable. While we expected to find that the effect of the Danmaku density on social presence is stronger for people with more Danmaku watching experiences than people with less Danmaku watching experiences. However, the interaction effect of the Danmaku density and viewer’s previous watching experience on social presence was not found.
Introduction
Traditionally, co-viewing experience refers to people watching TV together with their families or acquaintances and share watching experiences with each other. Nowadays, with the prevalence of the internet technology as well as the advanced electronic devices, it is common that people watch TV online by themselves instead of physically co-watching with other viewers (Harboe, 2009). However, for many online viewers, the co-viewing experience is still needed. Although, some online video platforms such as Tudou, Youku and YouTube allow users to communicate with each other by leaving comments below the videos, such interactions are not synchronous, because users are not able to share their feelings and get reactions instantly. In order to get the synchronous co-viewing experiences, viewers start to share their opinions and impressions by means of social media such as Twitter and WhatsApp when watching videos. A recent case study conducted by Nagy and Midha (2014) revealed that 72% of primetime-active Twitter users tweeted about TV shows while watching live on FOX TV, and 58% tweeted about TV shows while watching on other platforms such as Netflix, Hulu, iTunes or On Demand.
Recently, a novel commentary system called “Danmaku” which enables co-viewing experience emerged and became increasingly popular in China. Literally Danmaku means bullets in Japanese, because the overlaid comments are scrolling from right to left and then disappear on screen, resembling moving bullets. Danmaku commentary system was first introduced in 2006 by nicovideo.jp, the most famous video-sharing site in Japan, and then became increasingly popular in China (Johnson, 2013). Different from the traditional commentary system that only enables users to leave comments separately below the video, Danmaku allows users to watch videos and post comments synchronously (Shen, Chan, & Hung, 2014). Each posted comment is attached to the video’s timeline. When viewers watch a video to a certain timeline point, comments posted at that point of timeline will scroll on
screen (Ma & Ge, 2014). Once the comments are posted, they will be visible for both current and future viewers (Chu, 2009), see Figure 1 in Appendix A.
As the major Danmaku video site in China, Bilibili.com reported average monthly active users of 85 million, and the number is increasing with a 30 percent yearly (Yang, 2018). On November, 2018, the Alexa Rank of Bilibili.com was 72 in global and 18 in China
(Alexa.com, 2018). Alexa.com is a powerful database which provides a measurement of how a website ranks in a particular country relative to other sites. Due to Danmaku’s increasing popularity, next to Bilibili.com, most major video-sharing websites in China (e.g., Youku, Tudou, LeTV and Tencent) adopted Danmaku commenting in recent years. Watching
Danmaku video is particularly popular among young generations. Bilibili.com reports that 82% users are under 24 years old (Bilibili.com, 2018). For many Generation Z users in China,
watching Danmaku videos have already become a daily routine (He & Zhang, 2014). The wide spread of Danmaku commenting have recently gained attention from both researchers and the media industry. So far, studies on mediated co-viewing experiences mainly concentrate on social media co-viewing which typically includes studies about Twitter users’ behaviors and attitudes towards TV program Tweets while the program is playing (Nagy & Midha, 2014) as well as studies about the effect of TV viewers’ backchannel communication during sport event on their sports channel loyalty (Lim, Hwang, Kim, & Biocca, 2015). However, research into Danmaku video watching experience has been rather limited. Studies that looked into Danmaku viewing so far investigated topics such as viewer’s motivation of watching Danmaku videos (Chen, Gao & Rau, 2015) and perceived benefits of Danmaku in virtual customer environments (Cui, 2017). A study exploring the motivations for watching Danmaku videos (Chen et al., 2015) revealed that most viewers watch Danmaku videos to obtain information, for entertainment and social connectedness. However, the effect of Danmaku on viewer’s watching experience has not been examined thoroughly. Research
by Cui (2017) revealed that the effect of Danmaku on customer engagement is mediated by perceived benefits. But, this study focused exclusively on marketing implications rather than entertainment media. To address the gap, by taking individual differences into account, this study aims to examine the influence of Danmaku on viewer’s watching experience from the perspective of entertainment communication. This goal leads to the overarching research question of:
RQ: Does the presence of Danmaku affect viewer’s watching experience? Theoretical Background and Hypotheses
Enjoyment
Most communication researchers have defined media enjoyment as a pleasurable response or positive reaction to media use (Vorderer, Klimmt, & Ritterfiled, 2004; Zillmann & Bryant, 1994; Miron, 2003). Although studies on enjoyment employed different theoretical frameworks such as uses and gratifications theory (Katz, Blumber, & Gurevitch, 1974), disposition theory (Zillmann & Cantor, 1976) and sensitivity theory (Reiss & Wiltz, 2004), these approaches in general lack of consistent definition of media enjoyment. Alternative definitions have been used such as pleasure (Grodal, 2000), joy (Reiss & Wiltz, 2004) and liking (Zillmann & Cantor, 1976).
To address the issues related with the lack of conceptual clarification regarding media enjoyment, Tamborini and colleagues (2010) provide a broader definition of this construct in addition to mere pleasure seeking. By adopting self-determination theory (SDT) which was first postulated by Deci and Ryan (2000), Tamborini et al. (2010) define media enjoyment as the satisfaction of three main intrinsic psychological needs: autonomy, competence and relatedness. An association between media enjoyment and psychological well-being is recognized. This relationship indicates that enjoyment can be interpreted as a functional term (Ryan, Rigby, & Przybylski, 2006). The research conducted by Ryan et al. (2006) is the first
attempt to measure media enjoyment as the satisfaction of psychological needs, however, a formal validation of this conception is lacking in their study. Building upon the work of Ryan et al. (2006), Tamborini et al. (2010) conducted an experimental study to validate this
conception. Their findings offer scientific evidence for supporting the assumption of
conceptualizing media enjoyment as need satisfaction (Tamborini et al., 2010). In this respect, satisfaction of the functional needs can be used to understand the enjoyment of entertainment media.
SDT is a psychological theory of human motivation which concentrates on the degree to which individual behaviors are self-determined (Deci & Ryan, 2000). This theory
postulates that individuals are intrinsically motivated to conduct certain activities which satisfy their basic psychological needs. Suggested by Deci and Ryan (2000), psychological needs are defined as “innate psychological nutriments that are essential for ongoing
psychological growth, integrity, and well-being” (p. 229). Three main psychological needs are involved in self-determination: autonomy, the degree to which people feel in control of doing a task; competence, the degree to which people feel able to achieve goals; and relatedness, the willingness to interact and connect with others (Deci & Ryan, 2000).
The intrinsic need for relatedness can be satisfied when watching Danmaku videos. The co-viewing experience of Danmaku video provides viewers a feeling of connecting with anonymous others who are watching the same video at the present moment or probably have watched a couple of days ago (Chen et al., 2015). So it can be inferred that viewers are likely to enjoy Danmaku videos since their needs for relatedness are satisfied. In addition, media enjoyment was also defined as satisfying intrinsic needs by Nabi and Krcmar (2004), they suggested that entertainment media content can satisfy both hedonic and eudaimonic needs. Hedonic needs refer to the pleasure of entertainment, while content with meanings and drives people to think will fulfill eudaimonic needs (Oliver & Bartsch, 2010). Danmaku videos
contain both humorous comments that amuse viewers and thought-provoking expressions that make people think (Chen et al., 2015). Danmaku videos can satisfy both needs, and thereby lead to enjoyment of watching these videos. Bilibili.com allows users to adjust amount of comments. A large amount of Danmaku comments scrolling on screen indicates more viewers are watching the video, and then the feeling of being connected with other people is expected to be stronger. Also, a large number of Danmaku comments indicates more humorous and thought-provoking comments are displayed, thus the feeling of enjoyment is likely to be stronger. Based on these, this thesis hypothesizes:
H1: High level of Danmaku density will lead to higher levels of enjoyment compared to low level of Danmaku density.
Social Presence
Social presence theory was first developed by Short, Williams and Christie (1976) asserting that if a media can convey richer social cues, this medium would be considered as more personal, and thus could enhance social presence. Social presence was measured by the ability of a communications medium to project physical and emotional presence (Short et al., 1976). The concept of social presence was originally interpreted from cues-filtered-out perspective, face-to-face communication was seen as transmitting richest social cues, the importance of non-verbal cues such as facial expressions and eye contacts was stressed (Daft & Lengel, 1984). In general, earliest arguments of social presence emphasized characteristics of medium per se, however, with the rise of computer-mediated communication, social presence theory was forced to be adapted to the new medium (Tu, 2000).
The concept of social presence has been redefined by researchers as the feeling of a community a user experiences in computer-mediated environment (Tu & Mclssac, 2002). Other CMC scholars such as Biocca and Harms (2002) provide a more comprehensive definition by arguing that “social presence is the moment-to-moment awareness of
co-presence of a mediated body and the sense of accessibility of the other being’s psychological, emotional, and intentional states” (p.14). In line with this, Sivunen and Nordback (2015) pointed out that the degree of social presence does not solely depend on medium attributes, but the quality of communication affects social presence as well. Based on the findings of a 3D virtual environments study conducted by Sivunen and Nordback (2015), higher quality of communication fosters higher degree of social presence in virtual environment, and it is an ongoing process constantly varied time by time. The advanced communication technology is served as a solution to the problems caused by the lack of physical presence in computer-mediated environment.
Suggested by Kim, Kwon and Cho (2011), interactivity is an essential factor which determines the communication quality and thereby influences the degree of social presence. The evolved perception of social presence emphasizes the importance of interaction between people in their online environment (Dixson, Greenwell, Rogers-Stacy, Weister, & Lauer, 2016). Prior studies indicated that interaction plays an important role in achieving social presence especially when people are physically distant (Sivunen & Nordback, 2015). Social presence theory has been applied in distance education to obtain insights into user behaviors. Study conducted by Kim et al. (2011) indicated that students perceived higher levels of social presence when they are willing to actively engage and interact with other students in an online learning environment. In addition, Kendall and Kendall (2017) suggested that in online
classes, structuring collaborative learning activities can invigorate online interactions between students and consequently increase the level of social presence. Such empirical evidences support the assertion that interactivity takes an important role in affecting the level of social presence in computer-mediated communication process.
Social presence theory can be applied to understand Danmaku watching experience. Although Danmaku is a text-based application which is not transmitting non-verbal cues, the
advanced feature of Danmaku which is rarely seen from the traditional post-viewing
commenting videos facilitates interactivity among viewers. Danmaku comments are scrolling simultaneously with the video, enables users to actively interact with each other by sharing their opinions and getting instant responses. By posting comments, Danmaku viewers are able to engage in online conversation or observe interaction between other viewers either
synchronously or asynchronously (Ma & Ge, 2014). The scrolling comments capacitate viewers to sense the existence of others co-viewing the same video, therefore, it can be inferred that it’s possible for Danmaku viewers to perceive some degree of social presence through these interactions. Bilibili.com allows users to adjust the density of Danmaku comments, a large amount of Danmaku comments scrolling on screen may increase the presence of social cues and enhance the quality of social interaction. Thus, this thesis hypothesizes:
H2: High level of Danmaku density will lead to higher levels of social presence compared to low level of Danmaku density.
Danmaku technology enables users to add additional social commentary
simultaneously while watching videos so that users are able to feel like watching TV with others and then actively interact with each other (Shin, 2013). Prior studies suggested that the unique entertaining value of the socially enriched content is one important motivation for people to watch Danmaku videos (Chen et al., 2017). Most viewers mentioned that the interesting interactions among comments enable an entertaining experience for even boring videos (Chen et al., 2017). Thus, possibly viewers will perceive some degree of social
presence through interactions. The sense of existence of others makes viewers feel connected with other people, thus their intrinsic needs of relatedness will get satisfied, and therefore they are more likely to experience enjoyment while watching a Danmaku video. Considering these, this thesis hypothesizes:
H3: The effect of Danmaku density on enjoyment is mediated by social presence. A previous qualitative study conducted by Chen et al. (2015) revealed that Danmaku watching experience varies between viewers. Viewers who have less Danmaku video watching experience tend to find the amount of information distracting. However,
experienced Danmaku viewers find it less distracting and they find the co-viewing exciting because they feel like interacting with other viewers (Chen et al., 2015). This study takes viewer’s previous Danmaku video watching experiences into account, and regards this factor as a moderator for H2:
H4: The effect of Danmaku density on social presence is moderated by viewer’s previous experience of watching Danmaku videos.
The conceptual model for this study is as follows:
Figure 2. Conceptual Model
Method Design
An online experiment using survey software Qualtrics was conducted to investigate people’s Danmaku video watching experience. Participants were randomly assigned to two conditions (high density vs. low density Danmaku commentary). For the high density condition, participants received a Danmaku video clip with 70% of the total comments that scrolled from right hand side to left hand side on screen while participants were watching the
video clip. For the low density condition, participants received a Danmaku video clip with 30% of the total comments on the dense. This experiment was approved by the ethical committee of the communication science department of the University of Amsterdam.
Sample
A total of 155 participants were recruited to participate in this study. Since the
Danmaku comments were written in Chinese characters, all participants were native Mandarin speakers. This was an age-based sample, participants were in the range between 18 to 30 years old, and the briefing section also indicated that participants had to be between 18 to 30 years old. Because Bilibili.com (2018) reported that the average age of their users is 24 years old, this age range was considered appropriate for this study. 30 participants were dropped from the study because they did not complete the survey and additional 13 participants were excluded from the analysis because they failed to watch the Danmaku video. A total of 112 participants were retained in the final sample, the mean age was 26.29 (SD = 2.65) years old, 68 females (60.71%) and 44 males (39.29%) participated. Regarding to education background, the majority of the sample consisted of higher educated participants (64.3% had bachelor’s degree or equivalent and 29.5% had master’s degree or was currently enrolled in master’s program).
Procedure
Participants were recruited via WeChat platform with convenient sampling strategy with the researcher’s personal WeChat network as the starting point. Data were collected over a one-week timeframe starting in November 16th, 2018. The survey took approximately 7 to 10 minutes to complete. Participants started the online survey with a briefing, which
illuminated the purpose of this study, and the briefing was identical for all participants. They were assured that the data would be collected anonymously for scientific purpose only. After participants agreed to take part in this study and gave informed consent, a Danmaku video
clip followed by a short instruction was presented. Participants were asked to watch in full screen and informed to put on earphones or adjust the volume to get ready for watching the video. Each participant was randomly placed in one of the two conditions (high density vs. low density Danmaku commentary). After they were exposed to the video, participants were asked to fill in a questionnaire that measured the outcome variables, manipulation check and demographic information. After completing the questionnaire, each participant was thanked for their time with a WeChat red envelope worth 1 Euro as incentive.
Stimuli
The stimuli came from a Chinese TV drama called The First Half of My Life, a 42-episode drama about a group of ordinary people who are faced with the dilemma of choosing between love and career. This drama is an adaption of a novel by a reputed Hong Kong writer Isabel Nee Yeh-su. This drama is chosen as the stimuli because it captures ordinary people’s daily lives and therefore content wise is thought to be relatable to the participants. Moreover, this TV drama even though being still contemporary as it was released in 2017, has already receded in popularity. Therefore, it was assumed that participants would not be familiar with the content. This experiment chose a 3 minutes 37 seconds video clip that contained a complete scene as stimuli. This scene depicted a neutral conversion between the main characters Zijun and Han discussing about job selection.
The video clip was generated from Bilibili.com with accompanying Danmaku comments. Bilibili.com offers an advanced function which allows users to adjust the presentation of Danmaku, for instance, the speed of scrolling Danmaku, Danmaku text
transparency and Danmaku density. The original video clip contains 381 Danmaku comments in total. In order to manipulate the degree of Danmaku commentary, the Danmaku density was manipulated. In this study, high Danmaku density was defined as 70% of the total comments displayed when the video is playing, while 30% of the total comments displayed
was defined as low density. For instance, in certain scenes, this can lead to 21 comments scrolling simultaneously in the high Danmaku density condition, while only 3 comments appearing in the same scene in the low Danmaku density condition (see Figure 3 and Figure 4 in Appendix A). By setting the Danmaku density to 30% (low density) and 70% (high
density), 2 stimuli were created. Since the adjusted Danmaku video could not be downloaded directly from Bilibili.com, these two versions of stimuli were screen recorded by QuickTime Player. To ensure that no other factors would impact research constructs, the contents of the two videos were identical, and the only difference was the Danmaku density.
Measures
Enjoyment. This variable was measured by the enjoyment subset of the Intrinsic Motivation Inventory (IMI), a reliable and validated scale that measures self-reported media enjoyment (Tamborini et al., 2010). The original scale consists of 7 items on a 7-point Likert scale. The 7 items were adapted to fit the context of this study. The scale was shortened by deleting three items, because those items are identical in meaning with the remaining four items in Chinese expression. Participants were asked to indicate the level of agreement with the statements such as “I enjoyed watching this video very much” and “Watching this video was fun to do”. The answer scale ranged from 1 (strongly disagree) to 7 (strongly agree). Because two items were negatively framed (e.g. “I thought it was boring to watch this video”), they were both reverse coded. The reliability test showed a Cronbach’s alpha of .90. A scale was created to measure enjoyment by averaging the four items (M = 4.00, SD = 1.62). Social presence. Previous research has shown that social presence is a three dimensional construct (Ni, 2017). The first dimension of this construct is the perceived salience of other viewers, which is measured by a 4-items, 7 point Likert scale developed by Short et al. (2012). This dimension concentrates on a sense of engaging in face-to-face conversations between viewers (Lee & Shin, 2012). This scale has been applied to measure
social presence in a previous Danmaku watching study (Ni, 2017). Participants were asked to indicate the level of agreement with four statements, for example, “Having the comments appearing above the video, I felt as if other viewers were speaking directly to me”. The answer scale ranged from 1 (strongly disagree) to 7 (strongly agree). A scale was created by averaging the four items, resulting in Cronbach’s alpha of .86 (M = 4.57, SD = 1.23).
The second dimension focused on the degree to which viewers could understand others’ thoughts based on Danmaku comments presented. In addition, whether viewers could be affected by others’ emotions expressed through Danmaku comments was also addressed in this dimension. This dimension was measured with a 7-items 7 point Likert scale which was developed by Harms and Biocca (2004). Considering the context of this study as well as to ensure appropriateness to Chinese language, this scale was adapted into a 4-items 7 point Likert scale with the answer ranged from 1 (strongly disagree) to 7 (strongly agree). An example item is “Other viewers’ thoughts in their comments were clear to me”. The four items have a Cronbach’s alpha value of .80. By averaging the four items, a new scale was generated to measure this dimension (M = 4.92, SD = 1.09).
The third dimension is the sense of belongingness generated from watching Danmaku comments. It addresses whether viewers feel they are being part of a community. This measurement has 3 items on a 7-point Likert scale which was initially developed by Xu and Yan (2011). The scale has been applied to a prior Danmaku study measured to what extend viewers felt they belong to a large community (Ni, 2017). These three items were adapted based on the context of this study, for instance, “Watching this video gives me a strong sense of being in touch with other viewers”. The response categories ranged from 1 (strongly disagree) to 7 (strongly agree). A scale was created by averaging these three items, resulting in a Cronbach’s alpha of .89 (M = 4.26, SD = 1.46).
measured one concept. Eventually, a new scale was created to measure social presence by averaging these 11 items, resulting in a Cronbach’s alpha of .92 (M = 4.61, SD = 1.11). Previous Danmaku watching experience. This variable was operationalized by asking participants to identify how frequently they watch Danmaku videos using a 7-point Likert scale (1= never, 7= always) (M = 3.27, SD = 2.04).
Results Randomization Check
To check if participant’s gender was comparable across conditions with different Danmaku density (high density vs. low density Danmaku commentary), a Chi-square test was employed. The result showed that participants were equally distributed in both conditions regarding gender, χ2 (1, N = 112)= 2.40, p = .122. Moreover, to check if participant’s age was comparable across conditions with different Danmaku density, an independent sample t-test was conducted with Danmaku density (high density vs. low density Danmaku commentary) as the independent variable and age as the dependent variable. The result revealed that participants were equally distributed in both conditions in terms of age, t (106.48) = -0.78, p = .435. Additionally, results from Chi-square test showed that the randomization of
participants across conditions was successful regarding their educational level, χ2 (3, N = 112)= 3.46, p = .326.
Manipulation Check
Participants were asked to indicate how busy they found the level of Danmaku in the video they just watched, on a scale from 1 (Very much busy) to 7 (Not at all busy). An independent sample t-test was performed and the result showed that participants in the high Danmaku density condition found the level of Danmaku in the video much busier (M = 5.39, SD = 1.33) compare to participants in the low Danmaku density condition (M = 3.41, SD =
was successful.
Hypothesis Testing
To test the hypotheses of this study, Model 7 in the PROCESS macro for SPSS using a bootstrap estimation approach with 5000 samples was performed to conduct a moderated mediation analysis. Hypothesis 1 proposed that the high level of Danmaku density would lead to higher levels of enjoyment compared to the low level of Danmaku density. Model 7 gave results of this direct effect between the Danmaku density and enjoyment. Results indicated that the Danmaku density was a statistically significant predictor of enjoyment, b = 0.67, t = 2.21, p = .029, 95% CI [0.07, 1.26], for the high Danmaku density condition, the predicted enjoyment increased 0.67 compared to the low Danmaku density condition. Thus this
hypothesis was supported, see Table 1 in Appendix B. This result indicated that the Danmaku density indeed has an effect on enjoyment, the high level of Danmaku density leads to higher levels of enjoyment compared to the low level of Danmaku density.
To test Hypothesis 2 and Hypothesis 3, Model 7 was performed with the Danmaku density as the independent variable, social presence as the mediator variable and enjoyment as the dependent variable. All coefficients were unstandardized. Results indicated that the
Danmaku density was a statistically significant predictor of social presence, b = 0.78, t = 3.92, p = .000, 95% CI [0.39, 1.18], for the high Danmaku density condition, the predicted social presence increased 0.78 compared to the low Danmaku density condition. The result thus supported Hypothesis 2, which stated that the high level of Danmaku density would lead to higher levels of social presence compared to the low level of Danmaku density. It was also found that social presence was a statistically significant predictor of enjoyment, b= 0.58, t = 4.46, p < .001, 95% CI [0.32, 0.84]. The Danmaku density was no longer a statistically significant predictor of enjoyment after controlling for the mediator variable, social presence, b = 0.23, t = 0.79, p = .431, 95% CI [-0.34, 0.79], confirming the mediation effect.
Approximately 18% of the variance in enjoyment was explained by the predictors (R2 = 0.18). Additionally, the results revealed that the indirect effect of the Danmaku density on
enjoyment was statistically significant, b = 0.44, 95% CI [0.21, 0.79], the confidence interval did not include zero, so the effect could be considered statistically significant. The high Danmaku density was associated with approximately 0.44 points higher enjoyment scores as mediated by social presence, see Table 1 in Appendix B. The findings indicated that social presence fully mediated the relationship between the Danmaku density and enjoyment such that when viewers are in the high Danmaku density condition, they are more likely to perceive higher degrees of social presence, leading to greater enjoyment. Hypothesis 3 was thus
supported.
Hypothesis 4 proposed that the effect of the Danmaku density on social presence was moderated by viewer’s previous experience of watching Danmaku videos. Model 7 was performed with the Danmaku density as the independent variable, social presence as the dependent variable and viewer’s previous Danmaku watching experience as the moderator variable. Results showed that the regression model was significant, F (3, 108) = 6.46, p = .001, R2 = 0.17, approximately 17% of the variance in social presence was explained by the
predictors. A statistically significant main effect of the previous Danmaku watching
experience on social presence was revealed, b = 0.12, t = 2.24, p = .027, 95% CI [0.01, 0.23], for every one unit increase in the previous Danmaku watching experience, which runs from 1 (never) to 7 (always), the predicted social presence increased by 0.12. However, the
interaction effect between the Danmaku density and the previous Danmaku watching experience on social presence was not statistically significant, b = -0.02, t = -0.15, p = .882, 95% CI [-0.24, 0.20], see Table 2 in Appendix B. Based on the results, Hypothesis 4 was rejected. The findings suggested that the effect of the Danmaku density on social presence is not stronger for people with more Danmaku watching experiences than people with less
Danmaku watching experiences.
Figure 5. Model 7 Results
Discussion
The aim of the present study was to investigate whether the Danmaku density affects viewer’s feeling of enjoyment and the role of social presence as a mediator of that effect. Additionally, this study examined whether individual differences, specifically viewer’s previous watching experience would moderate the effect of the Danmaku density on viewer’s perceived social presence. This study assumed that the Danmaku density would have an effect on viewer’s enjoyment. The findings showed that the Danmaku density indeed affected viewer’s enjoyment. Viewers found it more enjoyable when they watched videos with the high level of Danmaku density compared with the low level of Danmaku density. This finding mirrors previous studies that found Danmaku comments provide viewers enjoyable interactive experiences (Chen, Cao, & Wang, 2013; Djamasbi, Hall-Phillips, Liu, Li, & Bian, 2016). Also, this finding is consistent with expectations based on research of media enjoyment in self-determination theory: people are likely to enjoy consuming entertainment media as long as their needs for relatedness are satisfied (Tamborini et al., 2010).
20s (Bilibili.com, 2018). It has been suggested by Arnett (2004) that when young adults step into emerging adulthood, they often move from one residence to another. This constant moving makes young adulthood an unstable time (Arnett, 2004). Since they are physically distant from real-life relatives and friends, they are eager to connect with others and establish new friendships (Park, Kee, & Valenzuela, 2009). Engaging with media becomes a common way for young adults to connect with other people. Their needs of relatedness can be satisfied through the co-viewing experiences generated by Danmaku videos. While watching Danmaku videos, they can easily find other co-viewers via the scrolling comments, and consequently creating a feeling of being connected with others (Chen et al., 2017). As their needs for relatedness are satisfied, they are more likely to feel enjoyment. In the present study, greater enjoyment was associated with higher level of Danmaku density, indicating that, a large amount of Danmaku comments makes viewers feel like they are connecting with numerous others, thus the feeling of enjoyment is stronger. In addition, Danmaku videos contain both humorous and meaningful comments that fulfill viewers’ hedonic and eudaimonic needs (Nabi & Krcmar, 2004). As these two intrinsic needs are satisfied, viewers are more likely to feel enjoyment of watching Danmaku videos. The high density Danmaku video contains more humorous and meaningful messages compared to the low density Danmaku video, so viewers are more likely to have greater enjoyment.
Another assumption of this study postulated that if the video was presented with the high level of Danmaku density, then viewers tended to perceive higher degree of social presence. The results indicated that there was indeed an effect of the Danmaku density on viewer’s perceived social presence. Participants who have watched the video with the high level of Danmaku density reported higher levels of social presence compared to those who watched the video with the low level of Danmaku density. This finding is in line with social presence theory stating that a media which conveys richer social cues can enhance social
presence (Short et al., 1976). The video with a large amount of Danmaku comments possibly increases the presence of social cues, and then viewers perceive higher degree of social presence consequently.
Additionally, the finding of this study is consistent with a prior study which suggested that people adopt internet technology to fulfill their social needs by means of both
synchronous and asynchronous interactions (Parker & Plank, 2000). Through Danmaku commentary system, viewers are allowed to communicate and collaborate with other viewers while watching videos (Zhao & Tang, 2016). The shared experience of watching and
commenting on the same Danmaku video enables viewers to seek company through interacting with others, and also to build a sense of belongingness to a virtual community (Chen et al., 2017). Although interactions facilitated by Danmaku comments are
asynchronous in the scenario of this present study (Danmaku comments presented in this study were generated previously by Bilibili.com users, rather than real-time comments), since Danmaku comments are attached to the initial video’ timeline, allowing viewers to see others’ comments when the scene is going on. Such presentation creates an illusion of synchronous social interactions, thereby fostering a sense of social presence of others.
The innovation of the present study is in placing social presence as a process through which the Danmaku density might have effects on viewer’s feeling of enjoyment. This study proposed that the effect of the Danmaku density on enjoyment is mediated by social presence. The findings indicated that the effect of the Danmaku density on enjoyment occurred through viewer’s perceived social presence. It is to say that when viewers are presented with a high density Danmaku video, they tend to perceive higher degree of social presence, and
consequently have greater enjoyment of watching that Danmaku video. Viewers tend to enjoy watching Danmaku video because Danmaku comments foster social presence of others, so that makes them feel like they are connected with other viewers (Chen et al., 2015). This
finding resonates with a prior Danmaku study suggesting that looking for a co-viewing experience is one major motivation for people choosing to watch Danmaku videos (Ma et al., 2015).
Chen and colleagues (2015) pointed out that individual differences, specifically viewer’s previous Danmaku video watching experience can possibly affect the way they watch Danmaku videos. This current study found some evidence about viewer’s previous Danmaku video watching experience in their perceived social presence. It turned out that viewer’s previous watching experience was positively correlated with their perceived social presence. This finding indicated that if viewers are more experienced with Danmaku videos, then they are more likely to perceive higher degree of social presence while watching Danmaku videos. This finding supported the prior study conducted by Chen et al. (2015) suggesting that non-Danmaku viewers differ from experienced Danmaku video viewers on attitudes toward Danmaku. Novice viewers tend to consider Danmaku comments as abundant information which distracts them from focusing on the video content, while experienced Danmaku video viewers find Danmaku comments interesting and tend to enjoy the co-viewing experience. However, this study failed to reveal the interaction effect between the Danmaku density and viewer’s previous Danmaku video watching experience on perceived social presence. Little work has been conducted formerly on the interaction effect by taking viewer’s previous watching experience into account. The possible explanation can be that frequent viewers are more experienced with Danmaku technology, they acquaint with the features and functions well. It can be inferred that experienced viewers are able to handle the amount of commentary. So when they are exposed to high density Danmaku videos, they tend to sense higher level of social presence. However, inexperienced viewers could be
overwhelmed by high density Danmaku commentary. Since they cannot attend to it, they are not likely to feel as much social presence as experienced viewers.
Theoretical and Practical Implications
This present study has the potential to extend the research on effects of Danmaku on viewer’s watching experience. Prior research on Danmaku video watching experience is scarce, only few study focused on the effect of Danmaku from the point of marketing and education rather than entertainment communication. Theoretically, by integrating
self-determination theory as well as social presence theory, this study proposed a model to explain the role of the Danmaku density in influencing viewer’s watching experience. The findings provide empirical evidences of the effects of Danmaku commentary on viewer’s watching experience from the perspective of entertainment communication. Additionally, this study also has practical implications. An increasing amount of users now actively engage in the computer mediated communication environment. Young people as the earliest adopters of emerging technologies, frequently share videos online, thus by knowing their video watching preferences will provide insights for video sharing websites to optimize their content and add tailored features. The findings of this study may enlighten video website designers with useful information that viewers’ communicative activities are separate from watching online videos. Danmaku-enabled video sites are served not only as platforms offering videos, but also communication tools for users to interact with each other. Thus, the design approach should not be limited, practitioners may consider adopting Danmaku function as creative value-adds in optimizing their existing conventional video sites.
Limitation and Future Studies
The present study has several limitations. Firstly, the findings were limited because as stimuli, a Chinese video site which implemented Danmaku commentary system was selected for this study. Accordingly, only native Mandarin speakers were recruited, which might limit generalizability. To the best of the researcher’s knowledge, except East Asia, there has been no adoption of Danmaku system in other cultures. It is very likely that cultural factors could
affect viewer’s watching experience. For instance, Danmaku commentary system is novel for most people in western countries, they are relatively inexperienced with this technology. So they might be overwhelmed by the scrolling comments, and consequently feel less enjoyment when watching Danmaku videos. Future research could examine whether viewer’s watching experience is different between cultures.
Secondly, viewer’s Danmaku video watching experience may vary with different types of videos. A prior exploratory study suggested that video content can possibly determine whether viewers choose to display Danmaku comments or not (Liu, Suh, & Wagner, 2017). This present study only opted for a drama depicted a neural conversion between two ordinary people, so the generalizability is somewhat limited. Further research can investigate Danmaku video watching experience in different video genre to see how viewers perceive Danmaku comments differently.
Thirdly, the active interaction that Danmaku enabled is the most unique feature attracts viewers, however, this study did not allow participants to either post comments or adjust the existing comments. The lack of engagement might also affect their Danmaku watching experience. Future studies can conduct a mixed-methods research approach to qualitatively explore why viewers actively post comments and quantitatively investigate how such behaviors affect their Danmaku watching experience.
Lastly, this study did not account for the content of the Danmaku comments. For instance, the valence of Danmaku comments could possibly influence viewer’s enjoyment. A prior study by van der Hoorn (2018) examining the effect of YouTube video comments on viewer’s entertainment experience suggested that viewers tend to feel more enjoyment when watching a video if the comments toward that video are positive. Future studies can take the valence of Danmaku comments into account by asking participants whether they perceive the comments about the video as positive or negative. Additionally, the low density Danmaku
video had less number of comments compared to the high density Danmaku video, so for some comments presented in the high density Danmanku video were not able to be displayed in the low Danmaku video. This might affect viewer’s watching experience as well.
Conclusion
In closing, this study sought to gain deeper understanding of the effects of Danmaku comments on viewer’s watching experience from an entertainment communication
perspective. The findings are consistent with the self-determination theory as well as social presence theory. The results provide empirical support that viewer’s watching experience differs by the Danmaku density. Viewers presented with the high density Danmaku video tend to perceive higher degree of social presence and feel more enjoyable of watching Danmaku video. In addition, the innovation in this study is postulating that the effect of the Danmaku density on enjoyment is mediated by viewer’s perceived social presence. The high Danmaku density is associated with higher degree of perceived social presence that is, in turn, associates with greater enjoyment of watching Danmaku video. Moreover, though this study failed to reveal the interaction effect of the Danmaku density and viewer’s previous watching experience on their perceived social presence, this perspective by considering the role of individual difference is broadened. Future studies may enhance the current study by
investigating the effect of personality traits on viewer’s Danmaku video watching experience. Further, this study may make contribution to video sharing websites to optimize their existing platforms by adding more interactive features, and thus to improve watching experiences.
Reference
Alexa.com. (2018). Site Overview: Bilibili.com. Retrieved from https://www.alexa.com/siteinfo/bilibili.com
Arnett, J. (2004): A longer road to adulthood. In: Emerging adulthood: The winding road from late teens through the twenties (chapter 1). Oxford: Oxford University Press. Bilibili.com. (2018). Retrieved from https://www.bilibili.com/blackboard/aboutUs.html Biocca, F. & Harms, C. (2002). Defining and measuring social presence: Contribution to the
Networked Minds theory and measure, Proceedings of PRESENCE 2002: 7-36. Chen, Y., Gao, Q., & Rau, P. L. P. (2017). Watching a movie alone yet together:
understanding reasons for watching Danmaku videos. International Journal of Human–Computer Interaction, 33(9), 731-743.
Chen, Y., Gao, Q., & Rau, P.-L. P. (2015). Understanding Gratifications of Watching Danmaku Videos–Videos with Overlaid Comments Cross-Cultural Design Methods, Practice and Impact (pp. 153-163): Springer.
Chen,Y., Cao, S., & Wang, T. (2013). A Close Study of Bullet Screen Website and Bullet Screen Users: From a Youth Subculture Perspective, Youth Exploration, (6), pp. 19-24.
Chu, D. (2009). Collective behavior in YouTube: a case study of ‘Bus Uncle’ on-line videos. Asian Journal of Communication 19(3): 337–353.
Daft, R. L., & Lengel, R. H. (1984). Information richness: A new approach to managerial information processing and organization design. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 6, pp. 191-234). Greenwich: JAI Press.
Dixson, M. D., Greenwell, M. R., Rogers-Stacy, C., Weister, T., & Lauer, S. (2016). Nonverbal immediacy behaviors and online student engagement: Bringing past
instructional research into the present virtual classroom. Communication Education, 66(1), 37-53. doi:10.1080/03634523.2016.1209222
Djamasbi, S., Hall-Phillips, A., Liu, Z., Li, W., & Bian, J. (2016). Social Viewing, Bullet Screen, & User Experience: A First Look’ in 2016 49th Hawaii International Conference on System Sciences (HICSS), 648-657. IEEE Xplore.
Grodal, T. (2000). Video games and the pleasures of control. In D. Zillman & P. Vorderer (Eds.), Media entertainment. The psychology of its appeal (pp. 197–213).Mahweh, NJ:Erlbaum.
Harboe, G. (2009). In search of social television. In P. Cesar, D. Geerts, & K. Chorianopoulos (Eds.), Social interactive television (pp. 1–13).
He, H., & Zhang, H. (2014). From the non-mainstream to AB websites: The succession of subculture features of the 80s, 90s, and 00s generations, and the Internet investment strategy. SWS Research.
Johnson, D. (2013). Polyphonic/Pseudo-synchronic: animated writing in the comment feed of nicovideo. Japanese Studies 33 (3), 297–313.
Katz, E., Blumler, J. G., & Gurevitch, M. (1974). Utilization of mass communication by the individual. In J. G. Blumler & E. Katz (Eds.), The uses of mass communications: Current perspectives on gratifications research (pp. 19–32). Beverly Hills, CA: Sage. Kendall, J. & Kendall, K. (2017). Enhancing Online Executive Education Using Storytelling: An Approach to Strengthening Online Social Presence . Decision Sciences: Journal of Innovative Education, 15 (1), 62-81. doi:10.1111/dsji.12121
Kim, J., Kwon, Y., & Cho, D. (2011). Invstigating factors that influence social presence and learning outcomes in distance higher education. Computer & Education, 57,
1512-1520.
moderates the effects of politicians' Twitter communication. Communication Research, 0093650212466407.
Lim, J. S., Hwang, Y. C., Kim, S., & Biocca, F. (2015). How social media engagement leads to sports channel loyalty: Mediating roles of social presence and channel commitment. Computer in Human Behavior, 46, 158-167.
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. Liu, L., Suh, A., & Wagner, C. (2017). Who is with You? Integrating a Play Experience into
Online Video Watching via Danmaku Technology. In: Kurosu M, ed. Human-Computer Interaction. Interaction Contexts: 19th International Conference, HCI International 2017, Vancouver, BC, Canada, July 9-14, 2017, Proceedings, Part II, Cham: Springer International Publishing, 63-73.
Ma, Z., & Ge, J. (2014). Riben Donghua de Danmu Pinglun Fenxi: Yizhong Zhunshehui Jiaowang de Shijiao [Analysis of Japanese Animation Overlaid Comment (danmu): A Perspective of Parasocial Interaction. Guoji Xinwen Jie, 8, 116-130.
Miron, D. (2003). Enjoyment of violence. In J. Bryant, D. R. Roskos-Ewoldsen & J. Cantor (Eds.), Communication and emotion: Essays in honor of Dolf Zillmann (pp. 445–472). Mahwah, NJ: Erlbaum.
Nabi, R. L., & Krcmar, M. (2004). Conceptualizing media enjoyment as attitude: Implications for mass media effects research. Communication theory, 14(4), 288-310.
Nagy, J., & Midha, A. (2014). The value of earned audiences: How social interactions amplify TV impact. Journal of Advertising Research, 54(4), 448–453.
doi:10.2501/JAR-54-4-448-453
Ni, Y. (2017), A Study of Danmaku Video on Attention Allocation, Social Presence,
Oliver, M. B., & Bartsch, A. (2010). Appreciation as audience response: Exploring entertainment gratifications beyond hedonism. Human Communication Research, 36(1), 53-81.
Park, N., Kee, K.F., & Valenzuela, S. (2009). Being immersed in social networking environment: Facebook groups, uses and gratifications, and social outcomes. Cyberpsychol. Behav. 12 (6), 729–733.
Parker, B. J., & Plank, R. E. (2000). A uses and gratifications perspective on the Internet as a new information source. American Business Review, 18, 43–49.
Reiss, S., & Wiltz, J. (2004).Why people watch reality TV. Media Psychology, 6(4), 363–378. Ryan, R. M., Rigby, C. S., & Przybylski, A. (2006). The motivational pull of video games: A
self-determination theory approach. Motivation and Emotion, 30(4), 344–360.
Shen, Y., Chan, H., & Hung, I. (2014). Let the comments fly: the effects of flying commentary presentation on consumer judgment. Paper presented at the Proceeding of 35th
International Conference on Information Systems, 14–17 Dec. Auckland, New Zealand.
Shin, D.-H. (2013). Defining sociability and social presence in Social TV. Computers in Human Behavior, 29(3), 939-947.
Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommunications. London: John Wiley & Sons, Ltd.
Sivunen, A., & Nordbäck, E. (2015). Social Presence as a Multi-Dimensional Group Construct in 3D Virtual Environments. Journal of Computer-Mediated Communication, 20, 19-36. doi:doi:10.1111/jcc4.12090
Tamborini, R., Bowman, N. D., Eden, A., Grizzard, M., & Organ, A. (2010). Defining Media Enjoyment as the Satisfaction of Intrinsic Needs. Journal of Communication, 60(4), 758-777. doi:10.1111/j.1460-2466.2010.01513.x
Tu, C. H. (2000). On-line learning migration: from social learning theory to social presence theory in a CMC environment. Journal of Network and Computer Application, 23(1), 27–37.
Tu, C. H., & McIssac, M. (2002). The relationship of Social Presence and interaction in online classes. The American journal of distance education, 16(3), 131-150.
van der Hoorn, N. (2018). Beyond the YouTube video: how comments affect our entertainment experience: the effect of social features on YouTube on the viewers entertainment experience (Master’s thesis). Retrieved from
http://scriptiesonline.uba.uva.nl/document/662634
Vorderer, P., Klimmt, C., & Ritterfield, U. (2004). Enjoyment: At the heart of media entertainment. Communication Theory, 14, 388–408.
Xu, H., & Yan, R.-N. (2011). Feeling Connected via Television Viewing: Exploring the Scale and Its Correlates. Communication Studies, 62(2), 186-206.
Yang, Y. (2018). Online entertainment platform Bilibili reports revenue surge but warns China content crackdown will hit costs. South China Morning Post. Retrieved from https://www.scmp.com/tech/article/2161614/online-entertainment-platform-bilibili-reports-revenue-surge-warns-china
Zhao, Y., &Tang, J. (2016) Exploring the Motivational Affordances of Danmaku Video Sharing Websites: Evidence from Gamification Design. In: Kurosu M. (eds) Human-Computer Interaction. Novel User Experiences. HCI 2016. Lecture Notes in Human-Computer Science, vol 9733. Springer, Cham.
Zillmann, D., & Bryant, J. (1994). Entertainment as media effect. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 437–461). Hillsdale, NJ:Erlbaum.
Chapman & H. Foot (Eds.), Humor and laughter: Theory, research, and applications (pp. 93–115). London: Wiley.
Appendix A
Figure 3. Screenshot of High Danmaku Density Condition
Appendix B
Table 1
Mediation Analysis Results of the Danmaku Density Predicting Enjoyment by Taking Social Presence as Mediator Paths R2 F df B SE t p 95% CI a 0.17 6.46 3, 108 0.78 0.20 3.92 .000 0.39, 1.18 b 0.18 11.38 2, 109 0.58 0.13 4.46 < .001 0.32, 0.84 c 0.04 4.87 1, 110 0.67 0.30 2.21 .029 0.07, 1.26 c’ 0.18 12.23 2, 109 0.23 0.29 0.79 .431 -0.34, 0.79 c’’ - - - 0.44 0.14 - - 0.21, 0.79
Note. Values represent unstandardized path coefficients. a = path from the Danmaku density to social presence;
b = path from social presence to enjoyment; c = the direct effect of the Danmaku density on enjoyment; c’ = path from the Danmaku density to enjoyment after controlling for social presence; c’’ = path from the Danmaku density to enjoyment through social presence (indirect effect).
Table 2
Moderation Analysis Results of the Interaction Effect of the Danmaku Density and Viewer’s Previous Watching Experience on Social Presence
Predictor R2 F df B SE t p 95% CI Danmaku Density 0.17 6.46 3, 108 0.78 0.20 3.92 .000 0.39, 1.18 Watching Experience 0.12 0.06 2.24 .027 0.01, 0.23 Interaction -0.02 0.11 -0.15 .882 -0.24, 0.20
Note. Values represent unstandardized coefficients, the regression model was statistically significant, p = .001.
