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
Naomi van der Hoorn – 11035552 Graduate School of Communication Master’s Program Communication Science Mastertrack Entertainment Communication Supervisor: Susanne Baumgartner
Word count: 7.309
Abstract
Since a couple of years, YouTube is taking over our lives. People spend more and more time on YouTube. But how does the social side of YouTube affect our experiences? There are lots of studies that have investigated the effects of peer evaluations on books, hotels, videos, and TV programs. However, there is little literature about the effect of comments on YouTube on enjoyment and transportation. The aim of this study was to investigate whether social features (e.g., amount of likes and comments) on YouTube affect our entertainment experience. An online experiment, with a 2 x 2 x 2 factorial design, was conducted to test these effects. The sample of this study consists of 143 participants, ranging from 17 to 57 years old. Results showed that there is a small direct effect of the valence of the comments on both enjoyment and transportation. Besides, it was also found that the attractiveness of the commenter has a small moderating effect on the relationship between the valence of the comments and enjoyment. Also, gender has a small moderating effect on the relationship between the valence of the comments and enjoyment and transportation. The current research shows that comments on YouTube videos does affect how people evaluate that specific video.
Limitations of the current study are discussed and recommendations for future research are given.
Introduction
Watching TV is being more and more replaced by online videos. The first YouTube video was uploaded in 2005. Ever since, it became the biggest video platform worldwide, where every amateurish video maker can upload a video. YouTube is currently used by more than 1.300.000.000 people and it has more than 30 million visitors a day. Every minute, more than 300 hours of YouTube videos are uploaded. Of all people worldwide aged between 18 and 49, 80% uses YouTube (YouTube for Press, n. d.). All these YouTube videos have social features where people can express their opinions about that video. These social features on
YouTube include user comments as well as the number of likes/dislikes. But what are the effects of these social features? How are these features related to a user’s own evaluation of that video?
The current research will examine potential conformity effects of social features on YouTube on the entertainment experience of users. Previous research has shown that peer evaluations of books have an effect on enjoyment and transportation of other readers
(Shedlosky-Shoemaker, Costabile, DeLuca & Arkin, 2011). These researchers found that peer evaluations influenced people’s enjoyment and transportation, and this effect was more pronounced when peers provided negative comments. However, this study focused only on books. Therefore, the effects of peer evaluations about online videos remain unclear. But there are some previous studies that did focus on online videos, e.g. Möller and Kühne (in press) who investigated the effect of user comments on hedonic and eudaimonic enjoyment.
The influence of social information might also depend on characteristics of the
commenter. For example, Wiegman (1987) investigated the effect of attractiveness of the live audience of a TV program on the evaluation of this program by the audience at home. It has been found that an attractive live audience who reacted positive to the interview with the Dutch politician, had a stronger effect on the viewers’ evaluation of that program. This indicates that the attractiveness of the commentator might be an important factor to take into account when investigating conformity effects.
These effects of social information on YouTube might also be moderated by sex differences. This means that the effects might be different for female viewers than for male viewers. Through a meta-analysis, Cooper (1979) found that there are contradictory results about sex differences in conformity effects. According to Cooper (1979), research has shown that in face-to-face communication, females conform more often than males do. However, Cooper (1979) also found in his meta-analysis that many studies that did not show any gender
differences when it comes to conformity effects. It remains unclear whether sex might be a moderator in the context of YouTube videos. Therefore, it might be an interesting factor to take into account when investigating the relationship between social features and the entertainment experience.
Several scholars investigated how evaluations of others influence people’s own evaluations (Shedlosky-Shoemaker, Costabile, DeLuca & Arkin, 20011; Wiegman, 1987; Walther, DeAndrea, Kim & Anthony, 2010; Möller & Kühne, in press), but none of them investigated how this works for comments on vlogs on YouTube. Besides, it has not been investigated how far the specific characteristics of the person who is commenting on the video affects the entertainment experience. Therefore, this research will focus on how positive and negative comments on YouTube from attractive and unattractive peers will influence people’s entertainment experience. The entertainment experience in this case is divided into two categories; enjoyment and transportation. As described above, it already has been found that peer evaluations about books influences how people evaluate that book. But does this also count for YouTube videos where viewers do not personally know the person who is commenting?
Altogether, this leads to the following research question: What is the effect of social features of others on YouTube on the entertainment experience of the viewer and how does the attractiveness of the commenters influence this effect?
Theoretical Framework The Social Features of YouTube
Nowadays YouTube is the most popular platform for uploading videos for both professional and amateurish video makers. YouTube gives the viewer the options to rate the video by giving likes, dislikes, and comments. Some researchers examined how these social features on YouTube affect the video retrieval effectiveness (Chelaru, Orellana-Rodriguez &
Altingovde, 2014). The researchers examined the effectiveness of every social feature of YouTube on how people retrieve a certain video and how this affects people’s ranking of a specific video. People rank a video by giving a like or dislike. They found that tags and video descriptions (basic features) are most effective when it comes to video retrieval. This means that the basic features are most important for people to rank videos.
It has also been found that some people have a bad feeling towards YouTube comments (Schultes, Dorner & Lehner, 2013). The reason for this is the high amount of inferior comments, which are defined as comments that include offensive statements without any relevant information about the video. However, people still make use of comments because of the discussion comments and substantial comments, which are defined as follows: ‘Contains comments without offensive statements that carry certain content information and are, ideally, directly related to the actual video content’ (Schultes, Dorner & Lehner, 2013, pp. 659). Thus, comments are not highly appreciated, unless they are not offensive and include useful information about the specific video.
Peer Evaluations
According to the social conformity theory from Asch (1951), people form their
behaviors and beliefs based on those of others. The theory states that people desire to fit into a group and they may achieve this goal by basing their opinions on what others say. Peer
pressure plays an important role for the development of people’s attitudes and behaviors. By putting a group of people in the same room and letting them answer questions out loud, Asch (1951) found that people will say what others say, also when it is the wrong answer. This means that people base their opinion on what others do or say. This might not always be the case, Asch (1951) found that 25% of the participants conformed always, 50% at least once, and only 25% conformed never. Thus, it can be said that, on average, 75% of all people do conform sometimes. Later on, Noelle-Neumann (1974) found that people only share their
opinion when they share the opinion of the majority. This was called the spiral of silence. This means that the majority influences people’s opinion, because the minority will not even share their opinion. The minority won’t share their opinion, because they have the feeling that they will not fit into a social group when they share it.
The conformity effects described in the previous paragraph could also occur during online experiences. This means that people base their behaviors and beliefs on what others do or say about online content. For example, it has been found that customer word of mouth affects people’s purchase intentions (Park, Lee & Han, 2007). Chevalier and Mayzlin (2006) examined whether online book reviews affected people’s purchase behaviors. They found that reviews on book webshops such as Amazon are commonly positive, but the number of stars they receive can influence the decision making of the customer. When previous buyers give a book a better evaluation, it is more likely that others will buy it. Other researchers
investigated the effect of hotel reviews on booking intentions (Tsao, Hsieh, Shih & Lin, 2015). These researchers asked the participants about their conformity, before they were exposed to the stimulus materials. With this, they were possible to divide the participants in people with either a high or a low level of conformity. Their results showed that booking intentions depended strongly on the valence of the review. When the review was positive, people with a higher conformity degree showed higher booking intentions. For people with a lower degree of conformity, the amount of reviews had a stronger effect.
Evaluations of others, as reflected in YouTube comments, may also affect the
entertainment experience of users. It has been found that the expression of others has an effect on how people experience media content (Ramanatha & McGill, 2007). The experience of seeing what someone else finds affects how people evaluate that experience. Ramanatha & McGill (2007) assume that this also happens when people consume a movie. Similarly, Tal-Or (2016) examined the role of co-viewing television. Through this research, it has been
found that co-viewing has an effect on the transportation into the movie. The presence of a co-viewer increased people’s transportation. The participants showed a higher degree of transportation when their co-viewer expressed positive reactions to the movie. Banjo, Appiah, Wang, Brown, & Walther (2015) examined the effects of in-group and out-group responses to racial comedy. The results showed that responses of in-group members had a stronger effect for black people, in comparison to responses of out-group members. When black people were viewing with other black people, they showed a more positive attitude, greater perceived similarity and identification. However, this effect did not occur for white viewers. A reason for this is the fact that the content they were exposed to, was all written, produced, and directed by black people. Therefore, the black viewers identified better with the characters in the program than the white viewers, especially when the black viewers watched together with a black co-viewer.
When co-viewers itself have an effect on people’s entertainment experience, it might be interesting to focus more explicitly on the evaluations of others on social media. For example, Sherman, Payton, Hernandez, Greenfield, and Dapretto (2016) examined the effects of peer influences on behavior on social media. Research shows that people are more likely to like a post on Instagram which already has many likes. This result occurred for both neutral and risky (i.e., drinking, smoking) photos. Instagram posts with many likes had a stronger effect on reward processing, social cognition, imitation, and attention. Edwards, Edwards, Qing, and Wahl (2007) found that students who receive positive evaluations via computer mediated communication about a teacher, evaluated that teacher more positively as well. However, it remains unclear whether these effects hold for all commentators. In addition, DeAndrea, Kim, and Anthony (2010) examined whether online comments had an effect on perceptions of antimarijuana public service announcements (PSA) on YouTube. This research has shown that both supportive and derisive comments had an effect on the evaluations of the
PSA, but not on marijuana attitudes. Thus, it might be true that comments of others on a video have an effect on how people perceive that video.
Because evaluations of others have an effect on people, user comments might have an effect on people’s entertainment experience. As described above, evaluations of others have an effect on how people perceive certain content. The entertainment experience can be divided into two components. At first, the entertainment experience consists of enjoyment. Enjoyment is a positive reaction that may occur when people consume media entertainment (e.g., Raney & Bryant, 2002; Tamborini, 2003). Enjoyment can be seen as a two-folded process; entertainment content can meet hedonic and eudaimonic needs. Hedonic needs are the pleasure of entertainment without further deeper meanings. Messages with more deep and meaningful content will fulfill eudaimonic needs (Oliver & Bartsch, 2010). Nabi and Krcmar (2004) defined enjoyment as satisfying intrinsic needs. Media messages can fulfill both needs, and thereby lead to enjoyment of this content.
Besides enjoyment, transportation is an important component of the entertainment experience. Green and Clark (2013, p. 477) describe transportation as follows: “Individuals who are transported into a narrative world are completely immersed in the story; they experience high levels of cognitive and affective engagement, and may form vivid mental images.”
Altogether, there might be a link between conformity effects and social features. As described above, evaluations of others have an effect on how people enjoy content and how people are involved into that content. Waddell and Sundar (2017) found that negative social media comments about a TV program lead to a decrease of enjoyment. Möller and Kühne (in press) investigated how user comments on a video affected people’s hedonic and eudaimonic experiences. Möller and Kühne (in press) found that user comments do have an effect on people’s entertainment experience. More specifically, this research shows that positive
comments about a certain video lead to an increase in both hedonic and eudaimonic
experiences, and vice versa. This means that entertainment experiences of online videos are not only predicted by its content, but also depend on user comments. Therefore, the first hypotheses of the current study are as follows:
H1.1: Positive comments lead to higher enjoyment than negative comments. H1.2: Positive comments lead to higher transportation than negative comments. Attractiveness
It has been found that attractiveness is an important determinant when people create opinions. This can be explained by the halo effect (Landy & Sigall, 1974). Landy and Sigall (1974) asked their participants in an experimental setting to evaluate essays that were either well or poorly written by either attractive or unattractive authors. The researchers found that the essays with attractive authors were evaluated as of a better quality compared to the essays that were written by less attractive authors. Later on, Kaplan (1978) called this the
attractiveness halo effect. He also found that attractive writers were evaluated as being more talented. Because attractiveness seems to have an effect on the evaluations of people, it might also be that the attractiveness of a commenter on a YouTube video has an effect on the
evaluation of that specific video. Although the commenter is not the one who made the video, the attractiveness of the commenter might have an effect on how people create their opinion.
It differs per person whether someone perceives someone as being attractive or not. However, there are some physical features that are often associated with physical
attractiveness. Women are perceived as more attractive when their face is feminine, this means small chins, large eyes, high cheekbones and full lips. The physical features that make men more attractive are masculine faces. Masculine faces have broad chins and facial
Attractive people may have a greater source credibility, and what they post online may thus have a stronger impact. Research has shown that attractive people are more independent and create their own opinion (Miller, 1970). Therefore, they are commonly seen as not being easily manipulated. Because their source credibility might be higher, attractive people might be more persuasive than unattractive people.
In line with this reasoning, it has been found that when it comes to politics, attractive people are seen as people with more knowledge (Palmer & Peterson, 2015). People will rather seek attractive individuals out than unattractive individuals and it is likely that they will accept their advice when it comes to politics. Therefore, attractive people are seen as more persuasive. Another study showed that attractiveness and opinion similarity of the source has an effect on how the message is perceived (Main, Aditya & Dahl, 2013). Participants scored higher on consumer satisfaction when the source was attractive and shared a similar opinion. This only applied to messages where the persuasion attempt was indirect. Indirect persuasion attempts are attempts that go further than just an explicit attempt where it is clear that the brand is trying to sell something. For example, giving the consumer compliments about how they look when they get into the store (indirect), instead of giving compliments when a consumer tries something on from the store (direct). Because several scholars found that attractive people are perceived as more credible, more successful, and more persuasive, the effectiveness of attractiveness of the commenter on YouTube will be tested in the following hypotheses:
H2.1: The positive effect of positive comments on enjoyment is stronger when the commenters are attractive than when the commenters are unattractive.
H2.2: The positive effect of positive comments on transportation is stronger when the commenters are attractive than when the commenters are unattractive.
H2.3 The negative effect of a negative comments on enjoyment is stronger when the commenters are attractive than when the commenters are unattractive.
H2.4 The negative effect of a negative comments on transportation is stronger when the commenters are attractive than when the commenters are unattractive.
Sex Differences
Eagly (1978) examined sex differences in influenceability, primarily focused on persuasion and conformity. There is some evidence that women are more easily influenced than men, but these findings are very marginal. It was found that sex differences were stronger in group pressure conformity studies. Besides, it has been found that there are more studies that show stronger conformity effects for females than studies that show stronger influenceability effects for females (Eagly, 1978). In addition, Cross, Brown, Morgan, and Laland (2017) investigated sex differences in the relationship between confidence and conformity effects. They found that when women are exposed to a mental rotation task (3D objects that rotate), women show lower levels of confidence. This leads to the fact that they are more likely to switch answers and therefore conform more than males. Therefore, there might be a difference between women and men in reaction to social influences.
Not all studies did find stronger conformity effects for women. For example, Cooper (1979) found in a meta-analysis that some studies found equal conformity effects for both males and females. Cooper (1979) found in his meta-analysis some evidence that females conform more than males. For example, Dean, Austin, and Watts (1971), Eagly and Telaak (1972), and Marquis (1973) found stronger conformity effects for females than for males. However, overall, the literature rather points towards stronger conformity effects for females than for males. This leads to the following hypothesis:
H3: The effect between social features on YouTube and the entertainment experience of people is moderated by the gender of the viewer. The effect of social features on YouTube will be stronger for females than for males.
Method Participants
Participants were recruited through Facebook. The final sample consisted of 143 participants aged between 17 and 57 (M = 26.37, SD = 8.51). Of all participants, 39.9% were male (M = 25.86, SD = 7.52) and 60.1% were female (M = 26.71, SD = 9.14). The biggest part of the sample was enrolled in higher education (55.2% was enrolled in university and 22.4% was enrolled in university of applied sciences). A smaller amount, 9.1%, reported that they were enrolled in vocational education and 8,4% only finished high school.
Procedure
To test the hypotheses, an online experiment was conducted. This experiment is a 2 x 2 x 2 factorial design with valence of social information (positive vs. negative), attractiveness of commenter (attractive vs. unattractive), and gender (male vs. female) as factors. The participants were randomly assigned to the experimental conditions. This experiment
consisted of eight conditions, these conditions will be explained in the next paragraph. Before the participants could start with the experiment, all participants had to fill in an informed consent. The participants were first asked about their demographics such as age, gender, and education level. The experimental conditions varied by different comments about the same YouTube video. All participants were exposed to the same YouTube video. After exposing the participants to the comments and the YouTube video, the participants were asked about how much they enjoyed the video and how much they felt transported into the video. In addition to that, they were also asked whether they perceived the commenter as attractive or not and whether they perceived the comments about the video either positive or negative.
Stimulus Materials
The participants were first exposed to the manipulated commenters, comments, amount of likes and views for 30 seconds. Subsequently, they all watched the same YouTube vlog. All stimulus materials can be found in Appendix I. Participants in the positive
conditions were exposed to two positive comments (e.g., ‘I really like your video! You’re such an inspiring person to me’ and ‘What a great video. Your editing skills are great!! I’m subscribed to your channel for more than a year now and it’s getting better every time’). Participants in the negative conditions were exposed to two negative comments (e.g., ‘I don’t like this video at all. You are super arrogant and nobody cares about how you clean your shoes and stuff’ and ‘Pffffff what a stupid vlog… It irritates me that people think this is a ‘job’’). Participants in the attractive conditions were exposed to two attractive commenters and participants in the unattractive conditions were exposed to two unattractive commenters.
Female participants were exposed to female commenters and male participants were exposed to male commenters. Research has shown that boys stronger identify with masculine males and girls stronger identify with feminine females (Hill & Lynch, 1983). This means that girls will feel more connection to female commenters and males will feel more connected with male commenters. Hoffner and Buchanan (2005) found that people stronger identify with same-sex others in TV programs. Therefore, the participants were exposed to same-sex commenters, which leads to eight different conditions.
For the experiment, a vlog of the famous Swedish vlogger Jon Olsson was used. The vlog was mainly about the fact that Jon went to a McLaren shop to look for a new car. The original video took 12 minutes, which was shortened to 5.30 minutes for the experiment. This vlog was chosen, because Jon Olsson is not well-known in the Netherlands. This makes the change that people had seen the vlog before less likely. Also, the vlog was very well-edited, but also has some rather uninteresting parts. For example, a few minutes of the vlog are about
how Jon cleans his shoes. This can be perceived as not interesting for many people. Because it will be measured whether positive or negative comments have an effect, a vlog with good and bad parts was needed. Therefore, this vlog was appropriate for this research.
Measurements
Enjoyment. To measure enjoyment, the enjoyment scale of Wirth, Hofer, and Schramm (2012) was used. This scale was developed to measure hedonic entertainment among a sad but meaningful movie. This scale consists out of three items (e.g., It gave me
pleasure to watch the video, watching the video amused me and the video was exciting). In
addition, a fourth item was added (e.g., I enjoyed the video). All items were measured with a 7-point Likert scale (e.g., 1 = strongly agree, 7 = strongly disagree). To test whether the scale was consistent, a factor analysis was conducted. This factor analysis showed that the scale is unidimensional, with one component with an Eigenvalue higher than 1.00. This dimension explained 90.51% of the total variance. To test whether the scale was reliable, a reliability analysis was conducted. This analysis showed that the scale was reliable with a Cronbach’s Alpha of .97. For the interpretability, all items were recoded so that higher values indicate a higher level of entertainment. The mean score of the scale was 3.55 (SD = 0.82).
Transportation. To measure transportation, the engagement scale of Buselle and Bilandzic (2009) was used. This scale was used for measuring narrative engagement.
Transportation means that someone is completely cognitive and affective engaged into a story and is fully immersed into the story (Green & Clarck, 2013). Therefore, the engagement scale of Buselle and Bilandzic (1009) was appropriate to measure transportation. This scale tests whether someone was fully engaged into the video or not. This scale consists out of six items and was adapted to the vlog that was used for this experiment (e.g., I was mentally involved
with the video, I was really pulled into the story, I was completely immersed in the video, The viewing experience was intense for me, I wanted to learn how the video ended, and I wanted
to know how the events would unfold). All items were measured with a 7-point Likert scale
(e.g., 1 = strongly agree, 7 = strongly disagree). To test whether the scale was consistent, a factor analysis was conducted. This factor analysis shows that the scale is unidimensional, with one component with an Eigenvalue higher than 1.00. This dimension explains 76.09% of the total variance. To test whether the scale was reliable, a reliability analysis was conducted. This analysis shows that the scale was reliable with a Cronbach’s Alpha of .94. For better interpretability, all items were recoded so that higher values indicate more transportation. The mean score of the scale was 2.91 (SD = 1.45).
Attractiveness. Attractiveness was measured by one question about whether the participants perceived the commenter as attractive or not. This was measured with a 7-point Likert scale (e.g., 1 = strongly agree, 7 = strongly disagree). For better interpretability, this item was recoded so that higher values indicate more attractiveness. The mean score of this item was 3.62 (SD = 1.81).
Valence of the social information. The valence of the video was measured with one question about whether the participants perceived the comments about the video as positive or negative. This was measured with a 7-point Likert scale (e.g., 1 = strongly agree, 7 = strongly disagree). For better interpretability, this item was recoded so that higher values indicate that the comments about the video were perceived as more positive. The mean score of this item was 3.97 (SD = 2.38).
YouTube use. A few questions were asked about whether the participants have seen the video before (e.g. 1 = yes, 2 = no), have seen another video of the same YouTuber before (e.g. 1 = yes, 2 = no) and how often they watch vlogs on YouTube (1 = very often, 5 = never). For better interpretability, this item was recoded so that higher values mean that participants watch YouTube vlogs more often. The mean score of this item is 2.38 (SD = 1.21). Only 2.1%
of all participants had seen the video before and 8.4% had seen another video of the same YouTuber before.
Results Randomization Check
Age. To test whether the participant’s age was evenly distributed over the conditions, an ANOVA was conducted, with experimental condition as independent variable and age as dependent variable. The ANOVA indicates that age does not differ across conditions, F(7, 134) = 1.47, p = .183
YouTube Use. To test whether the amount of YouTube use was evenly distributed over the conditions, an ANOVA was conducted, with experimental condition as independent variable and the frequency of watching vlogs as dependent variable. The ANOVA indicated that the mean frequency of how often the participants watch vlogs on YouTube does not significantly differ per condition, F(7, 134) = .813, p = .578. A Chi-square test was conducted to test whether the number of participants who had seen the vlog before were evenly
distributed over the conditions. This analysis showed that the frequency of people who did see the video before did not significantly differ from the people who did not, Chi-square(7) = 6.05,
p = 534. Another Chi-square test was conducted to test whether the participants who had seen
another vlog of the same YouTuber before was evenly distributed over the conditions. This analysis showed that the frequency of people who had seen another vlog of the same
YouTuber before did not significantly differ from the frequency of people who did not, Chi-square(7) = 7.17, p = .41.
Overall, these analyses show that the randomization across conditions was successful for both age and YouTube use. All participants were randomly assigned to one of the eight conditions. The conditions were positive attractive male (n = 17), positive unattractive male (n = 13), negative attractive male (n = 15), negative unattractive male (n= 12), positive
attractive female (n = 18), positive unattractive female (n = 21), negative attractive female (n = 23) and negative unattractive female (n = 24).
Manipulation check
Attractiveness. To test whether the attractiveness of the commenters were manipulated correctly, an independent-samples t-test was conducted, with experimental condition (attractive commenters/unattractive commenters) as independent variable and perceived attractiveness as dependent variable. The analysis showed that the perceived attractiveness of people in the attractive condition (M = 4.14, SD = 1.85) did not significantly differ from the perceived attractiveness of people in the unattractive condition (M = 4.60, SD = 1.78), t(128) = 1.44, p = .153, CI = [-1.09, .17], Cohen’s d = 0.02. Twelve people filled in that they did not remember the attractiveness of the commenter. To test whether there was a difference in this effect for males and females, an ANOVA was conducted, with experimental condition (attractive commenters/unattractive commenters) and gender as independent
variables and perceived attractiveness as dependent variable. The analysis showed that there is no interaction effect between attractiveness of the commenters and gender on perceived attractiveness of the commenters, F(1, 130) = .16, p = .686, η² = .00. However, the mean scales showed that the differences were higher for male participants than for female
participants. Figure 1 shows that male participants scored higher on perceived attractiveness in the attractive condition than female participants.
Valence. To test whether the valence of the comments was manipulated correctly, an independent-samples t-test was conducted, with experimental condition (positive/negative) as independent variable and perceived valence (measured with a 7-point Likert scale, ranging from strongly agree to strongly disagree) as dependent variable. The analysis showed that the perceived valence of the participants in the positive condition was significantly more positive (M = 5.94, SD = 1.22) than the perceived valence of the participants in the negative condition (M = 2.24, SD = 1.73), t(136) = 14.47, p < .000, CI = [-4.21, -3.20], Cohen’s d = 0.25. Four people filled in that they did not remember the valence of the comments.
Tests of Hypotheses
Main effects of social features on the entertainment experience. The first hypothesis stated that positive comments have a positive effect on people’s entertainment experience and negative comments have a negative effect. To test whether the valence of the comments have an effect on enjoyment, an ANOVA, with enjoyment as dependent variable and valence as independent variable, was conducted. The results showed that there was a small significant effect between the valence of the comments and enjoyment, F(1, 140) = 8.40,
p = .004, η² = .06. Results indicate that people in the positive condition (M = 3.75, SD = .87)
enjoyed the video significantly more than people in the negative condition (M = 3.36, SD 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 Attractive Unattractive
Figure 1. The interaction effect between attractiveness of the commenters and gender on perceived attractiveness of
the commenters.
= .73). To test whether the valence of the comments had an effect on transportation, an ANOVA, with transportation as dependent variable and valence as independent variable, was conducted. Results indicated that there is a small significant effect between the valence of the comments and transportation, F(1, 140) = 12.25, p = .001, η² = .08. Participants in the
positive condition (M = 3.33, SD = 1.55) felt more transported into the video than people in the negative condition (M = 2.51, SD = 1.24). Therefore, the findings support H1.1 and H1.2.
Interaction effects of social features and attractiveness on the entertainment experience. The second hypothesis stated that the effects of the valence of the comments on enjoyment and transportation are stronger when the commenter is attractive in comparison to when the commenter is unattractive. To test whether the attractiveness of the commenter moderated the effect of the comments’ valence on enjoyment, a two-way ANOVA with enjoyment as dependent variable and valence and attractiveness as independent variables, was conducted. Results indicate that there was a small, but approaching significant interaction effect between the attractiveness of commenter and the valence of the comments on
enjoyment, F(1, 138) = 3.34, p = .07, η² = .02. Figure 2 shows that the effect was stronger for participants in the attractive condition than for people in the unattractive condition. People in the attractive condition enjoyed the video more when they were exposed to positive
comments and enjoyed it less when they were exposed to the negative comments. This effect was less strong for participants who were exposed to comments from unattractive commenters. Therefore, H2.1 and H2.3 were supported.
To test whether the attractiveness of the commenter moderated the effect of the comments’ valence on transportation, a two-way ANOVA, with transportation as dependent variable and valence and attractiveness as independent variables, was conducted. Results indicate that there is no significant interaction effect between the attractiveness of the commenter and the valence of the comments on transportation, F(1, 138) = 1.65, p = .20. Therefore, H2.2 and H2.4 were rejected.
Interaction effects of social features and gender on the entertainment experience. Hypothesis 3 indicated that the effect between social features on YouTube and the
entertainment experience is moderated by the gender of the viewer. It was hypothesized that this effect was stronger for females than for males. To test whether the gender of the
participants moderated the effect of the comments’ valence on enjoyment, a two-way ANOVA, with enjoyment as dependent variable and valence and gender as independent variables, was conducted. Results indicate that there was a small, but approaching significant interaction effect between the valence of the comments and gender on enjoyment, F(1, 138) = 2.89, p = 0.09, η² = .02. Figure 3 shows that the effect was stronger for males than for females. Male participants enjoyed the video more than female participants when the valence of the
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 Attractive Unattractive
Figure 2. The interaction effect between attractiveness of the commenters and valence of the comments on
enjoyment.
comments was positive. This means that H3.1 was not supported, because it was hypothesized that this effect was stronger for females than for males, but the opposite was found.
H3.2 indicated that there is an interaction effect between valence of the comments and gender on transportation. To test whether the gender of the participants moderated the effect of the comments’ valence on transportation a two-way, with transportation as dependent variable and valence and gender as independent variables, was conducted. Results indicate that there is a small interaction effect between valence of the comments and gender on transportation, F(1, 138) = 4.44, p = .037, η² = .03. Figure 3 shows that the effect was stronger for males than for females. Therefore, H3.2 was not supported, because it was hypothesized that this effect would be stronger for females than for males, but the opposite was found. 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 Positive Negative
Figure 3. The interaction effect between gender and valence of the comments on ejoyment.
Additional analyses. In addition to the previous analyses, it was tested whether this effect also occurs when attractiveness was added as an additional moderator. To test whether gender has a moderating effect on the interaction effect of the valence of the comments, and the attractiveness of the commenter on enjoyment, a three-way ANOVA was conducted. In this analysis, enjoyment was used as a dependent variable and gender, valence, and
attractiveness of the commenter were used as independent variables. Results indicate that there was no significant interaction effect between gender, valence of the comments, and attractiveness on enjoyment, F(1, 134) = .89, p = .34.
To test whether gender has a moderating effect on the interaction effect of the valence of the video and the attractiveness of the commenter on transportation, a second three-way ANOVA was conducted. In this analysis, transportation was used as a dependent variable and gender, valence, and attractiveness of the commenter were used as independent variables. Results show that there is no significant interaction-effect between gender, valence of the comments and attractiveness on transportation, F(1, 134) = .82, p = .37.
Discussion
The aim of this study was to investigate whether social features on YouTube have an effect on the entertainment experience and how attractiveness and gender moderate this effect.
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 Positive Negative
Figure 3. The interaction effect between gender and valence of the comments on transportation.
Results show that social features have a small effect on both enjoyment and transportation. Participants who were exposed to positive comments reported more enjoyment and felt more transported into the video than people who saw negative comments. This is in line with the social conformity theory of Asch (1951), which states that people base their attitudes and beliefs on what others do or say. The current research shows that positive comments lead to a more positive experience and vice versa. This finding is also in line with previous research on the effects of peer evaluations on entertainment experiences (Möller & Kühne, in press; Sherman et al., 2016; Edwards, Edwards, Qing & Wahl, 2007). This means that comments of others affect the way users perceive videos on YouTube. The way others talked about the video is a predictor for both enjoyment and transportation. This is interesting, because this means that the video itself is not the only a predictor for how people evaluate it.
The findings of this study also showed that attractiveness of the commenters had a small moderating effect on the relationship between the valence of the comments and enjoyment. The effect of attractiveness of the commenters was stronger for people in the attractive conditions. This means that people in the positive attractive condition showed more enjoyment toward the video and people in the negative attractive condition showed less enjoyment. These differences were less in the unattractive conditions. This means that attractive people have a stronger effect on enjoyment than unattractive people. This could be explained by the attractiveness halo effect of Kaplan (1978), which states that attractive people make content which is perceived as of better quality. It might be true that this leads to more trustworthiness and therefore have a stronger effect than people who are unattractive.
This finding is surprising, because the manipulation check for attractiveness failed. This means that participants could not remember that the commenters were either attractive or unattractive. A reason for this might be a wrong question about attractiveness. People were asked about whether they perceived the commenters as attractive or not. This could be a
wrong question, because the participants were exposed to same sex commenters. Therefore, they might not think the commenters were attractive, but they might still have found them good looking. It could also be that the effect of attractiveness on enjoyment happens unconsciously. People may not remember how the commenters looked like, or they did not say they were attractive, but unconsciously they perceived them as attractive, and this influenced their subsequent entertainment experience.
The attractiveness of the commenters had no effect on the relationship between social features and transportation. The reason here might be that attractiveness only has an effect on direct responses to the video in light of whether it is good or bad. Transportation is a more elaborated process and therefore more complex. It might be true that people’s transportation into a specific video does not depend on the attractiveness of the commenters, but more on the narrative of that video. Van Laer, De Ruyter, Visconti, and Wetzels (2014) found that there are different aspects from a narrative, such as identifiable characters, attention and imaginable plot that lead to narrative transportation. Therefore, it might be true that other factors, such as the attractiveness of evaluators about a specific narrative, do not predict transportation, but this can still influence their enjoyment.
In addition, gender had a small effect on the relationship between social features and enjoyment and transportation. Surprisingly, the effect of the valence of the comments was stronger for males than for females. Therefore, H3 was rejected, because it was hypothesized that this effect would be stronger for females than for males. Gender had no effect on the relationship between social features and attractiveness on both enjoyment and transportation. Thus, females do not react stronger on attractiveness than males do. This is contradictive with previous research, where several scholars found that females conform more than males (Dean, Austin & Watts, 1971; Eagly & Telaak, 1972; Marquis, 1973). To investigate this in more
detail, future research is needed to examine whether there are sex differences in conformity effects when it comes to social features from YouTube videos.
Although these gender differences were contradictive to the expectations, there is a logical reason to explain this effect. The video that was used for this study was a vlog from a male YouTuber. The video was about cleaning sneakers and going to a McLaren shop. This might be more tailored to males than to females. Therefore, it might be logical that male participants liked the video more than females did and were therefore more influenced by other commenters. Another reason for this effect could be the fact that the sample consists of more females than males. This could have an effect on the results, because a small group of male participants might not be representative for all males in society.
In summary, social features on YouTube had a small effect on both enjoyment and transportation. This effect is moderated by attractiveness and gender, with these effects being stronger among males, and for attractive commenters. Therefore, it can be concluded that social features do have an effect on people’s entertainment experience. People react
differently when the valence of the comments and the number of likes differs in a YouTube video.
Limitations and Future Research
The present study has a few limitations that need to be addressed in future studies. First of all, the sample of the study was very small and not based on a representative sample. Moreover, there were more female than male participants in this study. In a future study, it is therefore better to use a more representative sample. This means that the sample should consist of as much male participants as female participants, and also of people from a wider range of different backgrounds and ages. This would make results more generalizable to a bigger group of people. This study can only conclude about the sample, but not about all males and females in society.
A second shortcoming is that the video that was used for this study was more tailored to men. In a future study, it would be better to use a vlog that is tailored to both men and women, or to use a broader range of different vlogs. This would increase the validity of this study, because it would make it more easily to measure what is meant to measure. It is not only meant to measure the entertainment experience of male participants, but also of female participants. Now the female participants did not like the video at all, even when the
comments were positive. The mean score was under the 4, which was the middle score on the scale. This effect might be different when the vlog is more tailored to them. Another option is to use more vlogs, to see the difference between different kinds of vlogs tailored to different audiences.
Third, the manipulation for attractiveness seemed to have failed. It is not clear whether the manipulation itself failed, in that the attractive and good-looking commenters were really not perceived as more attractive, or whether there was a problem with the wording in the manipulation check. In a future study, it would be better to use different wording, and use terms such as ‘good-looking’ instead of ‘attractive’. Ideally, a pretest is included to test whether the participants actually perceive the commenters as attractive or unattractive.
The current research was limited to only two moderators; attractiveness and gender. Some small effects were found, but it might be interesting to build up on this. For future research it might be interesting to add other moderators to see whether these effects are stronger for specific groups. It would be interesting to investigate whether other factors lead to stronger or weaker effects. For example, specific characteristics of the vlogger itself, such as gender, humor, and attractiveness, might lead to stronger or weaker effects. Since it was already found that the attractiveness of the commenters has an effect on the entertainment experience, it might be true that the attractiveness of the vlogger itself has an effect as well. Besides, it was found that males enjoyed the video more than females. Maybe vlogs with
same sex vloggers might lead to more identification and therefore might lead to more
enjoyment and transportation. Also, characteristics of the audience might lead to other effects as well, such as age, gender, and attractiveness.
References
Asch, S. E., & Guetzkow, H. (1951). Effects of group pressure upon the modification and distortion of judgments. Groups, leadership, and men, 222-236.
Banjo, O. O., Appiah, O., Wang, Z., Brown, C., & Walther, W. O. (2015). Co-viewing effects of ethnic-oriented programming: an examination of in-group bias and racial comedy exposure. Journalism & Mass Communication Quarterly, 92(3), 662-680.
Berndt, T. J. (1979). Developmental changes in conformity to peers and parents.
Developmental psychology, 15(6), 608.
Bussey, K., & Perry, D. G. (1982). Same-sex imitation: The avoidance of cross-sex models or the acceptance of same-sex models? Sex Roles, 8(7), 773-784.
Chelaru, S., Orellana-Rodriguez, C., & Altingovde, I. S. (2014). How useful is social feedback for learning to rank YouTube videos? World Wide Web, 17(5), 997-1025. Chevalier, J. A., & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book
reviews. Journal of marketing research, 43(3), 345-354.
Cross, C. P., Brown, G. R., Morgan, T. J., & Laland, K. N. (2017). Sex differences in confidence influence patterns of conformity. British Journal of Psychology, 108(4), 655-667.
Dean, R. B., Austin, J. A., & Watts, W. A. (1971). Forewarning effects in persuasion: Field and classroom experiments. Journal of Personality and Social Psychology, 18(2), 210. Denham, B. E. (2004). Toward an explication of media enjoyment: The synergy of social
norms, viewing situations, and program content. Communication Theory, 14(4), 370-387.
Eagly, A. H., & Telaak, K. (1972). Width of the latitude of acceptance as a determinant of attitude change. Journal of Personality and Social Psychology, 23(3), 388.
Edwards, C., Edwards, A., Qing, Q., & Wahl, S. T. (2007). The influence of computer-mediated word-of-mouth communication on student perceptions of instructors and attitudes toward learning course content. Communication Education, 56(3), 255-277. Gangestad, S. W., & Scheyd, G. J. (2005). The evolution of human physical attractiveness.
Annu. Rev. Anthropol., 34, 523-548.
Green, M. C., & Clark, J. L. (2013). Transportation into narrative worlds: implications for entertainment media influences on tobacco use. Addiction, 108(3), 477-484. Hoffner, C., & Buchanan, M. (2005). Young adults' wishful identification with television
characters: The role of perceived similarity and character attributes. Media psychology,
7(4), 325-351.
Kaplan, R. M. (1978). Is beauty talent? Sex interaction in the attractiveness halo effect. Sex
Roles, 4(2), 195-204.
Van Laer, T., De Ruyter, K., Visconti, L. M., & Wetzels, M. (2013). The extended
transportation-imagery model: A meta-analysis of the antecedents and consequences of consumers' narrative transportation. Journal of Consumer research, 40(5), 797-817. Landy, D., & Sigall, H. (1974). Beauty is talent: Task evaluation as a function of the
performer's physical attractiveness. Journal of Personality and Social Psychology,
29(3), 299.
Littlewood, W. (1996). “Autonomy”: An anatomy and a framework. System, 24(4), 427-435.
Main, K. J., Aditya, S., & Dahl, D. W. (2014). How source cues impact consumer evaluations of direct and indirect persuasion attempts. Canadian Journal of Administrative
Sciences/Revue Canadienne des Sciences de l'Administration, 31(1), 35-43.
Marquis, P. C. (1973). Experimenter-subject interaction as a function of authoritarianism and response set. Journal of Personality and Social Psychology, 25, 289 – 296.
Miller, A. G. (1970). Social perception of internal-external control. Perceptual and Motor
Skills, 30(1), 103-109.
Möller, A. M., & Kühne, R. (in press). The Effects of User Comments on Hedonic and Eudaimonic Entertainment Experiences When Watching Online Videos.
Communications.
Nabi, R. L., & Krcmar, M. (2004). Conceptualizing media enjoyment as attitude: Implications for mass media effects research. Communication theory, 14(4), 288-310.
Noelle-Neumann, E. (1974). The spiral of silence a theory of public opinion. Journal of
communication, 24(2), 43-51.
Oliver, M. B., & Bartsch, A. (2010). Appreciation as audience response: Exploring entertainment gratifications beyond hedonism. Human Communication Research,
36(1), 53-81.
Palmer, C. L., & Peterson, R. D. (2016). Halo effects and the attractiveness premium in perceptions of political expertise. American Politics Research, 44(2), 353-382. Park, D. H., Lee, J., & Han, I. (2007). The effect of on-line consumer reviews on consumer
purchasing intention: The moderating role of involvement. International journal of
electronic commerce, 11(4), 125-148.
Perry, S. D. (2001). Commercial humor enhancement of program enjoyment: Gender and program appeal as mitigating factors. Mass Communication & Society, 4(1), 103-116. Raney, A. A., & Bryant, J. (2002). Moral judgment and crime drama: An integrated theory of
enjoyment. Journal of communication, 52(2), 402-415.
Ramanathan, S., & McGill, A. L. (2007). Consuming with others: Social influences on moment-to-moment and retrospective evaluations of an experience. Journal of
Schultes, P., Dorner, V., & Lehner, F. (2013). Leave a Comment! An In-Depth Analysis of User Comments on YouTube. Wirtschaftsinformatik, 42, 659-673.
Shedlosky-Shoemaker, R., Costabile, K. A., DeLuca, H. K., & Arkin, R. M. (2011). The social experience of entertainment media: Effects of others’ evaluations on our
experience. Journal of Media Psychology: Theories, Methods, and Applications, 23(3), 111.
Sherman, L. E., Payton, A. A., Hernandez, L. M., Greenfield, P. M., & Dapretto, M. (2016). The power of the like in adolescence: effects of peer influence on neural and
behavioral responses to social media. Psychological science, 27(7), 1027-1035.
Tal-Or, N. (2016). How co-viewing affects attitudes: The mediating roles of transportation and identification. Media Psychology, 19(3), 381-405.
Tamborini, R. (2003). Enjoyment and social functions of horror. Communication and emotion:
Essays in honor of Dolf Zillmann, 417-443.
Tsao, W. C., Hsieh, M. T., Shih, L. W., & Lin, T. M. (2015). Compliance with eWOM: The influence of hotel reviews on booking intention from the perspective of consumer conformity. International Journal of Hospitality Management, 46, 99-111.
Waddell, T. F., & Sundar, S. S. (2017). # thisshowsucks! The overpowering influence of negative social media comments on television viewers. Journal of Broadcasting &
Electronic Media, 61(2), 393-409.
Walther, J. B., DeAndrea, D., Kim, J., & Anthony, J. C. (2010). The influence of online comments on perceptions of antimarijuana public service announcements on YouTube. Human Communication Research, 36(4), 469-492.
Wiegman, O. (1987). Attitude change in a realistic experiment: The effect of party membership and audience reaction during an interview with a Dutch politician.
Wirth, W., Hofer, M., & Schramm, H. (2012). Beyond pleasure: Exploring the eudaimonic entertainment experience. Human Communication Research, 38(4), 406-428. Youtube. (n.d.). Youtube voor de pers. Retrieved from
Appendix I - Stimulus material
YouTube video: https://www.youtube.com/watch?v=vT2cWylK9AM&t=30s
