The Affect of Social Media on Parasocial Relationships

Abstract

This study aims to understand how social media affects parasocial relationships between celebrities and normal, everyday people. The research conducted in this study aims to establish a better understanding of how people view celebrities based off of social media and how that is affected by what career paths the celebrities have as well as how people view different kinds of celebrities based off of the different parasocial relationships that one might have with different celebrities. A total of 107 people participated in the study, with 80 women (roughly 74.8%) and 25 men (roughly 23.4%). The participants were recruited through a network system, where those recruiting the participants asked people in their networks of family, friends, and acquaintances to take part in the study. The method that was used was a manipulated screenshot of a Twitter account presented to each of the 107 participants, and we tested to see if there was a difference in the parasocial relationship that the participants had with one celebrity versus the other kind of celebrity. When the results came in, the data showed that we could not reject the null hypothesis for either of the hypotheses that we tested, and so we were unable to support what was being tested through our hypotheses.

Literature Review

The focus of this study is to measure how well people relate to celebrities and public figures as well as how that might affect how those celebrities and public figures are viewed via social media. Previous studies have been done researching the ways that people identify with certain celebrities and how those relationships exist in the first place. In the first study that I found, Benson P. Fraser and William J. Brown conducted research on how people identify with Elvis Presley, and how closely people model themselves and their own characteristics based off of the characteristics of the mediated persona (Benson P. Fraser & William J. Brown (2002)). Their study found that “… people selectively integrate the perceived values and behaviors they see in celebrities they admire and adopt them into their own lives” and that “mass communications scholars do not have an adequate theoretical framework for explaining the role of celebrities in influencing values and behaviors” (Benson P. Fraser & William J. Brown (2002)).

Previous studies have focused on developing a foundation of research for parasocial interatctions and parasocial relationships. In their study, T. Phillip Madison and Lance V. Porter discuss the different ways in which parasocial interactions, parasocial relationships, and imagined interactions are formed, between both mediated figures and other human beings. They discussed how parasocial interactions and relationships have changed, as well as how imagined interactions have changed to include both mediated public figures as well as fictional characters. As they stated in their study, “II (imagined interaction) research may also provide useful concepts for media effects researchers, as we seek to explain human cognition, emotion, behavior, and media decisions as results of media exposure” (Madison, T. P., & Porter, L. V. (2015)). This statement, in conjunction with their findings that measures of imagined interactions can be “contextualized for measuring the attributes of PSI” (Madison, T. P., & Porter, L. V. (2015)) and “the different attributes of PSRs with mediated personae are significantly related to and highly correlated with one another” (Madison, T. P., & Porter, L. V. (2015)), can indicate that imagined interactions, parasocial interactions, and parasocial relationships can effectively take place on a human face-to-face level as well as with a one-sided relationship between a fan and a mediated celebrity persona.

Jihyun Kim and Hayeon Song completed a study titled Celebrity’s Self-Disclosure on Twitter and Parasocial Relationships: A Mediating Role of Social Presence, in which they researched the parasocial relationships that people had with celebrities based off of each person’s use of the social media site Twitter. In their study, they discuss how celebrities typically use social media and their Twitter accounts in order to disclose personal information, professional, information, or both to fans and the general public, which can affect how the fans and the public develop parasocial relationships with the celebrities. This study tested to see how participants’ perceptions of celebrities were affected by celebrities’ self disclosure on social media (particularly via Twitter), and the results were that celebrities’ self-disclosure on social media “enhances fans’ feeling of social presence” as well as the fact that “social presence is found to facilitate positive PSI experiences” (Kim, J., & Song, H. (2015)).

Following Jihyun Kim and Hayeon Song’s research study, as well as thinking about the other studies on celebrities and parasocial interactions and relationships, I began to think more and more about how social media would affect people’s interactions and relationships with mediated figures – specifically, would a person’s view of and parasocial interaction and relationship with one kind of celebrity affect their view and parasocial interaction and relationship with another kind of celebrity. Conducting research on this topic could be helpful in understanding parasocial interaction and parasocial relationships between people and celebrities better in terms of how they affect how celebrities are viewed and the different kinds of relationships that people may have with celebrities based off of what kind of persona the celebrity has through the media (and, more specifically, social media). Such a study could help to set the foundations to begin to practically and theoretically understand the different ways in which people interact with celebrities and their mediated presences.

The concept being tested in this study is parasocial relationships, which has been defined as “relationships that are typically one-sided relationship built between a fan and a celebrity, where the fan thinks and feels as though the celebrity is a friend despite the fact that the celebrity is actually not a friend and close personal acquaintance” for the sake of this study. Similarly, a fan is defined as “someone who takes a great liking to and interest of a mediated persona typically known as a celebrity,” and a celebrity is defined as “a person who is well-known and well-trained in the area of the media and has made a career out of being well-known and famous.” The relevant dimensions (variables) being tested in this study are how hardworking celebrities appear to be, and how relatable the celebrities are. The variables are meant to measure participants’ parasocial relationship with the celebrities through the use of social media. The social media site selected for this study was Twitter for similar reasons to that it was chosen for Jihyun Kim and Hayeon Song’s research study – celebrities typically use social media and their Twitter accounts in order to disclose personal information, professional, information, or both to fans and the general public (Kim, J., & Song, H. (2015)).

Building on the research of previous studies, the theory being researched in this study is whether or not people will have different opinions of and relationships with celebrities and public figures based off of the use of social media (specifically Twitter) and parasocial relationships. The independent variable being manipulated for each hypothesis is the exact celebrity status of each mediated persona. The first hypothesis being tested (hereafter known as H1) is as follows: Actors will more likely be seen as more hardworking than internet-famous people. The dependent variable is the characteristic of being hardworking. The rationale for this hypothesis is that actors tend to be more of a traditional mediated celebrity and are what people are used to seeing, so the public will be more inclined to understand how much time, effort, and skill someone typically tends to put into their career of acting, thus they are able to develop a stronger parasocial relationship with the celebrity actor.

The second hypothesis being tested (hereafter known as H2) is as follows: Internet-famous people will be seen as more relatable than actors. The dependent variable being presented is the characteristic of relatability. The rationale for this hypothesis is that video bloggers on YouTube tend to be seen as everyday, “average” people despite being a popular celebrity figure, and so they are able to connect and identify with the average person on an easier level than the famous actor or actress would, thus helping “normal” people create and maintain a parasocial relationship with such celebrities.

Methods Section

Participants

A total of 107 people completed the survey, ranging from 18 years of age to 80 years of age (M = 28.41, SD = 14.01) completed an online survey that was sent out through email. For the study, they were recruited through network sampling, through which those recruiting people asked people who they knew in person and online to participate in the study. The participants were randomly assigned which variation of the study they would receive, and they were split into two groups – one group consisting of 53 participants, and the other group consisting of 54 participants. Out of the 107 participants, 23.4% were male, 74.8% were female, and 1.8% identified as another gender or were missing from the response set. 74.8% of the participants identified as white, 7.5% identified as Hispanic/Latino, 1.9% identified as black/African American, 16.8% identified as Asian/Pacific Islander, 3.7% chose the option “other” for their ethnicity, and 0% identified as Native American. For the participants’ level of education, 20.6% graduated from high school, 35.5% received some college education without receiving a degree, 7.5% received an Associate’s degree, 23.4% received a Bachelor’s degree, 8.4% received a Master’s degree, and 4.7% received a Doctorate degree. When asked about which political party they belonged to, 20.6% of participants belonged to the Republican Party, 44.9% belonged to the Democratic Party, 33.6% were Independents or belonged to other political parties, and 0.9% of the participants were missing from this question. On a scale from 1 to 7, with 1 being labeled as “very liberal” and 7 being labeled as “very conservative,” the mean for ideology was 3.34, and the standard deviation was 1.46.

Stimuli

The stimuli used in this study were two pictures of a Twitter account belonging to a woman named Jill Jones. One of the manipulations showed that Jill Jones was an actress who was also a mother, and an activist/environmentalist going to an awards show and excited to see her fans there, which is the manipulation that was used alongside H1. The other manipulation of the picture showed that Jill Jones was a beauty video blogger on YouTube at a video blogger convention and excited to see her fans there; this manipulated version of the stimuli was used with H2. The stimuli were used to see if there would be a difference in how people viewed different kinds of celebrities and public figures with a focus on who would be seen as more “hardworking” and/or “relatable.”

Variables

For H1, the characteristic of being hardworking was tested with this study. There was a total of three questions used to measure this variable, and each question used a seven point scale, similar to a Likert-type scale, to determine how much the participants agreed with the statement being presented. The scale ranged from 1 being labelled as “strongly disagree” to 7 being labelled as “strongly agree” and 4 being labelled as “neutral.” The reliability for these questions was a Cronbach’s Alpha of 0.855. Overall, the questions used for the variable are reliable, and the index for the variable can help to better understand such a claim (M = 3.92, Median = 4.00, SD = 1.10). The questions were as follows:

  1. Jill Jones is hard-working.
  2. Jill Jones balances her career and other commitments.
  3. Jill Jones puts in the necessary effort and skill to succeed.

As for H2, a total of three questions were used to measure this variable and each question used a seven point scale, similar to a Likert-type scale, to determine how much the participants agreed with the statement being presented. The scale ranged from 1 being “strongly disagree” to 7 being “strongly agree” and 4 being neutral. The set of questions had a Cronbach’s Alpha of 0.780. However, the hypothesis had an alpha-if-item-deleted of 0.814 if the first question was deleted from the set of questions. While the alpha would have been higher if the first question was deleted, I have decided to keep the question included in the set due to it including content that was necessary to statistically understand. The index for this variable was a result of such reliability (M = 4.05, Median = 4.00, SD = 1.188). The questions were presented as follows:

  1. Jill Jones is an interesting person.
  2. Jill Jones is relatable.
  3. Jill Jones is able to relate to people who have the same interests as her.

Results Section

For H1, a t-test was performed. The t-value for this hypothesis was 0.600 and the p-value was 0.550. For the variable of the actress, the mean was 3.8526. For the variable of the YouTube beauty video blogger, the mean was 3.9815. As a result, the study failed to reject the null hypothesis, and we failed to support the hypothesis that was presented due to the fact that the null hypothesis could not be rejected.

For H2, a t-test was also performed. With this hypothesis, the t-value was .698 and the p-value was a 0.487. The mean for the variable of the actress was 3.9686.  The mean for the variable of the YouTube beauty video blogger was 4.1296. As a result, we failed to reject the null hypothesis. We failed to support the hypothesis that was presented due to the fact that the null hypothesis could not be rejected.

Hypothesis 1
T-value 0.600
P-value 0.550
Mean for the actress variable 3.8526
Mean for the video blogger variable 3.9815
Hypothesis 2
T-value 0.698
P-value 0.487
Mean for the actress variable 3.9686
Mean for the blogger variable 4.1296

 

Discussion Section

The main findings of the study indicated that there was no measurable distinction between how people viewed actors versus how they viewed beauty bloggers. The pattern of results indicated that the participants viewed the manipulations roughly the same, and that there generally didn’t seem to be any differences, which meant that the hypotheses couldn’t be supported and that we failed to reject the null – meaning that we failed to find enough evidence to say that we could reject the notion that there wasn’t any relationship between the variables tested in this study. The differences in the study mainly boiled down to one person not completing the survey related to the manipulated variable of the actress, one of the questions for the manipulated variable of the beauty video blogger having a slightly higher Cronbach’s alpha (if item deleted), and slightly different results for the t-tests that were run. The results were slightly unexpected, they did go against what was predicted in both of the hypotheses; in any case, the study went as well as anyone could have expected.

The implications of these findings are useful for both applied and theoretical approaches and purposes. The findings can help people better understand how different kinds of celebrities are viewed, and how that affects the way that people perceive the different celebrities via social media. Additionally, this can help social media users and content creators to better understand how they can reach people and how they might be viewed, especially via social media and Twitter, with fans and followers that have parasocial relationships with them. The implications of the findings of this study can help with theoretical research on the topic of social media and parasocial relationships in communications research by further explaining how people’s parasocial interactions and relationships with celebrities work, especially if they are being studied through the use of social media. Additionally, the findings presented in this study can aid communications researchers in their developing theoretical research on how social media affects communicative interactions and relationships between people as well as celebrities.

Despite the manipulated independent variable seeming to have little effect on the dependent variables, this can help researchers figure out how to better understand, question, and test the relationship between the variables presented in this study as well as how social media affects parasocial relationships and interactions. The theory presented in this study – whether or not people will have different opinions of and relationships with celebrities and public figures based off of the use of social media (specifically Twitter) and parasocial relationship – can be modified to better fit the research being conducted as well as to better connect the research already completed in this field of communications studies. Even with that, it seems to have been a bit of a stretch from past theories and theoretical research on this subject since little research and studies have been done on the cross-sections of social media and parasocial relationships, so this study could have missed some critical aspect of what to research and how to measure it more effectively. Pre-existing theories should have been slightly more modified to take into account how social media affects people and how people interact, especially with celebrities, and how people build parasocial relationships.

The shortcomings and flaws of this research study are that there are questions that should be reworded in order to better gauge the hypotheses in question. In particular, the question and statement measuring the second hypothesis of “internet-famous people will be seen as more relatable than actors” need to be worded more in a way that will more clearly and fully measure how the independent variable of internet-famous people interacts with and manipulates the dependent variable of relatability. The statements were not worded in a way conducive to this study, and thus affected the results and findings of the study. The question and statements failed to sufficiently measure the relationship presented in the second hypothesis by not accurately measuring what they were meant to measure, resulting in a negative and harmful effect on the study and potentially invalidating the findings.

For future studies attempting to replicate the research found in this study, further theoretical research needs to be completed on this topic, and the theory and hypotheses should be expanded upon so that they are better able to capture a more accurate statistical basis for support of or against this theory and the hypotheses. Make sure to have a better foundation for the research being conducted, and work on creating better ways of statistically testing and measuring the hypotheses presented in the study.

Citations and Further Readings

Benson P. Fraser & William J. Brown (2002) Media, Celebrities, and Social Influence: Identification with Elvis Presley, Mass Communication and Society, 5:2, 183-206, DOI: 10.1207/S15327825MCS0502_5

Brown, W. J. (2015), Examining Four Processes of Audience Involvement With Media Personae: Transportation, Parasocial Interaction, Identification, and Worship. Commun Theor, 25: 259–283. doi:10.1111/comt.12053

Kim, J., & Song, H. (2015). Celebrity’s Self-Disclosure on Twitter and Parasocial Relationships: A Mediating Role of Social Presence. SSRN Electronic Journal, 570-577. doi:10.2139/ssrn.2756744

Madison, T. P., & Porter, L. V. (2015). Cognitive and Imagery Attributes of Parasocial Relationships. Imagination, Cognition and Personality, 35(4), 359-379. doi:10.1177/0276236615599340

Young Min Baek, Young Bae, and Hyunmi Jang. Cyberpsychology, Behavior, and Social Networking. July 2013, 16(7): 512-517. https://doi.org/10.1089/cyber.2012.0510

Advertisements

Leave a Reply