Unveiling the Influence of Audio Type and TikTok Usage Frequency on Parasocial Interactions

Delve into our in-depth analysis of TikTok user experiences. Our study investigates how audio types, including speaking, voice-over effects, music, and control, affect the strength of parasocial interactions among female viewers. We also explore the fascinating relationship between audio type and TikTok usage frequency. Our findings reveal valuable insights into the dynamics of social media engagement, shedding light on what truly drives these digital connections. Join us on a journey through data-driven exploration, unlocking the secrets of TikTok's immersive world. Gain a deeper understanding of the factors that shape the bonds between content creators and their audience.

Problem Description

This sample addresses the question of how audio types in 'Get Ready with Me' TikTok videos impact the strength of parasocial interactions among female viewers. We also examine the role of TikTok usage frequency in moderating this relationship. By exploring these dynamics, we aim to provide insights into the intricate interplay between content, audience engagement, and the platform's audio features.

Main Question:To what extent does audio type (speaking, voice-over effect, music, control) in 'Get Ready with Me' TikTok videos predict the strength of female viewers' parasocial interaction, and is this relationship moderated by the frequency of time spent on TikTok?

Manipulation Check

To assess the effectiveness of the stimuli used in the experiment, a manipulation check was conducted. The primary objective of this check was to confirm that participants correctly identified the audio type of the TikTok videos they watched. This validation process ensured that the different audio types were clearly distinguishable to participants. The manipulation check aimed to establish the accuracy of participants' recognition of the audio type.

To address concerns related to the Chi-Square analysis, a recoding procedure was implemented. Following this adjustment, the manipulation check results confirmed the participants' accurate identification of the audio type in the TikTok videos, supporting the integrity of the experimental design.

Table 1: Chi Square Analysis Results

Measures

The independent variable for this study is audio type used in a ‘Get Ready with Me’ TikTok video. The dependent variable is the viewer's strength of Parasocial interaction. The moderator in the study is the frequency of time spent on TikTok. Several control variables are also listed.

'Get Ready with Me' Videos

'Get Ready with Me' or 'GRWM' videos are popular on platforms like YouTube and TikTok, where vloggers showcase their daily routines, often involving makeup or preparations. Notable influencers in this category were mentioned.

Audio Type

There were four different audio types used in the experiment: speaking, voice-over effect, music, and a control group with no audio.

Parasocial Interaction

Parasocial interaction was measured using the EPSI-Experience of Parasocial Interaction scale. The scale demonstrated good reliability and indicated that, on average, participants had a moderately strong parasocial interaction.

Table 2: Parasocial Interaction Scale Results

Frequency of TikTok Use

The frequency of TikTok use, a moderator for the study, was measured on a continuous level using a 7-point Likert scale.

Table 3: Frequency of TikTok Use Results

Control Variables

Control variables included time spent on social media in general, other social media apps used, rank order of social media platforms used, gender, and familiarity with the TikToker in the video.

Time Spent on Social Media in General

Participants' time spent on TikTok was measured as a control variable for this experiment. This was measured with the same 1-item 7-point Likert scale as frequency of time spent on TikTok. Only the question slightly changed from “In the past week, on average, approximately how much time PER DAY have you spent actively using Facebook?” to “In the past week, on average approximately how much time PER DAY have you spent actively using social media?” On the scale from one to seven, participants scored higher than average, M= 4.66, SD = 1.14. This means that in our sample, 34.1% of participants spend a self-reported average of 2-3 hours on social media per day, 26% of participants spend a self-reported average of more than 3 hours per day, and 25.2% of participants spend a self-reported average of 1-2 hours on social media per day.

Table 4: Standard deviation of the average time spent on social media

Other Social Media Apps Used

Data on what other social media apps participants used was also collected for this study. Participants were asked to select all social media apps they actively use, in a one item multiple response question (See Appendix 2). Social media platforms were selected based on global rankings of social media apps used but excluded any messaging apps (Kemp, 2022). By running a multiple response analysis in SPSS it was found that 95.9% of participants also actively use Instagram, 75.6% actively use Youtube, 67.5% actively use Snapchat, 60.2% actively use Facebook, 56.9% actively use LinkedIn, 49.6% actively use Pinterest, and 35.8% actively use Twitter.

Rank Order of Social Media Platforms Used

Participants in the study were also asked to rank the 8 most popular social media apps by 1= most used through 8= least used. If participants did not use the apps they were asked to place the app(s) at the bottom of the list. Using Friedman’s ANOVA in SPSS it was found that there was consistency in the manner in which participants ranked which social media apps they used most (χ2 = 382.37, p < .001). The results showed that the order from most to least used social media apps by participants goes Instagram (M= 1.93), TikTok (M= 2.33), Youtube (M= 4.40), Snapchat (M= 4.55), Facebook (M= 4.62), LinkedIn (M= 5.90), Pinterest (M= 6.11), and Twitter (M= 6.16).

Gender

Gender was also measured as a control variable in this research experiment. Gender was measured with one item “With what gender do you identify?” In the sample, 80.5% of participants identified as female, 17.1% identified as male, and 2.4% identified as other (M=1.85, SD= 0.42).

Familiarity with TikToker

The final control variable that was measured was the participants familiarity with the TikToker that was in the TikTok video. This means that it was measured whether or not the participants already knew how the TikToker in the video was before taking the survey. Familiarity of the TikToker was measured with 1 item “Did you recognize the TikToker in the video?” On a scale from one to three, the participants scored on average, M= 2.02, SD= 0.27. Therefore, in general 92.7 % participants were not familiar with the TikToker in the stimulus videos. Because so many participants did not know who the TikToker was, this was excluded from the final analysis.

Results

Statistical Analysis:

Version 29 of SPSS (the Statistical Package for the Social Sciences) was used to conduct the main data analysis for this study. Before running any analysis, the data was checked for missing values and although proven reliable by previous research, a reliability analysis of the scale for Parasocial Interaction was run to make sure that scale accurately measures Parasocial Interaction. Descriptive statistics were also run in SPSS to analyze the sample’s overall demographics, as well as a randomization check to evaluate the overall distribution of the test statistics.

For the main effect of audio type on strength of Parasocial Interaction moderated by frequency of time spent on TikTok, a regression with a dummy coded IV (audio type) and an interaction term was run. The significance level for all the relevant statistical tests were set at below .05.

Randomization Check

A randomization check was conducted to assess the balance of groups. The results revealed no significant difference in mean parasocial interaction scores across different audio types, supporting the assumption of successful randomization.

Table 5: Randomization Check Results

Analysis

The analysis did not find a significant difference in the strength of parasocial interaction among female TikTok users exposed to different audio types. The ANOVA results and multiple comparisons did not support H1.

Table 6: ANOVA Results

Table 7: Multiple Comparisons

Analysis

The regression model examining the association between audio type, TikTok usage frequency, and parasocial interaction did not provide strong evidence for an association. The overall regression model was not statistically significant, and none of the predictors, except for the variable "Song," had statistically significant coefficients, indicating limited associations between audio type, TikTok usage, and parasocial interaction.

Table 8: Regression Model Summary

Table 9: Regression Coefficients

Conclusion

In summary, the study successfully balanced groups through the randomization check, but the analyses did not provide strong evidence to support either hypothesis. The strength of parasocial interaction among female TikTok users did not significantly differ based on audio type, and the relationship between audio type, TikTok usage frequency, and parasocial interaction was weak.

These findings suggest that factors other than audio type and TikTok usage frequency may play a more prominent role in influencing the strength of parasocial interaction among female viewers. It is important to note that the study could benefit from a larger sample size and a more comprehensive approach to further explore these relationships. The tables presented here provide a detailed breakdown of the results and measurements conducted in the study.