INTRODUCTION

Tobacco use is a major cause of non-communicable diseases and death, with over 8 million people dying yearly from smoking-related causes1. Numerous studies consistently indicate that tobacco use is a significant risk factor associated with cancer2. Smoking and mental health problems are highly comorbid among young people3. Therefore, effective tobacco control is urgently needed.

Social media have been crucial in promoting health and educating about various diseases4. Social media can provide services for young adult smokers seeking smoking cessation help5. Research shows that interactive web-based smoking cessation interventions contribute to quitting smoking6. Since young people are the primary social media users, targeting smoking cessation promotion at young people through social media is considered a practical approach.

Among various social media platforms, online videos have become the preferred form of content consumption for people, especially young individuals. As of June 2022, China has seen significant growth in the user base of online videos, reaching 962 million users7,8. Bilibili, as one of the video platforms, has the youngest user base among Chinese internet users, with 90% being users aged <25 years and often associated with youth culture9,10. Previous studies have shown that Bilibili also plays a positive role in promoting health among young people11. Therefore, it is worth exploring whether Bilibili actively promotes smoking cessation.

The success of health promotion through the media often depends on the persuasiveness and impact of the transmitted information12. When measuring the effectiveness of video dissemination, studies have shown that both MSV (message sensation value) and message appeal are crucial indicators for assessing the persuasive strength of video information13. MSV refers to the extent to which the audio-visual characteristics of television messages evoke sensory, emotional, and arousal responses14. MSV captures their attention to the information by inducing emotional reactions in respondents, leading to persuasion15. A higher MSV results in a more significant persuasive impact16. Research has demonstrated the effectiveness of high MSV in antismoking advertisements targeting young people17.

Message appeal refers to the information used to attract consumers’ attention and influence their perception of the product15. Different types of message appeals, whether rational or emotional, can influence the persuasive effects of information recipients18. In smoking cessation research, the three most prominent message appeals are threat, social, and humor appeal19. Additionally, research indicates that economic and psychological effects also play a role in reducing harm from electronic cigarettes in videos20.

Therefore, this study aims to explore the influence of MSV and message appeal, as two indicators for evaluating information persuasiveness, on smoking cessation videos on Bilibili. It aims to provide specific strategies for social media, especially social video platforms, to enhance the communication effectiveness of smoking cessation promotion.

METHODS

Sample

Using the keyword ‘smoking cessation’, the research team searched for smoking cessation-themed videos on Bilibili in 2022, resulting in 423 videos. The research team manually filtered the data to exclude videos unrelated to smoking cessation or that did not meet the coding criteria. Ultimately, 337 videos were selected as the study sample. These videos contained information such as URLs, publishers, view counts, share counts, comment counts, like counts, danmaku (bullet comment) counts, and coin counts (a Bilibili feature where fans can reward video creators).

The sample data in this study exhibited a generally normal distribution. To further confirm the normality, the Kolmogorov-Smirnov test was employed with a significance level of p<0.05, considering the skewness and kurtosis values as references. The criteria set were absolute skewness values less than three and absolute kurtosis values less than 10, indicating a normal distribution21. For data that did not follow a normal distribution, appropriate transformations, such as logarithmic transformations, were applied to address the issue.

Coding

Dependent variable

Following the approach proposed by Chen et al.19, the communication effects of the videos were categorized into four dimensions: overall communication effect, communication breadth, communication recognition, and communication participation. The overall communication effect was measured using the logarithmic sum of seven indicators: shares, views, likes, favorites, coins, danmaku (bullet comment), and video comments11. The number of shares19 measured communication breadth. Communication recognition was assessed using the logarithmic sum of four indicators: views, likes, favorites, and coins19. The number of danmaku and comments19 measured communication participation.

Independent variable

This study referred to the Lang et al.22 definition of MSV and the category design in the MSV coding scheme by Paek et al.13. Additionally, considering the characteristics of the sample and referring to the category construction by Palmgreen et al.23, several variables were modified or removed during the development of the coding scheme. For instance, the category ‘Loud/Fast Music’ was removed from the audio type category. After preliminary observations of smoking cessation video samples on Bilibili, it was determined that elements such as evidence and norms in the persuasion message features played a more significant role in persuasion. As a result, the coding scheme by Paek et al.24 was adjusted, replacing ‘Acted out’ and ‘Unexpected format’ with ‘Norms’ and ‘Evidence’, respectively. The MSV dimension coding comprised three types and twelve items (Table 1). In addition to using threat, social, and humor appeals, we also included psychological and economic benefits in the coding of message appeals (Table 1). During coding, a binary approach was used to determine whether the five types of appeals were present in the samples, with 0 indicating the absence and 1 indicating the presence of each appeal25.

Table 1

Operational definitions of variables

VariablesPerreault and Leigh Index
Types of creators0.98
1) Individuals 2) Medical workers, 3) Media, 4) Corporate companies
Video type0.76
1) Vlog, 2) Showcase, 3) Animation, 4) Acting, 5) Documentary, 6) News, 7) Mashup video, 8) Oral speech, 9) Pictorial slideshow, 10) Interview
Video length0.93
1) <30 s, 2) 30–59 s, 3) 1 min –1 min 59 s, 4) 2 min – 2 min 59 s, 5) 3 min – 4 min 59 s, 6) ≥5 min
Message appeals
Threat appeal: the overall impression is that you will suffer in some way if you smoke by showing cancer patients, gross teeth or lungs, and scary images of people who suffer from smoking-related diseases.0.88
Social appeal: visuals and major headlines convey that you will have more friends, dates, and popularity if you do not smoke.1
Humor appeal: play on words, puns, or nonsensical statements.0.88
Psychological benefit: Visuals and major headlines may move beyond outward appearance to convey smoking cessation’s impact on self-esteem and self-confidence.0.93
Economic benefit: Overall, you will achieve career success and financial security or benefit from smoking cessation.0.93
Message sensation value
Video/images
Number of cuts: the number of times the camera cuts from one visual scene to the next. Converted to low (0–6), moderate (7–14), and high (≥15) levels and coded as 0, 1, or 2.0.83
Visual effects: anything beyond the range of human ability involving special visual effects.0.98
Slow motion: the slowing of real-life action through technical intervention.0.98
Bold or unusual colors: unusual colors outside the range of colors usually perceived in real life.1
Intense images: intense or horrifying images including needles going into arms, guns pointed at heads or death.0.85
Audio/music
Sound saturation: background sound throughout the video clip, including street noise or other sounds, rather than simply a person talking throughout the video clip.0.76
Background music: music to accompany the dialogue or action of the video clip.0.98
Sound effects: unusual sounds (those that could not have occurred in real life) heard in the video clip, including gongs and other noises.0.93
Content
Acted out: instead of being told about the dangers of smoking, viewers see actions corresponding to the point of the smoking cessation video clip.0.88
Surprise/twist ending: the presence of a climactic, shocking end to the smoking cessation video clip.0.83
Evidence: The use of evidence enhances the credibility of the message source.0.8
Norms: Messages can specify what ought to be done or avoided (injunctive norms) or what is typical (descriptive norms).0.83

Referring to the coding scheme by Stellefson et al.26, the citation of video creator types was incorporated as one of the coding indicators. Video creators were classified into four types: individuals, healthcare professionals, media, and companies (Table 1). We classified the video types by drawing on references such as Niederdeppe17 and Li et al.27 and considering the characteristics of the sample videos (Table 1)17,27.

Reliability test

Two coders jointly performed the coding for this study. Before the official coding process, the coders randomly selected 10% of the video samples for pre-coding and conducted a HOLSTI reliability test to assess the agreement between the coders28. After two training rounds, the inter-coder reliability coefficient between the two coders was 0.901, indicating good reliability.

Statistical analysis

Nominal variables were presented as frequency (%), and the chi-squared test was used to examine the associations between creator and video type, MSV, and message appeal. ANOVA was employed to analyze the relationships among creator types, video types, and MSV. In cases of missing values, the mode was used to fill in nominal variables, while the mean was used for quantitative variables. Linear regression analysis examined the relationships between the independent and dependent variables (communication effect). IBM SPSS Statistics 26.0 was utilized to process all coded data, and the statistical significance level was set at p<0.05 for all two-tailed tests.

RESULTS

Overview of videos related to smoking cessation

Among the 337 videos, individual users accounted for the highest proportion, approximately 48.1%, followed by media at 35.4%. Healthcare professionals were relatively fewer, accounting for 10.3%, while corporate creators were the least represented, comprising only 6.2% (Table 2). Oral speech and Vlog-style videos had higher proportions, accounting for 37.7% and 26.7%, respectively. Conversely, documentary and interview videos had lower proportions, with only 0.3% and 0.6%, respectively (Table 2). As shown in Table 2, the average MSV score was 4.49 (SD=2.23), ranging from 0 to 11. Upon closer observation, the videos exhibited the following distribution of characteristics: approximately 35% of the videos included a high level of editing (i.e. exceeding 15 cuts), around 55.8% of the videos had sound saturation, about 47.5% included background music, approximately 55.2% of the videos advocated for viewers to take action towards smoking cessation, around 33.2% of the videos had a surprise ending, and about 61.4% of the videos established message standards. The remaining elements were used in <20% of the videos. From Table 2, it can be observed that threat appeals were most commonly used (49%), followed by humor appeals (24.2%) and psychological benefit (22.1%), with economic benefit (8%) and social appeals (1.5%) being the least utilized.

Table 2

General characteristics of sample video clips, Bilibili 2022 (N=337)

CharacteristicsMarginal meanSDRange
MSV13,22-244.492.230 – 11
Communication breadth236.581377.633
Communication recognition79317.744337806.209
Communication participation165.384720.906
Message appeals25%Creators types26%Video types%Video length%
Social1.5Individuals48.1Oral speech37.7<30 s9.8
Threat49.0Medical workers10.3Vlog26.730–59 s24.3
Humor24.2Corporation6.2Showcase17.81 min – 1 min 59s29.1
Psychological benefit22.1Media35.4Acting7.32 min – 2 min 59s13.1
Economic benefit8.0Mashup video3.9min – min 59s13.9
Pictorial slideshow2.4≥5 min9.8

[i] MSV: message sensation value.

From the results of the chi-squared analysis, it can be seen that individual users are more inclined to create Vlog-style videos. Media and healthcare professionals are more inclined to produce oral speech videos, while corporate users are more inclined to create showcase-style videos (p<0.001) (Table 3).

Table 3

Number of video types according to creator types, Bilibili 2022 (N=337)

Video typesCreator types
Totalχ²p
MediaIndividualsMedical workersCorporation
Oral speech8722180127234.06<0.001
Vlog3851190
Pictorial slideshow06118
Acting4180325
Mashup video180413
Showcase231911760
News20316
Animation03115
Interview01012
Documentary00011
Total1201623520337

The chi-squared test revealed significant differences in the utilization of threat appeal (p<0.001), humor appeal (p<0.001), and psychological benefit (p<0.05) among different types of videos. Specifically, oral speech videos tended to employ threat appeal and psychological benefit, while acting videos were more likely to use humor appeal.

One-way analysis of variance (ANOVA) revealed significant differences in MSV levels among different types of smoking cessation videos. However, due to sample sizes of ≤2 for some video types (e.g. only one documentary video), the requirements for post hoc tests were not met in this study. Therefore, post hoc tests were not conducted on the study sample. Specifically, most video types, such as animations, news, slideshow-style videos, and mashup videos, tended to include music. Vlogs and showcase videos generally had higher levels of sound saturation. Performance-based, mashup-style, and interview-style videos often featured surprise/twist endings. On the other hand, oral speech and documentary videos typically incorporate the ‘evidence’ element.

The communication effect of smoking cessation topic videos and its influencing factors

To examine the communication effect of smoking cessation videos, we conducted a linear regression analysis with independent variables selected from the perspectives of video creation, MSV, and message appeal. The results of linear regression analysis for dependent variables such as overall communication effect, communication breadth, communication recognition, and communication participation indicated that MSV positively influenced the communication effect of the videos (p<0.001). When the videos included message appeal elements such as threat appeals, humor appeals, and economic benefits, the overall communication effect, communication breadth, recognition, and participation of the videos were enhanced (p<0.001) (Table 4). Additionally, the overall communication effect and communication recognition were influenced by the type of video creator (p<0.001), while the communication breadth was influenced by the type of video (p<0.001) (Table 5). Further details of the linear regression analysis can be found in the Supplementary file.

Table 4

The frequencies of message appeal used in different video types, Bilibili 2022 (N=337)

Message appealVideo types
Totalχ²p
Oral speechVlogPictorial slideshowActingMashup videoShowcaseNewsAnimationInterviewDocumentary
Total1279082513606521337
Social6.4860.690
None1268982513576521332
Yes11000300005
Threat42.769<0.001
None595741910133010166
Yes6833463473511171
Humor109.329<0.001
None11582633384311256
Yes1282221022221081
Psychological benefit20.4940.015
None917472513436311264
Yes361610017021073
Economic benefit15.8140.071
None1178662513506521311
Yes10420010000026
Table 5

The differences in message sensation value among different types of videos, Bilibili 2022 (N=337)

Video typesTotalMessage sensation value
Fp
Marginal meanSD
Oral speech1272.8431.35939.575<0.001
Vlog904.2561.458
Pictorial slideshow85.2500.886
Acting256.9601.695
Mashup video138.0001.354
Showcase606.2331.899
News64.8332.714
Animation56.0002.550
Interview26.5002.121
Documentary15.0000.000
Total3374.4962.228

[i] SD: standard deviation.

DISCUSSION

This study conducted a content analysis of smoking cessation videos on Bilibili and investigated factors influencing the communication effect of these videos. Based on social cognitive theory, individuals exposed to videos may learn about smoking and its consequences, which can influence their smoking behavior. Although it is challenging to measure the impact of watching videos on smoking behavior, viewers’ reactions to the videos can be used to evaluate the effectiveness of smoking cessation videos on a promotional and educational level. Understanding the factors that influence the communication effect of videos is crucial for promoting smoking cessation on social media platforms, particularly video platforms. The findings of this study indicate that MSV and message appeal elements such as threat appeals, humor appeals, and economic benefits significantly influence the communication effect of smoking cessation videos. In contrast, the impact of video types and creator types on the communication effect is limited.

MSV and smoking cessation promotion for youth groups

This study found that MSV significantly influences the communication effect of smoking cessation videos. Higher MSV scores are associated with better communication effects of smoking cessation videos. Additionally, the variance results showed significant differences in MSV among different video types. Mashup videos had the highest average MSV score (marginal mean=8.00, SD=1.35), although they had the lowest number of videos. On the other hand, oral speech videos had the lowest average MSV score (marginal mean=8.00, SD=1.35), but they had the highest number of videos.

The study also found that viewers are more likely to watch videos with impressive content, such as powerful images, actions, and fast-paced sequences22. Attractive visuals can create a stronger association with smoking, making people perceive significant benefits and lower risks29. Research suggests repeated and frequent exposure to smoking cessation messages containing MSV elements can change people’s attitudes toward smoking or maintaining a non-smoking status30. This study suggests that video creators on social video platforms can incorporate more MSV elements into their video production to enhance the persuasiveness of smoking cessation-themed videos.

Message appeal and smoking cessation promotion for youth groups

The results demonstrated that videos with message appeal elements such as threat appeals, humor appeals, and economic benefits have better communication effects31. In this study, 49% of smoking cessation-themed videos on Bilibili contained threat appeals. Video creators often use threat appeals to emphasize the irreversible harm caused by smoking to the body and the negative influence on individuals in the same environment. Threat appeals can change attitudes by arousing sufficient fear and enhancing people’s responses to the information15.

Humor appeals are also an essential factor influencing the communication effect of smoking cessation videos18. Humor appeals are often found in advertisement videos or narrative-based videos. When humor is combined with execution cues such as music, intriguing interludes, and charming models in advertisements or narrative videos, they evoke positive effects in the audience, such as excitement, joy, and warmth25,32. This study suggests that creators of smoking cessation-themed videos can consider incorporating humor elements to evoke positive emotions and responses among young audiences regarding smoking cessation.

While primary smoking prevention among youth is crucial, promoting smoking cessation directly benefits economic savings33. The study also found that although videos with economic benefit elements were relatively few, they still influenced the communication effect of the videos34. Therefore, when creating smoking cessation-themed videos, creators can consider incorporating elements related to economic benefits to emphasize the significance of smoking cessation in terms of financial savings.

Video content and smoking cessation promotion for youth groups

Certain video types on social media platforms receive better communication effects due to public preferences20. Therefore, regarding smoking cessation videos on Bilibili, video types influence the communication breadth. Further analysis using one-way analysis of variance revealed significant differences in the breadth of communication among different video types (p<0.001). Therefore, video creators must select appropriate types for creating smoking cessation-themed videos35.

Furthermore, the study results indicated that the type of video creator particularly impacts the overall communication effect and communication recognition. Specifically, smoking cessation-themed videos created by corporate users tend to have better overall communication effects and higher recognition. Although corporate users accounted for only 6.2% of the participation in creating smoking cessation-themed videos, their substantial resources in the workforce and finances allowed for well-produced videos, resulting in better overall communication effects and video recognition. Additionally, videos created by healthcare professionals also achieved sound overall communication effects and recognition. It is likely because healthcare professionals often possess high authority in health communication on social platforms, and the public trusts their statements36.

Limitations

Despite the significant findings of this study, several limitations should be considered. Firstly, this study only focused on smoking cessation videos in 2022, which may not fully represent the temporal dynamics. Secondly, the communication effects of smoking cessation videos on Bilibili only reflect the attitudes and perceptions of Chinese youth towards smoking cessation. In contrast, the communication effects of smoking cessation videos on other social media platforms warrant further exploration. Thirdly, future research should examine to what extent video effectiveness studies reflect the effectiveness of smoking cessation campaigns.

CONCLUSIONS

This study demonstrates that social video platforms can significantly promote smoking cessation among youth, serving as essential avenues for relevant organizations and professionals to reach out to young smokers. In terms of specific video production, video creators can consider the preferences of young audiences and incorporate high MSV elements, such as visual effects, background music, and evidence, into their videos. Additionally, message appeal factors such as threat appeals, humor appeals, and economic benefits are beneficial for achieving better results in promoting smoking cessation among youth.