The Siege Effect: Exploring the Characteristics of Youth Consumption Behavior and Its Influencing Factors in the Chinese Short Drama Market ()
1. Introduction
Short drama taking advantage of the rapid development of mobile Internet, big data, cloud computing and other information technology came into being, and with the advantages of new media communication, with short and compact, compact rhythm and other characteristics, personalized video recommendations for the audience, quickly attracted a large number of viewers, in 2023, China’s micro-short drama market reached 37.39 billion yuan in 2023, and in June 2024, the scale of China’s micro-short drama users has reached 576 million people. The youth demographic accounts for over 70% of the audience, establishing them as the core consumer base for short dramas. Researching their consumption behavior provides crucial insights for the industry to precisely identify user needs and optimize content offerings. Since 2022, Chinese micro-short dramas have begun to build up momentum in overseas markets and gradually become a new driving force for cultural overseas communication in the context of digital media. In 2023, ReelShrot (a micro-short drama platform for the international market) developed by a Chinese company, had a download volume of 21,120,300 times, with a total revenue of up to $28,770,000. At the end of February 2024, a total of more than 40 types of short drama apps entered overseas markets, with a cumulative download volume of nearly 55 million times. However, the rapid development of the short drama market has also brought some inevitable problems. Due to the imperfect regulatory system of the emerging market, it is difficult to resolve the problems of homogenization and low quality of the content of short dramas, coupled with the frequent problem of easy addiction, the research and standardization of the short drama market is imperative.
In this regard, in December 2020, the State Administration of Radio, Film and Television (SARFT) issued the Circular on Issues Related to the Audit of the Contents of Micro-Short Dramas in Online Film and TV Dramas, which for the first time defined online micro-short dramas, marking that online micro-short dramas were formally included in the scope of industry regulation. On January 12, 2024, the State Administration of Radio and Television (SARFT) issued the “Traveling with Micro-Short Dramas” creation plan, proposing that in 2024, the State Administration of Radio and Television (SARFT) will create and broadcast 100 excellent micro-dramas, and promote the formation of a new consumption model of “micro-drama + cultural tourism”. From the three dimensions of regulation, guidance and support, and ecological construction, the SARFT has made great efforts to promote the high-quality development of micro-short dramas and escort the healthy growth of the industry.
However, existing international studies predominantly adopt Western cultural frameworks, paying insufficient attention to Chinese youth cohorts. This research focusing on Chinese young adults thus serves dual purposes: addressing localized practical concerns while contributing differentiated perspectives to global digital addiction governance.
In this context, researching the consumption behavior of youth groups can not only provide reference for creators to produce high-quality short dramas in line with market demand and cultural value, but also take educational measures to cultivate correct values and aesthetics to promote the healthy growth of youth groups after understanding the impact of short dramas on youth groups. In addition, the production of micro-skit involves many aspects and requires diversified talents, and young people are a strong reserve army of talents, so it is important for this study to investigate the young people aged 18 - 22 to seek a win-win path for the skit industry and young students.
2. Literature Review
1) A study of the characteristics associated with the short drama market
The academic research on the characteristics of the short drama market focuses on three topics: profitability model, market performance and its negative impact.
In terms of the short drama market profit model, Dong & Wang (2025) pointed out the commercial profit path of short dramas in their study of online short dramas. Generally speaking, there are several short drama profit models in general: user payment, copyright distribution, advertising and marketing, and e-commerce bandwagon. Among them, advertising and marketing and paid recharge are its most important profit sources. The user-paid model allows viewers to directly pay for high-quality content through membership and single-episode payment, etc. The advertisement cooperation converts viewers’ attention into advertisement revenue through sharing with the platform and advertisement implantation. These diversified profit models together constitute the business ecology of the short drama industry. Later, Xu (2024) pointed out in his study on the quality improvement measures of online micro-drama that the micro-drama market should give full play to its advantage of “a small boat with a good head start” and develop diversified profit models. For example, by opening the corresponding reward service, to obtain additional income; the use of AICG technology to reduce the cost of setting, transitions, etc.; the same type of goods appearing in the micro short drama to create an exclusive window, give full play to the long tail effect and other measures. The future micro short drama market should create a multi-channel commercial operation form, develop a diversified profit model of network micro short drama, and improve the micro short drama industry chain.
In the performance dimension of the short drama market, Liu (2024) analyzed the economic benefits brought by the explosion of micro short dramas from a commercial perspective. For example, “After the flash marriage, Mr. Fu’s vest can’t be hidden” one-day recharge water flow broke 20 million yuan, “I was a stepmother in the 1980s” recharge on the day of the launch exceeded 20 million yuan, and “Wushuang” recharged more than 100 million yuan on the eight days of the launch… These astonishing figures not only highlight the high recognition of consumers to the micro-short dramas, but also highlight the fact that this new entertainment form development vigor. The measurement of Yue (2024) found that as of June 2024, the scale of China’s micro short drama users has reached 576 million people, accounting for 52.4% of the overall Internet users. Meanwhile, data from AiMedia Consulting shows that the market size of China’s micro short drama market reached 37.39 billion yuan in 2023, with a year-on-year growth of up to 267.65%, and it is expected that the market size of China’s micro short dramas will be able to reach 50 billion yuan in 2024, and in 2027 it will be expected to reach 100 billion yuan. Another survey shows that in the first half of 2023, China’s online platforms launched more than 480 short micro-dramas, equivalent to an average of 2.7 a day. This not only reveals the huge economic potential of the short drama market, but also maps out the deep love of contemporary young people for short dramas, and the new trend and vigor of the short drama market.
In terms of the negative impacts of short dramas and countermeasures, along with the rapid development of the Internet, digital addiction is particularly prominent among college students as the “digital natives”, and the potential problems of digital domination and media consumption follow the development of the short drama industry, and the negative impacts of addiction to short dramas have become more and more intense, which has triggered widespread concern among scholars in different scientific fields. scholars’ general concern. For example, Kumar et al. (2021) pointed out that with the advent of the digital age, online micro-sketch dramas have become the newest source of entertainment for the millennial generation, but its development has both advantages and disadvantages, as online skit dramas combine education and entertainment, and at the same time, it creates a new obstacle, “Internet addiction”. The so-called internet addiction not only keeps the youth engrossed in virtual skit episodes, but also has a significant impact on the physical and mental health of the youth. Elhai et al. (2025) investigated the mechanisms linking anxiety and fear of missing out (FoMO) to internet addiction, revealing significant correlations between these negative emotions and the severity of addictive behaviors. In parallel, Bhukya & Lakshmana (2025) conducted a bibliometric analysis examining internet usage patterns, cyber-addiction, and mental health in adult populations. The study established strong associations between internet addiction and deteriorating mental/physical health outcomes, ultimately advocating for targeted interventions and preventive measures to address this growing public health concern. Dhanuka et al. (2019) conducted a cross-sectional survey of 90 youth (18 - 25 years) from Mumbai and showed that excessive obsession with watching web series can lead to disruption in their personal lives and various health problems. It can have a negative impact on the emotional, physical, mental and social health of young people. The problem of addiction is an important negative effect of Internet micro-series. Similarly, Li et al. (2024) determined a significant negative correlation between social support and college students’ digital addiction through meta-analysis, in which the social support measurement tool and education level moderated the effect more strongly, i.e., based on certain social influences, specialists encountered unfair social prejudice, were prone to academic burnout, faced great employment pressure, and therefore had a higher likelihood of digital addiction, and the researchers suggested that family and family members should have a greater role in the development of digital addiction. In this regard, researchers have proposed targeted and enhanced interventions for digitally addicted college students that need to be taken by families and schools. After analyzing the nature of the addiction economy, Gao (2020) takes the short video industry as an example and analyzes the business model of digital addiction, i.e., the spurt development of the mobile Internet has provided a hotbed for it, and through algorithmic capture and free mode, consumers are dominated by data, which gradually reduces the sensitivity of budgetary constraints and jeopardizes the mental health of consumers.
In summary, for the difficulties of short drama governance, Cui et al. (2020) mentioned in the interview that it is not enough to rely on platform autonomy and governmental governance, and it is necessary to rely on the governance model of multiple subjects to achieve the combination of internal and external, and on the other hand, it is possible to use AI to create an efficient recognition mechanism to identify and manage vulgar, bloody and violent content, and identify sensitive phenomena such as imitation, remake and plagiarism, in order to protect the quality of original products and their intellectual property rights. Intellectual Property Rights. In their latest research report, Yu and Liu (2024) examined the media platform economy in the context of Internet technology, pointing out that the logic of flow under the algorithmic structure has invaded the public nature, authenticity, and creativity of production to a certain extent, and that in the platform economy, the “active user theory” has replaced the “passive audience theory”. In the platform economy, the “active user theory” replaces the “passive audience theory” as the new starting point for research, and “personalized consumption” has become the new mother topic of media consumption research. In this context, the human-machine cooperative industrial pattern and intelligent industrial regulation have become the new conception of intelligent media.
Therefore, Sun (2022) analyzes in detail the platform addictive technologies and damages from the perspective of fusion of jurisprudence and psychology, i.e., some digital platforms stimulate individual dopamine secretion by means of extreme personalized push, Zeigernike cycle, interface manipulation, etc., which in turn make users form addictions, which cause many harms to the economy and society such as information cocoon, violent desensitization, and so on. After comparing different rule of law programs to overcome digital addiction, he puts forward regulatory proposals to focus on the regulation of antitrust law, regulate the exploitative abuse of attention, and review the stifling of hybrid mergers and acquisitions, which provides a new perspective of disciplinary convergence for the governance of digital addiction, and reveals that we can explore from multiple perspectives and disciplines. At the governance level, only through a multifaceted and shared governance model can the governance of the short-form drama industry be prompted to achieve a certain degree of success. As Li (2025) points out, under the synergistic promotion of technological innovation, government support and multi-party cooperation, the chaos in the micro-short drama industry has been effectively curbed, and it is steadily moving towards a new stage of boutique development, shifting from purely chasing “high flow” to pursuing “high quality”. The new stage is shifting from the mere pursuit of “high traffic” to the pursuit of “high quality”.
2) Research on Media Consumption and Media Governance
Media consumption is an important theoretical basis for the current study of the short drama market. In terms of the expansion of this theory, a large number of research results have emerged in the academic community. Representative results such as Zhang (2024) argues that media consumerism, i.e., the embodiment of consumerist culture in the field of mass media. With the development of Internet technology, online short drama, as an emerging Internet technology product, has rapidly won the favor of consumers with its fast-paced and crisp narrative features, and has become a new media consumption form. The dissemination and profitability of short dramas rely to a certain extent on users’ media choices, and derive a series of new issues based on the economics of digital behavior. The competition among major platforms in the short drama market is essentially to increase user stickiness and interfere with users’ media choices by improving their own attractiveness. For instance, Amadi et al. (2024) examined public consumption of drag-themed short dramas on Facebook, revealing that entertainment value and social acceptance serve as primary determinants of audience engagement. Crucially, these two factors - entertainment and social validation - represent fundamental objectives in media selection, which collectively enhance user retention for drag performance content. Yu, Yang, & Qu (2020) cut from the perspective of media menu and broadly generalized media menu into seven types: tool-type, community-type, game-type, pan-entertainment-type, information-type, short-video-type, and affordable-type, and used principal component factor analysis to conclude that media menu preference is significantly affected by personal factors such as gender, age, marital status, family income, education level, and structural urban-rural distribution and media usage habits. This helps us to understand the homogeneity of media menus and the behavioral patterns of selective media use in the Internet era.
As for how to promote the media consumption of online short dramas and promote the development of the economy, we can draw on the ideas put forward by Hui & Nai (2024) in his study on how mainstream media can attract young overseas audiences: namely, deepening precise communication, strengthening the training of relevant talents, and innovating the market tools for short dramas. By doing so, audience attention can be effectively attracted, wider dissemination of short dramas can be realized, and the development of short drama economy can be promoted.
As of 2023, approximately 69% of U.S. adults and 81% of adolescents were active social media users. Aitken et al. (2024) investigated the impact of social media consumption on college students’ affect and executive functioning, revealing significant negative effects on cognitive control processes while underscoring the urgent need for interventions promoting healthier digital habits. In a complementary study, Keum (2025) conducted path analysis with 338 emerging adults to examine relationships between critical social media literacy, social support systems, flourishing, and mental health outcomes. The findings position critical social media literacy as a pivotal tool for developing effective support networks and fostering prosperous lifestyles among emerging adults. Collectively, these studies demonstrate that cultivating healthy digital practices and critical media literacy serves as an effective instrument for media governance.
In terms of media governance, the negative impact of short dramas is a direct manifestation of the exploitation of Internet users by digital capital. In the era of digital capitalism, the behavioral data generated by Internet users browsing web pages and social media are appropriated by digital platforms without compensation and sold to advertisers for profits, making users the “digital laborers” of capital. Based on the perspective of Marx’s critique of capital, Hou (2024) elaborates that digital capital realizes full-time, full-area and full-member domination over human beings through the privatization of data, the virtualization of life, and the invisibility of manipulation, and proposes a path to crack the problem, i.e., on the one hand, we should use the humanistic logic to regulate the “barbaric side” of digital capital to realize the return of human subjectivity. On the other hand, data sharing should be used to activate the “civilized side” of digital capital, so as to promote capital to better serve human development and let all people share the fruits of digital development. In addition, Zhang et al. (2023) believe that the scientific, technological, resource and capital nature of data reflects the fact that in the era of big data, the power derived from the possession of data puts the possessor in an advantageous position in social relations, and the concentration and monopolization of data even gives the possessor a dominant power. Digital capital carries out “circular monopolization” of traffic, data, technology and market through “looting” to produce data platforms, consolidates the development of digital power through complex and hidden labor exploitation, and uses finance to penetrate in order to realize the disorderly expansion of data power. expansion. Similarly, Zhang & Guan (2024), in her study of anti-digital reflection and critique of media in the digital age, points out that there is an urgent need to set up a mechanism for nonviolent communication in the current digital age, and that the digital media should be both on-line and capped. Digital media are only tools, it is always people who dominate digital communication, and products are ultimately only consumed by digitized people, not the digitization of people, so that digital serves people, not people serving digital.
A survey of the existing research literature found that the current relevant research mainly focuses on the explosive emergence of network micro short dramas, market performance and its contribution to the economy as an emerging product, etc., and there are not too many studies on the addiction phenomenon of network short dramas, nor is there any focus on the network short dramas as a form of new media consumption, and the study of the behavioral effects of this media consumption on consumers. It can be seen that the existing research on online short dramas is slightly fragmented.
Our study demonstrates distinctive contributions and innovations compared to existing research. Theoretically, it adopts a more innovative perspective. While Dhanuka et al. (2019) primarily focused on short-video addiction among youth without delving into its impact on college students’ academic and daily lives, our research not only examines addiction phenomena but also explores potential psychological and behavioral consequences, offering a more comprehensive viewpoint. Similarly, although Kumar et al. (2021) investigated short-videos’ influence on millennials, their work mainly centered on entertainment functions. Our study further reveals impacts on values and aesthetic perceptions by integrating characteristics of short-videos with college students’ media usage patterns.
Methodologically, this study employs a mixed-methods approach combining questionnaires, in-depth interviews, and multi-model analysis, demonstrating clear advantages over studies using singular methods. Many existing studies rely solely on either questionnaires or interviews, resulting in limitations in depth or breadth. For instance, Park et al. (2015) questionnaire-based analysis of user motivations and satisfaction failed to thoroughly examine addiction phenomena and their complex behavioral impacts. Our research systematically collects broad data through surveys, probes individual psychological states and behavioral motivations via in-depth interviews, and finally conducts quantitative analysis using logistic regression and grey relational models - achieving more comprehensive, accurate, and scientific results.
Substantively, this study not only provides holistic analysis of college students’ short-video consumption but also proposes practical pathways for translating academic findings into real-world applications. Moreover, our data sources present unique advantages compared to studies relying on secondary data or literature reviews. Although Yu et al. (2020) analyzed the short-video industry’s status and challenges through literature review, their conclusions’ reliability was somewhat compromised by lacking primary data support. In contrast, our first-hand survey and interview data directly reflect actual consumption behaviors, establishing a solid empirical foundation for the research.
3. Data Collection and Descriptive Analysis
1) Survey Implementation and Analysis Logic
1. Main content of the research
In this paper, according to the research purpose, the questionnaire is collected from four levels: user’s personal information, basic situation of watching, self-assessment of the impact of short plays and willingness to pay, and the specific items are shown in Table 1.
Table 1. Main survey items.
Personal Information |
Sex, grade, major |
Viewing |
Exposure mode, viewing platform, viewing frequency, single viewing duration, viewing period |
Impact assessment |
Purpose of viewing, type of preference, preference factors, self-assessment of the impact of the skit, reasons for the development of the skit, recommendations |
willingness to pay |
Consumption |
Whether or not you paid to watch the skit |
willingness to pay |
Willingness to pay to watch quality sketches |
price range |
The perceived willingness price range for a short play, given interest in a short play |
2. Research process
A small pre-survey was conducted to test the feasibility and validity of the questionnaire before officially conducting the large-scale survey. The pre-survey was conducted by means of an online questionnaire, and the target respondents were 30 undergraduate students in universities. The content of the survey covers the frequency, duration, platform selection, content preference, and spending of the undergraduate group on watching short dramas, as well as the manifestations and causes of addiction to short dramas and its impact on study and life.
Through the pre-survey, we analyzed the collected data for reliability using Alpha coefficient. Based on the feedback from the pre-survey, we made some modifications to the questionnaire. In the pre-survey, some students did not watch the skit and had difficulty in answering some questions, which affected the reliability of the data. Therefore, we optimized the skip logic of the questionnaire, for those students who do not watch the skit, they will directly skip from question 6 to the end when filling in the questionnaire to avoid invalid filling. Secondly, the questionnaire responses were in the form of choices, and some of the questions were not answered precisely enough, and at the same time, they were not conducive to quantitative analysis. Accordingly, we changed some of the questions to Likert scales to ensure the scientificity and accuracy of the data. In addition, some complex questions were simplified to ensure that the questionnaire is easy to read and easy to answer, and to improve the efficiency of questionnaire filling.
Based on the pre-survey, we optimized the design of the questionnaire. The questionnaire is divided into two parts: the first part is the basic information, including gender, age, grade, major, etc.; the second part is the short drama consumption behavior and addiction, covering the frequency, duration, platform selection, content preference, and spending of watching short dramas, as well as the manifestation of addiction to short dramas, its causes, and its impact on study and life. The questionnaire was based on a Likert scale to ensure the scientific and analyzable nature of the data.
Based on the preliminary survey results, we calculated the required sample size using the following parameters:
Confidence level (a) = 0.95
Absolute margin of error (d) = 5%
Estimated proportion (p) = 0.8 (derived from the pre-survey)
Design effect (Deff) = 1.35
Anticipated response rate (r) = 80%
(1)
(2)
(3)
The initial sample size calculation yielded 246, which was adjusted to 333 after accounting for the design effect. Further incorporating the expected response rate, the final target sample size was determined to be 417. To ensure data reliability, we ultimately collected 420 valid responses.
The formal survey distributed questionnaires through the Internet platform, covering college students in different regions of the country to ensure the diversity and representativeness of the sample. We actually distributed 420 questionnaires and finally collected 375 valid questionnaires, with an effective recovery rate of about 90%. At the same time as the formal survey, we also selected some representative samples for in-depth interviews. The purpose of the in-depth interviews is to obtain more detailed and effective information, and gain an in-depth understanding of the motives, habits, and addictions of college students’ short drama consumption behavior. Through the in-depth interviews, we gained a more comprehensive understanding of the characteristics and problems of the college student group in terms of short drama consumption.
3. Analytical framework
This study will conduct a two-part analysis focusing on user profiling and payment willingness, following rigorous reliability and validity testing of the questionnaire.
The analytical approach consists of three key components:
Descriptive statistical analysis of users’ viewing habits, encompassing fundamental factors including gender distribution, viewing duration, frequency, time slots, and content preferences;
Comprehensive examination of payment behaviors utilizing:
Chi-square tests to identify significant associations
Multinomial logistic regression models to determine predictive factors
Grey relational analysis to assess influence weights
Key independent variables include gender, viewing frequency, time slots, and viewing purposes as they relate to payment willingness;
Evidence-based conclusions and targeted recommendations derived from the analytical findings.
2) Reliability and Validity Analysis of the Questionnaire
1. Reliability analysis
Reliability refers to the degree of consistency in measurement results when the same instrument is repeatedly applied to measure the same variable. In the context of this study, it indicates whether the questionnaire can yield reliable and reproducible results when administered to the same subjects on multiple occasions. The higher the reliability, the more reasonable the design of the questionnaire and the higher the reliability of the questionnaire. The reliability analysis adopted this time is based on the Cronbach’s alpha coefficient , DeVellis (2017) believes that when the value is in the range of 0.60 - 0.65 it is the worst state, when the value is in the range of 0.65 - 0.70 it can barely reach the minimum, when the value is in the range of 0.70 - 0.80 it is a very desirable state, and when the value is in the range of 0.80 - 0.90 it is the most ideal state. Using spss software to analyze, the Cronbach’s alpha coefficient is 0.781, which is greater than 0.7, and the reliability of each dimension measured is greater than 0.7, therefore, the questionnaire is a questionnaire with high credibility and a reasonable design (Table 2).
Table 2. Reliability test table.
Reliability statistics |
Cronbach Alpha |
Normalized term-based clone Bach Alpha |
item count (of a consignment etc.) |
0.781 |
0.772 |
8 |
2. Validity analysis
Validity refers to the extent to which a measurement instrument accurately and precisely measures the intended theoretical constructed in this context, it evaluates whether our questionnaire effectively captures the true dimensions of short-video consumption behaviors we aim to investigate. The closer the measure is to the desired measurement variable, the higher the validity is; conversely, the lower the validity is. This questionnaire utilized SPSS KMO and Bartlett for validity analysis. According to Kaiser’s methodological framework, the interpretation of KMO statistics follows these guidelines: measurements below 0.5 must be discarded due to insufficient sampling adequacy; values between 0.6 - 0.7 are minimally acceptable though not ideal; the 0.7 - 0.8 range suggests moderate adequacy; 0.8 - 0.9 indicates good sampling appropriateness; and scores above 0.9 reflect superb suitability for factor analysis. The KMO value obtained from this survey is 0.808 with 28 degrees of freedom and the significance level of Bartlett’s test of sphericity is 0.000, which is smaller than 0.001, which indicates that the questionnaire data can be factor analyzed (Table 3).
Table 3. Validity test table.
KMO and Bartlett’s test |
KMO Bartlett’s Test of Sphericity |
0.808 |
|
approximate chi-square (math.) |
669.242 |
(number of) degrees of freedom (physics) |
28 |
significance |
0.000 |
3) User Profile
1. More balanced gender ratio
As shown in Figure 1, 57.33% of the survey respondents are female and 42.67% are male. There are slightly more women than men, and the gender ratio of consumers is more balanced.
2. Users’ stickiness is strong, Watching time is mostly concentrated in bedtime.
As shown in Figure 2, 44.27% of the college students studied seldom watched skits, 24.27% watched them 1 - 2 times a week, and 19.47% never watched them. The percentage of those who watched multiple times a day was 6.4%.
As shown in Figure 3, 42.76% watched short dramas for less than 30 minutes at a time, while 32.57% watched for 30 minutes to 1 hour. The percentages of those who watched 1 to 2 hours and more than 2 hours were 12.5% and 12.17% respectively.
As shown in Figure 4, the time period for the respondents to watch online short dramas is distributed throughout the day, with the peak viewing period before going to bed, accounting for 56.91%, followed by before and after meals, accounting for 31.91%. The percentages of recess and lunch break were 24.34% and 25% respectively.
Therefore, most of the college students have the habit of watching short online dramas in their leisure time, and the duration of watching is longer and the user stickiness is stronger.
Figure 1. Gender distribution of consumers.
Figure 2. Frequency of viewing.
Figure 3. Duration of viewing.
Figure 4. Viewing time period.
3. Users’ acceptance of all types of short dramas is high, mainly focus on the plot and actor’s face value.
As shown in Figure 5, campus youth short dramas are the most popular, accounting for 55.59%, followed by costume traversal and reverse inspirational categories, accounting for 43.09% and 40.46% respectively. Suspense thriller and sci-fi mystery category accounted for a lower percentage, 32.89% and 23.03% respectively. It can be seen that the college student group has a higher acceptance of different types of online short dramas.
As shown in Figure 6, plot attraction is the most important factor, accounting for 79.93%, followed by actor’s face value, accounting for 60.53%. The production and shooting quality and obtaining emotional value and emotional resonance accounted for 50.66% and 33.55% respectively. It can be seen that the college student group has a high demand for the overall quality of online short dramas, and the plot, actor face value, and filming quality are important influencing factors.
Figure 5. Preference types.
Figure 6. Preference factors.
4. Overall self-control is strong, with a small percentage showing addiction
The survey results show that 83.88% believe that the impact of short dramas is both positive and negative, 9.87% think it is very positive and 6.25% think it is very negative. This indicates that the group’s overall perception of online short dramas is good, and there is a large potential for consumption. Through users’ self-assessment of the impact of short dramas, it can be understood that the self-control of the researched group is generally strong, but there is a small number of the researched who are addicted to online dramas, which has caused non-negligible impacts on their daily study and life as well as their physical and mental health.
4. Willingness to Pay Analysis
1) Current Payment Status
As shown in Figure 7, 76.64% of the respondents did not pay for watching the skit, which is the majority, while 23.36% did.
Figure 7. Payments.
2) Gender and Willingness to Pay Chi-Square Test
As shown in Figure 8, 39.14% of people have a wait-and-see attitude towards high-quality short dramas, 20.39% are very reluctant to pay, and 18.09% are not willing to pay. The percentage of those who are more willing and completely willing to pay are 15.79% and 6.58% respectively. It can be seen that most people have the potential to pay for high-quality web sketches. Using SPSS to carry out chi-square test on the user’s gender and willingness to pay, as shown in Table 4 and Table 5, the minimum expected number is 7.70 greater than 5, and the p-value takes the asymptotic significance of 0.024, which is less than 0.05, indicating that there is a significant difference between men and women in their willingness to pay, and from the cross-tabulation table, it can be seen that women’s willingness to pay is greater than that of men in general.
Figure 8. Willingness to pay.
Table 4. Gender and willingness to pay chi-square test table.
|
(be) worth |
(number of) degrees of freedom (physics) |
Progressive significance (bilateral) |
Pearson’s chi-square (math.) |
11.200a |
4 |
0.024 |
likelihood ratio |
11.148 |
4 |
0.025 |
Number of active cases |
304 |
|
|
a. 0 cells (0.0%) have a desired count less than 5. The minimum desired count is 7.70.
Table 5. Cross-tabulation of gender and willingness to pay.
|
Are you willing to pay to watch high quality sketches? |
1 |
2 |
3 |
4 |
5 |
(grand) total |
distinguishing between the sexes |
male |
reckoning |
32 |
22 |
35 |
17 |
5 |
117 |
expectation count |
23.9 |
21.2 |
45.8 |
18.5 |
7.7 |
117.0 |
percentage |
27.4% |
18.8% |
29.9% |
14.5% |
9.4% |
100.0% |
daughter |
reckoning |
30 |
33 |
84 |
31 |
9 |
187 |
expectation count |
38.1 |
33.8 |
73.2 |
29.5 |
12.3 |
187.0 |
percentage |
16.0% |
17.6% |
44.9% |
16.6% |
4.8% |
100.0% |
total |
reckoning |
62 |
55 |
119 |
48 |
20 |
304 |
expectation count |
62.0 |
55.0 |
119.0 |
48.0 |
20.0 |
304.0 |
percentage |
20.4% |
18.1% |
39.1% |
15.8% |
6.6% |
100.0% |
3) Logistic Regression Analysis of Consumption Intention
In this part, we constructed an economic model between factors such as gender, grade, major, initial exposure to skits, frequency of viewing, length of each viewing, and college students’ online consumption, and collected primary data through online and offline questionnaires, and numbered the different influencing factors, obtaining Table 6. Based on this, we pre-processed the data, and used the formula (4) to normalization. Then, we constructed a logistic regression model and used matlab software to find the parameter values, accuracy, precision, recall, and F1 score.
(4)
Table 6. Table of impact factor numbers.
factor |
Classification of influencing factors |
serial number |
distinguishing between the sexes |
male |
0 |
women |
1 |
grade |
first-year university student |
1 |
second-year university student |
2 |
third-year university student |
3 |
Senior and above |
4 |
specialized field |
science and engineering as academic subjects |
1 |
liberal arts |
2 |
medical science |
3 |
Arts, Education |
4 |
(sth. or sb) else |
5 |
contact method |
Big Data Push |
1 |
Referral from friends |
2 |
Listening to classmates talk about post-search |
3 |
viewing frequency |
Multiple times per day |
1 |
1 - 2 times per day |
2 |
1 - 2 times per week |
3 |
Rarely watched |
4 |
never watched |
5 |
Duration of a single viewing |
Less than 30 minutes |
1 |
30 minutes to 1 hour |
2 |
1 to 2 hours |
3 |
More than 2 hours |
4 |
Figure 9. Logistic regression test plot.
Figure 9 shows the change of the loss value Loss with the number of iterations Iteration during the training process of a logistic regression model, and its specific value is shown in Table 6. From the figure, we can see that the loss value gradually decreases from the initial about 0.69 to about 0.45. This indicates that with the increase of the number of iterations, the loss of the model is decreasing, and the model is gradually learning and optimizing. The decreasing trend of the loss value is relatively smooth without drastic fluctuations, which indicates that the convergence rate of the model is relatively stable. The learning rate is set appropriately, neither too fast leading to oscillations nor too slow leading to slow convergence. After 1000 iterations, the loss value stabilized at about 0.45 or so. The lower this value, the smaller the gap between the model’s predictions and the true labels, and the better the model’s performance.
Figure 10. Normality tests.
As shown in Figure 10, the Anderson-Darling test (A2 = 1, p = 0.0005) rejects the null hypothesis of normally distributed residuals. The distribution exhibits negative skewness (γ1 = −0.88) and platykurtosis (γ2 = 2.13) relative to the normal distribution. These diagnostics, while informative, are not strictly relevant for logistic regression which makes no parametric assumptions about residual distributions.
Table 7. Logistic regression test table.
Iteration |
Loss |
100 |
0.6615 |
200 |
0.6315 |
300 |
0.6033 |
400 |
0.5769 |
500 |
0.5521 |
600 |
0.5289 |
700 |
0.5070 |
800 |
0.4865 |
900 |
0.4671 |
1000 |
0.4489 |
As shown in Table 7, the model accuracy (Accuracy) is 99.20%, Precision: is 100.00%, Recall is 100.00% and F1 Score is 100.00%. These metrics indicate that the model is performing at an extremely high level in all aspects. Accuracy is the percentage of samples that the model predicts correctly out of the total samples. 99.30% accuracy means that the model makes correct predictions in almost all cases. This is a very strong indicator that the model is very reliable overall. Accuracy is the proportion of samples predicted by the model to be in the positive category that are actually in the positive category. 100.00% accuracy means that the model did not incorrectly predict any samples in the negative category to be in the positive category, i.e., there were no false positives. This shows that the model is very accurate in predicting positive classes and does not introduce unnecessary false positives. Recall is the proportion of samples that are actually positive classes that are correctly predicted as positive classes by the model. A recall of 100.00% means that the model is able to identify all positive class samples and did not miss any positive class samples, i.e., there are no false negatives. This shows that the model is very effective in capturing positive class samples. The F1 score is the reconciled mean of precision and recall which tries to find a balance between precision and recall. The F1 score of 100.00% shows that the model performs very well in both precision and recall and achieves a perfect balance.
Table 8. Table of model test indicators.
Accuracy |
99.20% |
Precision |
100.00% |
Recall |
100.00% |
F1 Score |
100.00% |
Table 9. Table of regression intercept and weight values.
theta |
Final theta values |
Theta0 |
0.4516 |
Theta1 |
0.0771 |
Theta2 |
0.0323 |
Theta3 |
0.0113 |
Theta4 |
0.0098 |
Theta5 |
0.0107 |
Theta6 |
0.0106 |
Based on Table 8, we finally derived the formula for the sigmoid function of college students’ behavior of paying for web skits:
(5)
where z = 0.4516 + 0.0771 × Gender + 0.0323 × Grade + 0.0113 × Major + 0.0098 × Initial Exposure + 0.0107 × Viewing Frequency + 0.0106 × Viewing Duration
The results of the logistic regression analysis in Table 9 show that there is a significant influence of factors such as gender, grade, major, initial exposure to short dramas, frequency of viewing, and length of each viewing on whether college students pay for online short dramas. Specifically, female gender and senior grade had the most significant effect on paying behavior, with females and seniors more likely to pay for short dramas. This is generally related to women’s preference for entertainment content as well as the higher financial independence of senior students. In addition, students majoring in, for example, liberal arts and arts also show a higher willingness to pay, which is related to the higher acceptance of cultural and entertainment content among students in these majors.
In terms of viewing behavior, students who watched short dramas frequently and for a longer duration each time were more likely to pay for short dramas. This indicates that there is a positive relationship between active viewing behavior and willingness to pay. In addition, the way of initial exposure to short dramas also has a certain impact on paying behavior, and students who were exposed to short dramas through friends’ recommendation or search were more willing to pay, which reflects the role of social recommendation and active search in influencing paying behavior.
In summary, gender, grade, major, viewing behavior and initial exposure to skits are the main factors influencing whether college students pay for skits. These findings provide an important reference for understanding the online consumption behavior of college students and can provide a basis for related platforms to develop precise marketing strategies.
4) Gray Correlation Analysis of Influential Factors on Consumption Behavior
We constructed a gray correlation model to analyze the relationship between college students’ consumption behavior and the purpose of watching short dramas, the frequency of watching dramas, the duration of watching dramas, and the time period of watching dramas. In the process of model construction, we first pre-processed the data, applying the following formula to standardize and homogenize the collected data.
Data preprocessing is one of the essential steps in gray correlation analysis, which aims to eliminate the differences in the scale and order of magnitude between different indicators, so as to make the data comparable. We convert the raw data into dimensionless relative data through methods such as initialization or homogenization to ensure that the results of the analysis of each indicator will not be affected by differences in units or orders of magnitude in the calculation process. Given the different types of data involved in this analysis, we also normalized the data by converting all indicators to very large ones. This step not only improves the comparability of the data, but also lays the foundation for subsequent correlation calculations.
(6)
(7)
(8)
After normalization and homogenization, we obtain a data matrix that can be used for grey correlation analysis, and we get the reference series
, the
Compare the series
.
where the correlation coefficient between
and the kth number of
, and the correlation between
and
, we use Equations (9) and (10) to calculate.
(9)
(10)
Using Matlab software, we obtain the correlation of feature 1 Purpose of Watching Short Drama: 0.7258; feature 2 Frequency of Watching Drama: 0.7759; feature 3 Duration of Watching Drama: 0.7752; and feature 4 Time Period of Watching Drama: 0.7010. In the gray correlation model in Table 10, the closer the correlation is to 1, it means that the correlation of the feature with the consumption behavior is stronger. The closer to 1 in the gray correlation model, the stronger the correlation between the feature and consumption behavior.
Table 10. Gray correlation analysis table.
Feature 1 looks at the relevance of the purpose of the skit |
0.7258 |
Feature 2 looks at the relevance of the frequency of sketches |
0.7759 |
Feature 3 looks at the relevance of the length of the skit |
0.7752 |
Feature 4 looks at the relevance of the time period of the short play |
0.7010 |
We analyzed the relationship between the purpose of watching short dramas, the frequency of watching dramas, the length of watching dramas, the time period of watching dramas and consumption behavior, and obtained the following conclusions. The correlation between watching frequency and consumption behavior has the highest correlation, 0.7759, which indicates that the higher the frequency of watching episodes, the more likely users are to generate consumption behaviors, this is because high frequency of watching behavior reflects users’ high interest in episodes, which leads to greater probability of generating consumption activities such as purchasing peripherals of episodes and subscribing to memberships. The correlation of the correlation of the length of time spent watching dramas is also high, at 0.7752, which indicates that the longer the time a user spends watching dramas each time, the greater the likelihood that he or she will engage in consumption behaviors. Longer viewing times may mean that users are deeply invested in the content of the episodes and therefore have a higher acceptance of the associated consumption content. The correlation of the purpose of watching short dramas is 0.7258, which is relatively low but still at a high level, suggesting that there is some correlation between users’ preference for short dramas and their consumption behavior, but the degree of correlation is not as high as that of the frequency of watching dramas and the length of time spent watching them. This may be due to the relatively limited consumption scenarios of short dramas, or users’ lower willingness to consume short dramas. The correlation degree of time period of watching dramas is 0.7010, which has the lowest degree of correlation, but still has a certain degree of correlation, which may mean that users’ tendency to consume behaviors will be affected to some extent when they watch dramas in specific time periods, such as in the evening. All in all, from the correlation, the frequency of watching dramas and the duration of watching dramas are the two most important factors influencing consumption behavior, while the purpose of watching short dramas and the time period of watching dramas have relatively less influence.
Therefore, in our study of college students’ online short drama consumption behavior, we can focus on the frequency and duration of watching dramas, and increase the probability of consumption behavior by increasing the duration and frequency of watching episodes.
5. Conclusion and Recommendations
1) Conclusion of the study
1. The short drama market presents explosive growth and structural conflicts
In 2023, the market size of China’s micro short drama market exceeded 37.39 billion yuan, the user scale reached 576 million, and the youth group accounted for more than 70%, becoming the core audience. Platform competition is fierce, with platforms such as Jittery Voice and B Station dominating the market, and the “short drama+” model opening up a new path for traffic realization. At the employment level, more than 83,000 related enterprises have registered in 2024, covering the whole industry chain of script creation, filming and promotion, creating diversified jobs. However, the industry is facing deep-rooted problems such as homogenization of content, deviation of values, confusion of copyright and insufficient supervision, and is in urgent need of transformation from “quantitative change” to “qualitative change”.
2. College students’ consumption behavior is characterized by “entertainment and high activity”
Consumers of online short dramas are mainly students of science and engineering and liberal arts. There is a significant correlation between gender and the willingness to consume online short dramas, with women’s willingness to consume significantly higher than that of men. Bedtime (56.91%) and fragmented viewing (42.76% for single < 30 minutes) become typical scenes, and entertainment demand (86.51%) drives the popularity of campus youth (55.59%), ancient costume reversal, and short dramas, which shows that the youth group watches online short dramas out of the need for entertainment and relaxation and killing time, and for the factors of online short dramas, they place more importance on overall visual enjoyment, not limited by story topics. Meanwhile, multiple regression analysis shows that the willingness to pay for high-quality short films, gender and viewing frequency are the most important factors influencing users’ willingness to consume short films.
3. The two-way game of industry opportunities and risks
Gray correlation analysis shows that high-frequency viewing and long-time investment are the key triggers for paid conversion, revealing the potential conversion path from “viewing inertia” to “consumption behavior”, and we need to focus on guiding high-frequency and long-time users. The youth group recognizes both positive and negative impacts of short dramas, and 39.14% of them have a wait-and-see attitude towards paying for high-quality short dramas, which means that there is more potential space for this group to consume, but homogenization of content and rough production have become the problems that need to be solved for the development of short dramas nowadays.
4. High generalizability of research findings
With the advancement of science and technology and the diversification of entertainment methods, short-form dramas have emerged as a new form of entertainment and are gradually gaining popularity worldwide. This study primarily focuses on Chinese youth aged 18 - 22, and its findings are not only applicable to this specific age group in China but also provide valuable references for analyzing the dependence and consumption behaviors of youth in other countries regarding short-form dramas. Consequently, the results of this study demonstrate strong generalizability.
2) Policy recommendations
1. Strengthen technological innovation, optimize platform cooperation, and broaden profit models to promote the high-quality development of the short drama industry
While strengthening cooperation with high stickiness platforms such as Jittery and B Station, it is imperative for the short drama industry to create an independent platform to enhance user experience. It is necessary to promote technological innovation and content upgrading, improve production quality, and explore new forms such as interactive short dramas. The industry also needs to explore diversified profit modes such as IP derivative development and brand-customized short dramas to satisfy users’ low-cost viewing needs and optimize the revenue model.
In addition, the short drama should be deeply integrated with the field of education, develop content with educational significance, expand application scenarios, and enhance social value. Universities can utilize the form of short dramas to develop courses in ideology and politics and thematic cases, and enhance the attractiveness and infectiousness of learning. Through multi-party synergy, the short drama industry is expected to find a balance between technological innovation, content quality and social responsibility to achieve sustainable development. The industry should strengthen investment in technological research and development, utilize 5G and AI technology to improve production efficiency and quality, and promote short dramas towards boutique development.
2. Create a healthy environment for the consumption of short dramas through policy leadership, financial support and educational guidance
Reducing the design of “unlimited swiping” through algorithmic ethics, adding “anti-addiction tips”, and developing a “youth mode” to filter out undesirable content are effective mechanisms for optimizing the platform’s recommendation mechanism. The government should set up a short drama creation fund to promote cooperation between universities and producers to develop short drama works; the government needs to provide policy guidance and financial support; the media needs to publicize healthy consumption models, create a rational consumption atmosphere, and guide people to set up a correct view of consumption of short dramas.
Colleges and universities should provide relevant guidance and education to help young people in colleges and universities to form the correct values and aesthetics, and to promote the benign development of short dramas and young people in colleges and universities. Schools can introduce case studies of short dramas through curriculum design to improve students’ media critical power; they can also join hands with families to pay attention to their children’s emotional needs and reduce avoidance-type viewing motives; or they can carry out psychological counseling and provide one-on-one psychological counseling for students who have symptoms of addiction to help them rebuild their behavioral patterns.
3. Actively exploring the synergistic development path of benign interaction and win-win cooperation between the short drama industry and youth groups
The industry should explore social topics of common concern to youth groups, create works that are both entertaining and educational, incorporate positive values, satisfy users’ emotional resonance and social needs, and use AI technology to analyze user behavior data to provide customized intervention programs for over-watchers and guide reasonable viewing. Colleges and universities can offer short drama production courses, provide internship opportunities, transform students’ interests into professional abilities, and train professionals for the industry through industry-university-research cooperation. At the same time, the platform can introduce the “Achievement System”, which rewards users for controlling the length of their viewing time to ensure that their study and living time are not affected and to improve the efficiency of time utilization. In addition, by organizing short drama creation contests, campus screenings and other activities to enhance the sense of participation and creativity of college students, and realize the positive interaction and development of the short drama industry and youth groups, short drama can become part of campus culture.
4. Broadcasting administrative departments at all levels to earnestly fulfill regulatory responsibilities
Broadcasting administrative departments at all levels must deeply comprehend the decisive significance of establishing “Two Affirmations”, enhance “Four Consciousnesses”, strengthen “Four Confidences”, and achieve “Two Upholds”. They shall conscientiously implement the responsibility system for ideological work, further consolidate territorial management responsibilities, primary supervision responsibilities, and main entity responsibilities, strictly control the orientation, content, and broadcasting standards of short dramas, and truly fulfill their duties of territorial management. All television broadcasting institutions and online audiovisual program service providers shall rigorously implement pre-broadcast review systems. Relevant industry associations for television and online dramas should leverage their roles in industry services, management, and self-discipline to create a favorable development environment for promoting the healthy prosperity of short drama production.
5. Local governments to promote regional short drama development through localized approaches and facilitate nationwide exchange of outstanding micro short dramas
For instance, Beijing Municipal Broadcasting Bureau continues to advance the quality refinement and internationalization of micro short dramas through its “Light Chasing Program” International Micro Short Drama Competition. Additionally, Beijing will launch a “Greater Audiovisual” initiative to promote exchanges and dissemination of excellent micro short dramas.
Shanghai has announced plans to invest approximately 50 million yuan to roll out version 2.0 of its “Blossoms” micro short drama program, enhancing Shanghai’s influence in the national micro short drama industry.
Henan Province has proposed a “Micro Short Drama Plus” action plan, opening multiple integrated creation directions to develop a multi-layered micro short drama production model incorporating literature, law, technology and other fields.
Yunnan Province capitalizes on its unique ethnic cultural characteristics by introducing multiple “Micro Short Drama Plus” creative measures, aiming to better showcase Yunnan’s natural landscapes and cultural heritage through micro short dramas, thereby achieving cultural confidence and dissemination.
6. Limitations and Perspectives
1) Limitations
1. Smaller sample size
The respondents of this study were mainly concentrated in the youth group of 18 - 19 years old, accounting for a high percentage of 55.73%, while the percentage of youth groups of other age groups was relatively low, and youth groups other than those of 18 - 22 years old were not included in the scope of the survey. This uneven distribution of the sample may have weakened the generalizability and representativeness of the findings. Although youths aged 18 - 22 are one of the major groups of skit consumers, there may be significant differences in skit consumption behaviors and addiction tendencies among youths of different age groups. For example, the group of youths facing graduation from their senior year of college may have increased their consumption of short dramas due to reduced course pressure. In addition, the lack of sample capacity may also lead to insufficiently in-depth research on the short play consumption behaviors of certain marginalized groups, thus affecting the comprehensiveness of the study’s conclusions.
2. Lack of dynamism
The study is based on cross-sectional data, which is unable to capture the evolution of short drama consumption behavior over time. The rapid development of the short drama market and changes in the policy environment may bring important impacts on the consumption behavior of young groups. For example, the micro-short drama bidding system introduced by the SARFT in 2024 and the continuous optimization of platform algorithms may have significantly altered the viewing habits and willingness to pay of young people, and the inability of cross-sectional data to reflect these dynamic changes makes the ability to interpret the results in the time dimension limited. In addition, the content of short dramas is updated more quickly, and users’ preferences for different types of short dramas may change over time, and this dynamic is yet to be further verified and supplemented by subsequent tracking research.
2) Future Prospects
The booming development of the short drama market provides a multi-dimensional exploration space for academic research. In the future, we can introduce a variety of technical means to deepen the microscopic analysis of users’ attention mechanism and addictive behaviors: from the integration of technology and form innovation, such as AI-generated scripts, VR/AR immersive short dramas and interactive narrative mode of user experience research; social value excavation, to explore the potential of cross-boundary application of short dramas in the field of cultural heritage, public welfare dissemination, and mental health interventions, etc.
Localization strategy under the vision of globalization is to analyze the content adaptation mechanism and cross-cultural dissemination effectiveness of Chinese short dramas going overseas; data ethics and algorithm governance, for the accuracy of user profiles and privacy, constructing a dynamic regulatory framework to address the issues of user profile accuracy, privacy protection and algorithmic bias; research on sustainable business models, such as the development of the whole chain of IPs and synergistic effect of “short drama + e-commerce/cultural tourism”, etc., to provide a comprehensive analysis and research for the industry regulation and development. We will analyze and conduct research to provide empirical support for industry standardization and healthy user consumption.
Through the application of multidisciplinary intersection and cutting-edge methods, short drama research will help the industry break through the plight of homogenization and realize the organic unity of social and economic benefits.