1. Introduction

This study focuses on the high penetration of short videos among college students and the frequent occurrence of overspending. Platforms such as ShakeYin use algorithmic recommendation and immersive experience as the core social carrier, and tools such as 'chanting’ promote the prevalence of overconsumption. However, there is a lack of systematic research on the mechanism by which the intensity of short-video use influences overconsumption behaviour, especially the interaction between psychological traits and consumption motivation. This study is of great significance in deepening the understanding of the association between media use and consumption behaviour, and constructing a rational consumption guidance mechanism for college students.

This study focuses on three dimensions: the four dimensions of short-video dependence, such as withdrawal and avoidance, are quantified based on the SVAS scale; a self-developed scale is used to depict the overconsumption behaviours and motives; and six types of consumption motives, such as price sensitivity, are classified by CMMS to explore their influence on overconsumption.

The study adopted the questionnaire survey method as the core research method and obtained 124 valid samples. After the SPSS reliability test (total scale Cronbach α=0.929, KMO=0.761), the hypotheses were verified by descriptive statistics and correlation analysis, and the theoretical interpretation was carried out by combining with the literature method. The method supports variable correlation analysis with quantitative data to ensure empirical rigour.

The core objectives of this study are the following: 1) to reveal the causal association between short video dependence (loss of control, withdrawal, etc.) and overconsumption and the mechanism of psychological traits; 2) to identify the types of motives that significantly predict overconsumption, such as promotional addiction, and analyse their interactive effects in consumption decision-making; 3) to validate the theoretical model of 'short video dependence affects consumption decision-making through the mediation of psychological traits’, and analyse the impact of short video dependence on consumption decision-making through the mediation of psychological traits. decision-making’ and provide a replicable framework for related research and practical intervention.

2. Literature review

2.1. Impact of short video use on consumer psychology and behaviour

Existing studies have shown that the influence of short video usage intensity on consumption psychology and behavior is mainly reflected in psychological mediation effect, symbolic consumption drive, and technological convenience induction. For example, relevant studies have pointed out that the intensity of short video usage affects sleep quality through the chain mediating effect of loneliness and misplaced anxiety, and loneliness is significantly related to consumption impulses [1]; Internet usage (including short videos) affects credit consumption behavior through the mediating effect of social comparison and materialism, with online shopping having the most significant effect [2].

The cultural environment created by short videos is easy to promote symbolic consumption and comparison psychology. The Netflix culture and show-off of wealth in short videos can easily trigger college students to follow the trend of consumption, excessively pursue symbolic values such as brand names, and form the tendency of over-consumption [3]; short video use exacerbates depression in female college students through the chain mediation of self-objectification and body satisfaction, which reflects the deeper impact of visualized media on self-perception [4].

2.2. Causes and influences of overconsumption

The causes of college students' overspending involve multiple dimensions such as individual psychology, social needs and external environment. Relevant studies have shown that college students who lack social support (e.g., insufficient family financial education) are more likely to fall into credit card debt, and impulsive consumption tendency is positively correlated with debt [5]; low family income, lack of financial education, and peer pressure are the key social factors leading to college students' overconsumption, and the risk is higher for females and ethnic minority students [6]; the convenience of virtual payment in the Internet era and the live streaming of goods and services have intensified impulsive consumption, and some college students use online loans to satisfy their vanity needs due to the psychology of comparison [7]. The convenience of virtual payment and live bandwagon marketing in the Internet era exacerbate impulsive consumption, and some college students satisfy their vanity needs through online loans due to the psychology of comparison [7].

In terms of psychological mechanisms, school fatigue and social phobia significantly and positively predicted short-video addiction, which in turn led to decreased well-being and deterioration of parent-child relationship [8]; excessive use of social networks affects credit consumption behaviors through the mediation of money anxiety and money-power concepts, which highlights the association between psychological factors and consumption decision-making [9].

2.3. Research methodology and theoretical modelling

Existing literature generally adopts quantitative research methods, such as questionnaire survey combined with structural equation modeling (SEM). For example, through questionnaire data, structural equation modeling was applied to verify the multiple mediation path of “Internet use → social comparison → materialism → credit consumption” [2]. In addition, there is also related literature that confirms the chain-mediated effects of loneliness and misplaced anxiety between short video use and sleep quality through Bootstrap mediation test [1].

As for theoretical models, social comparison theory and objectification theory are widely used. For example, based on objectification theory, it reveals that short video use reduces body satisfaction through self-objectification, which in turn induces depression [4]; from the perspective of social media influence, it is proposed that college students' consumerism is subject to the dual roles of group identity and information cocoon [10]. However, most existing studies focus on a single theory and lack in-depth analysis of the technical characteristics of short video algorithmic recommendation and user-generated content [3].

3. Methodology

This study mainly adopts the questionnaire survey method, which means that the researcher designs relevant questions according to the purpose and object of the study, and then distributes and fills in the questions through the target group so as to obtain the first-hand data and information, and it is a method that is widely used in the social surveys at home and abroad at present.

This study was designed based on the Short Video Addiction Scale (SVAS), which was developed according to the Likert five-point scale method to reflect the intensity of college students' short video use from multiple dimensions. The short video use intensity module quantitatively assessed the degree of college students' dependence on short videos by examining the four dimensions of withdrawal, avoidance, loss of control, and inefficiency. The section on overconsumption behaviours focuses on the frequency of using credit tools, instalment purchases and consumption behaviours beyond affordability, which visually presents the current situation and extent of overconsumption and provides a data basis for subsequent correlation analysis [11]. The section of motivation for overspending aims to understand the internal driving force of college students' overspending, and the topics involve fashion trends, social needs, quality of life improvement and coping with emergencie [12]. Drawing on the PS consumption purchase motivation analysis method of the CMMS platform, this study also examines the six consumption traits of price sensitivity, promotion influence, quality pursuit, brand following, word-of-mouth influence, and advertisement drive, to analyse the tendency of college students to engage in different consumption purchase motivations, in order to more comprehensively and deeply explore their relationship with the intensity of short-video use and overconsumption.

4. Results and analyses of the questionnaire

4.1. Questionnaire reliability tests

In this study, SPSS27 was used to test the reliability of the questionnaire. The internal consistency coefficient (Cronbach alpha coefficient), which is a generally accepted index, was used for the reliability test. When the alpha coefficient is higher than 0.8, it indicates high reliability; when the alpha coefficient is between 0.7 and 0.8, it indicates good reliability; when the alpha coefficient is between 0.6 and 0.7, it indicates acceptable reliability; and when the alpha coefficient is less than 0.6, it indicates poor reliability. less than 0.6 indicates poor reliability.

As shown in Table 1, in the reliability test of the scale questions of this questionnaire, the coefficients of all dimensions are higher than 0.6 and the alpha coefficient of the total scale is 0.929, which indicates that the scale used in this study is of high quality of reliability and the obtained results are reliable and can be used.

The validity test was performed using KMO and Bartlett's spherical test, if the KMO value is >0.6 or more and the Bartlett's test level is <0.05, it means that the questionnaire has good significance and good construct validity.

As shown in Table 2, according to the results of the upper exploratory factor analysis, it can be seen that the coefficient of the KMO test is 0.761, which is greater than 0.7, and the significance of the test of sphericity is infinitely close to 0, which indicates that this questionnaire is suitable for factor analysis and has good validity. Through the reliability and validity test, it can be seen that this questionnaire can be used for empirical analysis.

4.2. Descriptive analysis

4.2.1. Basic information

In this study, 124 valid questionnaires were collected, of which 61 were from female students and 63 were from male students, with a balanced ratio of male to female. There was an even distribution of grades, with majors dominated by arts, history and medicine (each accounting for about 25%), economics and management majors accounting for the third largest share of 20%, arts accounting for 20%, and science and technology with other majors accounting for a total of 12%.

The main source of living expenses for the survey respondents is family support (76.6%), followed by part-time jobs (19.4%). Living expenses were mainly concentrated in the range of RMB 1,000-2,000 (about 73.39%), with a distribution of "high in the middle and low on both sides".

The daily duration of social media use of survey respondents is mainly concentrated in 1-5 hours (about 52.56%), with 12.96% exceeding 5 hours. The duration of a single swipe of short videos is mainly 15-30 minutes (about 23.10%), but a high proportion of those greater than or equal to 2 hours (about 21.05%), showing a polarized distribution. The maximum duration of a single swipe of short videos is dominated by the 15-30 minute range, accounting for about 23.10%, but the number of people in the greater-than-equal-to-2-hour range also accounts for a higher proportion, accounting for about 21.05%.

4.2.2. Shart video dependency module

The findings indicated that the college student population showed significant dependence on short video use. Specifically, when unable to use short videos, they generally feel anxious, irritable and uneasy (withdrawal, mean 3.94), and their attention is also affected, as well as a sense of boredom and emptiness. Short videos have become an important escape tool for them to cope with negative emotions and stress, and they may even neglect face-to-face communication as a result (avoidance, mean 3.73). In addition, college students generally have time management problems when using short videos, often exceeding the expected duration, repeatedly trying to reduce the use but hardly succeeding (uncontrollability, mean 3.81), and even disrupting plans and delaying affairs. Short video use also had a negative impact on their life efficiency, resulting in insufficient time, wasting time, and generally perceived reduced life efficiency and even fatigue (inefficiency, mean 3.67). Overall, college students showed high dependence, emotional avoidance, loss of control of time and inefficiency in short video use.

4.2.3. Analysis of college students’ concepts and motives for overspending

The survey shows that college students have relatively high acceptance of overspending, are generally willing to take on some debt to meet short-term needs (mean 3.98), and believe that under the premise of having a clear repayment plan, overspending is a reasonable financial strategy (mean 3.95), and moreover, it is a necessary way of adapting to the modern society (mean 4.16), and it can improve the quality of life to a certain extent ( The mean value is 4.02.) Nevertheless, there is still disagreement among students as to whether or not overspending is sensible.

In terms of overspending motives, advertisements and promotional activities have a strong guiding effect on college students' consumption behavior (mean 4.11). To fulfill basic living needs (mean 4.05) and study needs (mean 4.04) are important overspending drivers. In addition, when faced with emergency expenditures, college students show the highest tendency to overspend (mean 4.24). Peer influence in social environments (mean 3.96) and the pursuit of electronics or luxury goods (mean 4) also drove some college students to choose overspending.

Analysis of college students' Consumption and Purchase Motives

When shopping, college students exhibit the complex characteristics of price sensitivity and the pursuit of quality. They pursue cost-effectiveness, are easily attracted by promotional information such as "low price" and "save money", and may even buy non-essential products due to promotions. At the same time, they also pay attention to the quality of goods, are willing to pay a higher price for high-quality goods, pay attention to material performance and environmental certification, and pay attention to the evaluation of goods. Famous brands can bring a sense of reliability and face, brands with good reputation are favored, and limited edition products are also attractive. Recommendations from friends and netizens, reviews on social media platforms and trends in the social circle have a significant impact on their purchasing decisions, and they are more likely to trust real user reviews. Creative advertisements, visual effects, music, limited time offers and celebrity/web celebrity endorsements also inspire them to buy. All in all, college students' shopping is easily driven by promotions and word-of-mouth.

4.3. Correlation analysis

4.3.1. Hypothesis formulation

Before data analysis, this study proposes the following hypotheses:

Hypothesis 1: The longer the longest duration of a single short video session for college students, the higher the likelihood of Premature consumption motivation and Premature consumption behavior.

Hypothesis 2: The longer the Average daily time spent on social media by college students, the higher the likelihood of Premature consumption motivation and Premature consumption behavior.

Hypothesis 3: The higher the Withdrawal symptoms and Loss of control over short videos among college students, the higher the likelihood of Premature consumption motivation and Premature consumption behavior.

Hypothesis 4: Among college students, Price - sensitive consumers have a higher likelihood of Premature consumption motivation and Premature consumption behavior.

Hypothesis 5: Among college students, Promotion - addicted consumers have a higher likelihood of Premature consumption motivation and Premature consumption behavior.

4.3.2. Normality test

Since the sample size of the questionnaire is greater than 50, this study employs the Kolmogorov - Smirnov (K - S) test.

After the Kolmogorov - Smirnov (K - S) test, the significance levels of both the Kolmogorov - Smirnov and Shapiro - Wilk tests for all variables are less than 0.05. This indicates that the overall sample of the questionnaire shows a non - normal distribution. Accordingly, the Kendall correlation coefficient is used for the correlation analysis in the following text.

4.3.3. Correlation analysis between the degree of short video dependence and premature consumption behavior

As shown in Table 3, Withdrawal symptoms have a significant positive correlation with both Premature consumption behavior (τ = 0.197, p = 0.003) and Premature consumption motivation (τ = 0.258, p < 0.001). This implies that consumers with higher levels of Withdrawal symptoms may be more easily attracted to and engage in premature consumption.

Escapism also has a significant positive correlation with Premature consumption behavior (τ = 0.240, p < 0.001) and Premature consumption motivation (τ = 0.236, p < 0.001), indicating that consumers with a tendency towards Escapism are more likely to relieve stress or escape from reality through premature consumption.

Inefficiency has significant positive correlations with Withdrawal symptoms (τ = 0.273, p < 0.001), Escapism (τ = 0.247, p < 0.001), and Loss of control (τ = 0.652, p < 0.001). This shows that consumers with Inefficiency are more likely to get caught in a cycle of unhealthy consumer psychological characteristics.

**. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed).

4.3.4. Correlation analysis between types of consumption purchase motivations and premature consumption behavior

**. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed).

As shown in Table 4, Quality-pursuing consumers have significant positive correlations with Premature consumption behavior (τ=0.146, p=0.026), Premature consumption motivation (τ=0.103, p=0.114), Price-sensitive consumers (τ=0.496, p<0.001), and Promotion-addicted consumers (τ=0.447, p<0.001). This suggests that consumers who prioritize quality may be more active in premature consumption while also paying higher attention to prices and promotions.

Word-of-mouth influencers exhibit significant positive correlations with multiple variables, including Premature consumption behavior (τ=0.088, p=0.184), Premature consumption motivation (τ=0.067, p=0.308), Price-sensitive consumers (τ=0.218, p<0.001), Promotion-addicted consumers (τ=0.237, p<0.001), Quality-pursuing consumers (τ=0.334, p<0.001), and Brand-following consumers (τ=0.393, p<0.001). This highlights the critical role of word-of-mouth in consumers' decision-making processes.

Advertising-driven consumers show significant positive correlations with Premature consumption behavior (τ=0.080, p=0.223), Premature consumption motivation (τ=0.095, p=0.147), Price-sensitive consumers (τ=0.267, p<0.001), Promotion-addicted consumers (τ=0.221, p<0.001), Quality-pursuing consumers (τ=0.358, p<0.001), Brand-following consumers (τ=0.434, p<0.001), and Word-of-mouth influencers (τ=0.422, p<0.001). This indicates that advertising has a broad impact on shaping consumers' behaviors and attitudes.

5. Discussion

This study empirically explores and analyzes the correlation between the intensity of short video use and college students' Premature consumption, A total of 124 college students, including 50.8% males and 49.2% females, were involved in the study. Based on five hypotheses, the Premature consumption behavior of college students and its related influencing factors were deeply explored through data analysis. The results of the study are as follows:

First, the length of short video use and social media use correlate weakly with Premature consumption behavior and motivation and do not reach statistical significance levels. This may imply that the diversity and complexity of short video content makes its influence on consumption behavior not direct or strong, and it may also be because college students who use social media for a long time may be more exposed to consumption information and advertisements, but they may also thus enhance their rational cognition and self-control of consumption behaviors, which may inhibit the impulse to overspend. Therefore, hypotheses one and hypotheses two were not significantly supported by the data in this study.

Second, psychological traits and premature consumption behavior and motivation. The results show that withdrawal symptoms and premature consumption behavior and motivation are both significantly positively correlated, indicating that the higher the withdrawal symptoms of short videos, the stronger the premature consumption behavior and motivation of college students. The results show that there is a significant positive correlation between withdrawal symptoms and premature consumption behavior and motivation, indicating that the higher the withdrawal symptoms are, the stronger the premature consumption behavior and motivation are. While Loss of control has a significant positive correlation with premature consumption motivation, the correlation with premature consumption behavior is weaker and does not reach the significant level. This may indicate that withdrawal symptoms and Loss of control of short videos have more direct and significant effects on premature consumption motivation, while the effects on actual behavior may be moderated by other factors. Overall, hypothesis three is supported by the data to some extent, especially for the effect of withdrawal symptoms.

Third, consumption preferences and premature consumption behavior and motivation. For Promotion-addicted consumers, there was a significant positive correlation between them and premature consumption behavior, indicating that Promotion-addicted college students are more likely to exhibit premature consumption behavior. however, the correlation with premature consumption motivation did not reach a significant level, which may be due to the fact that promotion -addicted consumers are more likely to be driven by immediate promotional offers rather than intrinsic premature consumption motivation. Therefore, hypothesis four was significantly supported in terms of premature consumption behavior, but the evidence was insufficient in terms of premature consumption motivation. For brand-following consumers, there is a significant positive correlation between them and premature consumption motivation, i.e. brand-following college students are more likely to have stronger premature consumption motivation. however, the correlation with premature consumption behavior does not reach a significant level, which may be because although Brand-following consumers have stronger motivation, they may be less motivated in their actual consumption behavior. However, the correlation with premature consumption behavior does not reach a significant level, which may be because brand-following consumers, although having stronger motivation, may be affected by a combination of brand value, personal financial status and other factors in their actual consumption behavior, and thus may not necessarily exhibit higher premature consumption behavior. Therefore, hypothesis 5 is significantly supported in terms of premature consumption motivation, while there is insufficient evidence in terms of premature consumption behavior.

6. Conclusion

This study aimed to understand the new trend of college students' consumption behavior in the digital media era, and examined in depth their characteristics in terms of short video use, Premature consumption concepts and behaviors, and consumption purchase motives. The findings reveal that college students' high dependence on short videos is not only reflected in Withdrawal symptoms, but also in the utilization of it for Escapism avoidance and uncontrollable use leading to lower efficiency. Meanwhile, college students are relatively open to Premature consumption and are driven by multiple factors such as advertising promotions, basic needs and social environment. Although the direct correlation between the total duration of short video/social media use and Premature consumption is not significant, it suggests that specific content types, individual psychological traits, and the context of use may play a more critical role.

Based on the above findings, future research should focus on exploring the following key directions: first, to deeply analyze how different types of consumption preferences influence college students' Premature consumption behavior through psychological and behavioral pathways; second, given that the association between total duration of use is not significant, future research should focus on specific content types of social media and short videos to explore the role of fragmented information and deep social interactions in stimulating college students' Premature consumption. information and deep social interactions in stimulating consumption desire and influencing consumption decisions, in order to more accurately assess the drivers of digital media content. Third, we further examine the dynamic changes of the dimensions of short video dependence in different consumption contexts and their impact on consumption behavior, in order to gain a deeper understanding of the complex interactions between individual psychological states and consumption behavior.

Through in-depth exploration of these research directions, future research is expected to reveal the complex relationship between digital media use and college students' consumption behavior in a more comprehensive and refined way, providing educators, parents, platform operators and policy makers with a more solid theoretical foundation and practical guidance, in order to help college students establish a rational concept of consumption in the era of digital media, cultivate healthy consumption habits, and effectively cope with consumption risks.

Authors contribution

All the authors contributed equally and their names were listed in alphabetical order.