The blind box concept first emerged in Japan. It describes a sealed, non-transparent small container with identical external packaging that houses figurines in various designs. In the Chinese market, blind boxes primarily feature character dolls like Molly and Pucky, as well as collaborative editions with animated films and franchises, such as Toy Story and Disney. In recent years, companies like POP MART have driven the trend by frequently releasing diverse and intricately designed products. The widespread presence of blind box vending machines in major shopping centers, coupled with the appeal of artistic aesthetics and the thrill of unpredictability, has made these products highly popular among young consumers. Research confirms that blind boxes have become a dominant trend in the fashion consumption preferences of China's younger generation.

Industry studies indicate that China's blind box market is projected to reach a value of 25 billion yuan by 2025 (Zeng, 2021). Additionally, spending by young consumers surged by 73% between 2016 and 2020, surpassing expenditures by those aged over 35 by the end of that period, establishing them as the primary consumer demographic. Compared to previous generations, post-90s consumers benefit from significantly better economic conditions, leading to a stronger preference for unique and exciting shopping experiences that emphasize novelty and individuality (Wang and Zhou, 2021).

Previous research in consumer psychology has highlighted that individuals differ significantly in their inherent tendencies toward novelty seeking and fashion orientation. Novelty seeking refers to a stable personality trait that reflects a consumer's intrinsic desire to pursue new, unique, and stimulating experiences (Zuckerman, 1994). Individuals high in novelty seeking tend to prefer unfamiliar products, enjoy exploratory behaviors, and are more inclined to take risks in consumption. Similarly, a strong orientation toward fashion and trend-following creates a cognitive preference for products that symbolize uniqueness, freshness, and social expressiveness (Workman & Studak, 2006). These traits together often lead consumers to make rapid, emotional decisions when encountering novel consumption stimuli.

Existing studies have shown that novelty-seeking traits are closely related to impulsive buying behavior. Novelty-seeking consumers tend to respond more strongly to external stimuli and have difficulty exercising self-control in environments characterized by newness and excitement (Rook & Fisher, 1995). When exposed to products that embody uncertainty, collectability, and limited availability, these individuals experience heightened emotional arousal, which increases the likelihood of impulsive purchases. Moreover, fashion-oriented consumers, driven by the desire for social identity and symbolic expression, are more likely to exhibit frequent purchasing behaviors to maintain alignment with current trends (Tigert et al., 1976). Thus, high novelty and fashion sensitivity can directly translate into repetitive and spontaneous buying patterns in emerging consumption contexts.

In the context of blind box consumption, such personality traits play an amplified role. Blind boxes integrate uncertainty, novelty, surprise, and collection incentives—features that are particularly appealing to novelty-seeking consumers. Recent marketing research indicates that when consumers with strong novelty-seeking tendencies encounter uncertain outcomes, the element of surprise intensifies emotional engagement and stimulates reward-seeking motivations (Hill et al., 2021). Additionally, the social and cultural symbolism embedded in blind box collections, such as exclusivity and trendiness, further strengthens the purchasing drive among fashion-oriented individuals. These psychological tendencies can ultimately transform into high-frequency purchasing behaviors, as consumers repeatedly buy blind boxes to obtain novel designs, complete collections, or keep pace with evolving trends.

Although prior studies have acknowledged the relationship between novelty-seeking traits and impulsive buying, few have examined how this trait specifically shapes consumer behavior within the blind box market. Existing research on blind boxes remains relatively limited and mainly focuses on consumption motivations or emotional responses (Yan and Wu, 2021; Wang et al., 2022). There is still a lack of systematic empirical evidence explaining how personality-driven novelty pursuit influences impulsive purchase intentions under conditions of reward uncertainty. Therefore, taking blind boxes as the research context, this study further explores the internal mechanism through which novelty-seeking and fashion-oriented traits trigger impulsive tendencies, and how these tendencies ultimately lead to frequent purchasing behaviors. This research not only enriches theoretical understanding of personality-based consumption processes but also provides meaningful insights for enterprises seeking to design marketing strategies that align with consumers' psychological needs in the blind box industry.

1. Theoretical background
   1.  Impulsive Buying Tendency

Impulsive buying is defined as a sudden, compelling, and unplanned decision to purchase a product, driven largely by emotional arousal rather than deliberate cognitive processing (Rook, 1987). Individuals with high impulsivity tend to act quickly under emotional stimulation and are more likely to make purchases triggered by excitement, curiosity, or perceived rewards. Prior research indicates that both novelty seeking and fashion orientation are significant predictors of impulsive buying because they amplify consumers' emotional responses in highly stimulating consumption environments. When consumers perceive a product as new, unique, or socially desirable, the resulting emotional impulse can override rational evaluation and directly lead to purchasing behavior.

1. 2.  Information gap theory

The information gap is the gap between what individuals already know and what they want to get. When people realize that they have an information gap, they will be curious, which will further motivate them to take action to fill the information gap (Loewenstein, 1994). When faced with a huge information gap, people will pay attention to the information they know, but as the gap gradually narrows, the focus will turn to the information gap, thereby strengthening individual curiosity.

Menon and Soman (2002) demonstrated that advertising tactics designed to arouse curiosity can enhance user engagement and learning motivation. When consumers encounter partial rather than complete promotional content, they are more inclined to actively seek additional details. Brod and Breitwieser (2019) suggested that curiosity arises when individuals attempt to predict whether their internal expectations will be fulfilled. As participants formulate these predictions, their personal information gap widens while anticipating the correct answer, thereby intensifying their curiosity. In an experiment by Singh and Manjaly (2021), participants were asked to fill in missing letters (the second, fourth, and seventh positions) in an incomplete word. Findings indicated that both the presence of an information gap (the omitted letters) and participants' uncertainty about the complete word heightened their curiosity to learn it. Once the missing information was provided, participants experienced satisfaction.

In summary, the information gap theory examines how individuals engage with uncertain rewards. When exposed to incomplete rather than fully disclosed information, people tend to have a more engaging experience, which fosters more positive attitudes or stronger behavioral inclinations (Ruan et al., 2018). Drawing on this theory, the current study investigates how perceived uncertainty influences curiosity fulfillment and, subsequently, impulsive buying intentions.

1. 3.  Stimulus-organism-response theory

The Stimulus-Organism-Response (SOR) framework, introduced by Mehrabian and Russell (1974), describes how external stimuli affect an individual's cognitive, emotional, and behavioral reactions. Over time, this model has been widely applied in psychological and consumer behavior research.

In the context of blind boxes, each series typically includes 12 standard items and 1 hidden item, with the specific product remaining unknown until the box is opened. This element of surprise is treated as an external stimulus in this study. Prior SOR-based research has often measured external stimuli through perceptual factors like information quality and interactivity (Thomas & Mathew, 2018; Zhu et al., 2020). Accordingly, this study employs perceived uncertainty as a key stimulus variable.

The "O" (Organism) component of the SOR model typically represents cognitive and emotional states, such as perceived trust or enjoyment (Huang et al., 2020). In this research, consumer curiosity serves as the organism factor. Meanwhile, the "R" (Response) component reflects behavioral outcomes shaped by cognitive and emotional processes. As purchase intention has frequently been used as a response variable in prior studies, this paper adopts impulsive purchase intention as the behavioral outcome. Thus, the study examines the connections between perceived uncertainty, curiosity, and impulsive buying behavior within the SOR framework.

2. Research design
   1.  Questionnaire Design

To strictly ensure validity and internal consistency, the questionnaire design utilized core constructs directly cited from mature and authoritative scales, in addition to basic demographic information. These included the Big Five Inventory-10 (BFI-10) and the Chinese Big Five Personality Inventory (CBF-PI) for personality measurement, the Consumer Styles Inventory (CSI) (Sproles & Kendall, 1986) for assessing consumption styles, and the Barratt Impulsiveness Scale (BIS) (Rock & Fisher, 1995) for measuring impulsivity.

1. 2.  Data Collection

This study adopted an online questionnaire survey method for primary data collection. The electronic questionnaires were generated primarily through online survey platforms such as "WJX" and distributed via social network channels including WeChat Moments and university student communities. The survey respondents were restricted to Chinese university students with independent consumption capabilities to ensure the sample accurately reflected the characteristics of the study's target audience.

1. 3.  Data Processiong

During the data collection period, a total of 32 original questionnaires were recovered. To ensure data quality, we conducted a rigorous manual review of all samples, eliminating 1 invalid sample that showed obvious irregularities, such as an excessively short response time or patterned answering. Ultimately, 31 high-quality valid questionnaires were included in the statistics, resulting in an effective response rate of 96.9%.

3. Empirical result
   1.  Demographic Profile of Respondents

Judging from the distribution of the chart data, the age distribution of the sample in this survey is highly concentrated in the 18 to 25 age range. The vast majority belongs to the 20-year-old group, which aligns perfectly with our pre-set target group of university student subjects. In terms of gender composition, the number of female samples is approximately twice that of males. This skew in the gender ratio may suggest a higher willingness among females to participate in specific consumption topics (such as blind boxes or novelty fashion items).

In the economic dimension, respondents with monthly living expenses in the 2000-2999 RMB and 3000-3999 RMB ranges accounted for more than half of the total. This distribution excludes extreme financial constraints associated with subsistence consumption while also filtering out the noise caused by excessive wealth, ensuring the sample has good homogeneity in the key variable of "disposable idle funds". Although the charts show that some males have higher living expenses, the overall economic strength curves between genders overlap significantly. This implies that if significant differences in consumption behavior appear in subsequent analyses, we have stronger grounds to attribute them to psychological traits rather than simple disparities in purchasing power.

Fig1. Age and Gender Distribution of Respondents

Fig2. Distribution of Respondents' Monthly Living Expenses

Fig3. Comparison of Monthly Living Expenses Between Genders

Fig4. Distribution of Monthly Living Expenses by Gender

1. 2.  Data Analysis

First, descriptive statistics and difference tests for each variable were conducted. Second, Independent Samples T-tests and One-way ANOVA were performed on each variable to determine differences caused by gender, purchase frequency and purchasing power, respectively. Then, the correlation between variables was analyzed to derive a correlation matrix. Finally, regression models between variables were constructed to identify potential predictive relationships.

1. 3.  Descriptive Statistics and Difference Tests of Variables

First, Table 1 presents the basic descriptive statistics for the research variables in this study.

Table 1. Descriptive Statistics of Research Variables

Next, we examined the differences in research scales regarding the demographic variable of gender. The results (as shown in Table 2) indicate that there are no significant differences in any of the research variables regarding gender.

Table 2. Independent Samples T-test for Gender

Then, we examined differences in research variables regarding monthly living expenses. The six tiers ranging from "<1000 RMB" to ">5000 RMB" were assigned values from 1 to 6 based on purchasing power and divided into two groups for inter-group difference testing: purchasing power levels 1-3 as the first group, and levels 4-6 as the second group.

Table 3. Independent Samples T-test for Purchasing Power

Note：*p<.05

The results (as shown in Table 3) indicate that there are no significant differences in Neuroticism, Openness, Decision Confusion Orientation, or Novelty & Fashion Cognitive Orientation regarding purchasing power. However, Impulsive Consumption (t=-1.817, p<.05) showed a significant difference based on purchasing power; individuals with higher purchasing power scored significantly higher on Impulsive Consumption than those with lower purchasing power.

Finally, we examined the differences in research variables regarding purchase frequency. The results (as shown in Table 4) indicate that Neuroticism, Novelty & Fashion Cognitive Orientation, and Impulsive Consumption all show significant differences based on purchase frequency. Further post-hoc tests revealed that individuals who purchase frequently scored significantly higher on Neuroticism and Impulsive Consumption than those who never purchase. Additionally, frequent purchasers scored significantly higher on Novelty & Fashion Cognitive Orientation than both occasional purchasers and those who never purchase.

Table 4. One-Way ANOVA for Purchase Frequency

Note：*p<.05

1. 4.  Correlation Between Research Variables

The correlation analysis results (as shown in Table 5) indicate significant correlations between Openness and Neuroticism, Neuroticism and Impulsive Consumption, and Novelty & Fashion Cognitive Orientation and Impulsive Consumption. Specifically, Neuroticism has a low positive correlation with Impulsive Consumption (r=.247, p<0.05); Openness has a low negative correlation with Neuroticism (r=-.383，p<.05); while Novelty & Fashion Cognitive Orientation has a relatively high positive correlation with Impulsive Consumption (r=.447，p<0.05).

Table 5. Correlation Matrix of Research Variables

Note：*p<.05

1. 5.  Investigation into Factors Influencing Impulsive Consumption Trait

To examine the influence of personality traits on impulsive consumption traits, this study used multiple linear regression analysis to explore the effects of Neuroticism, Openness, Decision Confusion Orientation, and Novelty & Fashion Cognitive Orientation (independent variables) on Impulsive Consumption (dependent variable). To ensure the interpretability of the results, we first tested the variables for multicollinearity.

Table 6. Multicollinearity Check

The results show that the Variance Inflation Factor (VIF) for all variables is less than 10, indicating no severe multicollinearity. Next, a linear regression model was constructed with the total score of Impulsive Consumption as the dependent variable and the total scores of Neuroticism, Openness, Decision Confusion Orientation, and Novelty & Fashion Cognitive Orientation as independent variables.

Variance analysis showed that the model's overall predictive power did not reach the standard for statistical significance (F(4,26)=2.222, p=.094). However, its explanatory power (R²=.255) indicates that the four variables can still jointly explain 25.5% of the variance in impulsive consumption, suggesting some reference value.

The multiple linear regression analysis found that Novelty & Fashion Cognitive Orientation has a significant positive predictive effect on Impulsive Consumption (β=0.385, p<0.05). The standardized coefficient indicates that after controlling for other variables, for every 1 standard deviation increase in Novelty & Fashion Cognitive Orientation, Impulsive Consumption increases by approximately 0.385 standard deviations. This result suggests that "the higher the consumer's cognitive level regarding fashion and novelty, the more likely they are to engage in impulsive buying behavior."

Table 7. Multiple Linear Regression Results

Note：*p<.05

However, the predictive effects of Neuroticism, Openness, and Decision Confusion Orientation were not significant.

Based on the above analysis, to ensure the statistical significance of the model and the interpretability of the results, a stepwise regression method was subsequently used to screen potential predictor variables.

Table 8. Multiple Linear Regression Results

Note：*p<.05

Table 9. Excluded Variables

Note：*p<.05

This model added independent variables stepwise, retaining only the statistically significant model. The results showed that the model retained only one variable: "Novelty & Fashion Cognitive Orientation." Variance analysis indicated that this simplified model has statistical significance (F(1,29)=7.226, p=0.012) and can stably explain 19.9% of the variance in impulsive consumption (R² =.199). Compared to the original model which contained four predictors but was not overall significant (p =.094), the simplified model sacrificed a small amount of explanatory power but achieved higher statistical robustness and parsimony.

The regression coefficient shows that Novelty & Fashion Cognitive Orientation is a strong positive predictor of Impulsive Consumption (β= 0.447, p=0.012). This result remains robust after controlling for other non-significant variables.

Neuroticism, Openness, and Decision Confusion Orientation were excluded during the stepwise regression process (p > 0.05). Their low partial correlation coefficients and high tolerance indicate that these variables have weak independent explanatory power for impulsive consumption. This suggests that broad personality traits (Neuroticism, Openness) or decision-making styles may influence impulsive consumption indirectly by affecting specific consumption cognitions (such as Novelty & Fashion Cognition), rather than through a simple direct linear relationship.

1. 6.  Investigation into Factors Influencing Impulsive Consumption Behavior

To further explore the factors influencing impulsive buying behavior, this study constructed a multiple linear regression model with Purchase Frequency as the dependent variable, and Impulsive Consumption traits, Gender, and Living Expenses as independent variables. The ANOVA results showed that the model overall has statistical significance (F =3.395, p<0.05), indicating that the three independent variables included in this model, as a combination, can effectively predict changes in impulsive buying behavior (purchasing blind boxes).

Table 10. Multiple Linear Regression Results

Note：*p<.05