How Do Personal Attributes Shape AI Dependency in Chinese Higher Education Context? Insights from Needs Frustration Perspective

Wenjun Zhong; Jianghua Luo; Ya Lyu

doi:10.1371/journal.pone.0313314

Abstract

Objective

The adoption of Generative AI in education presents both opportunities and challenges, particularly regarding its potential to foster student dependency. However, the psychological drivers of this dependency remain unclear. This study addresses this gap by applying the Interaction of Person-Affect-Cognition-Execution (I-PACE) model and Basic Psychological Needs (BPN) theory to explore how specific personality traits—neuroticism, self-critical perfectionism, and impulsivity—contribute to AI dependency through needs frustration, negative academic emotions, and reinforced performance beliefs.

Method

Data were collected from 958 university students (M_age = 21.67) across various disciplines. Structural equation modeling (SEM) was used to analyze the relationships among the variables.

Results

Neuroticism, self-critical perfectionism, and impulsivity were found to be significantly associated with increase needs frustration and negative academic emotions, which in turn reinforced students’ positive beliefs about performance of AI tools, deepening their dependency. The study also uncovered complex serial mediation effects, highlighting intricate psychological pathways that drive maladaptive AI use.

Conclusions

This research provides a critical insight into the interplay between personality traits and technology use, shedding light on the nuanced ways in which individual differences influence dependency on generative AI. The findings offer practical strategies for educators to promote balanced AI use and support student well-being in educational settings.

Citation: Zhong W, Luo J, Lyu Y (2024) How Do Personal Attributes Shape AI Dependency in Chinese Higher Education Context? Insights from Needs Frustration Perspective. PLoS ONE 19(11): e0313314. https://doi.org/10.1371/journal.pone.0313314

Editor: Orlando Manuel da Costa Gomes, Instituto Superior de Contabilidade e Administracao de Lisboa - Instituto Politecnico de Lisboa (ISCAL-IPL), PORTUGAL

Received: September 6, 2024; Accepted: October 23, 2024; Published: November 1, 2024

Copyright: © 2024 Zhong et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The dataset generated and analyzed during this study, along with the full-length questionnaire, are available on the Open Science Framework (OSF) platform and can be accessed via the following link: [https://osf.io/rys9q/?view_only=93ffef1955fc4ebdbe9a55093b77ca4c].

Funding: This work was supported by the Key Project of the National Social Science Fund of China [grant number ACA210010].

Competing interests: No authors have competing interests.

Introduction

The rapid integration of digital technologies into daily life has significantly influenced how individuals, particularly students, interact with information and educational tools. This transformation has garnered the attention of personality psychologists, who are increasingly focused on understanding the role of individual differences in shaping these interactions [1, 2]. A notable development in this digital landscape is the emergence of Generative AI tools like ChatGPT, which have revolutionized the way students and professionals engage with educational and professional contents [3].

While Generative AI tools offer advanced capabilities, such as personalized learning experiences, immediate feedback, and the potential to enhance productivity [4, 5]. Their rapid adoption has also raised concerns, particularly regarding over-reliance in educational contexts [6]. This dependency raised concerns about academic integrity problem, as compromising academic integrity can significantly impair the reputation of higher education institutions and diminish students’ intrinsic motivation [7]. Moreover, reliance on AI can impede the development of critical thinking skills, as students may become accustomed to seeking quick answers rather than engaging in deeper analytical processes [8, 9]. The conveniences offered by these tools can also foster procrastination, as students may defer tasks by relying on AI to provide last-minute assistance [10].

Despite the growing use of ChatGPT in educational settings, most research has focused on the technical attributes and educational benefits of these tools, while overlooking the psychological mechanisms that drive dependency [11]. This technology-centric approach neglects critical user-centered factors—particularly personal attributes—that could contribute to maladaptive technology use. Although recent studies have begun to explore contextual influences such as workload, time constraints, as well as stress level [10, 12], the role of individual psychological states and personality traits in fostering dependency remains underexplored. Addressing this gap is crucial for understanding the broader implications of generative AI in education and ensuring its responsible use [13].

Traits such as neuroticism, perfectionism, impulsivity, and low self-esteem have been consistently linked to maladaptive behaviors in digital contexts, including excessive social media use and internet addiction [14–16]. Individuals with these traits often experience psychological distress, which can intensify in high-pressure academic environments. To investigate how these personality traits influence the development of dependency on ChatGPT, this study applies the Interaction of Person-Affect-Cognition-Execution (I-PACE) model and Basic Psychological Needs (BPN) theory. Specifically, the study aims to answer the research question: “How does the interplay between personality traits and cognitive-emotional experiences shape the tendency to develop dependency on ChatGPT in educational environments?” By exploring the interaction between personality traits and cognitive-emotional experiences, this research aims to fill the existing gap in the literature on AI dependency in educational settings. The findings are expected to contribute to the theoretical discourse in personality psychology and provide practical insights for educators and technology developers. More broadly, our research offers new perspectives on the dynamics of human‒machine interactions in the era of Generative AI.

Theoretical framework and hypothesis formulation

I-PACE framework

The Interaction of Person-Affect-Cognition-Execution (I-PACE) model, proposed by Brand [17], offers a comprehensive framework for understanding the development of maladaptive behaviors in digital contexts. This model suggests that the interplay between personal characteristics (e.g., personality traits), affective responses, cognitive processes, and executive functions drives such behaviors. In this study, we focus on personality traits as key personal characteristics that influence how individuals experience and respond to their environment. Specifically, needs frustration, shaped by these personality traits, acts as a subjective experience that triggers affective and cognitive responses. Cognitive responses, such as performance expectations of ChatGPT, play a crucial role in this process. These expectations represent biases about the tool’s effectiveness in enhancing academic performance. When students develop high performance expectations, they may become overly reliant on ChatGPT, believing that it can help them achieve better results with minimal effort [18, 19]. This excessive reliance can lead to maladaptive patterns of technology use, consistent with the pathways described by the I-PACE model.

Needs frustration and ChatGPT dependency

According to Basic Psychological Needs (BPN) theory, the frustration of needs for autonomy, relatedness, and competence can lead to negative psychological outcomes, such as increased stress, reduced motivation, and reliance on external coping mechanisms [20]. Przybylski [21] found that individuals with frustrated psychological needs tend to turn to video games as a source of competence and relatedness, suggesting a similar pattern may apply to other forms of technology. Research by Kardefelt [22] indicated that individuals with higher levels of needs frustration are more likely to develop problematic behaviors with technology use since internet addiction can be considered compensatory to alleviating discomfort raised from state of frustration. In line with previous research findings, ChatGPT can serve as a coping mechanism for unmet psychological needs by providing immediate responses and simulated interactions, thereby fulfilling students’ needs for competence and autonomy [23, 24]. Thus, we hypothesize that:

H1: Needs frustration is positively associated with ChatGPT dependency.

Personality as an antecedent

Personality traits significantly influence how individuals perceive and respond to their environments, particularly in academic settings. Traits such as neuroticism, self-critical perfectionism, and impulsivity predispose individuals to heightened needs frustration and negative academic emotions, which can lead to increased dependency on ChatGPT.

Neuroticism is characterized by a tendency to experience negative emotions, such as anxiety, depression, and irritability [25]. Individuals with high neuroticism are more sensitive to stressors and often perceive situations as threatening, leading to frustration with unmet psychological needs, particularly autonomy, relatedness, and competence [26]. This sensitivity can exacerbate emotions like anxiety, fear, and guilt, further contributing to negative academic experiences [27].

Self-Critical Perfectionism involves setting excessively high personal standards and being overly critical of one’s performance [28]. This trait is associated with chronic self-doubt and fear of failure, resulting in a persistent sense of inadequacy and unmet needs for competence and autonomy [29]. Such individuals are prone to anxiety, depression, and guilt, particularly in academic contexts where they often perceive themselves as falling short of their high expectations [30].

Impulsivity is defined by a tendency to act without forethought, often leading to difficulties in self-control and decision-making [31]. Impulsive individuals may struggle to meet their basic psychological needs due to their inability to plan or sustain effortful activities, leading to frustration [32]. This trait is also linked to increased feelings of frustration, anger, and regret, especially when academic setbacks occur as a result of poor planning [33].

H2: Personality traits (neuroticism, self-critical perfectionism, and impulsivity) are positively associated with needs frustration.
H3: Personality traits (neuroticism, self-critical perfectionism, and impulsivity) are positively associated with negative academic emotions.

Negative academic emotions and performance expectations

Negative academic emotions, such as anxiety and hopelessness, are common in educational settings and have been shown to impair attention and academic performance [34, 35]. These emotions are often a direct consequence of needs frustration [36] and may drive students to seek external coping mechanisms like ChatGPT to manage their stress. Additionally, performance expectations refer to the extent to which individuals believe that using technology enhances their performance. It is recognized as a crucial cognitive factor influencing attitudes and behavioral intentions to use technology, as justified by the Unified Theory of Acceptance and Use of Technology (UTAUT) model [37]. The I-PACE model proposes that cognitive bias and illusion, the belief that technology can provide efficient and accurate answers, can bridge personal and psychological factors and problematic technological behavior. Therefore, we hypothesize that:

H4: Needs frustration positively influences negative academic emotions.
H5: Negative academic emotions are positively associated with ChatGPT dependency.
H6: Needs frustration is positively associated with performance expectations.
H7: Negative academic emotions are positively associated with performance expectations.
H8: Performance expectations are positively associated with ChatGPT dependency.

Serial mediation effects

To understand the comprehensive relationship between personality traits, needs frustration, negative academic emotions, performance expectations, and ChatGPT dependency, we propose a serial mediation model (Fig 1). This model hypothesizes that personality traits (neuroticism, self-critical perfectionism, and impulsivity) indirectly influence ChatGPT dependency through a sequence of mediators—needs frustration, negative academic emotions, and performance expectations. Specifically, we posit that the three proposed personality traits, including neuroticism, self-critical perfectionism, and impulsivity, predict higher levels of needs frustration. This subjective psychological state increases negative academic emotion and performance expectations, consequently leading to ChatGPT dependency.

Download:

Fig 1. Proposed research model.

https://doi.org/10.1371/journal.pone.0313314.g001

Materials and methods

Participants and procedure

Participants were recruited online through a reputable survey platform named “Sojump,” a prominent platform in China specializing in online questionnaires. We obtained ethical approval from IRB of Xinjiang Normal University (No. 1076220240523003). The recruitment started on 5^th June 2024 and ended on 12^th June 2024. Upon expressing interest in the study, the participants were provided with a written informed consent form outlining the study’s purpose, procedures, potential risks, and benefits. All participants provided written informed consent before participation. The participants were assured of the confidentiality and anonymity of their responses. Full-time undergraduate and postgraduate students with experience using ChatGPT for academic work were targeted. To ensure data quality, duplicate submissions (identified by IP address) were excluded. Of the 1,000 students invited, 958 provided valid responses. Participants ranged from 18 to 28 years old (M = 21.67, SD = 2.46). The sample was representative of the broader student population in terms of gender and academic background (Table 1).

Download:

Table 1. Demographic statistics of participants (N = 958).

https://doi.org/10.1371/journal.pone.0313314.t001

Measurements

In addition to demographic information, including gender, age, study discipline, degree level, and frequency of ChatGPT use, the questionnaire encompassed all constructs outlined in the proposed research model. The items were derived from instruments that demonstrated robust reliability and validity in prior studies. Minor modifications were made to some items to better fit our research aim, and these modifications were agreed upon by the three researchers involved in this study. To ensure measurement precision within the Chinese context, one researcher translated the original English instruments into Chinese, followed by back translation by another bilingual researcher.

Neuroticism was assessed using the 12-item subscale of the NEO Five-Factor Inventory (NEO-FFI) [38]. The neuroticism subscale comprises 12 items that evaluate participants’ proneness to psychological distress and emotional instability. Participants responded to each item on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The scale exhibits strong reliability (McDonald’s ω = .94).

Impulsivity was measured by the Barratt Impulsiveness Scale-Short Form (BIS-15) [39]. The scale consists of 15 items (McDonald’s ω = .967) evaluating three factors of impulsivity: Non-planning impulsivity (McDonald’s ω = .906), which measures tendency to act without planning; motor impulsivity (McDonald’s ω = .908), which reflects spontaneous and unpremeditated action; and attentional impulsivity (McDonald’s ω = .905), which measures the inability to focus attention and tendency to make careless decisions. The responses were recorded on a 4-point Likert scale ranging from 1 (rarely/never) to 4 (almost always) according to the original version of the scale. To simplify the structural model while retaining the integrity of the measurement, the scores for each of the three impulsivity subfactors were averaged and rescaled to a 5-point scale. This process created composite scores for each sub-factor, which were used as indicators in the structural model.

Self-critical perfectionism was assessed via the 6-item subscale of the Big Three Perfectionism Scale-Short Form (BTPS-SF) [40], evaluating individuals’ concern about mistakes and high self-imposed standards. All items are graded on a 5-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). In this study, the scale of self-critical perfectionism demonstrated acceptable reliability (McDonald’s ω = .876).

Needs frustration was assessed via the Basic Psychological Needs Satisfaction and Frustration Scale (BPNSFS) [41]. The scale measures the degree to which individuals feel that their psychological needs are unmet, aligning with the self-determination theory. The needs frustration scale comprises 12 items (McDonald’s ω = .942), measuring autonomy frustration (McDonald’s ω = .846), relatedness frustration (McDonald’s ω = .847), and competence frustration (McDonald’s ω = .838). Each factor was assessed using four items on a 5-point Likert scale ranging from 1 (not all true) to 5 (very true). Composite scores for autonomy, relatedness, and competence frustration were computed and used as indicators in the structural model.

Negative academic emotion was measured using the modified 12-item Achievement Emotions Questionnaire-Short Form (AEQ-S) [42], covering irritation, anxiety, and hopelessness. It demonstrated good reliability (McDonald’s ω = .94). All items were rated on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). We averaged the item scores based on the factor structure to form composite scores for irritation, anxiety, and hopelessness in the structural model.

The performance expectation of ChatGPT was evaluated with a modified 6-item scale based on perceived service quality from mobile providers [43]. The Revised scale specifically targets students’ perceptions of ChatGPT performance in supporting their academic activities (McDonald’s ω = 0.895). It consists of six items measured on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Sample statements include “The information provided by ChatGPT is accurate and reliable for my academic work,” and “ChatGPT provides me with prompt assistance for my study-related queries.”

ChatGPT dependency was measured with 6 items adapted from prior studies [12, 44]. Sample items included “I tried to reduce my use of ChatGPT to help with my assignment, but I failed,” and “I feel that I relied on using ChatGPT to help complete my academic task more and more.” Each item is measured on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The reliability index indicated acceptable internal consistency (McDonald’s ω = 0.863).

Statistical analysis

Data analysis was conducted using SPSS and AMOS to rigorously evaluate the hypothesized relationships within the theoretical framework. The analysis commenced with comprehensive descriptive statistics, including means, standard deviations, and zero-order correlations of all study variables. To assess the potential common method bias, Harman’s single-factor test was performed by conducting an exploratory factor analysis (EFA) on all scale items without rotation to determine the amount of variance explained by a single factor. An acceptable result is that no single factor accounts for more than 50% of the total variance [45]. Additionally, the unmeasured common latent method factor technique was integrated into our confirmatory factor analysis (CFA), allowing each item to load onto both its theoretical construct and common latent method factor, further mitigating method-induced variance [46].

Construct validity was supported by factor loadings above 0.70, Average Variance Extracted (AVE ≥ 0.50), and Composite Reliability (CR ≥ 0.70) [47]. Discriminant validity was confirmed using the Heterotrait-Monotrait ratio (HTMT < 0.85) [48]. The overall structural model fit was assessed using several fit indices with the following standards: Values for the Comparative Fit Index (CFI) of ≥ 0.95, Goodness of Fit Index (GFI) of ≥ 0.90, Root Mean Square Error of Approximation (RMSEA) of ≤ 0.06 and Standardized Root Mean Square Residual (SRMR) of ≤ 0.08 indicate a good fit [49]. Path coefficients were considered significant at p < 0.05. Mediation analysis was conducted to evaluate the indirect effects. Robustness tests, including bias-corrected bootstrapping with 5,000 iterations for indirect effects, were performed to ensure the stability and reliability of the findings [50].

Results

Measurement model

Descriptive statistics and common method bias.

S1 Table shows the Pearson correlation coefficients between all constructs. All correlations were significant (p < .01). Harman’s single-factor test indicated that common method bias was not a concern (variance = 23.902%). Additionally, the comparison with the CLF technique revealed that the addition of the common latent factor did not significantly improve the model fit (Δχ²/Δdf = 43.27/1, p > .05) confirming absence of common method bias.

Test of measurement validity.

Table 2 presents the results of the fit indices for all the constructs in the unidimensional item-level CFA model. These results confirm that the constructs, both first- and second-order, are well represented by their respective items. Confirmatory Factor Analysis supported the validity of the constructs, with all factors loading greater than .70 (S2 Table), all CR value were greater than .70, and AVE were greater than .50 (Table 3). Discriminant validity was confirmed with all HTMT ratio was less than .85 (Table 4).

Download:

Table 2. Fit indices for item-level CFA.

https://doi.org/10.1371/journal.pone.0313314.t002

Download:

Table 3. CR and AVE of structural constructs.

https://doi.org/10.1371/journal.pone.0313314.t003

Download:

Table 4. Heterotrait-Monotrait (HTMT) ratio of structural constructs.

https://doi.org/10.1371/journal.pone.0313314.t004

Structural model analysis

Test of model fit.

The structural model demonstrated an acceptable fit (χ²/df = 1.139, CFI = .996, GFI = .96, TLI = .96, RMSEA = .012[.007-.016], SRMR = .047). Despite the data not exhibiting multivariate normality, maximum likelihood estimation was used. Bias-corrected bootstrapping with 5,000 resamples addressed potential non-normality biases, showing no significant change in standard errors (S3 Table).

Direct path analysis.

SEM revealed significant relationships between personality traits and needs frustration, with neuroticism (β = .169), self-critical perfectionism (β = .224), and impulsivity (β = .158) (Table 5 and Fig 2). These traits also significantly influenced negative academic emotion, with neuroticism (β = .214), self-critical perfectionism (β = .113), and impulsivity (β = .083). Needs frustration directly affected negative academic emotion (β = .284), performance expectations (β = .331), and ChatGPT dependency (β = .136). Performance expectations were positively related to ChatGPT dependency (β = .242), while negative academic emotion did not have a direct effect. The model explained 14.6% of the variance in needs frustration, 22.7% in negative academic emotions, 32.5% in performance expectations, and 12.3% in ChatGPT dependency.

Download:

Fig 2. Structural model results.

https://doi.org/10.1371/journal.pone.0313314.g002

Download:

Table 5. Structural path analysis.

https://doi.org/10.1371/journal.pone.0313314.t005

Mediation and serial mediation analysis.

Table 6 presents the results of the indirect analyses of each personality trait. The indirect effects of neuroticism (β = .023, 95% CI [.009-.047]), self-critical perfectionism (β = .03, 95% CI [.012-.056]), and impulsivity (β = .021, 95% CI [.008-.043]) on ChatGPT dependency through needs frustration were significant. Serial mediation effects were also found significant through needs frustration and performance expectations for neuroticism (β = .013, 95% CI [.007-.024]), self-critical perfectionism (β = .018, 95% CI [.01-.03]), and impulsivity (β = .013, 95% CI [.002-.016]). Finally, the serial mediation involving needs frustration, negative academic emotion, and performance expectations significantly mediated the influence of neuroticism (β = .004, 95% CI [.002-.008]), self-critical perfectionism (β = .005, 95% CI [.003-.01]), and impulsivity (β = .004, 95% CI [.002-.008]) and ChatGPT dependency.

Download:

Table 6. Indirect path analysis.

https://doi.org/10.1371/journal.pone.0313314.t006

Exploratory analysis.

During the review process, we were requested to conducting an additional exploratory analysis to examine the influence of demographic variables on needs frustration and ChatGPT dependency, aiming to explore nuanced differences in participants’ demographic backgrounds. The analysis revealed that only gender and degree level had a statistically significant impact on needs frustration, while other demographic factors, such as academic discipline and frequency of ChatGPT usage, showed no significant effects on either needs frustration or ChatGPT dependency. Gender demonstrated a significant influence, with females reporting higher levels of needs frustration compared to males (β = 0.095, 95% CI [0.067, 0.290], p = 0.002). In terms of degree level, there was a negative relationship between academic progression and needs frustration (β = -0.098, 95% CI [-0.220, -0.050], p = 0.001), indicating that students at higher academic levels (e.g., Ph.D.) experienced lower needs frustration than those at lower levels (e.g., undergraduate). Other variables, including study discipline and ChatGPT use frequency, did not significantly predict needs frustration or ChatGPT dependency.

Discussion

This study demonstrates that neuroticism, self-critical perfectionism, and impulsivity are key personality traits contributing to dependency on ChatGPT among university students, primarily through their impact on needs frustration, cognitive biases, and performance expectations. These findings emphasize the importance of addressing both emotional and cognitive factors to mitigate AI dependency in educational contexts. By applying the Interaction of Person-Affect-Cognition-Execution (I-PACE) model and Basic Psychological Needs (BPN) theory, this study advances the understanding of how unmet psychological needs drive maladaptive dependencies, aligning with prior research linking needs frustration to technology overuse [51, 52].

Personality and AI dependency

Our findings show that neuroticism emerged as the strongest predictor of negative academic emotions, suggesting that individuals high in neuroticism are particularly vulnerable to academic stressors. The AI tool’s ability to provide instant feedback and solutions offers a temporary sense of control and competence, making it especially appealing to those with high neuroticism. This finding aligns with previous research linking neuroticism to digital dependency, where neurotic individuals use tools such as social media to manage emotional distress [53, 54]. However, our study extends this understanding by highlighting the unique academic context of ChatGPT use. Unlike social media, which primarily provides emotional gratification, ChatGPT directly impacts students’ perceived competence and control over academic tasks. This creates a specific cognitive bias where neurotic individuals may feel that their academic success depends on the tool, potentially leading to overreliance as a crutch to avoid confronting academic insecurities and emotional distress.

Self-critical perfectionism also plays a significant role, characterized by fear of underperforming and tendencies toward procrastination [55], Such individuals may find ChatGPT particularly appealing because it offers immediate feedback and reassurance, helping them manage procrastination and alleviate the fear of failure. However, this reliance can prevent the development of resilience and problem-solving skills, reinforcing feelings of inadequacy and fostering a cycle of dependency [56].

Impulsivity, though a less consistent predictor, was still significantly associated with needs frustration and negative emotions. Impulsive individuals, who often act without forethought, may turn to AI tools as a quick fix for academic challenges, aligning with previous findings that link impulsivity to maladaptive technology use [57]. This reliance on ChatGPT is further reinforced by avoidance behavior characteristics since they prefer easier routes to complete tasks and avoid the efforts associated with thorough planning and critical thinking. This avoidance behavior intensifies under high academic pressure, making ChatGPT an attractive option because of its ability to deliver straightforward answers quickly.

Cognitive factors as key drivers

A key finding is that negative academic emotions alone does not mediate the relationship between personality traits and ChatGPT dependency. Instead, the combined influence of negative emotions and performance expectations is crucial in fostering dependency. This aligns with cognitive-behavioral models, suggesting that negative self-appraisal, coupled with cognitive biases, can lead to maladaptive technology use [58]. Previous research by Casale [59] and Brand [17] has shown that positive metacognitions about technology use mediate the link between emotional dysregulation and technology reliance, supporting our findings. Thus, students experiencing frustration may hold negative self-appraisals and doubts about their self-efficacy, leading them to develop a strong belief in ChatGPT as a coping mechanism. Furthermore, the finding that degree level significantly predicts needs frustration, but not ChatGPT dependency directly, suggests that students in more advanced academic programs tend to develop greater cognitive resilience and problem-solving skills. This resilience may buffer them from frustration when facing academic challenges, thus reducing their likelihood of becoming overly dependent on tools like ChatGPT. Gender differences further contribute to this dynamic. Female students reported higher levels of needs frustration, indicating that they are more susceptible to emotional distress in academic settings. However, gender did not directly influence ChatGPT dependency. This suggests that while women may experience greater academic frustration, other cognitive factors—such as cognitive resilience and self-efficacy beliefs—likely play a more critical role in determining technology dependency.

Technologically, the perceived usefulness and ease of use of ChatGPT are well-documented antecedents of technology acceptance in educational settings [60]. The interactive nature of ChatGPT, providing quick answers with minimal input, makes it easily accessible to students. Although this study did not directly measure perceptions of ChatGPT, the high frequency of use reported (42.2% daily, 40.6% very often) suggests its perceived value. From a broader perspective, research shows that people often have unrealistic expectations of a technology’s ability to bring about positive change, akin to a placebo effect observed in digital interventions [61, 62]. In the context of ChatGPT, students may develop an illusion of efficacy, fostering dependency.

Implications

This study contributes to personality psychology by demonstrating that specific traits—neuroticism, self-critical perfectionism, and impulsivity—affect dependency on Generative AI tools like ChatGPT. By applying the I-PACE model, the study validates the model’s emphasis on the interaction between individual differences and affective-cognitive processes in understanding technology dependency. Additionally, the integration of BPN theory highlights how needs frustration mediates the relationship between personality traits and AI use, providing a nuanced understanding of the psychological mechanisms driving dependency. These findings suggest a refinement of current models to account for the specific role of cognitive expectations in fostering AI dependency.

Practically, these findings suggest the need for tailored interventions that address both the emotional and cognitive aspects of AI dependency. For individuals prone to neuroticism, self-critical perfectionism, and impulsivity, educators could implement strategies that focus on emotional regulation skills, such as mindfulness-based stress reduction, and cognitive restructuring techniques to challenge maladaptive thought patterns. Technology developers could design AI tools with built-in features that promote self-regulation and provide feedback on healthy usage patterns. Additionally, psycho-educational programs that help students understand the realistic capabilities and limitations of AI tools could reduce overreliance and dependency. Moreover, cognitive-behavioral strategies, such as cognitive restructuring, could also help challenge negative self-appraisals and enhance coping mechanisms. Interventions targeting performance expectations could help students develop more balanced and realistic appraisals of their abilities, reducing the perceived need to rely on AI tools as a crutch. Such approaches may reduce needs frustration, manage performance expectations, and promote a healthier, more effective use of AI tools in educational contexts.

Limitations and future directions

This study’s cross-sectional design limits the ability to draw causal inferences, meaning that the directionality of the relationships observed cannot be definitively established. Future research employing longitudinal approaches could better capture the dynamic evolution of AI dependency over time. Additionally, the study’s focus on Chinese university students may limit the generalizability of the findings to other cultural contexts. Variations in educational practices, cultural attitudes toward technology, and personality trait expressions suggest that cross-cultural replication studies are needed to assess the robustness of these findings. Cultural variables, such as the norms and values of different countries and educational institutions, may moderate the relationship between personality traits and AI dependency. Exploring the interaction between personality traits and cultural factors in future studies could offer more nuanced insights into how these relationships manifest in various cultural settings. Moreover, expanding the range of personality traits examined could provide a more comprehensive understanding of how different personality dimensions interact with technology use behaviors. Additionally, researchers could investigate the specific technical features of AI tools that contribute to dependency, as understanding both personal characteristics and technological attributes may deepen insights into the factors driving dependency. By combining these perspectives, future research could provide a more holistic view of how individual, technological and cultural factors interact in shaping AI dependency.

Conclusion

This study highlights the significant influence of personality traits, including neuroticism, self-critical perfectionism, and impulsivity on dependency of Generative AI tools like ChatGPT, primarily through their impacts on needs frustration, negative emotional, and performance expectations. By uncovering these psychological drivers, we provide a basis for developing interventions aimed at fostering healthier interactions between students and AI technologies, while also supporting their academic well-being. Future research should expand on these findings by examining these relationships over time and within diverse cultural contexts to further understand the broader implications of AI dependency.

Supporting information

S1 Table. Zero-order correlations.

https://doi.org/10.1371/journal.pone.0313314.s001

(DOCX)

S2 Table. Factor loadings, reliability and validity.

https://doi.org/10.1371/journal.pone.0313314.s002

(DOCX)

S3 Table. Bootstrapping bias correction for the structural model.

https://doi.org/10.1371/journal.pone.0313314.s003

(DOCX)

Acknowledgments

We would also like to acknowledge all the participants for their participation and cooperation in this study.

References

1. Fontana A, Benzi IMA, Cipresso P. Problematic internet use as a moderator between personality dimensions and internalizing and externalizing symptoms in adolescence. Current psychology. 2023;42(22):19419–28.
- View Article
- Google Scholar
2. Harari GM, Vaid SS, Müller SR, Stachl C, Marrero Z, Schoedel R, et al. Personality sensing for theory development and assessment in the digital age. European Journal of Personality. 2020;34(5):649–69.
- View Article
- Google Scholar
3. Mhlanga D. Digital transformation education, opportunities, and challenges of the application of ChatGPT to emerging economies. Education Research International. 2023;2023(1):7605075.
- View Article
- Google Scholar
4. Singhal K, Azizi S, Tu T, Mahdavi SS, Wei J, Chung HW, et al. Large language models encode clinical knowledge. Nature. 2023;620(7972):172–80. pmid:37438534
- View Article
- PubMed/NCBI
- Google Scholar
5. Ansari AN, Ahmad S, Bhutta SM. Mapping the global evidence around the use of ChatGPT in higher education: A systematic scoping review. Education and Information Technologies. 2023.
- View Article
- Google Scholar
6. Chan CKY. A comprehensive AI policy education framework for university teaching and learning. International journal of educational technology in higher education. 2023;20(1):38.
- View Article
- Google Scholar
7. Krou MR, Fong CJ, Hoff MA. Achievement motivation and academic dishonesty: A meta-analytic investigation. Educational Psychology Review. 2021;33(2):427–58.
- View Article
- Google Scholar
8. Hess BJ, Cupido N, Ross S, Kvern B. Becoming adaptive experts in an era of rapid advances in generative artificial intelligence. Medical Teacher. 2024;46(3):300–3. pmid:38092006
- View Article
- PubMed/NCBI
- Google Scholar
9. Imran M, Almusharraf N. Analyzing the role of ChatGPT as a writing assistant at higher education level: A systematic review of the literature. Contemporary Educational Technology. 2023;15(4):ep464.
- View Article
- Google Scholar
10. Abbas M, Jam FA, Khan TI. Is it harmful or helpful? Examining the causes and consequences of generative AI usage among university students. International Journal of Educational Technology in Higher Education. 2024;21(1):10.
- View Article
- Google Scholar
11. Almogren AS, Al-Rahmi WM, Dahri NA. Exploring Factors Influencing the Acceptance of ChatGPT in Higher Education: A Smart Education Perspective. Heliyon. 2024.
- View Article
- Google Scholar
12. Zhang S, Zhao X, Zhou T, Kim JH. Do you have AI dependency? The roles of academic self-efficacy, academic stress, and performance expectations on problematic AI usage behavior. International Journal of Educational Technology in Higher Education. 2024;21(1):34.
- View Article
- Google Scholar
13. Chaudhry IS, Sarwary SAM, El Refae GA, Chabchoub H. Time to revisit existing student’s performance evaluation approach in higher education sector in a new era of ChatGPT—a case study. Cogent Education. 2023;10(1):2210461.
- View Article
- Google Scholar
14. Karim NSA, Zamzuri NHA, Nor YM. Exploring the relationship between Internet ethics in university students and the big five model of personality. Computers & Education. 2009;53(1):86–93.
- View Article
- Google Scholar
15. Marciano L, Camerini A-L, Schulz PJ. Neuroticism and internet addiction: What is next? A systematic conceptual review. Personality and individual differences. 2022;185:111260.
- View Article
- Google Scholar
16. Sariyska R, Reuter M, Bey K, Sha P, Li M, Chen Y-F, et al. Self-esteem, personality and Internet Addiction: A cross-cultural comparison study. Personality and Individual Differences. 2014;61–62:28–33. https://doi.org/10.1016/j.paid.2014.01.001.
- View Article
- Google Scholar
17. Brand M, Young KS, Laier C, Wölfling K, Potenza MN. Integrating psychological and neurobiological considerations regarding the development and maintenance of specific Internet-use disorders: An Interaction of Person-Affect-Cognition-Execution (I-PACE) model. Neuroscience & Biobehavioral Reviews. 2016;71:252–66. pmid:27590829
- View Article
- PubMed/NCBI
- Google Scholar
18. Nan D, Lee H, Kim Y, Kim JH. My video game console is so cool! A coolness theory-based model for intention to use video game consoles. Technological Forecasting and Social Change. 2022;176:121451. pmid:34980931
- View Article
- PubMed/NCBI
- Google Scholar
19. Dwivedi YK, Rana NP, Jeyaraj A, Clement M, Williams MD. Re-examining the Unified Theory of Acceptance and Use of Technology (UTAUT): Towards a Revised Theoretical Model. Information Systems Frontiers. 2019;21(3):719–34.
- View Article
- Google Scholar
20. Ryan RM, Deci EL. Self-determination theory: Basic psychological needs in motivation, development, and wellness: Guilford publications; 2017.
21. Przybylski AK, Weinstein N, Murayama K, Lynch MF, Ryan RM. The ideal self at play: the appeal of video games that let you be all you can be. Psychol Sci. 2012;23(1):69–76. Epub 20111214. pmid:22173739.
- View Article
- PubMed/NCBI
- Google Scholar
22. Kardefelt-Winther D. A conceptual and methodological critique of internet addiction research: Towards a model of compensatory internet use. Computers in Human Behavior. 2014;31:351–4.
- View Article
- Google Scholar
23. Morris RR, Kouddous K, Kshirsagar R, Schueller SM. Towards an Artificially Empathic Conversational Agent for Mental Health Applications: System Design and User Perceptions. J Med Internet Res. 2018;20(6):e10148. Epub 20180626. pmid:29945856; PubMed Central PMCID: PMC6039770.
- View Article
- PubMed/NCBI
- Google Scholar
24. Jabagi N, Croteau A-M, Audebrand L, Marsan J. Unlocking perceived algorithmic autonomy-support: Scale development and validation. 2021.
- View Article
- Google Scholar
25. Costa PT, McCrae RR. Normal personality assessment in clinical practice: The NEO Personality Inventory. Psychological assessment. 1992;4(1):5.
- View Article
- Google Scholar
26. Shiner RL. Negative emotionality and neuroticism from childhood through adulthood. Handbook of personality development. 2018:137.
- View Article
- Google Scholar
27. Ormel J, Riese H, Rosmalen JG. Interpreting neuroticism scores across the adult life course: immutable or experience-dependent set points of negative affect? Clinical psychology review. 2012;32(1):71–9. pmid:22172577
- View Article
- PubMed/NCBI
- Google Scholar
28. Blatt SJ. The destructiveness of perfectionism: Implications for the treatment of depression. American psychologist. 1995;50(12):1003.
- View Article
- Google Scholar
29. Boone L, Vansteenkiste M, Soenens B, der Kaap-Deeder V, Verstuyf J. Self-critical perfectionism and binge eating symptoms: a longitudinal test of the intervening role of psychological need frustration. Journal of counseling psychology. 2014;61(3):363. pmid:25019540
- View Article
- PubMed/NCBI
- Google Scholar
30. Richardson CM, Rice KG. Self-critical perfectionism, daily stress, and disclosure of daily emotional events. Journal of counseling psychology. 2015;62(4):694. pmid:26167649
- View Article
- PubMed/NCBI
- Google Scholar
31. Whiteside SP, Lynam DR. The five factor model and impulsivity: Using a structural model of personality to understand impulsivity. Personality and individual differences. 2001;30(4):669–89.
- View Article
- Google Scholar
32. Shafiepour Motlagh F, Torabi Nahad M. Relationship between academic stress, educational impulsivity and negative academic self-concept with sense of social unconsciousness based on the mediation of academic frustration. Journal of Instruction and Evaluation. 2018;11(43):45–68.
- View Article
- Google Scholar
33. Gröndal M, Ask K, Winblad S. The distinction between irritability and anger and their associations with impulsivity and subjective wellbeing. Scientific Reports. 2023;13(1):10398. pmid:37369760
- View Article
- PubMed/NCBI
- Google Scholar
34. Pekrun R, Goetz T, Titz W, Perry RP. Academic emotions in students’ self-regulated learning and achievement: A program of qualitative and quantitative research. Educational psychologist. 2002;37(2):91–105.
- View Article
- Google Scholar
35. Fredrickson BL. Cultivating positive emotions to optimize health and well-being. Prevention & treatment. 2000;3(1):1a.
- View Article
- Google Scholar
36. Teixeira DS, Silva MN, Palmeira AL. How does frustration make you feel? A motivational analysis in exercise context. Motivation and Emotion. 2018;42:419–28.
- View Article
- Google Scholar
37. Venkatesh V, Morris MG, Davis GB, Davis FD. User acceptance of information technology: Toward a unified view. MIS quarterly. 2003:425–78.
- View Article
- Google Scholar
38. Costa PT, McCrae RR, Psychological Assessment Resources I. Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI): Psychological Assessment Resources; 1992.
- View Article
- Google Scholar
39. Patton JH, Stanford MS, Barratt ES. Factor structure of the barratt impulsiveness scale. Journal of Clinical Psychology. 1995;51(6):768–74. pmid:8778124
- View Article
- PubMed/NCBI
- Google Scholar
40. Feher A, Smith MM, Saklofske DH, Plouffe RA, Wilson CA, Sherry SB. The Big Three Perfectionism Scale–Short Form (BTPS-SF): Development of a Brief Self-Report Measure of Multidimensional Perfectionism. Journal of Psychoeducational Assessment. 2019;38(1):37–52.
- View Article
- Google Scholar
41. Chen B, Vansteenkiste M, Beyers W, Boone L, Deci EL, Van der Kaap-Deeder J, et al. Basic psychological need satisfaction, need frustration, and need strength across four cultures. Motivation and Emotion. 2015;39(2):216–36.
- View Article
- Google Scholar
42. Bieleke M, Gogol K, Goetz T, Daniels L, Pekrun R. The AEQ-S: A short version of the Achievement Emotions Questionnaire. Contemporary Educational Psychology. 2021;65:101940.
- View Article
- Google Scholar
43. Oghuma AP, Libaque-Saenz CF, Wong SF, Chang Y. An expectation-confirmation model of continuance intention to use mobile instant messaging. TELEMATICS AND INFORMATICS. 2016;33(1):34–47. WOS:000362304200005.
- View Article
- Google Scholar
44. Hu B, Mao Y, Kim KJ. How social anxiety leads to problematic use of conversational AI: The roles of loneliness, rumination, and mind perception. Computers in Human Behavior. 2023;145:107760. https://doi.org/10.1016/j.chb.2023.107760.
- View Article
- Google Scholar
45. Harman D. A single factor test of common method variance. Journal of Psychology. 1967;35(1967):359–78.
- View Article
- Google Scholar
46. Podsakoff PM, MacKenzie SB, Podsakoff NP. Sources of Method Bias in Social Science Research and Recommendations on How to Control It. Annual Review of Psychology. 2012;63(Volume 63, 2012):539–69. pmid:21838546
- View Article
- PubMed/NCBI
- Google Scholar
47. Fornell C, Larcker DF. Evaluating structural equation models with unobservable variables and measurement error. Journal of marketing research. 1981;18(1):39–50.
- View Article
- Google Scholar
48. Henseler J, Ringle CM, Sarstedt M. A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science. 2015;43(1):115–35.
- View Article
- Google Scholar
49. Lt Hu, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural equation modeling: a multidisciplinary journal. 1999;6(1):1–55.
- View Article
- Google Scholar
50. Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior research methods. 2008;40(3):879–91. pmid:18697684
- View Article
- PubMed/NCBI
- Google Scholar
51. Chamarro A, Oberst U, Cladellas R, Fuster H. Effect of the frustration of psychological needs on addictive behaviors in mobile videogamers—the mediating role of use expectancies and time spent gaming. International journal of environmental research and public health. 2020;17(17):6429. pmid:32899364
- View Article
- PubMed/NCBI
- Google Scholar
52. Rokosz M, Poprawa R, Barański M, Macyszyn P. Basic psychological need frustration and the risk of problematic internet use: the mediating role of hedonistic and compensatory outcome expectancies. Alcoholism and Drug Addiction/Alkoholizm i Narkomania. 2023;36(4):235–56.
- View Article
- Google Scholar
53. Bowden-Green T, Hinds J, Joinson A. Understanding neuroticism and social media: A systematic review. Personality and Individual Differences. 2021;168:110344.
- View Article
- Google Scholar
54. Shen J, Brdiczka O, Liu J. A study of Facebook behavior: What does it tell about your Neuroticism and Extraversion? Computers in Human Behavior. 2015;45:32–8.
- View Article
- Google Scholar
55. Hewitt PL, Flett GL. Perfectionism in the self and social contexts: conceptualization, assessment, and association with psychopathology. J Pers Soc Psychol. 1991;60(3):456–70. pmid:2027080.
- View Article
- PubMed/NCBI
- Google Scholar
56. Ayadi N, Pireinaladin S, Shokri M, Dargahi S, Zarein F. Investigating the mediating role of procrastination in the relationship between positive and negative perfectionism and Mobile phone addiction in gifted students. Iranian journal of psychiatry. 2021;16(1):30. pmid:34054980
- View Article
- PubMed/NCBI
- Google Scholar
57. Baumeister RF, Vohs KD. Self‐Regulation, ego depletion, and motivation. Social and personality psychology compass. 2007;1(1):115–28.
- View Article
- Google Scholar
58. Davis RA. A cognitive-behavioral model of pathological Internet use. Computers in human behavior. 2001;17(2):187–95.
- View Article
- Google Scholar
59. Casale S, Caplan SE, Fioravanti G. Positive metacognitions about Internet use: The mediating role in the relationship between emotional dysregulation and problematic use. Addictive Behaviors. 2016;59:84–8. pmid:27077964
- View Article
- PubMed/NCBI
- Google Scholar
60. Granić A, Marangunić N. Technology acceptance model in educational context: A systematic literature review. British Journal of Educational Technology. 2019;50(5):2572–93.
- View Article
- Google Scholar
61. Torous J, Firth J. The digital placebo effect: mobile mental health meets clinical psychiatry. Lancet Psychiatry. 2016;3(2):100–2. pmid:26851322.
- View Article
- PubMed/NCBI
- Google Scholar
62. Li H, Zhang R, Lee Y-C, Kraut RE, Mohr DC. Systematic review and meta-analysis of AI-based conversational agents for promoting mental health and well-being. npj Digital Medicine. 2023;6(1):236. pmid:38114588
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Fontana A, Benzi IMA, Cipresso P. Problematic internet use as a moderator between personality dimensions and internalizing and externalizing symptoms in adolescence. Current psychology. 2023;42(22):19419–28.
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. Harari GM, Vaid SS, Müller SR, Stachl C, Marrero Z, Schoedel R, et al. Personality sensing for theory development and assessment in the digital age. European Journal of Personality. 2020;34(5):649–69.
View Article
Google Scholar

[5] View Article

[6] Google Scholar

[ref3] 3. Mhlanga D. Digital transformation education, opportunities, and challenges of the application of ChatGPT to emerging economies. Education Research International. 2023;2023(1):7605075.
View Article
Google Scholar

[8] View Article

[9] Google Scholar

[ref4] 4. Singhal K, Azizi S, Tu T, Mahdavi SS, Wei J, Chung HW, et al. Large language models encode clinical knowledge. Nature. 2023;620(7972):172–80. pmid:37438534
View Article
PubMed/NCBI
Google Scholar

[11] View Article

[12] PubMed/NCBI

[13] Google Scholar

[ref5] 5. Ansari AN, Ahmad S, Bhutta SM. Mapping the global evidence around the use of ChatGPT in higher education: A systematic scoping review. Education and Information Technologies. 2023.
View Article
Google Scholar

[15] View Article

[16] Google Scholar

[ref6] 6. Chan CKY. A comprehensive AI policy education framework for university teaching and learning. International journal of educational technology in higher education. 2023;20(1):38.
View Article
Google Scholar

[18] View Article

[19] Google Scholar

[ref7] 7. Krou MR, Fong CJ, Hoff MA. Achievement motivation and academic dishonesty: A meta-analytic investigation. Educational Psychology Review. 2021;33(2):427–58.
View Article
Google Scholar

[21] View Article

[22] Google Scholar

[ref8] 8. Hess BJ, Cupido N, Ross S, Kvern B. Becoming adaptive experts in an era of rapid advances in generative artificial intelligence. Medical Teacher. 2024;46(3):300–3. pmid:38092006
View Article
PubMed/NCBI
Google Scholar

[24] View Article

[25] PubMed/NCBI

[26] Google Scholar

[ref9] 9. Imran M, Almusharraf N. Analyzing the role of ChatGPT as a writing assistant at higher education level: A systematic review of the literature. Contemporary Educational Technology. 2023;15(4):ep464.
View Article
Google Scholar

[28] View Article

[29] Google Scholar

[ref10] 10. Abbas M, Jam FA, Khan TI. Is it harmful or helpful? Examining the causes and consequences of generative AI usage among university students. International Journal of Educational Technology in Higher Education. 2024;21(1):10.
View Article
Google Scholar

[31] View Article

[32] Google Scholar

[ref11] 11. Almogren AS, Al-Rahmi WM, Dahri NA. Exploring Factors Influencing the Acceptance of ChatGPT in Higher Education: A Smart Education Perspective. Heliyon. 2024.
View Article
Google Scholar

[34] View Article

[35] Google Scholar

[ref12] 12. Zhang S, Zhao X, Zhou T, Kim JH. Do you have AI dependency? The roles of academic self-efficacy, academic stress, and performance expectations on problematic AI usage behavior. International Journal of Educational Technology in Higher Education. 2024;21(1):34.
View Article
Google Scholar

[37] View Article

[38] Google Scholar

[ref13] 13. Chaudhry IS, Sarwary SAM, El Refae GA, Chabchoub H. Time to revisit existing student’s performance evaluation approach in higher education sector in a new era of ChatGPT—a case study. Cogent Education. 2023;10(1):2210461.
View Article
Google Scholar

[40] View Article

[41] Google Scholar

[ref14] 14. Karim NSA, Zamzuri NHA, Nor YM. Exploring the relationship between Internet ethics in university students and the big five model of personality. Computers & Education. 2009;53(1):86–93.
View Article
Google Scholar

[43] View Article

[44] Google Scholar

[ref15] 15. Marciano L, Camerini A-L, Schulz PJ. Neuroticism and internet addiction: What is next? A systematic conceptual review. Personality and individual differences. 2022;185:111260.
View Article
Google Scholar

[46] View Article

[47] Google Scholar

[ref16] 16. Sariyska R, Reuter M, Bey K, Sha P, Li M, Chen Y-F, et al. Self-esteem, personality and Internet Addiction: A cross-cultural comparison study. Personality and Individual Differences. 2014;61–62:28–33. https://doi.org/10.1016/j.paid.2014.01.001.
View Article
Google Scholar

[49] View Article

[50] Google Scholar

[ref17] 17. Brand M, Young KS, Laier C, Wölfling K, Potenza MN. Integrating psychological and neurobiological considerations regarding the development and maintenance of specific Internet-use disorders: An Interaction of Person-Affect-Cognition-Execution (I-PACE) model. Neuroscience & Biobehavioral Reviews. 2016;71:252–66. pmid:27590829
View Article
PubMed/NCBI
Google Scholar

[52] View Article

[53] PubMed/NCBI

[54] Google Scholar

[ref18] 18. Nan D, Lee H, Kim Y, Kim JH. My video game console is so cool! A coolness theory-based model for intention to use video game consoles. Technological Forecasting and Social Change. 2022;176:121451. pmid:34980931
View Article
PubMed/NCBI
Google Scholar

[56] View Article

[57] PubMed/NCBI

[58] Google Scholar

[ref19] 19. Dwivedi YK, Rana NP, Jeyaraj A, Clement M, Williams MD. Re-examining the Unified Theory of Acceptance and Use of Technology (UTAUT): Towards a Revised Theoretical Model. Information Systems Frontiers. 2019;21(3):719–34.
View Article
Google Scholar

[60] View Article

[61] Google Scholar

[ref20] 20. Ryan RM, Deci EL. Self-determination theory: Basic psychological needs in motivation, development, and wellness: Guilford publications; 2017.

[ref21] 21. Przybylski AK, Weinstein N, Murayama K, Lynch MF, Ryan RM. The ideal self at play: the appeal of video games that let you be all you can be. Psychol Sci. 2012;23(1):69–76. Epub 20111214. pmid:22173739.
View Article
PubMed/NCBI
Google Scholar

[64] View Article

[65] PubMed/NCBI

[66] Google Scholar

[ref22] 22. Kardefelt-Winther D. A conceptual and methodological critique of internet addiction research: Towards a model of compensatory internet use. Computers in Human Behavior. 2014;31:351–4.
View Article
Google Scholar

[68] View Article

[69] Google Scholar

[ref23] 23. Morris RR, Kouddous K, Kshirsagar R, Schueller SM. Towards an Artificially Empathic Conversational Agent for Mental Health Applications: System Design and User Perceptions. J Med Internet Res. 2018;20(6):e10148. Epub 20180626. pmid:29945856; PubMed Central PMCID: PMC6039770.
View Article
PubMed/NCBI
Google Scholar

[71] View Article

[72] PubMed/NCBI

[73] Google Scholar

[ref24] 24. Jabagi N, Croteau A-M, Audebrand L, Marsan J. Unlocking perceived algorithmic autonomy-support: Scale development and validation. 2021.
View Article
Google Scholar

[75] View Article

[76] Google Scholar

[ref25] 25. Costa PT, McCrae RR. Normal personality assessment in clinical practice: The NEO Personality Inventory. Psychological assessment. 1992;4(1):5.
View Article
Google Scholar

[78] View Article

[79] Google Scholar

[ref26] 26. Shiner RL. Negative emotionality and neuroticism from childhood through adulthood. Handbook of personality development. 2018:137.
View Article
Google Scholar

[81] View Article

[82] Google Scholar

[ref27] 27. Ormel J, Riese H, Rosmalen JG. Interpreting neuroticism scores across the adult life course: immutable or experience-dependent set points of negative affect? Clinical psychology review. 2012;32(1):71–9. pmid:22172577
View Article
PubMed/NCBI
Google Scholar

[84] View Article

[85] PubMed/NCBI

[86] Google Scholar

[ref28] 28. Blatt SJ. The destructiveness of perfectionism: Implications for the treatment of depression. American psychologist. 1995;50(12):1003.
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref29] 29. Boone L, Vansteenkiste M, Soenens B, der Kaap-Deeder V, Verstuyf J. Self-critical perfectionism and binge eating symptoms: a longitudinal test of the intervening role of psychological need frustration. Journal of counseling psychology. 2014;61(3):363. pmid:25019540
View Article
PubMed/NCBI
Google Scholar

[91] View Article

[92] PubMed/NCBI

[93] Google Scholar

[ref30] 30. Richardson CM, Rice KG. Self-critical perfectionism, daily stress, and disclosure of daily emotional events. Journal of counseling psychology. 2015;62(4):694. pmid:26167649
View Article
PubMed/NCBI
Google Scholar

[95] View Article

[96] PubMed/NCBI

[97] Google Scholar

[ref31] 31. Whiteside SP, Lynam DR. The five factor model and impulsivity: Using a structural model of personality to understand impulsivity. Personality and individual differences. 2001;30(4):669–89.
View Article
Google Scholar

[99] View Article

[100] Google Scholar

[ref32] 32. Shafiepour Motlagh F, Torabi Nahad M. Relationship between academic stress, educational impulsivity and negative academic self-concept with sense of social unconsciousness based on the mediation of academic frustration. Journal of Instruction and Evaluation. 2018;11(43):45–68.
View Article
Google Scholar

[102] View Article

[103] Google Scholar

[ref33] 33. Gröndal M, Ask K, Winblad S. The distinction between irritability and anger and their associations with impulsivity and subjective wellbeing. Scientific Reports. 2023;13(1):10398. pmid:37369760
View Article
PubMed/NCBI
Google Scholar

[105] View Article

[106] PubMed/NCBI

[107] Google Scholar

[ref34] 34. Pekrun R, Goetz T, Titz W, Perry RP. Academic emotions in students’ self-regulated learning and achievement: A program of qualitative and quantitative research. Educational psychologist. 2002;37(2):91–105.
View Article
Google Scholar

[109] View Article

[110] Google Scholar

[ref35] 35. Fredrickson BL. Cultivating positive emotions to optimize health and well-being. Prevention & treatment. 2000;3(1):1a.
View Article
Google Scholar

[112] View Article

[113] Google Scholar

[ref36] 36. Teixeira DS, Silva MN, Palmeira AL. How does frustration make you feel? A motivational analysis in exercise context. Motivation and Emotion. 2018;42:419–28.
View Article
Google Scholar

[115] View Article

[116] Google Scholar

[ref37] 37. Venkatesh V, Morris MG, Davis GB, Davis FD. User acceptance of information technology: Toward a unified view. MIS quarterly. 2003:425–78.
View Article
Google Scholar

[118] View Article

[119] Google Scholar

[ref38] 38. Costa PT, McCrae RR, Psychological Assessment Resources I. Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI): Psychological Assessment Resources; 1992.
View Article
Google Scholar

[121] View Article

[122] Google Scholar

[ref39] 39. Patton JH, Stanford MS, Barratt ES. Factor structure of the barratt impulsiveness scale. Journal of Clinical Psychology. 1995;51(6):768–74. pmid:8778124
View Article
PubMed/NCBI
Google Scholar

[124] View Article

[125] PubMed/NCBI

[126] Google Scholar

[ref40] 40. Feher A, Smith MM, Saklofske DH, Plouffe RA, Wilson CA, Sherry SB. The Big Three Perfectionism Scale–Short Form (BTPS-SF): Development of a Brief Self-Report Measure of Multidimensional Perfectionism. Journal of Psychoeducational Assessment. 2019;38(1):37–52.
View Article
Google Scholar

[128] View Article

[129] Google Scholar

[ref41] 41. Chen B, Vansteenkiste M, Beyers W, Boone L, Deci EL, Van der Kaap-Deeder J, et al. Basic psychological need satisfaction, need frustration, and need strength across four cultures. Motivation and Emotion. 2015;39(2):216–36.
View Article
Google Scholar

[131] View Article

[132] Google Scholar

[ref42] 42. Bieleke M, Gogol K, Goetz T, Daniels L, Pekrun R. The AEQ-S: A short version of the Achievement Emotions Questionnaire. Contemporary Educational Psychology. 2021;65:101940.
View Article
Google Scholar

[134] View Article

[135] Google Scholar

[ref43] 43. Oghuma AP, Libaque-Saenz CF, Wong SF, Chang Y. An expectation-confirmation model of continuance intention to use mobile instant messaging. TELEMATICS AND INFORMATICS. 2016;33(1):34–47. WOS:000362304200005.
View Article
Google Scholar

[137] View Article

[138] Google Scholar

[ref44] 44. Hu B, Mao Y, Kim KJ. How social anxiety leads to problematic use of conversational AI: The roles of loneliness, rumination, and mind perception. Computers in Human Behavior. 2023;145:107760. https://doi.org/10.1016/j.chb.2023.107760.
View Article
Google Scholar

[140] View Article

[141] Google Scholar

[ref45] 45. Harman D. A single factor test of common method variance. Journal of Psychology. 1967;35(1967):359–78.
View Article
Google Scholar

[143] View Article

[144] Google Scholar

[ref46] 46. Podsakoff PM, MacKenzie SB, Podsakoff NP. Sources of Method Bias in Social Science Research and Recommendations on How to Control It. Annual Review of Psychology. 2012;63(Volume 63, 2012):539–69. pmid:21838546
View Article
PubMed/NCBI
Google Scholar

[146] View Article

[147] PubMed/NCBI

[148] Google Scholar

[ref47] 47. Fornell C, Larcker DF. Evaluating structural equation models with unobservable variables and measurement error. Journal of marketing research. 1981;18(1):39–50.
View Article
Google Scholar

[150] View Article

[151] Google Scholar

[ref48] 48. Henseler J, Ringle CM, Sarstedt M. A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science. 2015;43(1):115–35.
View Article
Google Scholar

[153] View Article

[154] Google Scholar

[ref49] 49. Lt Hu, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural equation modeling: a multidisciplinary journal. 1999;6(1):1–55.
View Article
Google Scholar

[156] View Article

[157] Google Scholar

[ref50] 50. Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior research methods. 2008;40(3):879–91. pmid:18697684
View Article
PubMed/NCBI
Google Scholar

[159] View Article

[160] PubMed/NCBI

[161] Google Scholar

[ref51] 51. Chamarro A, Oberst U, Cladellas R, Fuster H. Effect of the frustration of psychological needs on addictive behaviors in mobile videogamers—the mediating role of use expectancies and time spent gaming. International journal of environmental research and public health. 2020;17(17):6429. pmid:32899364
View Article
PubMed/NCBI
Google Scholar

[163] View Article

[164] PubMed/NCBI

[165] Google Scholar

[ref52] 52. Rokosz M, Poprawa R, Barański M, Macyszyn P. Basic psychological need frustration and the risk of problematic internet use: the mediating role of hedonistic and compensatory outcome expectancies. Alcoholism and Drug Addiction/Alkoholizm i Narkomania. 2023;36(4):235–56.
View Article
Google Scholar

[167] View Article

[168] Google Scholar

[ref53] 53. Bowden-Green T, Hinds J, Joinson A. Understanding neuroticism and social media: A systematic review. Personality and Individual Differences. 2021;168:110344.
View Article
Google Scholar

[170] View Article

[171] Google Scholar

[ref54] 54. Shen J, Brdiczka O, Liu J. A study of Facebook behavior: What does it tell about your Neuroticism and Extraversion? Computers in Human Behavior. 2015;45:32–8.
View Article
Google Scholar

[173] View Article

[174] Google Scholar

[ref55] 55. Hewitt PL, Flett GL. Perfectionism in the self and social contexts: conceptualization, assessment, and association with psychopathology. J Pers Soc Psychol. 1991;60(3):456–70. pmid:2027080.
View Article
PubMed/NCBI
Google Scholar

[176] View Article

[177] PubMed/NCBI

[178] Google Scholar

[ref56] 56. Ayadi N, Pireinaladin S, Shokri M, Dargahi S, Zarein F. Investigating the mediating role of procrastination in the relationship between positive and negative perfectionism and Mobile phone addiction in gifted students. Iranian journal of psychiatry. 2021;16(1):30. pmid:34054980
View Article
PubMed/NCBI
Google Scholar

[180] View Article

[181] PubMed/NCBI

[182] Google Scholar

[ref57] 57. Baumeister RF, Vohs KD. Self‐Regulation, ego depletion, and motivation. Social and personality psychology compass. 2007;1(1):115–28.
View Article
Google Scholar

[184] View Article

[185] Google Scholar

[ref58] 58. Davis RA. A cognitive-behavioral model of pathological Internet use. Computers in human behavior. 2001;17(2):187–95.
View Article
Google Scholar

[187] View Article

[188] Google Scholar

[ref59] 59. Casale S, Caplan SE, Fioravanti G. Positive metacognitions about Internet use: The mediating role in the relationship between emotional dysregulation and problematic use. Addictive Behaviors. 2016;59:84–8. pmid:27077964
View Article
PubMed/NCBI
Google Scholar

[190] View Article

[191] PubMed/NCBI

[192] Google Scholar

[ref60] 60. Granić A, Marangunić N. Technology acceptance model in educational context: A systematic literature review. British Journal of Educational Technology. 2019;50(5):2572–93.
View Article
Google Scholar

[194] View Article

[195] Google Scholar

[ref61] 61. Torous J, Firth J. The digital placebo effect: mobile mental health meets clinical psychiatry. Lancet Psychiatry. 2016;3(2):100–2. pmid:26851322.
View Article
PubMed/NCBI
Google Scholar

[197] View Article

[198] PubMed/NCBI

[199] Google Scholar

[ref62] 62. Li H, Zhang R, Lee Y-C, Kraut RE, Mohr DC. Systematic review and meta-analysis of AI-based conversational agents for promoting mental health and well-being. npj Digital Medicine. 2023;6(1):236. pmid:38114588
View Article
PubMed/NCBI
Google Scholar

[201] View Article

[202] PubMed/NCBI

[203] Google Scholar

Figures

Abstract

Objective

Method

Results

Conclusions

Introduction

Theoretical framework and hypothesis formulation

I-PACE framework

Needs frustration and ChatGPT dependency

Personality as an antecedent

Negative academic emotions and performance expectations

Serial mediation effects

Materials and methods

Participants and procedure

Measurements

Statistical analysis

Results

Measurement model

Descriptive statistics and common method bias.

Test of measurement validity.

Structural model analysis

Test of model fit.

Direct path analysis.

Mediation and serial mediation analysis.

Exploratory analysis.

Discussion

Personality and AI dependency

Cognitive factors as key drivers

Implications

Limitations and future directions

Conclusion

Supporting information

S1 Table. Zero-order correlations.

S2 Table. Factor loadings, reliability and validity.

S3 Table. Bootstrapping bias correction for the structural model.

Acknowledgments

References