簡易檢索 / 詳目顯示
| 研究生： |
陳以潔 Chen, Yi-Chieh |
|---|---|
| 論文名稱： |
情境的不確定性與焦慮程度對人智互動中的歸因方式之影響 The Impact of Situational Uncertainty and Anxiety on Attribution in Human-AI Interaction |
| 指導教授： | 陳宜秀 |
| 口試委員: |
侯宗佑
余能豪 |
| 學位類別： |
碩士
Master |
| 系所名稱： |
創新國際學院 - 全球傳播與創新科技碩士學位學程 Master’s Program in Global Communication and Innovation Technology |
| 論文出版年： | 2026 |
| 畢業學年度： | 114 |
| 語文別： | 英文 |
| 論文頁數： | 89 |
| 中文關鍵詞： | 生成式人工智慧 、情境不確定性 、狀態焦慮 、合理化 、人機互動 |
| 外文關鍵詞： | generative AI, situational uncertainty, state anxiety, rationalization, human-AI interaction |
| 相關次數： | 點閱：15 下載：0 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
隨著生成式人工智慧（generative artificial intelligence, GenAI）越來越常被用於模糊、不確定的決策情境，使用者如何理解 GenAI 回應已成為人機互動研究中的重要問題。使用者並非單純以理性方式接收回應，其情緒狀態與情境感受亦可能影響其對回應的解讀與評價。現有研究已廣泛評估生成式人工智慧系統及其輸出的效能、品質與可解釋性，並探討這些因素如何影響使用者的理解與反應。然而，較少研究將使用者視為主動的詮釋者，探討其如何理解並賦予 GenAI 回應意義。本研究將此過程稱為合理化，意指使用者對 GenAI 回應中未明確說明的意圖、邏輯或條件進行補充性解釋。本研究旨在探討情境中的不確定性是否會透過狀態焦慮，提高使用者對 GenAI 回應進行合理化解讀的傾向。本研究採用單因子組間實驗設計，受試者被隨機分派至高不確定性或低不確定性的職涯選擇情境。所有參與者接著閱讀相同的標準化 GenAI 回應，該回應被設計為中立、模糊，且不直接建議受試者選擇特定工作。本研究最終納入 67 份有效樣本，並透過量化量表與開放式回答分析，檢驗情境不確定性、狀態焦慮、與合理化行為之間的關係。
研究結果顯示，受試者感受到的情境不確定性能顯著預測狀態焦慮，但狀態焦慮未能顯著提高受試者對 GenAI 回應的正面評價，並對回應可靠性的評價呈現負向預測。然而，操弄結果顯示，隨機分派至高、低不確定性情境的兩組受試者，在情境不確定性感受上未達顯著差異，因此本研究無法根據實驗結果支持情境不確定性與合理化行為之間的關係。狀態焦慮也未能顯著預測合理化行為，其中介效果同樣未獲支持。開放式回答進一步顯示，合理化行為在兩組參與者中皆相當普遍。多數參與者會主動推論 GenAI 回應背後的意圖、邏輯或未明確說明的條件，顯示合理化可能是使用者面對模糊且保有解讀空間的回應時，普遍出現的意義建構過程。此外，高不確定性組較常將 GenAI 回應描述為具有工具性效益，例如協助整理資訊、建立初步思考框架與釐清問題。
整體而言，Gen AI 回應的意義不僅來自系統輸出的內容，也來自使用者依據自身情境、情緒狀態與資訊需求所進行的主動詮釋，即使回應未被視為完全可靠，使用者仍可能將其作為整理思緒與輔助決策的資源。本研究提出以使用者意義建構作為分析視角，突顯在不確定的情境中，理解使用者如何詮釋及運用 GenAI 回應的重要性。
As generative artificial intelligence is increasingly used in ambiguous and uncertain decision-making contexts, how users understand GenAI responses has become an important issue in human–AI interaction research. Users do not simply receive these responses in a purely rational manner. Their emotional states and perceptions of the situation may also shape how they interpret and evaluate them. Existing research has extensively examined the performance, quality, and explainability of GenAI systems and their outputs, as well as how these factors influence users’ understanding and responses. However, less attention has been paid to users as active interpreters who construct and assign meaning to GenAI responses. The study refers to this process as rationalization, defined as users’ supplementary explanations of the intentions, logic, or conditions that are not explicitly stated in a GenAI response.
The study aimed to examine whether situational uncertainty in a career decision-making context would increase users’ tendency to rationalize generative AI responses through state anxiety. Using a one-factor between-subjects experimental design, participants were randomly assigned to either a high-uncertainty or low-uncertainty career choice scenario. All participants then read the same standardized GenAI response, which was designed to be general, neutral, and non-directive. The final sample included 67 valid responses. Quantitative scales and open-ended responses were used to examine the relationships among situational uncertainty, state anxiety, evaluations of the AI response, and rationalization.
The results showed that participants’ perceived situational uncertainty significantly predicted state anxiety. State anxiety did not significantly increase positive evaluations of the GenAI response. Instead, it negatively predicted participants’ perceived reliability of the response. However, the high- and low-uncertainty conditions did not differ significantly in situational uncertainty. Therefore, the experimental results could not support a relationship between situational uncertainty and rationalization. State anxiety also did not significantly predict rationalization, and the mediating role of state anxiety was not supported.
Open-ended responses further showed that rationalization was common across both conditions. Most participants actively inferred the intentions, logic, or unstated conditions behind the GenAI response, suggesting that rationalization may be a common meaning-making process when users encounter responses that are ambiguous and open to interpretation. In addition, participants in the high-uncertainty condition more frequently described the GenAI response as instrumentally useful, such as helping them organize information, build an initial framework, and clarify the problem.
Overall, the meaning of a GenAI response arises not only from the content generated by the system, but also from users’ active interpretations based on their situations, emotional states, and information needs. Even when a response is not perceived as fully reliable, users may still use it as a resource for organizing their thoughts and supporting decision-making. This study takes a meaning-making approach to understanding users, highlighting the need to examine how they interpret and use GenAI responses in uncertain situational contexts.
1. Introduction 1
2. Theoretical Background 4
2.1 Human-AI Interaction 4
2.1.1 Why People Ask Questions Under Uncertainty 4
2.1.2 Why People Turn to GenAI for Answers 5
2.2 System and User Factors Affecting AI Interaction 5
2.2.1 System Design Factors 5
2.2.2 User Factors 6
2.3 Situational Factors 7
2.3.1 Situational Uncertainty 7
2.3.2 Situational Uncertainty and Anxiety 8
2.4 Social Cognition towards GenAI Responses 10
2.4.1 Sensemaking under Uncertainty 10
2.4.2 Attribution as a Sensemaking Mechanism 11
2.4.3 Attribution in Interpersonal Communication 11
2.4.4 Attribution in Human-Agent Interaction 12
2.5 How Uncertainty and Anxiety May Shape Rationalization of GenAI Responses 13
2.6 Research Questions and Hypotheses 14
3. Methodology 16
3.1 Research Design 16
3.2 Pilot Study 17
3.2.1 Purpose of the Pilot Study 17
3.2.2 Pilot Study Participants 18
3.2.3 Pilot Study Procedure 18
3.2.4 Measures in the Pilot Study 19
3.2.5 Pilot Study Results 20
3.2.6 Conclusion of the Pilot Study 23
3.3 Main Experiment 25
3.3.1 Participants 25
3.3.2 Experimental Procedure 27
3.3.3 Experimental Materials and Scenario Manipulation 29
3.4 Dependent Measures 31
3.4.1 Manipulation Check 31
3.4.2 State Anxiety 32
3.4.3 Trust in the GenAI Response 33
3.4.4 GenAI Response Evaluation: Validity, Reliability, Usefulness, and Personalization 33
3.4.5 Open Ended Questions and Coding Procedure 33
4. Results 35
4.1 Scale Reliability and Descriptive Statistics 35
4.2 Manipulation Check 37
4.2.1 Situational Uncertainty 37
4.2.2 Scenario Realism 38
4.2.3 Information Clarity of Job A and Job B 39
4.3 Main Hypothesis Testing 41
4.3.1 H1: Higher Situational Uncertainty Increases the Likelihood of Rationalizing GenAI Outputs 41
4.3.2 H2: Higher Situational Uncertainty Increases State Anxiety 42
4.3.3 H3: Higher State Anxiety Increases the Likelihood of Rationalizing GenAI Outputs 43
4.4 Situational Uncertainty, State Anxiety, and GenAI Response Evaluation 44
4.4.1 Situational Uncertainty and State Anxiety 44
4.4.2 State Anxiety and GenAI Response Evaluation 45
4.4.3 Reliability of the GenAI Response and Trust in AI 46
4.5 Analysis of Open-Ended Response 46
4.5.1 First Open-Ended Question: Theme Definitions and Distribution 47
4.5.2 Group Comparisons of Open-Ended Themes 50
4.5.3 Second Open-Ended Question: Rationalization 52
4.5.4 Mixed Evaluations: Criticizing AI Quality While Still Recognizing Its Instrumental Value 53
4.5.5 Decision-Aid Usefulness and Trust in AI 54
4.6 Summary of Results 55
5. Discussion 57
5.1 Overview of Findings 57
5.2 Why the Situational Uncertainty Manipulation Did Not Produce Clear Group Differences 57
5.3 Subjective Situational Uncertainty, Manageability, and State Anxiety 58
5.4 State Anxiety and Critical Evaluation of GenAI Reliability 59
5.5 Rationalization as a Common but Varied Meaning-Making Process 60
5.6 From Whether Users Rationalize to How They Rationalize the Response 61
5.7 The Value of GenAI Response Depends on Users’ Needs 62
6. Conclusion 64
6.1 Limitations and Future Directions 65
References 67
Appendix 73
Appendix A: Experimental Materials 73
Appendix B: Pilot Study Questionnaire 75
Appendix C: Main Study Questionnaire 80
Alquist, J. L., & Baumeister, R. F. (2024). Dealing with uncertain situations. The Journal of Positive Psychology, 19(6), 923–946. https://doi.org/10.1080/17439760.2023.2282781
Arbona, C., Fan, W., Phang, A., Olvera, N., & Dios, M. (2021). Intolerance of Uncertainty, Anxiety, and Career Indecision: A Mediation Model. Journal of Career Assessment, 29(4), 699–716. https://doi.org/10.1177/10690727211002564
Bromberg-Martin, E. S., & Monosov, I. E. (2020). Neural circuitry of information seeking. Current Opinion in Behavioral Sciences, 35, 62–70. https://doi.org/10.1016/j.cobeha.2020.07.006
Carleton, R. N. (2016). Into the unknown: A review and synthesis of contemporary models involving uncertainty. Journal of Anxiety Disorders, 39, 30–43. https://doi.org/10.1016/j.janxdis.2016.02.007
Chang, Y., Wang, X., Wang, J., Wu, Y., Yang, L., Zhu, K., Chen, H., Yi, X., Wang, C., Wang, Y., Ye, W., Zhang, Y., Chang, Y., Yu, P. S., Yang, Q., & Xie, X. (2024). A Survey on Evaluation of Large Language Models. ACM Transactions on Intelligent Systems and Technology, 15(3), 1–45. https://doi.org/10.1145/3641289
Chatterji, A., Cunningham, T., Deming, D., Hitzig, Z., Ong, C., Shan, C., & Wadman, K. (2025). How People Use ChatGPT. https://doi.org/10.3386/w34255
Chen, S., Yao, N., & Qian, M. (2018). The influence of uncertainty and intolerance of uncertainty on anxiety. Journal of Behavior Therapy and Experimental Psychiatry, 61, 60–65. https://doi.org/10.1016/j.jbtep.2018.06.005
Croes, E. A. J., Antheunis, M. L., Van Der Lee, C., & De Wit, J. M. S. (2024). Digital Confessions: The Willingness to Disclose Intimate Information to a Chatbot and its Impact on Emotional Well-Being. Interacting with Computers, 36(5), 279–292. https://doi.org/10.1093/iwc/iwae016
De Graaf, M. M. A., & Malle, B. F. (2019). People’s Explanations of Robot Behavior Subtly Reveal Mental State Inferences. 2019 14th ACM/IEEE International Conference on Human-Robot Interaction (HRI), 239–248. https://doi.org/10.1109/HRI.2019.8673308
De Graaf, M. M., & Malle, B. F. (2017). How People Explain Action (and Autonomous Intelligent Systems Should Too). AAAI Fall Symposia, 8.
Emma, R. (2025, July 22). OpenAI says ChatGPT users send over 2.5 billion prompts every day. https://www.theverge.com/news/710867/openai-chatgpt-daily-prompts-2-billion
Filiz, I., Judek, J. R., Lorenz, M., & Spiwoks, M. (2023). The extent of algorithm aversion in decision-making situations with varying gravity. PLOS ONE, 18(2), e0278751. https://doi.org/10.1371/journal.pone.0278751
Grice, H. P. (1975). Logic and conversation. In Speech acts (pp. 41–58). Brill.
Grupe, D. W., & Nitschke, J. B. (2013). Uncertainty and anticipation in anxiety: An integrated neurobiological and psychological perspective. Nature Reviews Neuroscience, 14(7), 488–501. https://doi.org/10.1038/nrn3524
Gu, Y., Gu, S., Lei, Y., & Li, H. (2020). From Uncertainty to Anxiety: How Uncertainty Fuels Anxiety in a Process Mediated by Intolerance of Uncertainty. Neural Plasticity, 2020, 1–8. https://doi.org/10.1155/2020/8866386
Gulati, A., & Oliver, N. (2026). Why do we Trust Chatbots? From Normative Principles to Behavioral Drivers (arXiv:2602.08707). arXiv. https://doi.org/10.48550/arXiv.2602.08707
Hall, D. (2019, December 10). The ELIZA Effect—99% Invisible. https://99percentinvisible.org/episode/the-eliza-effect/
Hartley, C. A., & Phelps, E. A. (2012). Anxiety and Decision-Making. Biological Psychiatry, 72(2), 113–118. https://doi.org/10.1016/j.biopsych.2011.12.027
Jones, E. (1908). Rationalization in every-day life. The Journal of Abnormal Psychology, 3(3), 161–169. https://doi.org/10.1037/h0070692
Kosmyna, N., Hauptmann, E., Yuan, Y. T., Situ, J., Liao, X.-H., Beresnitzky, A. V., Braunstein, I., & Maes, P. (2025). Your brain on ChatGPT: Accumulation of cognitive debt when using an AI assistant for essay writing task. arXiv Preprint arXiv:2506.08872, 4.
Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind:" Seizing" and" freezing.". Psychological Review, 103(2), 263.
Lee, E., Pataranutaporn, P., Amores, J., & Maes, P. (2024). Super-intelligence or Superstition? Exploring Psychological Factors Influencing Belief in AI Predictions about Personal Behavior (arXiv:2408.06602). arXiv. https://doi.org/10.48550/arXiv.2408.06602
Lee, H.-P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025). The impact of generative AI on critical thinking: Self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers. Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems, 1–22.
Loewenstein, G. (1994). The psychology of curiosity: A review and reinterpretation. Psychological Bulletin, 116(1), 75.
Maitlis, S., Vogus, T. J., & Lawrence, T. B. (2013). Sensemaking and emotion in organizations. Organizational Psychology Review, 3(3), 222–247. https://doi.org/10.1177/2041386613489062
Malle, B. F. (2022). Attribution Theories: How People Make Sense of Behavior. In D. Chadee (Ed.), Theories in Social Psychology, Second Edition (1st ed., pp. 93–120). Wiley. https://doi.org/10.1002/9781394266616.ch4
Marteau, T. M., & Bekker, H. (1992). The development of a six‐item short‐form of the state scale of the Spielberger State—Trait Anxiety Inventory (STAI). British Journal of Clinical Psychology, 31(3), 301–306. https://doi.org/10.1111/j.2044-8260.1992.tb00997.x
McBain, R. K., Cantor, J. H., Breslau, J., Diliberti, M., Zhang, L. A., Zhang, F., Burnett, A., Kofner, A., Rader, B., Pataranutaporn, P., Stein, B. D., Mehrotra, A., & Yu, H. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015
McGrath, M. J., Lack, O., Tisch, J., & Duenser, A. (2025). Measuring trust in artificial intelligence: Validation of an established scale and its short form. Frontiers in Artificial Intelligence, 8, 1582880. https://doi.org/10.3389/frai.2025.1582880
Miehling, E., Nagireddy, M., Sattigeri, P., Daly, E. M., Piorkowski, D., & Richards, J. T. (2024). Language Models in Dialogue: Conversational Maxims for Human-AI Interactions. Findings of the Association for Computational Linguistics: EMNLP 2024, 14420–14437. https://doi.org/10.18653/v1/2024.findings-emnlp.843
Morriss, J., Tupitsa, E., Dodd, H. F., & Hirsch, C. R. (2022). Uncertainty Makes Me Emotional: Uncertainty as an Elicitor and Modulator of Emotional States. Frontiers in Psychology, 13, 777025. https://doi.org/10.3389/fpsyg.2022.777025
Nourani, M., King, J., & Ragan, E. (2020). The Role of Domain Expertise in User Trust and the Impact of First Impressions with Intelligent Systems. Proceedings of the AAAI Conference on Human Computation and Crowdsourcing, 8, 112–121. https://doi.org/10.1609/hcomp.v8i1.7469
Okay-Somerville, B., & Scholarios, D. (2022). Focused for Some, Exploratory for Others: Job Search Strategies and Successful University-to-Work Transitions in the Context of Labor Market Ambiguity. Journal of Career Development, 49(1), 126–143. https://doi.org/10.1177/08948453211016058
Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1(4), 515–526. https://doi.org/10.1017/S0140525X00076512
Rosen, N. O., & Knäuper, B. (2009). A Little Uncertainty Goes a Long Way: State and Trait Differences in Uncertainty Interact to Increase Information Seeking but Also Increase Worry. Health Communication, 24(3), 228–238. https://doi.org/10.1080/10410230902804125
Skjuve, M., Følstad, A., & Brandtzaeg, P. B. (2023). The User Experience of ChatGPT: Findings from a Questionnaire Study of Early Users. Proceedings of the 5th International Conference on Conversational User Interfaces, 1–10. https://doi.org/10.1145/3571884.3597144
Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A. (1983). Manual for the State-Trait Anxiety Inventory. (Original work published Consulting Psychologists Press)
Swaroop, S., Buçinca, Z., Gajos, K. Z., & Doshi-Velez, F. (2024). Accuracy-Time Tradeoffs in AI-Assisted Decision Making under Time Pressure. Proceedings of the 29th International Conference on Intelligent User Interfaces, 138–154. https://doi.org/10.1145/3640543.3645206
Waytz, A., Gray, K., Epley, N., & Wegner, D. M. (2010). Causes and consequences of mind perception. Trends in Cognitive Sciences, 14(8), 383–388. https://doi.org/10.1016/j.tics.2010.05.006
Webster, D. M., & Kruglanski, A. W. (1994). Individual differences in need for cognitive closure. Journal of Personality and Social Psychology, 67(6), 1049.
Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. (2005). Organizing and the Process of Sensemaking. Organization Science, 16(4), 409–421. https://doi.org/10.1287/orsc.1050.0133
Whitson, J. A., & Galinsky, A. D. (2008). Lacking Control Increases Illusory Pattern Perception. Science, 322(5898), 115–117. https://doi.org/10.1126/science.1159845
Wu, C. M., Schulz, E., Pleskac, T. J., & Speekenbrink, M. (2022). Time pressure changes how people explore and respond to uncertainty. Scientific Reports, 12(1), 4122. https://doi.org/10.1038/s41598-022-07901-1
何瑞萍. (2014). Dervin與Weick意義建構理論之分析與比較. 大學圖書館, 18(1). https://doi.org/10.6146/univj.18-1.05
兒童福利聯盟. (2025). 2025年台灣青少年心理健康調查報告. 兒童福利聯盟. https://www.children.org.tw/publication_research/research_report/mentalhealth2025
鍾思嘉. (1982). 修訂情境焦慮量表簡介. 測驗與輔導, 10(6), 789–790.
