人工智能工作績效期待與滿意度關係之研究 -期望落差或安慰劑效果;Study on the Relationship between Performance Expectations and Satisfaction with Artificial Intelligence: Expectancy Disconfirmation or Placebo Effect

NCUIR > School of Management at National Central University > Graduate Institute of Business Administration > Electronic Thesis & Dissertation > Item 987654321/97509

Please use this identifier to cite or link to this item: https://ir.lib.ncu.edu.tw/handle/987654321/97509

Title:	人工智能工作績效期待與滿意度關係之研究 -期望落差或安慰劑效果;Study on the Relationship between Performance Expectations and Satisfaction with Artificial Intelligence: Expectancy Disconfirmation or Placebo Effect
Authors:	游凱亘;Yu, Kai-Hsuan
Contributors:	企業管理學系
Keywords:	期望落差;安慰劑效應;人工智能;期望;感知品質;滿意度;Expectancy Disconfirmation;Placebo Effect;Artificial Intelligence;Expectations;Perceived Quality;Satisfaction
Date:	2025-06-12
Issue Date:	2025-10-17 11:27:38 (UTC+8)
Publisher:	國立中央大學
Abstract:	本研究探討了⼈⼯智能（AI）在⼯作績效中的期望與滿意度之間的關係，並特別關注期望落差（Expectancy Disconfirmation）與安慰劑效應（Placebo Effect）。在現今社會中，AI 技術應⽤範圍廣泛，從簡單的⽇常任務到複雜的專業領域，⽣成式 AI 更是顯著改變了⼈們的⽣活。然⽽，社會對 AI 的期望及其滿意度存在差異，尤其是在相同的績效表現下，⼈們對 AI 和⼈類的評價標準往往不同。本研究的主要⽬的是探討在不同情境下，受試者對 AI 與⼈類的期望、感知品質及滿意度的關係，並檢驗期望落差和安慰劑效應的存在。研究⽅法包括設計實驗問卷，將受試者分為兩組，分別告知績效的完成對象為⼈類或是 AI，在兩種情境中進⾏測試。研究結果顯⽰，在兩種情境中，當受試者對 AI 的期望較⾼時，能顯著提⾼其感知品質，這反映了安慰劑效應的存在；然⽽，對⼈類專家的⾼期望則未能顯著提升感知品質，未觀察到安慰劑效應，進⼀步⽀持了期望落差的理論。此外，本研究也發現，期望對滿意度的直接影響在所有情境下均不顯著，但感知品質對滿意度的影響則具有顯著的正向關聯。這些發現有助於更好地理解 AI 技術應⽤中的⼼理機制，並為未來改進 AI 技術應⽤提供了有價值的⾒解。;This study investigates the relationship between expectations and satisfaction regarding the work performance of Artificial Intelligence (AI), with a particular focus on Expectancy Disconfirmation and the Placebo Effect. In today’s society, AI technologies are widely applied across various domains, ranging from simple daily tasks to complex professional fields. The emergence of generative AI has significantly transformed people’s lives. However, a gap remains between societal expectations of AI and the satisfaction derived from its performance—especially when AI and humans achieve similar performance outcomes, yet are evaluated by different standards. The main objective of this study is to examine how participants’ expectations, perceived quality, and satisfaction differ between AI and human performers under different scenarios, and to test the presence of expectancy disconfirmation and the placebo effect. An experimental questionnaire was designed in which participants were randomly divided into two groups. Each group was told that the performance in the scenario was completed either by a human or an AI, and their responses were measured across two different tasks. The results indicate that in both scenarios, high expectations of AI significantly increased perceived quality, suggesting the presence of the placebo effect. Conversely, high expectations of human experts did not significantly enhance perceived quality, and no placebo effect was observed—supporting the theory of expectancy disconfirmation. Additionally, the study found that while expectations had no significant direct effect on satisfaction in any scenario, perceived quality had a significant positive effect on satisfaction. These findings contribute to a better understanding of the psychological mechanisms behind AI application and offer valuable insights for improving the integration and acceptance of AI technologies in various fields.
Appears in Collections:	[Graduate Institute of Business Administration] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	22	View/Open

社群 sharing

Loading...