在高科技產業中，創新工作行為被視為組織維持競爭優勢的關鍵。本研究以資源保存理論為基礎，建構一個涵蓋工作倦怠、工作投入、心理抽離與創新工作行為之日內資源互動模型，並區分上午與下午兩個時段，探討資源於不同時間點的動態變化歷程。本研究採用經驗取樣法，以87位高科技產業全職員工為對象，連續十個工作天每日施測四次。研究變項包括工作倦怠（獨變項）、工作投入（中介變項）、心理抽離（調節變項）與創新工作行為（依變項），控制變項則涵蓋恢復狀態、工作負荷與組織創新氛圍等。資料分析採用階層線性模式進行假設驗證。研究結果顯示：（1）工作倦怠與創新工作行為之關係具有時段差異，上午不具顯著關聯，下午則呈現顯著正向效果；（2）工作投入於上午不具中介效果，在下午顯著中介工作倦怠與創新工作行為之間的關係；（3）心理抽離於任一時段皆未顯著調節工作倦怠與工作投入之關聯。結果顯示，在日內資源逐步耗損的歷程中，員工可能透過工作投入啟動補償性資源配置以維持創新表現，而心理抽離的恢復效能則可能受限於實際執行程度或心理抽離水準。本研究假設未獲實徵支持，顯示相關理論在特定情境下可能呈現不同作用模式，需進一步釐清其作用條件與可能的調節因素。實務層面亦應重視情境差異，避免將單一方案設計套用於工作環境。最後，針對研究內容提出研究限制與後續研究建議。

In the high-tech industry, innovative work behavior is regarded as a key factor for maintaining organizational competitiveness. Grounded in Conservation of Resources (COR) theory, this study proposes a within-day resource interaction model involving job burnout, job involvement, psychological detachment, and innovative work behavior, while distinguishing between morning and afternoon time segments to explore the dynamic patterns of resource fluctuation across the workday. Using the Experience Sampling Method (ESM), data were collected from 87 full-time employees in the high-tech industry, who completed four surveys per day across ten consecutive workdays. The main study variables included job burnout (independent variable), job involvement (mediator), psychological detachment (moderator), and innovative work behavior (dependent variable), with recovery state, job demands, and organizational innovation climate as control variables. Multilevel modeling was employed to test the proposed hypotheses. The results revealed that: (1) the relationship between job burnout and innovative work behavior varied by time of day—non-significant in the morning but significantly positive in the afternoon; (2) job involvement did not mediate the relationship in the morning, but served as a significant mediator in the afternoon; (3) psychological detachment did not moderate the relationship between job burnout and job involvement at either time point. These findings suggest that as resources are gradually depleted throughout the day, employees may adopt compensatory strategies by investing engagement to sustain innovative performance, while the effectiveness of psychological detachment as a recovery mechanism may be limited by its level or feasibility of implementation. The hypotheses in this study were not supported by empirical evidence, suggesting that the underlying theoretical mechanisms may operate differently under specific contextual conditions. Further research is needed to identify the conditions under which these theories hold and to explore possible moderating factors. In practice, it is also important to account for contextual differences and avoid applying uniform solutions across diverse work settings. Limitations and future research directions are discussed.