隨著組織環境快速變遷，科技日新月異以及資訊的快速流通，組織將面臨更多的互相衝突的挑戰，尤其是在人力資源管理上。舉個例子來說，部屬期待領導者能依據他們的能力和專業來分派任務，但部屬同時也期待領導者能分配一樣的工作量給所有的部屬，而能夠化解這樣的矛盾衝突的領導者，被稱作矛盾領導人，未來組織將會需要更多能夠解決互相矛盾衝突挑戰等的領導人。在2015年由Zhang et al. 等人提出衡量矛盾領導者行為的量表，但此量表是採用雙管問題 (Double-barreled question) 的設計，一般在建立量表時，會避免使用雙管問題，因為此設計容易混淆填答並產生謬誤。同時，這樣的設計在研究矛盾領導者對於部屬行為影響上，也有可能忽略了 “either/or” 領導者對於部屬的影響，在2018年Wang et al. 等人發現 “both/and” 領導者和”either/or”領導者對於部屬在績效的影響效果是一樣，這與Zhang et al. 等人做出的研究是不一樣的。 本研究的目的是在於解決雙管問題量表所產生的謬誤問題、重新檢驗矛盾領導領為對於部屬熟練度行為的影響及探討 “both/and” 領導者和 “either/or” 領導者者對於部屬熟練度行為的領導，究竟是 “both/and” 領導者對於部屬的熟練度行為影響比較大呢？還是 “either/or” 領導者對於部屬的熟練度行為影響比較大呢？本研究將會把 Zhang et al. 等人所建置雙管問題，拆成單一行為問題，並透過多項式回歸將領導者行為分成 “both/and” 領導者、 “either/or” 領導者及 “neither/nor” 領導者，這三種領導風格對於部屬熟練度行為的影響，然後在運用反應曲面分析法進一步探討哪一種領導風格會對部屬的熟練度行為產生比較大的影響 。最後，我們發現 “both/and” 領導者對於部屬熟練度行為的影響比 “neither/nor” 領導者大， “either/or” 領導者有的時候對於部屬熟練度行為的影響會比 “both/and” 領導者還要好。 ;As organizational environments are becoming increasingly more dynamic, complex, and competitive, leaders are often confronted with contradictive demands or tension in people management (Y. Zhang, Waldman, Han, & Li, 2015). The leaders, who can resolve these tensions, are paradox leader. In prior research, Zhang et al. (2015) has established a double-barreled questionnaire to measure paradoxical leadership behaviors. Due to nature of double-barreled questions, the questionnaire may cause confusion and ambiguity to raters which may ignore the effect of “either/or” perspective while examine the effect on subordinates’ proficient behavior. Another research also shown that “both/and” leadership and “either/or” leadership has the same influence on the performance of subordinates. The purpose of this research is to resolve the issue cause by double-barreled question items, reexamining the effect of paradoxical leadership behavior on proficient behavior of subordinates, and explore the effect of “both/and” leadership and “either/or” leadership on subordinates’ proficient behavior. We detach the double-barreled question items in Zhang et al. (2015)’s measurement and employ polynomial regression with response surface to reexamine the effect of paradox leader on subordinates’ proficient behavior. We also further to compare the effect of each leadership style on proficient behavior of subordinates on response surface. As result, we found “both/and” leadership has better effect on subordinates’ proficient behavior than “neither/nor” leadership and “either/or” leadership sometimes may have greater influence than “both/and” leadership on subordinates’ proficient behavior.
