人才選拔乃是透過各種不同測驗媒介，吸引有興趣者前來應試，藉以選拔出合於業務需要之人才，以達適才適所目的之一系列活動。因此，如何善用不同遴選方法以選用不同種類人才；如何將人員的特性(如能力、經驗等)與工作職位的要求，做最適切的搭配，亦即「選賢任能」，考試方式之選用為主要關鍵。鑑於我國現行公務人員考選方式計有筆試、口試、測驗、實地考試、審查著作發明等多種，其中筆試及口試二種更是最常用的方法，前者係在測量應考人之學識能力，其實施過程只須命擬試題，試畢評閱試卷計分，雖簡便，惟無法測出一個人的人格、修養、品性、應對能力及處世態度。而「口試」適能彌補筆試無法測量的缺點，故對部分職務除施以筆試外，另須輔以口試，方能測出應考人的真正能力，選出所欲人才。目前我國公務人員考試實施口試者，係參照民國七十六年訂頒之「公務人員考試口試規則」相關規定辦理口試有關事宜，實施以來，作業尚稱嚴謹，惟社會大眾對於口試之評定因素、分數所佔比重及其實施過程等方面，仍多所批評，為確保口試之客觀及公平，以發揮其功能，遂依以口試之原理，探討我國公務人員考試加採口試者，其口試實施規定予以通盤檢討，同時參酌外國採用口試選才制度之特點及長處，以收「他山之石，可以攻錯」之旨，期使我國口試制度益臻完善。

應徵者處於甄選面談的情境或多或少都會經驗到一定程度的焦慮，過去研究指出焦慮的確會對應徵者行為表現，以及面試官評價產生影響，本研究回顧文獻證據提出焦慮情緒狀態或許應視為造成面試官評價偏誤的來源之一；由情感事件理論的觀點，探討「氣氛營造」與「溫暖行為」能否降低應徵者焦慮情緒狀態對於應徵者行為的影響；以及探討面試官的「面談經驗」和「歸因傾向」對於應徵者行為和面試官評價的調節效果。本研究採取模擬面談法（mock interview），邀請一年內即將謀職的應徵者，以及具面談經驗的業界人士擔任面試官，進行一對一的甄選面談，蒐集應徵者和面試官雙方的配對資料，共計94筆。統計分析結果顯示「面談焦慮情緒狀態」對於「面試官評價」的影響，係完全透過應徵者「焦慮行為」所中介。並未如預期發現面試官「溫暖行為」與「氣氛營造」等行為，削弱應徵者焦慮情緒對其行為的影響，以及「面談經驗」和「歸因傾向」等面試官個人特性調節焦慮行為與面試官評價的負向關係。本研究分別針對應徵者和面試官雙方提出實務建議，以及提出未來可能的研究方向。 / Prior studies found that applicants’ anxious state influence applicants’ behaviors and in turn interviewer evaluations. The present study argues that anxiety may be one of transient errors biasing interviewer evaluations. Taking interpersonal skills and trait anxiety specific to employment interview into account, this study examines the relationship among anxious emotional state, applicants’ behaviors, and interviewer evaluation. Based upon affective event theory as overarching framework, I propose that rapport building and warm behavior of interviewers as affective events may generate applicants’ positive emotion mitigates the negative influence of anxiety state on applicants’ behavior. In addition, prior studies suggested that interview experience and attribution preferences influence information processes of interviewers. These interviewers’ characteristics may minimize the negative impact of applicants’ anxious behavior on interviewer evaluation. Mock interview with 94 perspective applicants at the time in pursuing jobs and 37 EMBA students as interviewer were recruited from 9 universities. Results of regression analyses showed that applicants’ anxious behavior mediate the relationship between anxiety state and interviewer evaluation. However, the moderating effects of rapport building and warm behavior on the relationship between interviewees’ anxiety state and behaviors did not receive support. The effects of applicants’ anxious behaviors on interviewer evaluation do not vary as function of interview experience and attribution preference. Discussions and future suggestions were provided.