本文主旨為探討桃園國際機場過去9 年來準點率 (On-Time Performance) 的變化。本文蒐集民國98-106 年間跑道、滑行道施工關閉次數、顯著危害天氣、航管因素等資料，進行實證分析。在利用多元迴歸法進行初步分析後，再以固定效果模型 (Fixed Effect Model) 和隨機效果模型(Random Effect Model) 進行分析，並從中選出適合者。結果顯示，跑道、滑行道施工關閉次數、顯著危害天氣、航管因素與準點率均呈負相關。最後以單跑道作業開始作為結構差異檢定的時間點，使用鄒檢定 (Chow Test)進行檢定。在縱橫資料迴歸分析 (Panel Data Regression) 中，Hausman Test檢定後以固定效果模型為佳，航管因素則影響機場準點率最嚴重。

The main purpose of this thesis is to investigate the changes in on-time performance of Taoyuan International Airport over the past nine years. The empirical analysis uses the data on runway / taxiway construction closure, significant meteorological information and air traffic control delay from 2009 to 2017. After a preliminary analysis using the multi-factor linear regression method, our panel data analysis shows that the notice to airmen, significant meteorological information and air traffic control delay were significantly negatively correlated with on-time performance. The Chow test is used to examine whether there is a structural change at the beginning of the single runway outage. The Hausman test selects the fixed effects model compared with the random effects model. And the air traffic control delay is found to be the most significant factor that influences the airport on-time performance.