本研究針對大學入學考試中心所發展的標準化空間能力測驗常模樣本資料，進行試題探索與模式驗證，以瞭解此測驗中的試題品質與因素結構。本研究樣本為臺灣地區公私立高中1838位高一學生，以隨機方式分割為建模（或探索）樣本與驗證樣本。研究中使用項目分析、內部一致性皆析、探索性因素分析以及驗證性因素分析等統計方式進行分析。 　　本研究結果顯示，探索後的建議版測驗不但具有良好的信度與效度，測驗的測量模式與觀察資料獲得適配，也能以更少的試題獲致更好的測量精確度。本研究的測驗建模與驗證程序可為後續相關測驗發展的參考程序。

　　The study explored and confirmed the standardized spatial ability test developed by the College Entrance Examination Center (CEEC) to understand the test’s item quality and factor structure. The norming sample consisted of 183810th graders from Taiwan public and private high schools, and was randomly divided into calibration (exploration) sample and validation sample in the study. Item analysis. Research results showed that the suggested version of the spatial ability test can improve test reliability and validity with fewer test items, and the measurement model fits well. The model exploration and validation steps implemented in the study can provide a referential procedure for future similar test constructions for the purpose of model reduction.