本研究採Rasch測量模式，分析教師自編科學成就測驗，以檢驗測驗試題的品質，並提供 有用的評量回饋訊息，作為教師調整教學與修正評量之參考。研究對象為高雄市8,936 位（男 生4,617 位、女生4,319位）七年級學生。研究工具為教師自編生物科成就測驗。測驗資料採 試題反應理論的Rasch模式進行分析。結果發現：整體而言，教師自編科學成就測驗能區分出 學生的能力程度。然而，有些試題的內容對女生較不公平。本研究也發現，教師對於部分試 題難易度的認知與學生實際的作答反應存有若干差異。最後，本研究透過評量回饋圖表，包 含學生能力與試題難度對照圖、學生作答診斷圖等，提出辨識學習基礎發展區和學習可能發 展區的範圍之方式，作為教師規劃學生學習進程之參考。

This study examined the quality of a teacher-made achievement test using Rasch analyses and provided useful assessment feedback information for improving teaching and testing. The participants were 8,936 seventh-grade students (4,617 boys and 4,319 girls) from the Kaohsiung area in Taiwan. Data were collected through a teacher-made biology achievement test and analyzed using the Rasch model. The results indicated that the test clearly discriminated students’ abilities. Some questions, however, were found to have unfair content for girls. Additionally, some mismatches were identified between teachers’ expectations and students’ assessments of question difficulty. This paper also provides assessment feedback diagrams, including a person–item map and diagnostic map, and a method of identifying “the zone of fundamental learning development” and “the zone of proximal learning development” is proposed to enable teachers to plan students’ learning progression.