透過您的圖書館登入
IP:18.222.182.105
  • 學位論文

文書編輯技能測驗試題難度之估計方法

指導教授 : 何榮桂
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


摘 要 本研究以電腦軟體應用丙級技術士檢定的術科試題為例,探討評估樣品實作評量試題難度的方法,研究目的有二: 一、分析試題難度與測驗元素數量的相關情形。 二、根據研究結果,建構乙套可行之實作評量的試題難度評估方法,以作為教師教學與研究之參考。 本研究之教學實證採準實驗研究法,以高職二、三年級學生為樣本,方便取樣選取94名曾參加丙級檢定的三年級學生。實驗初期以「文書編輯測驗元素實作題」進行文書編輯應用測驗作為研究基礎;另取樣即將參加丙級檢定之157名二年級學生,進行十八週之「文書編輯軟體應用教學」。於課程結束後,以「電腦丙檢術科第二站題庫」進行測驗,並填寫「試題難度調查問卷」,來瞭解實作評量應用於文書編輯技能測驗的實際情況。 研究結論如下: 一、電腦丙檢術科中30題文書編輯實作試題的難度值相近。 二、實作試題難度值與測驗元素個數呈高度負相關。 三、實作試題難度值與測驗元素預試的難度平均值零相關。

並列摘要


Abstract The objective of this work was to study and evaluate the difficulty estimation methodology of work sample performance assessment by examining items of computer software application skill test (CSAST) for level-3 technician. Two major purposes of this study were, 1. To analyze the relationship between the difficulty of items and numbers of test components. 2. To develop the difficulty estimation methodology for items of article edit skills for a practical performance assessment to help teachers’ instruction and research. The quasi-experimental approach was adopted in this study and samples of study were based on 2nd and 3rd year of senior high school students. First of all, 94 3rd year students who had taken the level-3 CSAST were chosen based on sample of conveniences to take work sample performance test with “article edit skill test components”. Secondly, 157 2nd year students who planned to take the level-3 CSAST were sampled and trained with “article edit software” course for 36 hours. More over, those 157 students continued to take all 30 CSAST items and answer the questionnaires of “difficulty estimation”. All test results were evaluated and analyzed statistically to understand the impact of work sample performance assessment to article edit skill test. Three conclusions are resulted from above tests and analyses were, 1. 30 CSAST items showed the similar difficulty. 2. A significantly negative correlation was observed between “difficulty of work sample performance items” and “numbers of test components”. 3. A zero correlation was observed between “difficulty of work sample performance items” and “pretest difficulty of test components”.

延伸閱讀