「再結晶」技能學習達成階層之鑑別性選擇式試題設計

A Design of Multiple-Choice Test for Assessing Levels of Attaining Recrystallization Technique

王澄霞(Cheng-Hsia Wang) ； 洪志明(Jhy-Ming Horng)

《化學》 48卷4期 (1990/12) Pp. 327-336

https://doi.org/10.6623/chem.1990042

摘要

目前一般教師對於化學實驗操作能力之評價，大部分憑教師本人之主觀印象給分，缺乏客觀的評量標準，因而往往對同一實驗操作能力，不同評量者所評得的分數有相當大的差異。本研究的目的在探討如何根據一些特定的評量準則，來設計再結晶實驗技能的選擇式評量試題。這些評量的準則指出學生應該做些什麼，如何做，何時做及為何如此做，因此可正確地評量出學生的能力並診斷學生所學到的和沒學到的，同時指出如何評量學生達到具體、辨識和歸納三個學習達成階層中的那一階層。再結晶是一種多重工作，不能只靠一次指定的實驗活動就能評量出來。它必需用一套具代表性的合理數目之特別設計的評量實驗活動來完成。所以再結晶技能的評量通常要包含多種評量模式：量表、實作測驗、面談、紙筆測驗和實驗報告等。這種統合評量方式費時費力，且執行時相當困難。本研究設計的選擇式試題，省時省力，易於執行，可以測出學生的操作之正確性和其心智運作，而用來替代上述的評量方法。此選擇式評量試題，經實測後，與(1)量表成績（再歸納階層有85%）和(2)實驗的成果有高的正相關。測試成績高者，所得的晶體產率和純度都高，其正相關程度，在具體階層有54.5%，在辨識階層有68.2%，在歸納階層為100%，顯示此試題具有鑑別性，能評量出學生所達到的學習階層。

並列摘要

Existing current teachers' assessments of students' manipulative abilities are largely based on impression grading without reference to specific assessment criteria, and lead to considerable differences in grades awarded by teachers for the same practical performance. The purpose of this work is to show how to construct a multiple-choice test items based on specific assessment criteria which point out what, how, why. and when the learner must do, so that the tool can give an accurate grade, diagnose what has or has not been learnt, and assess students' levels of attaining recrystallization technique. i.e., concrete, identity, and classificatory levels. Recrystallization is a multiple task which cannot be measured by only one assignment. It must be measured by a rational set of new assignments. The assessment instrument for recrystallization usually contains several assessment modes: check-list, practical test, clinical interview, and paper-pencil test. Such assessment takes too much time and is hard to administer. In this study a multiple-choice test is designed for use in the assessment as a substitute of the instrument mentioned. The multiple-choice test is inexpensive, less time consuming, and good for testing mental discrimination. This study reveals that students' performances in the multiple-choice test designed are found highly related to (1) per- formances in the check-list (85% in classificatory level) and (2) performances in the assigned task. Students with high scores in the multiple-choice test indeed obtained the crystals in high purity and high yield in the assigned task. The degrees of positive correlation are 54.5% in concrete level, 68.2% in identity level, and 100% in classificatory level.