資料處理及下結論技能之評量模式分析研究 作者:許榮富(國立臺灣師範大學物理學系)

摘要：

測驗之平行(parallel)條件，必須滿足高相關及無差異性，本研究在此設計要求下，旨在探討不同模式(format)之評量工具，對於資料處理及下結論等二項技能之評量的表現(performance)。研究者選擇標準化之中文TIPS(II) (The Test of Integral Process Skills II)，作為外在效標(external criterion)，修正「科學過程技能學習層次測驗」（選擇及問答方式之）Cronbach α值分別為0.62及0.68），並據以發展平行之開放式及引導式之實作測驗工具。
本研究以大臺北地區之國中學生為研究對象。第一次抽取2,705人，施以TIPS(II)測驗(α=0.84)，並以該測驗之表現將學生分為上、中、下三 群，進行二次取樣，得開放式實驗組、引導式實驗組及控制組等三組各129人，分別先後實施問答、實作及選擇等測驗。三組之TIPS(II)平均值分別 為：27.28、26.84、26.89。
根據受試學生群於外在效標(TIPS(II))之表現，定義出預期的差異，又配合各模式測驗表現的差異及相關情形，發現資料處理技能於論文方式或開放式實 作測驗中表現最為一致；下結論技能於問答方式測驗中的表現較為一致。該結果與我國國小及國中科學實驗教材之結構分析(SLSI-Science Laboratory Structual Inventory)分析結果間具有一致性。
根據本研究之發現及解釋，針對資料處理及下結論技能的評量、科學實驗教材及進一步之研究，提出多項建議。

《詳全文》

An Analysis of the Assessment of Science Laboratory Skills: The Performance in Parallel Format Instruments of Data Processing and Concluding Skills Author: Rong-Fu Hsu

Abstract：

The different instruments in parallel format were constructed in assessing the students performance in laboratory skills in this study. The parallel tests meet the requirements of high correlation and with no difference.
The standardized TIPS(II) was chosen as external criterion. Also, we revised "the Learning Hierarchy Test in Science Process Skill" (Cronbach's a is 0.62 and 0.68 for choice and essay forms respectively). Furthermore, we developed the corresponding practical tests: open-ended and guided format.
The subjects were sampled from junior high school students in Taipei. At first, 2,705 pupils were selected by using stratified sampling. According to the students' performance on TIPS(II) (α=0.84), we divided the subjects into three groups: upper, middle, and lower group. Finally, we got 129 pupils for each group: open-ended, guided and controlled. Each group took the essay format test, practical test, and choice format test. The performances of these three groups on TIPS(II) with average were 27.28, 26.84, 26.89 respectively.
We defined the predicted differences, and the differences and correlations of these tests with the external criterion (TIPW(II)). We found the most consistent in the essay and open-ended practical test among the four different formats in data processing skill. However, we found more consistent in the essay format than in the others in the concluding investigation skill. The findings were consistent with SLSI (Science Laboratory Structure Inventory) of the current science laboratory test. The recommendations for assessing the science laboratory skills were then suggested.