基本科學過程技能測驗 簡化版之信度與效度分析

Abstract

本研究主要目的是在發展基本科學過程技能測驗的精簡版，根據採用項目分析結果將原本60題的基本科學過程技能測驗刪除部分鑑別度較差的題目，並修改少數題意不夠明確的題目。精簡版包括觀察、分類、溝通、測量、推論、預測六項基本科學過程技能，每項過程技能有4個題目，共有24題，雖然題數僅有原測驗的40%，但以正式樣本國中二年級學生的答題資料所得到的α值為.798，且驗證性因素分析結果顯示主要整體模式適配度指標(RMSEA=0.026，CFI=0.959，TLI=0.952)均達理想標準，且與國中生物及理化在校成績之相關(均為.58)都有達顯著之水準，明顯高於與國文及數學的相關(分別為為.23及.19)，顯示精簡版測驗仍具有良好的信效度。精簡版測驗施測所需時間甚短，信效度也可接受，更具實用性。

The main purpose of this study is to develop a simplified version of the Basic Science Process Skills Test. According to the results of item analysis to remove the original 60 questions of the basic scientific process skills test to eliminate some of the poorly identified questions and modifies a few questions that are not clearly defined. The simplified version includes observation, classification, communication, measurement, inference, and prediction of six basic science process skills, each with four questions and a total of 24 questions. Although the number of questions is only 40% of the original test, the alpha value obtained from the second-grade students in junior high schools is .798. The confirmatory factor analysis showed that the main overall model fit index (RMSEA=0.026, CFI=0.959, TLI=0.952) reached the ideal standard. And the correlation with biological scores and physical and chemical achievements (both .58) has reached a significant level, which is significantly higher than that related to chinese and mathematics (.23 and .19, respectively), showing that the simplified version test still has good reliability and validity.Because the test takes a very short time, it is more practical