統整科學過程技能測驗簡化版之信度與效度分析

Integrated Science Process Skills Test Reliability and Validity of the Simplified Version

葉豐榮(Feng-Rung Yeh) ； 蔣光明(Kuang-Ming Chiang) ； 吳裕益(Yuh-Yih Wu)

《金門特教年刊》 1卷2期 (2018/07) Pp. 16-31

https://doi.org/10.6854/KASE.201807_1(2).0002

摘要

本研究主要目的是以統計分析方法探討精簡統整科學過程技能測驗之可行性，根據「答對率」、「校正題目與分測驗總分相關」及「刪題後分量表α係數」三項統指標來刪除答對率低於隨機猜測理論答對率及鑑別度較差之試題，再分別以「內部一致性α係數」、「驗證性因素分析」及「精簡版與國中學業成就之相關」來檢驗統整科學過程技能測驗簡化版之信度與效度。預試研究對象為高雄市某公立國民中學，普通班學生共計53位，驗證性因素分析對象為高雄市11所國民中學的200位學生。研究結果發現，刪除原有36題的統整科學過程技能測驗中鑑別度較差的題目而精簡成19題後，雖然題數僅有原測驗的50％，但仍具有良好的信效度，驗證性因素分析結果也顯示五因素結構之模式適合度良好。由於測驗時間縮短，更具實用性。未來為建構完整簡化版的學生科學過程技能測驗，建議依此統計方法加入簡化版的基本科學過程技能測驗。將可提供教師更多有關學生所具備的科學過程技能訊息，做為科學教學參考。

關鍵字

科學過程技能 ； 統整科學過程技能測驗 ； 項目分析 ； 信度分析 ； 效度分析

並列摘要

The main purpose of this study is to explore the feasibility of streamlining and integrating the scientific process skills test based on statistical analysis methods. The indicator is used to delete the test questions whose answer rate is lower than the random guess theory and the poorly-recognized test, is based on the three indicators: ＂answer rate＂、＂Correction of the correction problem and the total score of the subtest＂ and ＂The α coefficient of the subscale after the question is deleted＂. The reliability and validity of the simplified version of the integrated science process skills test will be tested by the ＂internal consistency α coefficient＂, ＂deterministic factor analysis＂ and ＂relevant correlation with junior high School Achievements＂. The research object is 53 students in the regular class of a public junior high school in Kaohsiung City. The confirmatory factor analysis target was 200 students from 11 public junior high schools in Kaohsiung City. The research found that after deleting the problem of poor discrimination in the original 36-question scientific process skill test and simplifying it into 19 questions, although the number of questions is only 50% of the original test, it still has good reliability and validity. The results of confirmatory factor analysis also show that the five-factor structure is well suited to the model. Due to the shortened test time, it is more practical. In the future, in order to construct a complete and simplified version of the Student Science Process Skills Test, it is recommended to add a simplified version of the Basic Science Process Skills Test based on this statistical method. It will provide teachers with more information about the scientific process skills of students as a reference for science teaching.

並列關鍵字

Science Process Skills ； Integrated Science Process Skills Tests ； Item analysis ； Reliability Analysis ； Validity Analysis