National Academy for Educational Research

　　為瞭解新、舊課綱學生在社會素養的表現情形，並找出影響學生學力成長變化的因素，本研究以108課綱的社會領域學習表現和學習內容為基礎，研發社會素養評量架構，社會素養評量架構擬包含探究、情意、學科訓練等三面向，研究者也依照評量架構研發社會素養的線上多媒體試題。正式調查的追蹤對象第一群為2019年9月入學的七年級學生（舊課綱最後一屆）的八年級，第二群為2019年9月入學的七年級學生（新課綱第一屆），抽到該批學生後，分別連續追蹤這兩批學生在國中三年的表現，抽樣方式為兩階段分層群集抽樣（two stage stratified cluster sampling），抽樣人數一屆約為6000人。正式施測時，有數個題本，每位學生作答的題目不完全相同，一次施測的總題數約為180小題。最後，逐年回收的資料，將進行信、效度分析，並以部分給分模式（partial credit model），估計與母群體和次群體的分數，瞭解新、舊課綱學生的成長情形和表現，並賦予學生成長量尺教育意義，回饋給課綱改革參考。

　　In order to understand the performances of social srudies of junior high school students under the old and new Curriculum Guidelines, and to find out the factors that affect students’ academic abilities, this study, based on the performance of social studies and subject contents, intends to develop a social literacy assessment framework. This assessment framework includes three dimensions, which are inquiry, affection, and subject training. The researchers also develop online multimedia test items for social literacy in accordance with the assessment framework. The first cohort of study subjects for the present study is the eighth graders, who were seventh graders when enrolled in September 2019 (the last year of the old Curriculum Guidelines). The second cohort of study subjects are seventh graders enrolled in September 2019 (the first year of the new Curriculum Guidelines). These two cohorts of students will be followed by their performance for three consecutive years. We will employ two stage stratified cluster sampling with a sample of about 6,000 students. At the formal test, there will be more than one version of the test booklets. The questions each student answers are not exactly the same. The total number of the test items is about 180 questions. The data collected annually will be analyzed for reliability and validity. We will use partial credit model to estimate the scores with population and subgroups to find out students’ growth and performance under the old and new Curriculum Guidelines, and also to give the educational implications of growth scale. The results of the