The analysis of validation of the new scoring formula for the Basic Competence Test (BCTEST) in 2009

• Main Authors: Guo-chin Huang, 黃國清
• Other Authors: Yuh-yih Wu
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/91125208856963460781

博士 === 國立高雄師範大學 === 教育學系 === 99 === The two main purposes of this study were to: 1. Construct a predictive model for the BCTEST score, and analyze the validation for the new scoring formula, 2. Develop the method to increase validation for the new scoring formula. In order to reach these two proposes, this study divided participants into two groups: 2763 junior-high students in Tainan and 5000 students in Taiwan, being analyzed for purposes 1 and purposes 2, respectively. The statistical methods, including descriptive statistics, correlation analysis, structure equation modeling (SEM), multiple-group analysis (MG) and bootstrapping were adopted to analyze the research data. The major findings concluded from this research are as follows: 1. The data of the 2763 junior-high students in Tainan well fitted the predictive model (subject’s intelligence, learning adjustment and behavioral disturbances predict the score). According to this model, there were many significant relations among subject’s intelligence, learning adjustment and behavioral disturbances, and these subject’s variables predict the score significantly. 2. Compared to the old scoring formula, the new scoring formula was more effective for the subjects. 3. The new scoring formula was proved to be more effective than the old one for high-ability subjects. 4. For low-ability subjects, the validation of the new scoring formula was not supported. 5. There were two ways to improve validation in writing tests and the procedure of score calculation. On the basis of the above findings, this article provides some suggestions for junior-high schools, educational authorities and other related studies in the future.