Apply Grey System Theory to the Score Prediction and Relation of the Basic Competence Test for Junior High Students

• Main Authors: Chun-Yuan Tseng, 曾俊元
• Other Authors: 許寶東
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/41066833334757811766

碩士 === 樹德科技大學 === 經營管理研究所 === 99 === Under the pressure of the Basic Competence Test, junior high students, besides regular school curriculums, will take the simulation tests arranged by school to get reference scores to predict their future performance in the BC Test. This method not only can check students’ learning effect but also provide teachers with the bases of remedial teaching. Various statistical methods are used to investigate the correlation between simulation test scores and the BC Test scores. Therefore, this research uses Grey System Theory to explore the prediction and relation between the BC Test scores and students’ academic achievement. The purpose of this research is mainly to use junior high students’ academic achievement, periodic test scores during three years, to predict their BC Test scores and compare with their actual scores through prediction models. Moreover, this research, which can be the references for students, parents, and school, also use Grey Relation Analysis to explore the relational grade between the BC Test scores and the total scores of periodic tests and to rank the total scores of the BC Test and then compare with the traditional rank. The results of this research are as follows: 1.The average of all the periodic test scores has the highest Grey Relational Grade with Chinese, followed by English, social studies, science, and math. 2.Using Grey System Analysis to rank the total scores of the BC Test has no obvious difference from the traditional rank. 3.Through GM（1.1）Model, Simple Regression Analysis, Exponential Smoothing, Simple Moving Average, this research uses 18 periodic test scores during the three years, including Chinese, English, social studies, science, and math, to predict the total scores of the BC Test. GM（1.1）Model has the best prediction effect. 4.Through GM（1.1）Model, using periodic test scores of Chinese, which has the highest grey relational grade with the average of all the periodic test scores, to predict Chinese scores of the BC Test has good prediction effect.