LOGISTIC REGRESSION MODEL FOR PREDICTING FIRST SEMESTER STUDENTS GPA CATEGORY BASED ON HIGH SCHOOL ACADEMIC ACHIEVEMENT

Abstract

(ProQuest: ... denotes formulae omitted.)INTRODUCTION:One of the important stakeholders in university is students. Before entering Duta Wacana Christian University (DWCU), students have had different educational and environment backgrounds. The different environments background means having a secondary school they have taken or take. For example, those coming from senior high school in Java and outside Java. And also they maybe come from ordinary high school or vocational high school. The last, they also maybe come from public high school or private high school high school. After prospective new students are admitted as college students at Faculty of Information Technology, DWCU and after learning process in the first semester, the results of their study are reflected in their first semester Grade Point Average (GPA). The GPA of a student can be regarded as evaluation of learning outcomes in higher education. There are some students who secured high GPA whereas some obtained low GPA. At the first semester, these students can sometimes move to another faculty. In addition, there are some students who have not registered for the next semester, so that they are no longer active as college students or even resign as a student from the Faculty of Information Technology DWCU.There are many different external factors that may affect the first semester GPA beside existing internal factors such as interests, talents, abilities, diligence, self-motivation. External factors that are used in this research are the location of high school, categories of high school, status of high school, and level of English test result. The location of high school can be categorized into high schools in Java and outside Java and the high schools are classified by usual high school and vocational high school which also thought to be investigated. Similarly, the status of high school is categorized into high school with the status of State and Private status. In addition, the level of English among the students is also a possible factor affecting students' first semester GPA. In this research we only limit on factors as mentioned above. Actually there are many more factors that may affect the GPA categories that is not covered, such as: gender, interest in studies, family economic level, employment background of parents / guardians and others.In a previous research, Santosa & Setiadi (2015) have investigated about the factors that affect students' GPA significantly, but they have not observed a relationship between GPA and these factors. This research is the next step of previous research which essentially tries to find mathematical models such as logistic regression to predict the categories of students' first semester GPA at the Faculty of Information Technology, DWCU.LITERATURE REVIEW:DATA MINING:Due to the internet development, there is a rise in global data usage. Data is a collection of facts that have been valid and ready to be used for the purposes of seeking information. Information can be obtained in various ways, either manually or automatically. If the amount of data is too large, then the information search becomes increasingly difficult, especially if it is performed manually. A search for information from a large number of datasets can be done automatically using Data Mining. Data Mining (Knowledge Discovery from Data) is a discipline in Information Technology that is interdisciplinary and is able to perform mining of data in large numbers in order to obtain various kinds of information in the form of certain patterns that are hard to find by searching manually (Rud, 2010).Stages of data mining in knowledge discovery process is as follows (Han & Kamber, 2011):1. Data collection phase (either from a database or other data source): In this research, the data retrieved from the achievements of high school students who come from the path of achievement in the Faculty of Information Technology UKDW act as the data source. …