Model Regresi Logistik Biner untuk Identifikasi Faktor-Faktor yang Mempengaruhi Siswa Masuk Jurusan IPA/IPS

Abstract

One direction is desired for students and parents is the science classes. On the one hand, this course allows the student has a choice of majors in universities more than other majors and a lot of work to only accept students from science classes. Ironically, many students who choose to major in science but does not support the student's ability. Dealing with the issue, it must be examined factors that influence students major in science. One method used to analyze the factors that affect students entering the science classes of binary logistic regression. Response variables used in this study categorical form of data on the status of the incoming students are students majoring in science or social studies. The data used in this study is population data class XI of SMAN 2 Bogor, West Java, the academic year 2012/2013. The number of total population is 279 students. Results of analysis of this study is binary logistic regression models produce values overall classification accuracy of 92.8% for the model classification table and 92.7% for the ROC curve. Explanatory variables used in this study a total of 18 explanatory variables and the explanatory variables are eight ties associated with variable responses to the real level of 5%. Wald test results obtained by the three explanatory variables that significantly influence the status of the students choose a major motivation, student aspirations and students are happy with the lesson or not exact.