Perbandingan Distribusi Binomial dan Distribusi Beta Binomial Menggunakan Struktur Perkalian Distribusi pada Analisis Kelulusan Mahasiswa Jurusan Matematika FMIPA UM

Abstract

he choice of two distributions with biner response, where the first response is with binomial distribution and the second with beta binomial, is often difficult because the mathematical decrease process is complicated. This research attempts to determine binomial and beta binomial distributions using distribution multiplication structure (SPD) applied to Markov Chain Monte Carlo (MCMC) assisted by Win BUGS 1.3 package to determine the parameter value of SPD. In this research the result of computing Bayes Factor shows 3.38, which means that beta binomial distribution can be used as an alternative to binomial distribution. In other words, the passing percentages given by a lecturer to a certain subject differ among groups of Exact 1, Exact 2, and social language subjects.