Linear Mixed Models to Analyze Indonesia^s PISA Reading Literacy Score Vera Maya Santi (a*), Irsyad Hasari (a), Dian Handayani (a)
a) Program Studi Statistika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Negeri Jakarta, Jl. Rawamangun Muka, Kota Jakarta Timur, DKI Jakarta, 13220, Indonesia
*vmsanti[at]unj.ac.id
Abstract
The Programme for International Students Assessment (PISA) is a periodic survey program to evaluate the quality of education of a country based on the development of student literacy. The results of the 2018 PISA survey showed that Indonesia occupied the position of 72 out of 79 countries. This shows that the quality of education in Indonesia is still low. This study aims to analyze the factors that affect the reading literacy score of PISA Indonesia quantitatively, which until now is still very rarely done. The model used to analyze PISA reading literacy scores as well as schools as a random effect is linear mixed models (LMM). The feasibility of the model is reviewed based on the model^s goodness criteria, namely the estimation of random effect variance, parameter significance tests and model diagnostics. The results showed that the significant factors that influenced PISA reading literacy scores included education level, father^s education, internet access at home, dictionary at home, number of (TVs, cellphones, computers, ebook tabs, and books at home), behavior skipping school and late in coming to school, not listening to teacher explanation, the age of entering kindergarten and elementary school, and having stayed in class during elementary school
Keywords: Linear Mixed Models- model Diagnostics- parameter significance tests- random effect- reading literacy score of PISA Indonesia
