In some cases, education research often involves the latent variables that have a causal relationship as well as a spatial effect. Therefore, it requires a statistical analysis technique called spatial structural equation modelling (spatial SEM). In this research, a spatial SEM was developed to model the quality of education in high schools in Sumenep Regency. This model was improved after the evaluation of an outer and inner model of the model scheme centroid, factor and path since some indicators were not valid. The path scheme model showed better results compared to the other schemes since all of its indicators were valid and its value of R-square increased. Furthermore, only the model of path scheme was tested for spatial effects. The result of the identification test of spatial effects on the inner model using a robust Lagrange multiplier test (using queen contiguity) showed that the education quality model leads to a spatial autoregressive model (SAR in SEM) with a significance level α of 5%, while the model of school infrastructure has no significant spatial effects. The improved model of SAR in SEM, the R2 value obtained was 47.33%, so that it is clear that data variation can be explained by the model of SAR in SEM for the quality of education in high schools.

Original languageEnglish
Article number012094
JournalJournal of Physics: Conference Series
Issue number1
Publication statusPublished - 21 Sept 2017
Event1st International Conference on Applied and Industrial Mathematics and Statistics 2017, ICoAIMS 2017 - Kuantan, Pahang, Malaysia
Duration: 8 Aug 201710 Aug 2017


Dive into the research topics of 'Modelling of the education quality of a high schools in Sumenep Regency using spatial structural equation modelling'. Together they form a unique fingerprint.

Cite this