To achieve the objectives of this research, namely:
a) To know the factors that affect poverty-stricken districts
in Java based on ordinal logistic regression models of spatial.
It can be done by the following procedures:
• Selection of explanatory variables that will be used in
the analysis, by looking at the closeness of the relationship of
these variables.
• Response variables to be used is a hotspot poverty
districts in the island of Java in the form of ordinal data with
6 categories from research that has been done by
Kusumaningrum (2010)[6].
• Looking at the closeness between districts in Java.
• Creating a spatial weighting matrix W which contains 1
if adjacent, and 0 if not directly adjacent.
• Create a new variable result multiplication of matrix W`
with Y, namely X spatial (W Y)
• Checking the assumption of logistic regression
International Journal of Statistics and Appl