Spatio-Temporal Neighbourhood-Level House Price Index Incorporating

SPATIO-TEMPORAL NEIGHBOURHOOD-LEVEL HOUSE PRICE INDEX INCORPORATING GEOGRAPHIC INFORMATION SYSTEM AND GEOGRAPHICALLY WEIGHTED REGRESSION

Dr. Ibrahim Atan Sipan Professor

Dr. Abdul Hamid Mar Iman

Universiti Teknologi Malaysia (UTM)

ABSTRACT

Although Malaysian house price index (MHPI) is meant to be a general price indicator, the heterogeneity of property attributes will naturally result in spatio-temporal price differentials Because of its generality, MHPI is often not reflective of the local price for it does not cater for street level price differentials. Therefore, neighbourhood-level house price index is needed for a better understanding of the local housing market. The main objective of this study is to develop spatio temporal neighbourhood-level house price index (STNL-HPI) incorporating Geographic Information System (GIS) functionality that can be used to improve house price indexation system. As many as 4,819 out of 28,817 parcels of transacted property were used for mapping STNL-HPI via a three-step process, namely OLS regressions using a statistical software, Geographically Weighted Regression (GWR) using ArcGIS software, and GIS-based analysis of STNL-HPI through an application called LHPI Viewer v.1.0.0. A stand-alone GIS-statistical application for STNL-HPI was successfully developed in this study. Testing and evaluation of this application was conducted among test-group users, namely real estate students, who fulfilled the requirements as 'test candidates' Respondents' evaluations and feedbacks were analysed to identify the strengths and weaknesses of the application. The overall results have shown that the modelling and GIS application were able to help users understand the visual variation of house prices across a particular neighbourhood They also appreciated the usefulness of the application especially to understand the dynamics of property market in the local context. Notwithstanding this, some shortcomings of the application were identified in terms of methodology and functionality, and further improvements have been suggested.

Key words: Spatial-temporal neighbourhood-level house price index, Geographic Information System, Geographically Weighted Regression, house price Index.