![Sample our Built Environment journals, sign in here to start your access, Latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5926/TOC-Subject-Sample-2021-Built-Environment.jpg"})
Abstract
This paper assesses the relative importance of structural and locational attributes on residential property values in Metropolitan Kuala Lumpur. A computer-intensive algorithm known as LMG method was used in the analysis. Results suggest that Metropolitan Kuala Lumpur property values are mainly determined by the structural attributes which explain 62% of price variations, and the locational attributes explain only 17%. Results further suggest that floor area, two or more storeys, property types, property with freehold status, lot size, greater number of bedrooms, and two locational attributes i.e. closeness to city centre and closeness to forest are the most important variables affecting property values. The result of this analysis offers deeper understanding on how structural and locational attributes interact with property values in an urban setting, and could further increase the degree of objectivity of the professional valuer’s in the predictive model.
Notes
1. A more extensive review of the methods is provided in Bi (Citation2012) and Braun and Oswald (Citation2011).
2. As has been argued in the literature, one of the limitations of HPM is that this method is found to be less sensitive to spatial variation, as the method assumes that the effect of locational amenities on residential property values is fixed across a geographical area. Fotheringham, Brunsdon, and Charlton (Citation2002) postulate that the relationship between locational amenities and residential property values across geographical areas in a stationary fashion may not be representative of the situation in any particular part of the study area and may hide some interesting and important local differences in the determinants of residential property values.
3. For discussion of hedonic regression theory, see Freeman (Citation1979b) and Kain and Quigley (Citation1970a).
4. Ordinary least-squares (OLS) regression is a generalised linear modelling method that may be used to model a single response variable which has been recorded on at least an interval scale. The method may be applied to single or multiple explanatory variables and also categorical explanatory variables that have been appropriately coded (Hutcheson, Citation2011, p. 224).
5. Only 2009 and 2010 data were available for analysis and this was appropriate set of data since that was the period of economic recovery in Malaysia, witnessing vibrant property transactions in a stabilising property market.
6. Our sample consists of 421 (10.41%) units of semi-detached house, 160 (3.96%) units of developer’s design bungalow house, 227 (5.61%) units of owner’s design bungalow house, 275 (6.79%) units of corner lot terrace house, 172 (4.25%) units of end lot terrace house, 2505 (61.93%) units of intermediate terrace house, 6 (0.15%) units of corner lot cluster house, 3 (0.07%) units of end lot cluster house, 53 (1.31%) units of intermediate lot cluster house and 223 (5.51%) units of townhouse. Our sample thus captures over 90% of landed residential properties in the study area that are transacted in 2009 and 2010.
7. GIS can be defined as a branch of computer technology that deals with geographical information (Maguire, Citation1991); or alternatively as the set of computer-based decision support systems containing data, hardware, software and organisational for capturing, storing, managing, manipulating, analysing and visualising a special type of information, namely spatially referenced data (for alternative definitions of GIS, see Dueker, Citation1979; Burrough, Citation1986). An attempt to employ GIS in hedonic house price study has started in the US, but in the late 1990s, much attention has been given by UK researchers in employing GIS for the purpose of their research and now GIS has widely been employed in hedonic house price study. GIS has obtained greater attention from researchers because of its capabilities, particularly in handling and organising large spatial data sets from various sources.
8. Spatial analysis techniques have been defined as those “whose results are dependent on the locations of the objects or events being analysed” (Goodchild, Haining, & Wise, Citation1992). Bailey (Citation1994) noted that spatial analysis is a general ability to manipulate spatial data into different forms and extract additional meaning as a result. Most importantly, spatial analysis involves the analysis of patterns in spatial data, relationships between patterns and other attributes, or the modelling of such relationships for the purpose of the understanding or prediction of certain phenomenon within the study region.
9. A household may perceive a locational amenity to be within a certain distance of the property, even though natural or man-made barriers may prevent a pedestrian from travelling along the shortest straight-line route to reach the locational amenity.
10. In Malaysia, primary school is intended for pupils aged 7–12, whilst secondary school is intended for pupils aged 13–18.
11. The identification of high-performing primary schools and high-performing secondary schools for this study is done based on the list produced by the Ministry of Education Malaysia.High-performing schools are defined as schools with ethos, character and a unique identity which enable the schools to excel in all aspects of education. These schools have strong and excellent work cultures and dynamic national human capital for holistic and continuous development in addition to being able to compete in the international arena, hence becoming the school of choice. There are six criteria were used to select high-performing schools; excellent academic achievement, towering personalities, national and international awards, linkages with institutions of higher learning, strong network and nationally and internationally benchmarked.
12. Multicollinearity exists when two or more of the predictors in a regression model are moderately or highly correlated. There are six ways to deal with highly correlated predictors when developing a linear regression model; manual variable selection (i.e. the variance inflation factor or VIF), tree-based automatic variable selection, regression-based automatic variable selection (i.e. stepwise regression methods), variable reduction via principal components analysis or PCA, variable reduction via partial least squares or PLS and parameter estimation via “shrinkage” methods.
13. Sequential means that the regressors are entered into the model in the order they are listed. The sequential sums of squares of all regressors do sum to the model sum of squares (Grömping, Citation2006).