HYPOTHESIS, EMPIRICAL MODEL AND EXPECTED RESULT
6.4 Variables Selection
The Hedonic Price Model comprises different housing attributes. In order to achieve the targets of the study, it is important to choose the appropriate variables of the model. This section is a practical section which discusses the selection of variables for the model.
Here are the objectives of the model being used in this study.
i ) To reveal the effect of chimney view on private residential property price
ii ) To confirm the negative impact of graveyard view on private residential property price
6.4.1 Dependent Variable
Dependent variable is the variable which changes with the variation of the independent variables. Property price is dependent variable used in the study. The property price can be classified as nominal price (NP) or deflated price (DP). Nominal price (NP) is the price of the property at the data of transaction, but deflated price (DP) is the deflated nominal price that can reflect the “real” price of the property at a particular time.
In the model of this study, the deflated price (DP) is the dependent variable. It is because the economic factors have been eliminated in the deflated price. The detail of the deflation method will be described in a later chapter.
6.4.2 Independent Variables
There are three common independent variables, which are the structural, locational and neighbourhood variables. Data samples for the model should be chosen carefully in order to minimize the effects of the location and neighbourhood. It is because the effects of location and neighbourhood are difficult to be qualified. Chapter 5 has discussed that the details of dataset selection. As focus is put on the view of chimney and graveyard, view variables will be included in the model.
6.4.2.1 Structural Variables
The structural variables include in the model are property age (AGE), floor level (FL) and property unit size in term of Gross Floor Area (GFA). It is very common to use these three structure variables in hedonic price model in real estate researches, as they are seen as the major determinants of property prices.
Age of the property (AGE)
Property age is defined as how old the property was at the date of that the Sale and Purchase Agreement (S&P) was signed. In Hong Kong, the property market is very active and it fluctuates very often. Thus, it is the most suitable to use “day”, “quarter” or
“month” for counting the age of the property. By using the “day” approach, the accuracy of the measurement is able to be enhanced, thus a more reliable result can be generated.
In this dissertation, the age of the property is defined as the number of days elapsed from the date that the occupation permit was issued to date that the Sale and Purchase Agreement (S&P) was signed. For instance, the occupation permit was issued at 1 October 1992 and the S&P was signed at 3 January 2002, then the age of the property is 3381 days.
Floor Level (FL)
This variable is used for indicating the position of that apartment in term of storey. It can be easily identified, for example, if the property is situated at the 11th floor of the residential building, the variable of FL is valued 11.
Gross Floor Area (GFA)
Gross Floor Area (GFA) represents the size of the property. It is measured in term of square feet. Similar with FL, it can be easily identified. If the Gross Floor Area of the property is 980 square feet, it is regarded that the GFA value of the property is 980.
6.4.2.2 Dummy Variables
Dummy variable is the numerical variable used in regression analysis. It is commonly used in the hedonic price model analysis for representing some discontinuous factors, such as building types. Usually, 0 or 1 dummy variable is used. The value of 1 is used when the discontinuous factor is available. On the contrary, the value of 0 is used when the discontinuous factor are not available. For example, Mok et al. (1996) demonstrate that 1 is given to a dummy variable if the housing unit is possessing sea view, otherwise the value of 0 is given. Dummy variable is also used in measuring the effect of variables that numbers cannot be used to quantify the variable. Chau Ma and Ho (2001) use 9 dummy variables to verify the relationship between lucky floor number and property price, which lucky floor number if regarded as variable that cannot be quantified. After defining dummy variable, the next step is to construct it into the hedonic price function. If linear relationship is assumed, the equation will be
∑
+∑
+∑
+ +ε+
=a a L b S c N fD
P 0 j j k k q q
In this model, “D” is used to represent the dummy variable. Its magnitude or effect on P will be shown as “f”. In the model of this study, dummy variables are used for the views of the properties. There are totally three dummy variables using in this study, which are Lucky Floor (LF), Graveyard View (GV) and Chimney View (CV).
Lucky Floor (LF)
The floor number which contains the lucky number “8” is regarded as lucky floor number.
In that case, the 8/F, 18/F, 28/F and 38/F of the residential buildings are South Horizons are regarded as lucky floor. “1” will be assigned to the floor level contains number 8, 0 otherwise.
Graveyard View (GV)
The Aberdeen Chinese Permanent Cemetery nearby the South Horizons provides the graveyard view for some of the apartments of South Horizons. In this study, the quality of graveyard view has not been quantified, as the magnitude of the influence of graveyard view is not the objective of this study. In the model, those properties with a graveyard view, regardless to the quality of the view will be valued “1”. For those properties without any graveyard view, “0” will be assigned.
Chimney View (CV)
Lamma Power Station is located in Lamma Island. The three chimneys of the Lamma Power Station provide chimney view to some of the apartments in South Horizons, especially for the blocks which are near the costal line. In this study, the quality of chimney view has not been quantified as the magnitude of the influence of chimney view is not the objective of this study. In the model, those properties with a chimney view,
regardless the quality of the chimney view will be valued “1”. For those properties without any chimney view, “0” will be assigned.