Estimating of Hedonic Price Models using Robust Regressions: Solving the Outlier Problem
Marmara University, Faculty of Economics, Department of Econometrics, Goztepe Campus, 34722, Kadıkoy – Istanbul, Turkey. e-mail: email@example.com.
This paper analyses the hedonic price model for Istanbul real estate market, in order to determine the relationship between house prices and housing features by using OLS. Since the OLS suffers from heteroscedasticity, non-normality and outliers, we estimate the hedonic price model by using robust regression which is called L1 and M regressions to get rid of these problems of OLS. For the purpose of investigating the relation between house prices and features of houses in Istanbul, the survey data of Ҫağlayan and Arikan’s work (Quality and Quantity, 2011) were used. Results of the study indicate that while having a central heating system, natural gas, a certain level of security, garage and four faces, have an effect that increase house prices, being located in a site or on a street have a decreasing effect on the same.
JEL Classification: C12; C13; R13.
Key Words: Outlier; Robust Regression; Hedonic Price Model.
Full Text (.pdf)