Robust and Resistant Estimations of Hedonic Prices for Second Hand Cars: an Application to the Istanbul Car Market

Abstract

In this study, the ‘hedonic price model' has been used in order to determine car characteristics that affect second hand car prices. The hedonic price model is estimated with 1032 observations involving Istanbul second hand car market data. In such a heterogeneous data set, there is often a problem of outliers. Standardized residuals, studentized residuals and DFBETAS criteria were applied to detect them. The hedonic price model is estimated by using robust and resistant methods because ordinary least squares (OLS) method is sensitive to outliers. According to the results of the study, automobile features such as kilometer, fuel type, drivetrain type, body type, engine cylinder volume, transmission type, cruise control, sunroof, trip computer, rain sensor and changing parts affect car prices.Keywords: Hedonic Price Model, Outlier, Robust Estimation, Resistant EstimationJEL Classifications: C12, C13, L62