Influential PointModel SelectionMulticollinearityRegression AnalysisOutlierResiduals AnalysisSecond-hand Car
This paper examines the relationship between the price of used cars in San Jose on December 23, 2017 from Carfax website. We use the multiple regression analysis to investigate 9 explanatory variables and use Stepwise Selection approach to select the best fitted model. The collected data shows the mileage which has been driven has negative correlation with second-hand cars’ price. The drive wheel types such as All Wheel Drive (AWD), Front-Wheel Drive (FWD) and Rear-Wheel Drive (RWD) indicate the negative correlation as well. From the perspective of customers, they suggested that they pay more attention to the year of model and the number of images that sellers upload to the website. Second-hand cars, which are new models and have more pictures may have higher price. The quantity of cylinders that the engine has, engine displacement, the amount of miles when consuming a gallon of gasoline (MPG), whether it is for personal or business use and gearbox have strong positive correlation among our 196 randomly selected observations.