Analys av vinpriser genom regression.

Abstract:   —We investigate which factors affect the price of wine. Among these are amount of alcohol, age, packaging and sealing. The investigation is divided into two parts. The country in which the wine is produced is examined in one model and the second model examines the effect grapes contained in the wine has on the price. In order to produce the two models data from Systembolaget AB’s assortment is used. When the data has been treated, a linear regression analysis is conducted. The results of the two regressions are then discussed in order to examine the reliability of the models. This discussion is based on the assumptions made when conducting a linear regression, and economic theory. It appears that the model with countries is better than the model with grapes, and that factors such as alcohol content, age, packaging and sealing are correlated with price. Finally the causes of these correlations are discussed.