Model selection and Akaike Information Criteria: An example from wine ratings and prices

Abstract

The effect of wine ratings on pricing has been a question for wine consumers for some time. Ultimately, wine preference, and thus how one judges a wine, is a subjective endeavor. Wine professionals have long rated wines and those published ratings have some effect on consumer sales. Previously, wine studies have found that there is a connection between rating and price. This study looks to try to verify that connection through insuring that best fit model development is used. For the first time in wine research, the authors have utilized Akaike Information Criteria (AIC) to compare different models and more dynamic hypothesis testing to explore the relationship between ratings and prices of wines. In the end, it was confirmed that there is a link, and the use of AIC also helped to not only confirm previous findings, but also to identify a new concern in wine ratings.

Keywords

• Wine
• Wine rating
• Model testing
• Akaike Information Criteria