A project to analyze consumer survey data.
The purpose of this project was to better understand factors that contribute to consumer ratings for particular beverages, namely, beer. The beer industry is booming in the American market, with consumers switching slowly from drinks with a high alcohol percentage to lighter and more refreshing beverages. Consumer trends can be seen in overall beverage ratings and in ratings for subcategories such as palate, aroma, or Alcohol By Volume. By focusing on these essential categories, key predictors can be identified and used to direct further development.
The initial dataset consists of 1.5 million consumer reviews provided by BeerAdvocate.com, a site where “beer geeks and industry professionals” can submit numeric and written reviews on a wide array of beverages. The focus of the project was larger trends in consumer preferences accross products so this data was filtered to exclude products with fewer than 500 reviews, aggregated by style and then filtered again to exclude any style with fewer than 5 products. The resulting dataset consisted of 25 styles, 275 individual products, and 270,323 individual consumer reviews.
| Name | Description | Type |
|---|---|---|
| beer_data | the project dataset | .CSV |
| beverage_pref | the project markdown | .RMD |
| beverage_pref | the knitted document | |
| consumer preferences report | the accompanying research paper | .DOCX |