Link to my data viz: https://public.tableau.com/app/profile/micaiah.davis/viz/IceCreamFlavors_17272250945740/Dashboard1#1

I did my data visualization in Tableau Desktop since that was what I was most familiar with, then saved it to Tableau Public for everyone to access. I haven’t used it in a while, and I figured the refresher would be helpful since I plan to use it for my capstone project. I wanted to have some fun with this assignment, so I looked at ice cream data. Originally, I wanted to look at popular ice cream flavors and how they might have changed over the years, but I couldn’t find a suitable dataset for that. The closest I found was a dataset on Kaggle that looked at the rating count, ingredients, reviews, and descriptions for different ice cream flavors across four brands.

The hardest part of the process was the data cleaning which I was hoping to avoid since the original dataset had symbols in some of the ice cream names and the brands were written as acronyms, but I wanted them spelled out. The description of each flavor also included a lot of symbols, but I left those alone since some were long and I wasn’t using that field in my viz. I cleaned up the names in Excel Power Query before using it in Tableau.

I found “Against Cleaning” particularly interesting because it gave a new perspective on data cleaning. Whenever I thought of data cleaning, it meant ensuring a standard was used across all fields(data). Are the titles of the ice cream flavors all in caps or is only the first letter of each word capitalized? Are ingredients separated by commas or colons? Are brand names abbreviated or spelled out? It never occurred to me that from another perspective, data cleaning could be seen as “wiping away what is different” which depending on the context prevents outliers from being explored, a sentiment shared by Lini in one of her annotations of the reading. If the data doesn’t tell the expected narrative, then it needs to be “cleaned” until it does. This line from the reading also stood out to me “To ask research questions, we needed to create our own dataset…” because it’s something I likely will have to do for my capstone project.