In 2017, researchers asked study participants via MTurk to list three happy moments they had in the past 24 hours. The researchers released the data as HappyDB, which provides 100,000 happy moments along with demographics. For the Pudding, Alvin Chang made an abstract map that places the moments based on less agency to more agency on the x-axis and immediate to long-term on the y-axis.
A Gemini LLM was used to classify the data into continents, countries, and states, based on type of happiness. You can zoom in to the individuals to read the moments and you can filter by location, age, sex, parental status, and marital status.
Having worked with this data a bit, I wonder if this map looks different if the ultrarich answered. Or maybe a yacht with five people on it just travels through the waterways.
Visualize This: The FlowingData Guide to Design, Visualization, and Statistics (2nd Edition)
