Assignment 1: Create a dataset

Due: 9 am, Fri Oct 16 → Submit to Canvas.

Mimi Onuoha’s art project “The Library of Missing Datasets” illuminates what we value and what we ignore in an increasingly data-saturated world. For example, there is little data that exists about financial transactions in the art market (especially between gallerists and artists) and on US judges and case outcomes. Think of a dataset that doesn’t exist and create it.

This could be as grandiose or as mundane as you would like: it could be the temperature of a patch of grass outside your room for a week, FitBit or other health tracking data, or more substantial research on social or political phenomena. (For example, the Data Against Feminicide project required scores of volunteers to track feminicide throughout Latin American by reading through innumerable news articles to collect data that governments did not.) There is no requirement around time span or number of observations in this dataset; the “Dogs of Cambridge” dataset is not bounded by time the way that temperature is. This dataset does need to demonstrate clear inclusion and exclusion criteria, require some kind of measurement apparatus, and provide you with enough information to do some analysis about what the data shows. To complete the assignment, you will submit the dataset, a write-up about the data collection process / what the data shows, and a data visualization of your choice.

Here is the handout for the data collection protocol which will be completed in class.

How this will be graded

• A: Each choice in the dataset (e.g., what counts as data, what is excluded) is deliberate and well-explained. The work presents an argument that is supported by the data and the account of how the dataset is constructed is reflective and well-reasoned. The visualization makes an argument that the prose could not do alone.
• B: The choices are sound and mostly explained; the claims are generally supported by the data. However, the reasoning is uneven and the write-up does not fully account for how the measurement instrument shaped the results or does not make a clear, complete argument.
• C: The write-up asserts things the protocol could not have captured; construction of dataset is incomplete or is not fully explained or reflected upon in the write-up.
• Below C: The dataset or write-up doesn’t follow the assignment instructions.

You can see a longer explanation on this assessment sheet. This is not a rubric, but a set of guidelines to help guide your writing.