Wrap Up
Key Takeaways
-
Context is crucial in data science; ignoring it can lead to misleading interpretations and reinforce existing power imbalances.
-
Ethical considerations in data science extend beyond the data itself to how it is framed and communicated, especially in terms of avoiding deficit narratives and being transparent about limitations.
-
Data often comes from environments influenced by power differentials, and understanding this can help in identifying what is missing or misrepresented in a dataset.
-
There are emerging tools and methods for adding context to data, but these need to be more widely adopted and funded to become the norm rather than the exception.
Exercises
-
Analyze a dataset of your choice and write a “data biography” for it, answering questions like: Where did it come from? Who collected it? When? How was it collected? Why was it collected?
-
Find a data visualization online and critique it. Does it consider context? Does it avoid deficit narratives? Is it transparent about its limitations?
-
Create your own data visualization based on a dataset, making sure to provide context and to consider ethical implications. Write a brief explanation of the choices you made in terms of framing and communication.
