Author: Ben Schmidt
Publisher: Ben Schmidt (Personal Website)
Publication Year: 2015
Summary: The interactive visualization here allows users to input words to query 14 million Rate My Professor reviews and see the splits in occurrence for gender across different school departments, as well for positive and negative reviews. Negative words generally occur more often for female professors, such as ‘horrible’, ‘awful’, and ‘terrible’, especially in STEM departments like engineering, physics, or computer science. The FAQ, also linked, contains some additional info and considerations. This visualization is an example of how implicit biases can affect our data, as if these RMP reviews were just pulled into a model blindly, there would be no accounting for this bias in reviews. It is important for us to consider how bias manifests in our language, and how that may affect data, particularly in text analytics.