1. Plot a few descriptive or more "neutral" words such as "class," "the," and "teach." What do you see? Now plot a few normative or more "loaded" words to investigate potential biases in the dataset. Possible examples: "funny," "mean," "fair," "unfair," "genius," or "brilliant." What do you see? Include at least three of the words you explored and describe what you saw in the data.

2. This dataset presents a binary classification of gender based on students' beliefs as to the gender of the professor. There are some people in the dataset whose gender is misdescribed, and others, such as non-binary people, who do not have a category that fits them at all. If you could design the ratings website, how might you address this problem?

3. In this assignment, we created a well-defined problem for you to work on; we told you what to do at each step and what counts as success at the end. As we discussed in class, problem formulation is one of the ways in which you embed values in your work as a computer scientist. Formulate a different problem related to the topic of professor evaluation or visualization of patterns in language use, ideally one you could solve with your current skills. Explain a) what values are being embedded in this problem formulation, and b) who this problem is being solved for.
