Ben Schmidt, the director of digital humanities and clinical associate professor of history at New York University, created the Gendered Language in Teacher Reviews project. His primary goal was to explore gender bias in reviews from RateMyProfessor.com, specifically the top 20 departments. Additionally, Schmidt was interested in monitoring what words other visitors chose to search for in the professor evaluations visualizations. In Scheinfeldt’s sense of an argument, this project does not have an argument in that it was not created to give a concrete, holistic answer to a specific question. This program was not designed with a claim in mind. Rather, this Digital Humanities project is a tool for exploring gender bias in teacher reviews. Essentially, this project is an instrument for entertainment and casual investigation—there are too many variables and shortcomings within the program to use it as a sole, scholarly source. In his background explanation, Schmidt states that this project is not peer reviewed and not academically rigorous enough to appear on his CV. However, as Scheinfeldt concluded, this project does promote experimentation, and has an underlying assumption: that there is gender bias teacher reviews, likely sexist towards women.
Ben Schmidt’s intended audiences for this project seems to be for himself and other digital humanities scholars. He states in his background post that he would be “pleased if [this project] shook loose some further study by people in the field.” While unclear, it is probable that he is referring to his own field of digital humanities. But it is also possible that he meant it to refer to universities. Either way, Schmidt intended this project for a relatively small audience.
The RateMyProfessor project uses one type of source: text. Schmidt pulled the raw text from RateMyProfessor.com, “in accordance with their robots.txt protocol.” From his post, it seems that taking this text is permissible if he does not redistribute the “underlying text.” Next, these texts were processed through Lincoln Mullen’s gender package. This package auto-assigned the gender based on names and dates of birth, using the American Social Security Administration’s data set or the Census Bureau data. After this, it appears Schmidt used the Bookworm project to shape the data. He does not describe this process, but there is a link to “using Bookworm” at the top of the site. The project is presented as a visualization of the evaluations in graphs which separate the genders by color, and chart the usage of each word by number of occurance. The project is interactive as users input their own search terms.
On the main page of the site, Schmidt concludes that not all words will have a gender split, but many will. The project does not have a distinctive conclusion evaluating sexism or the lack thereof in student evaluations, but its purpose was more for exploration than a definitive answer to this question. It is unclear how long the project took from start to finish. Schmidt mentions that scraping the text from the website took a couple months, and there is a post about the February 2015, but the link no longer works. There is an alert stating that the project is back online as of September 25, 2019. This project employs and enables comparing and illustrating as it visualizes data and prompts the user to compare terms and usage of terms based on gender. The main weaknesses of this project are that the gender of the reviewer is unavailable, it is not a rigorous “study,” and RateMyProfessor excludes terms like “sexist,” “racist,” and profanity. Finally, the website contains several broken links, revealing maintenance errors.
