This is a final project completed for DATASCI 112: Principles of Data Science class at Stanford University.
Instructor quality is paramount to students’ educational attainment, especially in higher education settings where tuition is ever-increasing and the rise of online resources and AI question the value of a university degree. In light of this reality, our research aims to evaluate the perceived quality of instruction at Stanford. In particular, we investigate the following questions:
- Within Stanford, what factors influence an instructor’s rating?
- What instructor characteristics are associated with disparities in these ratings?
We conclude that Stanford instructor ratings indicate that Humanities instructors tend to receive higher overall evaluations compared to other departments. While differences in subject matter and course structure may partly explain this trend, these findings suggest potential opportunities for improving STEM instruction. Targeted efforts such as enhancing clarity in presenting complex material, incorporating more interactive teaching methods, refining grading and feedback practices, and ensuring manageable workloads could also support more effective teaching. Finally, examining the approaches used by highly rated instructors, particularly in the Humanities, may offer transferable insights for instructional development across disciplines.
Please view the notebook files through the following links as the plots are not displayed through Github:
-
Link to the data exploration file (to make plotly.express plots visible):
-
Link to the data analysis file (to make plotly.express plots visible):
-
Link to poster: https://docs.google.com/presentation/d/1uKV22eYZe-k53oW4KcWnK8TLA_SNbXOf_RpW2ICKrsA/edit#slide=id.p