Questrom Professor Evaluation Dashboard

A way to simplify choosing professors

It’s always a hassle when it comes to selecting professors for courses. I saw how my friend and I had to open multiple Rates My Professor’s tabs or use the rating data tables from Questrom which are annoying to interpret. So I created a dashboard as a simple one-stop shop for selecting professors.

— A data-driven approach when it comes to selecting professors —

A sneak peak into the methods behind the madness:

  • The data is the aggregation of professors' scores rated by students from each end of each semester survey from 2012 to 2023. I copied and pasted all of the undergrad data into an Excel sheet.

    The data is available for BU students at:

    https://questromapps.bu.edu/qstnet/courseeval/coursegrid.aspx

  • I convert the Excel file into CSV to operate on Python.

    Then during the data exploratory phase, I found that there are course IDs that changed their name and the same course name changed their ID.

    The Python code on the size is a snippet of how I decided to clean up the data. There are three key criteria in cleaning the data:

    1. Find the most updated name and course ID in the dataset (2023)

    2. Then replace the old names and ID with the most recent name & ID.

    3. Ensure the updated name and ID refer to the same course.

    Fun Fact: Management Organization (MO) used to be called Organization Behavior (OB)

  • I divided the reports into 2 dashboards.

    The first dashboard allows students to search individual professors and see their performance over time by course.

    The second dashboard helps the student select and compare multiple professors for the same course or department.

    Most of the variables used in the graph are created in DAX.

    The most challenging but fun one is calculating the professors' percentile rank while ignoring the filters. (RANKX + CALCULATE + ALL)

    The most satisfying one is grouping professors with above and below-average ratings by dept & course allowing for the split of color when visualized. (IF statement)

  • Initially I create a single dashboard with multiples complexes graphs.

    However, after collecting feedbacks from alumni and students I decided to switch my approach.

    I focus more on key metrics students care abouts and make it as simple to digest as possible. And in the same spirit I also decided to split the reports into 2 specialize dashboards simplifying the report in either a search or compare.