Welcome to Gradebird!

Gradebird is a simple web app for student self-assessment using the alight workflow. After submitting their homework, the students:

  • Assess their own work according to a rubric
  • Learn from any mistakes
  • Tune their skills for the next assignment
Gradebird makes it easy—for both students and instructors.

After submitting each assignment, students use a web form assess their work according to an itemized rubric prepared by the instructor, supporting their choices by uploading an annotated version of their original submission. The instructor reviews the assessments, leaving feedback that students can incorporate for the next assignment. Point scores are assigned automatically (with optional instructor modifications), such that grading becomes a seamless by-product of a learning-focused workflow.

The alight workflow has been used in the Department of Physics at the University of Arizona since 2021, developed initially by Professor Drew Milsom and later adopted by Professor Sam Gralla. Finding traditional learning management systems too clunky for this approach, Prof. Gralla designed Gradebird in 2024.

Gradebird has been incredibly popular in graduate and upper-division undergraduate physics courses at the University of Arizona. When students receive credit for finding their own mistakes, they take this task seriously and end up learning the material much better. It also makes the homework assignments less stressful, since the students know that they can still learn a lot (and receive good grades) for problems they do not manage to solve. We have not yet tried gradebird in a large introductory course, but we expect the same response.

Anecdotally, yes, but like anything in education research, the gains are difficult to quantify. Prof. Milsom is publishing his first paper on the topic, and we hope that Gradebird will help us gather more data to achieve statistically valid conclusions.

Gradebird makes grading much more pleasant and fair, while drastically reducing the workload. It is infinitely easier to determine an accurate grade when the students have already found their own mistakes and described them to you. Then you decide if you agree with their assessments, and the numerical grades are automatically filled in.

This will free up time to perform other tasks related to the course. For example, teaching assistants normally devoted to grading can be asked to hold office hours, prepare problem solutions, or even come up with some of the problems to assign. As the head instructor, you will notice significantly more time available to actually teach the material, as opposed to assessing the students.

We initially employed this workflow using a traditional learning management system, simply asking students to upload an annotated version of their original submission. It worked fine from the student perspective, but didn't save us much time because of all the overhead. After getting immediate gains from a barebones version of Gradebird in piloted in Fall 2024, Prof. Gralla decided to develop a full-fledged tool. (You can also blame him for the relentless bird puns.)

We eventually plan to make Gradebird available for general use. For now, we are primarily interested in working with dedicated instructors whose feedback will help us improve the workflow and associated tool. If you fit that description, please email us at info@gradebird.com.

Please contact us at info@gradebird.com with any questions or suggestions.