STAR Assessments for Mastery Learning

Be sure you enroll in the correct course number

In this special topics course, small teams (2-4) of graduate and undergraduate students will develop and rigorously evaluate rich, machine-gradable assessments that would address learning goals that might arise in typical EECS courses. The assessments will promote mastery learning (aka proficiency learning) Links to an external site. by following the acronym STAR:

• Specific to a learning goal in a particular domain or topic. Examples from EECS could include timing diagrams, graph labeling, connect the circuit elements, etc, but not generic-format multiple-choice/numeric-answer/fill-in-the-blanks questions. Developing and evaluating novel question formats and types are the main goals.
• Tagged to specific learning outcomes, skills to be demonstrated (competencies), etc. within the context of a full or partial concept inventory for the course, so that evaluation can be focused on whether the learning outcome is in fact reinforced by the assessment
• Autogradable with instant or near-instant automatic feedback
• Randomized, so each exercise has a large or very large number of variants and can therefore be used for practice (formative assessment for mastery learning)

In addition to developing assessments, student teams will evaluate them by using the methods of HCI and education research to run either informal or formal pilot studies. We will encourage students to make any resulting artifacts available as open educational resources, and if appropriate, submit results for publication.

See example content and projects from inaugural Spring 2023 and other offerings.

Spring 2025 logistics

• Time and location: Mondays 14:00-15:30, room 606 Soda Hall
• First class meeting: January 27

Format & Prereqs, How to Enroll

• Instructors: Armando Fox, Dan Garcia, Narges Norouzi
• In person, regular attendance and participation required to pass

• Permission of instructors required to enroll, based on these conditions:

1. Experience as a student, academic staff, or both, for one or more EE/CS/Data Science/Info Science courses that could serve as the basis for you to think of and develop novel assessment types for those course(s).
2. Reasonable proficiency in Python: B+ or better in CS61A, or we might ask you to do a brief diagnostic interview
3. Pre-proposal for a STAR exercise you could create based on your experience in a specific course. You’re not committing yourself to do this specific one, but it will serve as a starting point to make sure you have some good concrete ideas about what you might work on. The instructors will be available for informal consultation/discussion to help you prepare your proposal.

• Grading: 3 units, letter graded; CS194 units count toward upper division EECS technical electives.
• Grading based on:
• Regular synchronous attendance, engagement, and participation in discussions. Contact instructors prior to a class meeting if it will be a problem for you to attend that meeting or meet these criteria generally.
• Project: In a small group, produce and evaluate a STAR assessment that works within the PrairieLearn assessment authoring system, tagged to concepts in a (partial) concept map for an appropriate course. 
• Team presentations: Paper summaries, project proposal, project progress (regular checkpoint assignments/deliverables), user studies, final presentation
• A few readings from the literature, which may be verified via periodic microquizzes.
• Important. Regardless of other considerations, to get an A for the project, you must package and deploy your work in a manner that is readily usable by others who use PrairieLearn. We will give details of how to do this, but since contributing to the ecosystem is a main goal of the course, it is not possible to get an A without completing this step. This will involve code reviews, pull requests, etc. just as in the usual GitHub open source contribution process.
• No midterm or final. A poster session will be held during the final class meeting time, participation in which is required. Everyone gets until the end of RRR week to finish final coding, make pull requests, etc. If there are appropriate conference deadlines coming up where a poster or short paper would be appropriate, students will be strongly encouraged to submit their work, with guidance from the instructors.

To enroll: please read the FAQ first to make sure the class is a good fit. At the bottom of that page is a link to fill out a short form to enroll.