Prototype a web-based tool for creating and executing task-delineated, collaborative, AI-assisted assignments
Background
This project is to design and develop a working prototype of a web tool for college instructors to create assignments that are AI-assisted, task-delineated, and collaborative in nature, and for students to execute those assignments.
The project comes from a Senior Instructor of Entrepreneurship and Strategy at the OSU College of Business (COB), Sanjai Tripathi, who is a non-technical project sponsor but one with experience in product management and tech startups, and who has identified a critical set of challenges and problems in higher ed, listed here briefly:
- Students must learn to beneficially use emerging generative AI (GenAI) tools, especially in an “active” learning mode; However, GenAI tempts students to execute tasks quickly rather than thoughtfully, or cheat outright, and in doing so threatens to undermine the learning process.
- The ideal design of of a college course, in addition to active tasks, should have transparency (clarity on task responsibilities and evaluative criteria), sufficient complexity and challenge, and ample opportunities for peer interaction and collaborative learning; however, the structure and features of assignment creators in existing Learning Management Systems (LMS) such as Canvas are limited in all these areas.
Target Users
To address these challenges, this project aims to design and demonstrate a new concept for the assignment creation and execution process, embodied in a web tool for use by two key user types:
- University instructors, focusing on those teaching undergraduate courses that incorporate active learning and teamwork, and an active interest but low to moderate expertise with GenAI.
- University students, focusing on those enrolled in the types of classes described above.
Envisioned Solution
The envisioned tool is very similar to the assignment engines of popular LMS tools (such as Canvas), and should emulate those in style and flow, but with key enhancements:
- Task delineation. Many college assignments have a single listed due date and rubric even though they are actually multi-step in nature (e.g., discussions, research reports, team project deliverables). Embodying assignments as a structured sequence (or IFTT process) of tasks allows for enhanced specification, guidance, transparency, accountability and assessment of complex assignments.
- Peer and group interaction. When it is well structured, peer and group interactions make college more fun and engaging, and enhance learning tremendously. The envisioned tool would enable instructors to specify interaction tasks as part of assignments, which current popular LMS systems do not do well.
- AI-assisted tasks. Instructors want students to beneficially utilize GenAI on assignments, but often do not know how to instruct students on prompting, create appropriate tools for assignments (e.g., custom GPTs with pre-set prompts that reference relevant files), or otherwise facilitate beneficial use while deterring uses that bypass the learning process and violate academic integrity standards. The envisioned tool would enable instructors with low to moderate proficiency to specify which tasks should incorporate GenAI assitance and select or create an appropriate tool for that task, while creating a record of how students use the tool for the purposes of assessment and assurance.
In simpler terms, this proposal is to give instructors a web-form-like tool to easily build complex assignments with specified tasks, and each task may have its own instruction set, due dates, assignee (individual or group), and prescribed GenAI-assistance resource, and the instructor can view a record that shows details of when and how the assignment was executed rather than just the final product (summarized as an aggregated report with links). For students, the experience should be like a “learning journey” through a set of instructed steps, with transparency at each step and on the process as a whole, seamlessly integrating individual, peer-evaluation and group assigned tasks to arrive at the final product.
Objectives
This project will be similar to innovative projects in industry, whereby the “business” side presents the outline of a solution and the “technical” team is given leeway and responsibility for determining the technical approach, negotiating the scope, and dealing with challenges as they arise to deliver a working solution by the end of a project period.
An “envisioned solution" is described above and the project sponsor can also provide use cases, user stories, and crude mockups and other forms of specification and illustration of desired features and value to be created. The sponsor will also be available for regular consultation throughout the process and to facilitate live demonstrations of the prototype tool in the classroom.
The primary goal of this project is to create a working prototype of a tool for instructors to create task-delineated, AI-assisted, collaborative assignments, and for students to be able to efficiently execute those assignments.
Objectives
Specific objectives for this project are negotiable. However, at a high level, the project must include at least the following stages and objective results.
Design
In the design stage, students will consult with the sponsor and conduct independent research to understand user goals and identify possible technical approaches, then engage in a design process that includes specifying a technical approach, creating architectural schematics, creating user flow diagrams and mockups, a Scrum-style Product Backlog of functional and non-functional features, and/or other embodiments of a proposed plan for review by engineering faculty and the project sponsor for feedback, followed by updates.
Prototype
In the prototype stage, students will develop a working solution in Agile fashion. This includes setting up a development environment and developing the initial features, followed by successive rounds Sprint development with a release and review at each stage.
Demonstration and Revision
In the Demo and Revision stage, the project sponsor will arrange a live demonstration of the prototype with a population of business students, and together sponsor and students will monitor, review, and analyze results, make a few updates, and create a recommended roadmap for the future (beyond the Capstone period).
Motivations
I have a hypothesis that structuring complex assignment as task-based “learning journeys" will allow us to radically improve students' experience with those assignments and beneficially incorporate AI into higher education. My motivation is to prove that hypothesis to be true, using the outcome of this capstone project for live demonstrations in the classroom that lead to case studies … and ultimately a paradigm shift in how we think about assignments in higher education.
Qualifications
Minimum Qualifications:
- Basic web development
- Knowledge of agile development
- Knowledge of AI systems
- Interest in education
Details
Project Partner:
Sanjai Tripathi
NDA/IPA:No Agreement Required
Number Groups:2
Project Status:Accepting Applicants
Keywords:WebArtificial Intelligence (AI)ResearchMachine Learning (ML)EducationNew Product or Game
