Canvas Study Bot

MY ROLE

UI/UX Designer , UX Researcher

SOFTWARE & TOOLS

Figma , Wireframing, User Surveys

TIMELINE

October 2021 - November 2021

Define

Project Overview

In an online schooling environment, getting assistance doing work can be difficult. How many times have you tried to find a meeting time to get assistance from your professor when you couldn’t make their office hours or they never responded to your email for an assignment? To solve the problem of the lack of academic support for students in an online school- specifically in STEM courses, we propose to develop a built-in AI “Study Buddy” to learning platforms that a teacher can use to assist students in an online learning environment. The course instructor could control what resources the AI makes available and when it is available to the students.

Problem Statement

In a college environment, getting assistance when completing work may not always work in a student’s favor. In many instances, college students cannot attend office hours and are not able to receive the proper help they need in time of completing their assignments. Hence, a lack of access to help and office hours at any moment may hinder a student’s ability to complete work to their best ability.

Foreseeable Challenges & Solutions

Problem: Students may rely on an AI study bot rather than actually engaging in the learning process.

  • The professor would have full control of the AI bot — they would monitor the questions and responses that students would have access to.

  • The bot would only supply practice problems to enable students to further practice learning material.

Problem: The bot does not have the full capacities to replace a teacher when students are completing work.

  • The AI bot acts to aid in an active learning style that relied on person-to-person studying due, to Covid-19’s transition weakened that learning style.

Problem: How will this help students who are taking more open-ended classes? (for example: English, History, etc)

  • Currently, the scope of this project is to find a solution to helping students taking STEM related courses, as they will likely utilize features such as practice problems or checking work.

Research

Quantitative Research

As a team, we created a user survey to determine our target audience their current use of the Canvas application. We surveyed over 20+ individuals and discovered details about their current habits with Canvas.

  • Most of our respondents were college students in the class of 2023

  • Around 87% of survey respondents had taken a STEM class at Rutgers

Our Competitors

Our competitors: understanding what students are currently using to help them with classwork. We’ve noticed a vast majority of respondents Chegg and YouTube to be the most helpful resources, while in person office-hours to be the least.

What students are feeling/saying:

When asked, how effective is office hours currently?

Many students felt that office hours were not as effective and had mentioned details about the difficulty of making appointments with students and teacher’s assistants when completing work.

Students currently often felt dismissed during office hours and felt that were often not helpful to their education. Hence many of them would resort to completing work on their own or by referring to other sources.

Students were asked how they could better prepare for assigned materials in their STEM related classes— most felt that having more practice problems and readily available resources would benefit them.

Design

Based on survey insights we felt that our design needed encompass four main features:

  • It needed to be readily available

  • It needed to have practice problems

  • It needed to be easy to use and efficient so students were able to study smarter and not harder

  • Allow students to review each session for further feedback for professors using the app

From there, we were able to take the next steps in incorporating these features into sketches and wireframes.

USER FLOW

In this particular scenario,

  1. A student is able to review answers or practice problems

  2. From there, a student can select practice problem and be given a practice problem that corresponded to their particular module/lesson. When answering a problem, a student has two options - to manually answer the question or to upload their own answers.

  3. After uploading their own work, the bot detects that their answer is not quite correct and directs them to lessons to review as well as outside material and sources that can help them with their practice problem.

  4. Then, the student can choose to exit the session. Once they exit, they see a feedback screen which allows them review their session with the Study-bot, which would later be given to teachers as feedback.

High-Fidelity Design & Prototype

We created a style guide to allow for cohesion with Canvas’ current user-interface

Final Prototype

Try out the prototype here

Takeaways:

I had a fun time working on this project. This project was essentially for a group project for a Project Management class. Prior to this assignment, I approached many case studies through a design perspective. However, with this project, I was able to tackle this problem through a project management lens and spent more time understanding concepts like MVPs and User Research. It was great being able to hear different perspectives and compromising with others when working with other students. Moreover, if I had more time, I would have been interested in conducting user testing with my team on students at Rutgers to see if this feature would have been a promising element to their studying process.

Study Bot Features

  • Practice & Review — based off of user research we felt that that student would benefit most from reviewing module materials and having practice problems. We wanted to implement that into our bot and allow students to have the choice of the two when using the study bot.

  • Uploading your own work — For STEM related classes, the process and workflow is often as crucial as the answer. We implemented a button for students to upload their own work, where the AI would analyze the student’s process based off of an image upload.

  • Feedback & Review — After a student finishes their session with the study bot, they are left with a prompt to rate and comment on how the feel about module material’s and the helpfulness of the bot. This feedback would be sent back to the teachers, (the users who manage this platform for students) so they can continue to improve the course for future students.