A Google Collab notebook with instructions to succesfully train the model used here: https://www.youtube.com/watch?v=5xzQFB-I5XQ
Corresponding proposal for the Futures Forum on Learning: Tools Competition is below
Inertia Education is founded by Alexander Duffy a Data and Analytics Consultant at EY with a background in Computer Engineering and Machine Learning. He is passionate about using technology to improve education as it is the solution to most any problem. Matt Feldman is an experienced Digital Marketer who has helped B2B SaaS firms, High-End Retailers, and Influencers grow their business. Sophie Callaway is a UX Strategist at Facebook with a focus on Human-Centered Design. Julia Foerster is a middle school math teacher for Teach for America at Green Dot Public Schools in Los Angeles and serves as an advisor. --- 100 words (team description)
Inertia Education helps educators save time by grading paper multiple choice exams with their phone and creating pointed insights on students. Using custom object recognition, paper exams can be graded, and scores broken down by topic; no bubble-sheets needed. Teachers can easily see performance on an individual or classroom basis within the Analytics Module, but assessments only tell part of the story. Coupled with a Presentation Module that uses live student feedback and knowledge checks, the Analytics Module can link how information was taught and tested, visualizing how students best learn. ---- 91 words (solution summary)
Every year there are tens of millions of paper multiple choice assessments administered globally, all with invaluable insights into a student’s knowledge and progress. These insights are lost due to a lack of time and tools available to convert them into actionable data. Current market analysis demonstrates an unsatisfied need for analytics from paper exams. InertiaED is meant to be a foundation upon which learning engineering can be applied no matter the infrastructure, subject, or student. Ultimately Inertia will help students identify their subject strengths and weaknesses, but more importantly, provide tangible data on how they learn; applicable to any challenge in life.
Multiple choice quizzes, tests, and homework assignments will be built using the intuitive web-based test builder. Each question is easily labeled using new or existing tags. Educators would be encouraged to use available learning assessment templates as a base to efficiently analyze current strengths and weaknesses. These tests can be easily printed and administered before being graded and uploaded via smartphone as easily as scanning a QR code without the need for a scantron. A working proof of concept using a custom machine learning model has already been created. This erases the need for expensive scanning equipment while also improving the medium by removing bubble sheets known to lead to errors and frustration and reducing paper along the way.
Teachers would build presentations with the easy to use web-tool briefly tagging what subjects are covered in which slides and how; or use approved samples from the community. Knowledge checks and feedback slides that students respond to in-class would be easy to include. Tags, responses, and additional information such as time and day of the week are then communicated to the analytical backbone.
These tags and the analytics hub are powerful mechanisms with which to empower educators. Knowledge gaps are easily identified allowing teachers to keep students on track by addressing critical needs on an individual or class wide basis. Tags from presentations are related to the meta cognitive aspect of learning engineering, providing suggestions on strategies for how students take in and retain information. The analytics hub could also identify student leaders and suggest groups with complementary strengths and weaknesses for activities. This enables students to learn from their peer’s unique knowledge while solidifying areas of their own strength by teaching. Additionally, educators will be encouraged to use tags creatively and in relation to their own unique needs. For example, test question tags with information about not only topic covered but context may help paint a story of students who succeed when presented with more relatable or interesting problems.
Revolutionizing learning laboratories that rely on paper multiple choice exams is an actionable short-term goal. Implementation would directly increase the number of students who are reading by 3rd grade and the number of students on track in middle-school math. Use by a strong educator could expand the number of students gaining data and computer science skills in high school by increasing interest and relatability. Students would be equipped with tools for college if that is their goal but also entrepreneurship or life in general.
This solution is modular and can be scaled quickly. Paper multiple choice exams are step one, other methods of testing including online exams or even third party-tools could be integrated within the testing portion. External technology such as the writing dashboard given as an example in this competition could be internally developed or integrated into presentations. Finally, the same technology could be applied for businesses. Multiple choice analytics could be applied to surveys. Additional inputs could be created, and employee characterizations could be drastically improved, as opposed to relying heavily on outdated broad metrics such as number of monthly sales.
With a focus on the ease of use, empowering users, and community, Inertia Education could eventually become the recognizable and reputable face for modern learning.