STEM Education Initiatives

(Research Under Development)

Tumblehome is always actively engaged in exploring new projects with various partners in areas of interest for future our work. A few projects we are engaged in currently…

The Knowledge Seeker Identity (KSI) Project

For the past three years, Tumblehome and University of Virginia have been collaborating on a project to study the early factors that influence the long term STEM identies of young people. Using the FOCIS instrument, developed at the Tai Lab within the Curry School of Education, researchers are evaluating a multitide of factors ranging from books to movies to parental/mentor and after school program influences on the STEM mindsets of K-12 children. Drawing primarily from thousands of science fair alumni, as well as students who have independently published their own scientific research in student journals, the research is focusing on students who are motivated to conduct self-guided science & engineering projects that go above and beyond the typical students. KSI seeks to identify what motivates students to develop this mindset, and to develop approaches to help other students form and maintain their own KSI identities. The project is currently funded with internal UVA funding and is seeking outside additional funding for further scaling.

AI-enhanced qualitative research methodology development

As part of the joint work between Tumblehome and the Tai Laboratory at UVA, we are engaged in a product development effort to create a workflow/methodology as well as a light app that will assist qualitatative researchers (especially those in STEM ed fields) to conduct code analysis on qualitative data sources (such as transcribed interviews) using the assistance of AI tools. Currently, we are utilizing Large Language Models (such as ChatGPT) to determine semantic equivalence between transcripts and expertly-developed codes, to assist in the analysis of our some of our own research and have recently agreed to work with outside partners to test our methodologies with external datasets as well. We expect this will greatly reduce the time and expense involved in qualitative analyses and will open up new possibilities for the addition of other media types to supplement and enhance qualitative research. Read more about this concept in papers written/co-authored by Tumblehome’s Barnas Monteith here: Unleashing the Economic Potential of Large Language Models: The Case of Chinese Language Efficiency and here: Use of Large Language Models to Aid Analysis of Textual Data

Project AIMAS

In partnership with the Dundee County Illinois Boys & Girls Club and the Boys & Girls Club of America, Tumblehome (Barnas Monteith, pictured here with University of Virginia Professor Rob Tai) is developing a curriculum series, primarily targeting middle school students, focused on the future of technology. This includies quantum (QISE), AI, blockchain (DLT) and classical computing semitech. Full of fun interactive hands-on activities, combined with AI-enhanced narrative digital experiences, students develop artistic/musical works of their own with the knowledge gained during their courses. Debuting in late 2023, the course encourages students to maximize creativity while at the same time learning the very latest stuff in computing tech.

More coming soon…