Paul, M. P. (2025, July), Full Paper: Leveraging real-time testing data to assess and predict student success in a team-based first-year engineering design project  Paper presented at FYEE 2025 Conference, University of Maryland - College Park, Maryland. 10.18260/1-2--55248

\bibitem{asee_peer_55248}
  Paul, M. P. (2025, July), \emph{Full Paper: Leveraging real-time testing data to assess and predict student success in a team-based first-year engineering design project}
  Paper presented at FYEE 2025 Conference, University of Maryland - College Park, Maryland.
  10.18260/1-2--55248

Matthew Patrick Paul.  "Full Paper: Leveraging real-time testing data to assess and predict student success in a team-based first-year engineering design project".  FYEE 2025 Conference, University of Maryland - College Park, Maryland, 2025, July.  ASEE Conferences, 2025.  https://peer.asee.org/55248 Internet.  14 Aug, 2025

\bibitem{asee_peer_55248}
  Matthew Patrick Paul.
  "Full Paper: Leveraging real-time testing data to assess and predict student success in a team-based first-year engineering design project".
  \emph{FYEE 2025 Conference, University of Maryland - College Park, Maryland, 2025, July}.
  ASEE Conferences, 2025.
  https://peer.asee.org/55248 Internet.  14 Aug, 2025

@INPROCEEDINGS{asee_peer_55248
author = "Matthew Patrick Paul"
title = "Full Paper: Leveraging real-time testing data to assess and predict student success in a team-based first-year engineering design project"
booktitle = "FYEE 2025 Conference"
year = "2025"
month = "July"
address = "University of Maryland - College Park, Maryland"
publisher = "ASEE Conferences"
note = {https://peer.asee.org/55248}
number = {10.18260/1-2--55248}
}

TY  - CPAPER
AB  - Purdue’s Course Signals project demonstrated that behavioral data such as system logins and time-on-task can serve as early predictors of student performance in large engineering courses [4]. This study explores how real-time behavioral data can be used to predict team performance in ENES100, a required first-year engineering design course at the University of Maryland. Student teams were tasked with designing and programming autonomous Over Terrain Vehicles (OTVs) to complete a mission in a standardized testing arena. Throughout the build phase, team testing behavior was automatically logged via WiFi connections to an overhead vision tracking system, which recorded time-stamped connection events for each team.

A dataset of over 36,000 connection events was processed in Excel using custom macros to extract four behavioral metrics: total number of connections, earliest connection time, number of distinct testing sessions, and cumulative testing time. Testing sessions were identified based on five-minute gaps between connection events, and total testing time was computed by summing session activity. These metrics were then compared to each team’s Milestone 7 (MS7) score, a comprehensive 240-point performance assessment of OTV functionality and mission execution.

Of the metrics analyzed, cumulative testing time showed a strong linear correlation with MS7 score (R² = 0.8677), while other variables showed little to no predictive value. High-performing teams demonstrated consistent weekly testing habits and averaged over 30 hours of cumulative testing within 7 weeks of construction. These results offer instructors quantitative benchmarks for identifying under-engaged teams during the semester. Future work includes developing an instructional tool that tracks team testing in real time and integrates Bayesian updating and machine learning methods to support early intervention and adaptive teaching strategies.
AU  - Matthew Patrick Paul
CY  - University of Maryland - College Park, Maryland
DA  - 2025/07/27
PB  - ASEE Conferences
TI  - Full Paper: Leveraging real-time testing data to assess and predict student success in a team-based first-year engineering design project
UR  - https://peer.asee.org/55248
DO  - 10.18260/1-2--55248
ER  -