Towards a Physiological Computing Infrastructure for Researching Students' Flow in Remote Learning

Maximilian Xiling Li, Mario Nadj, Alexander Maedche, Dirk Ifenthaler, Johannes WöhlerTechnol. Knowl. Learn 2022 · DOI · OpenReview

Introduction

In the field of learning analytics, "flow"—a state of total immersion and focused energy—is considered the pinnacle of student engagement. However, measuring flow in real-time presents significant challenges. Traditional methods rely on self-report surveys, which are inherently obtrusive and often break the very state they aim to measure. Furthermore, while physiological sensors offer a non-intrusive alternative, most research has been confined to controlled laboratory environments. Transitioning these studies to "the wild" of remote learning introduces hurdles related to infrastructure variance, data privacy, and sensor reliability.

Previous research into flow and engagement has largely focused on lab-based settings using games or work tasks as the primary context. As shown in the comparison below, the current study differentiates itself by moving into the field of remote learning, albeit with a smaller initial sample size typical of pilot infrastructure tests.

Comparison of study designs, sample sizes, and contexts across related research.

The SEN-Pi Infrastructure

To address the technical and ethical challenges of remote monitoring, researchers developed SEN-Pi (Sensor Raspberry Pi). This infrastructure is designed to be platform-independent and privacy-preserving. The system consists of three core components:

  1. Hardware Hub: A Polar H10 ECG chest belt paired with a Raspberry Pi. The Pi acts as a dedicated data hub, handling Bluetooth communication and local encrypted storage. This bypasses the student's personal computer, ensuring the system works regardless of the user's operating system.
  2. Privacy-First Browser Extension: A custom Chrome extension triggers Experience Sampling Method (ESM) surveys and logs browsing behavior. Crucially, it uses a "whitelist" approach, only tracking activity on pre-defined educational domains to protect student privacy.
  3. Participant Dashboard: Served directly from the Raspberry Pi, this interface allows students to view their own physiological data, such as heart rate timelines and Poincare plots, serving as both a tool and an incentive for participation.

Technically, the SEN-Pi approach offers a cost-effective and accessible alternative to proprietary lab setups, as detailed in the following comparison of sensing platforms.

Technical comparison of physiological sensing devices, IT platforms, and data access models.

Preliminary Results and Insights

A pilot study involving eight medical students demonstrated the feasibility of the SEN-Pi system. Key findings include:

  • Infrastructure Stability: The system proved stable for remote field use, with minor software bugs addressed via mid-study updates.
  • Flow Estimation: Using a proxy classifier trained on a previous dataset of software developers, the researchers estimated that students spent between 63.7% and 69.1% of their study time in a state of flow.
  • Behavioral Patterns: 86.2% of activity was concentrated on the primary university portal, with study sessions averaging 50 minutes.
  • Compliance: Maintaining engagement remains a challenge; only three of the seven students who returned equipment provided complete HRV data, suggesting that voluntary interest alone may be insufficient for long-term data collection.

Conclusion and Future Work

The SEN-Pi infrastructure represents a successful proof-of-concept for non-intrusive flow measurement in remote learning. Future work will focus on increasing sample sizes to train student-specific machine learning models and expanding the hardware to include environmental sensors (e.g., CO2 and temperature) to better understand how the physical study environment impacts cognitive states.

Summary updated

Abstract

With the advent of physiological computing systems, new avenues are emerging for the field of learning analytics related to the potential integration of physiological data. To this end, we developed a physiological computing infrastructure to collect physiological data, surveys, and browsing behavior data to capture students’ learning journey in remote learning. Specifically, our solution is based on the Raspberry Pi minicomputer and Polar H10 chest belt. In this work-in-progress paper, we present preliminary results and experiences we collected from a field study with medical students using our developed infrastructure. Our results do not only provide a new direction for more effectively capturing different types of data in remote learning by addressing the underlying challenges of remote setups, but also serve as a foundation for future work on developing a less obtrusive, (near) real-time measurement method based on the classification of cognitive-affective states such as flow or other learning-relevant constructs with the captured data using supervised machine learning.
BibTeX
@article{li2022towards,
  author = {Maximilian Xiling Li and Mario Nadj and Alexander Maedche and Dirk Ifenthaler and Johannes Wöhler},
  title = {{Towards a Physiological Computing Infrastructure for Researching Students' Flow in Remote Learning}},
  journal = {Technol. Knowl. Learn.},
  year = {2022},
  doi = {10.1007/s10758-021-09569-4},
}
Figure