Designing the Finch: Creating a Robot Aligned to Computer Science Concepts

AAAI 2010

TL;DR

This paper introduces the Finch, a low-cost, interactive robot designed to improve student engagement and retention in introductory Computer Science (CS1) courses. By addressing key weaknesses of previous robot-based learning tools, like high cost and delayed feedback, the Finch aims to make programming concepts more tangible and motivating for students.

What Problem Does This Paper Address?

The authors tackle the persistent problem of declining enrollment and retention in Computer Science programs. Traditional CS1 courses often struggle to engage students, leading to high dropout rates. While robots have been explored as a way to increase motivation, previous attempts faced significant hurdles: robots were often too expensive for individual student ownership, and the physical nature of robot programming led to delayed feedback and increased debugging time, which can be frustrating for novice programmers. The CSbots program, and specifically the Finch robot, aims to create a robot-based learning tool that retains the motivational benefits of robotics while overcoming these practical and pedagogical limitations.

Methods

The design of the Finch was an iterative process driven by feedback and collaboration with CS educators. Key methods included:

  • Needs Assessment: Initial evaluation of the CS1 landscape through surveys and interviews with educators.
  • Iterative Design: A cycle of designing the robot and curriculum, piloting it in courses, and evaluating the results to inform the next design iteration.
  • Partnerships: Deep collaboration with CS1 educators to ensure the robot and curriculum aligned with real-world teaching challenges and learning goals.
  • Backwards Design: Focusing on learning goals first, then designing instruction and assessment to meet those goals, extending this to the design of the educational tool itself.
  • Constraint-Driven Design: The Finch's design was heavily influenced by evidence-based constraints derived from prior studies and pilot evaluations, including low cost, suitability for home use, aesthetic appeal, and multi-modal interactivity.

Datasets

Not discussed in the paper. The paper focuses on the design and evaluation of a robot platform and its integration into CS1/CS2 curricula, rather than on training machine learning models on specific datasets.

Key Insights and Ideas

  • Cost and Ownership are Crucial: The high cost of previous robots was a major barrier to individual student ownership, limiting practice time and increasing reliance on lab access. The Finch was designed to be affordable (under $100) for individual student purchase.
  • Tethered Design Solves Multiple Problems: By tethering the Finch to a computer via USB, the authors eliminated the need for batteries, reduced on-board processing requirements, and simplified software support across different platforms. This also made the robot less expensive and more robust.
  • Multi-modal Interaction is Key to Engagement: It's not just about having sensors and actuators, but how they are combined. The Finch is designed to engage students through visual, auditory, and kinesthetic senses, allowing for richer and more interactive programming experiences.
  • Alignment with Existing Curriculum is Essential: For widespread adoption, educational tools must integrate with existing CS1 structures and learning goals, rather than requiring a complete overhaul. The Finch is positioned as an additional tool for assignments.
  • Student Ownership and Home Use Boost Retention: Pilot studies showed that when students could take the robot home, retention rates improved significantly, suggesting that increased access and personal connection to the tool are vital.

Key Results

  • Improved Retention: In the fall 2009 CCAC pilot, retention rates (excluding early drops) reached 80%, significantly outperforming previous years and the four-year average. The spring pilot showed a slight underperformance but still improved over the previous year.
  • Student Interest: Incoming students generally preferred programming a robot over a computer. Outgoing students overwhelmingly reported showing the Finch to others, and a notable portion engaged in recreational programming with the robot outside of assignments.
  • Robustness: The Finch demonstrated high hardware reliability, with very few failures reported despite extensive home use. Most failures were manufacturing defects, not due to student use.
  • Cost-Effectiveness: The Finch is estimated to cost under $100 commercially, addressing a major barrier from previous robot platforms.

Limitations and Future Work

  • Spring Pilot Underperformance: The spring pilot's retention rates were lower than expected, highlighting the importance of proper assignment alignment and instructor experience with the tool. The authors suggest that while the spring rates were slightly below average, the overall year's retention was better than average, and performance may improve with instructor experience.
  • Further Language Support: While currently programmed in Java, the authors intend to develop support for additional programming languages in the future.
  • Deeper Curriculum Integration: While the Finch is designed to integrate with existing curricula, further research could explore more profound curriculum redesigns that leverage the robot's capabilities.
  • Broader Evaluation: The paper presents preliminary findings from one community college pilot. A full analysis of this and other pilots is planned for future publication.

Summary updated

BibTeX
@inproceedings{2010designing,
  title = {{Designing the Finch: Creating a Robot Aligned to Computer Science Concepts}},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year = {2010},
}
Figure