Date
Publisher
arXiv
Early detection of struggling student programmers is crucial for providing
them with personalized support. While multiple AI-based approaches have been
proposed for this problem, they do not explicitly reason about students'
programming skills in the model. This study addresses this gap by developing in
collaboration with educators a taxonomy of proficiencies that categorizes how
students solve coding tasks and is embedded in the detection model. Our model,
termed the Proficiency Taxonomy Model (PTM), simultaneously learns the
student's coding skills based on their coding history and predicts whether they
will struggle on a new task. We extensively evaluated the effectiveness of the
PTM model on two separate datasets from introductory Java and Python courses
for beginner programmers. Experimental results demonstrate that PTM outperforms
state-of-the-art models in predicting struggling students. The paper showcases
the potential of combining structured insights from teachers for early
identification of those needing assistance in learning to code.
Study at a glance
- Question: Can an AI model using teacher-developed skill categories help identify beginner programmers who will struggle with new tasks?
- Setting and data: Two datasets from introductory Java and Python courses for beginner programmers.
- Method: Development of a predictive model with educators and evaluation across two student programming datasets.
Why it matters
Early detection of struggling student programmers is crucial for providing them with personalized support. However, existing AI-based approaches do not explicitly reason about students' programming skills in the model.
Key findings
- In introductory Java and Python courses, a new AI tool outperformed existing models at predicting which beginner programmers would struggle on upcoming tasks.
- The AI tool simultaneously learned students' coding skills from their past work and predicted whether they would struggle on a new task.
- Combining structured insights from teachers with AI demonstrated strong potential for identifying learners who need assistance in learning to code.
Summary drafted with AI from the paper’s abstract and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Study design
