Chris Martens's research while affiliated with North Carolina State University and other places

Publications (10)

Conference Paper
Full-text available
Regardless of skill level and background, programming can be challenging for all students. However, in the early stages of learning, challenges may particularly lead to a decrease in students’ sense of self-efficacy and interest in computer science. Hence, finding the moments when novices struggle during programming will help us provide support and...
Conference Paper
Full-text available
Positive student self-efficacy has been linked to undergraduate computer science students' improved retention rates and success in the major, with self-efficacy in programming being particularly important. To improve poor self-efficacy in programming, especially for novices, we must understand the moments that affect students' self-perceived progra...
Preprint
Full-text available
Open-ended programming increases students' motivation by allowing them to solve authentic problems and connect programming to their own interests. However, such open-ended projects are also challenging, as they often encourage students to explore new programming features and attempt tasks that they have not learned before. Code examples are effecti...
Article
Full-text available
Using machine learning to classify student code has many applications in computer science education, such as auto-grading, identifying struggling students from their code, and propagating feedback to address particular misconceptions. However, a fundamental challenge of using machine learning for code classification is how to represent program code...

Citations

... This may have resulted from survey fatigue, with too many questions having previously been administered; consequently, results that deviated from "Same as me" potentially had undue sway in final analysis results. A larger number of participants may allow for more interesting data mining analysis, which would be a worthwhile future research topic; other work has related to the automatic detection of these self-assessment moments through trace log data [12], and consequent work could use more survey results in conjunction with more programming data to draw more conclusions. ...
... The difficulties of programming education for novice learners have been confirmed through many studies [11][12][13]. The difficulty of the programming process is that when an error occurs, the cause is not known, so even if the problem is solved, the procedure is not correctly recognized. ...
... This could be implemented by asking students to add code comments. Alternatively, such struggling students might need an intervention that directs them to plan their work, as in the PlanIT system [24]. ...
... For example, Wang et. al. found that a large number of code features can lead models to overfit to the training data [18]. This has likely happened with D05 and L02, their LSTM input dimensions increased at least 4-fold from D02 and L01 respectively, making it possible that the models would overfit. ...