Classroom Impact
The HINTS Lab has conducted many studies in authentic learning contexts to evaluate our tools, better understand how students learn, and help guide pedagogy. Our studies have found:
- Block-based programming can be significantly more effective than text-based coding for students’ first coding experience, even when controlling for other factors (e.g. type of learning activity).
- Frame-based programming can also be an effective introductory experience, though it may provide less benefit compared to text-based programming for novices with little programming experience.
- Students complete Parsons problems significantly faster than traditional code writing problems, while performing just as well on subsequent assignments and assessments.
- Automated, data-driven hints, like those from iSnap, can improve students’ performance on programming assignments, both in terms of speed and correctness, and some results suggest this benefit may extend to future assignments without hints (though more research is needed).
- Adaptive and encouraging feedback, like that in our AIF System, can improve students’ performance on programming assignments, as well as their engagement and intentions to persist in Computer Science.