I think a project on Moodle Quizes would be good. There's a lot you can do here and the rabbit hole can go deep. There's a few discussions on where the quiz should go and qunatative analysis would be good.
People usally think of a question and want to make a question type based on that. Now my 2 cents.
- You could do a literature review on the type of questioning there is (more physiology then programs) then compare how Moodle's questions fit the models. If there is a gap you could recommend what new question types could cover, changes to current questions, etc.
- Do a literature review on the type of questioning there is and then compare the models to teachers questions to see if they are covering all the bases. If not recommend changes and write a new question type that could help.
- Analyses all the questions in Moodle (+contributed) see the similarities, differences and gaps. Interview Moodle developers, Teachers, Students to see what is missing or they like. Write a new question type based on this.
All of this may seem a long way from programing but research is not just making something, it's saying why, filling in a gap of knowledge and then creating.