Improvements to feedback on interactions

[quevivasbien]

Some ideas for better feedback on "quiz" type interactions:

• When a student submits a multiple choice (or any other type of question with an unambiguous correct answer), they should get immediate feedback if it was correct, instead of having to wait for the AI to generate a response before knowing whether they did it correctly. If the student's answer is correct, we probably should not have the AI provide feedback on it unless the student requests an explanation.
• Particularly on the survey-type interactions, we need some sort of acknowledgement if a student has already successfully submitted a response. (Even something like a green check mark would be sufficient.)
• Essay questions are currently very easily gameable. This will probably always be an issue to some extent (and shouldn't be an issue assuming students are actually trying to learn and not just complete a course for a grade), but there are maybe some easy safeguards we could put in place.

[leesjensen]

This sounds great! If I remember right the survey-questions are not graded and there is no invalid response, but giving better feedback that the response was acknowledged is a good idea.

I meant the survey to be feedback to the instructor. For example: where are you currently with the course objectives? or Is this instruction helpful? type questions. In the future I want to change these to a likert style question array. I was also hoping to use the survey for in class "clicker" type activities.

I love your essay trick. I don't have much experience "gaming it". Whatever we do there to avoid it is good, but in the end I hope MasteryLS will be for people who actually want to learn and so I'm not worried about anyone who is trying to quickly rack up points. In fact we could not give percentage points. Perhaps a star rating or simple Good, Better, Best response would help keep the focus on the learner to "own their learning".