The big question for our project is:

“How do engagement mechanisms affect learning from humans in a teaching
scenario?”

For the purpose here, our definition of engagement can be thought of as positive
engagement, which is inversely proportional to stress and frustration, while
being directly proportional to enjoyment and attention.  Engagement mechanisms
denote features of the robot that we predict would raise engagement, thus
leading to better interactive experiences between the robot and the user in a
scenario where the user is trying to teach the robot some task.

We believe that this is an important question to ask because engaging robots are
more entertaining to interact with and more likely to hold a person’s attention
as result. A robot that is more engaging to teach would also likely receive more
teaching, which would in turn facilitate robot learning from humans.