I've done my own experimentation on college students:
Publish slides without any words, telling them to fill in the words from what I show and say
Publish slides with words and additional annotations in the "notes" field.
For low level classes that require little more than regurgitate and fill in the bubble, grades are higher in the second set. For higher level classes that require synthesis and inference, grades are higher in the first set.
It's not a formal study by any means (nor approved by IRB), but I did it for myself to figure out how to best teach for myself and my classes.
There are fMRI studies that show that writing by hand engages more of the brain than typing, which can explain at least some of the effect.
