active learning triangle / how reliable are its predictions?

by P

I found a mention of the active learning triangle (in this slideshare presentation on education in Web 2.0, which references “Audio-Visual Methods in Teaching” by Holt Rinhart and Winston). It posits that the more we engage / internalise / transform what we learn (or act on what we learn) the more of it we remember after a period of time.

It seems like a useful model to think about rip-mix-learn practices, which are all further towards the “active” side of the triangle than the traditional lecture style of teaching and learning.

However, I am wondering to what extend the model has been tested and how much empirical evidence exists for the statements implicit in the triangle. It makes specific statements about a “2 week” timeframe, and assigns percentages (we remember x % of something) to different types of learning (reading, hearing about something, speaking about it, etc.). I wonder how reliable those percentages are, and how they were arrived at.