The retention data goes back, as far as we can tell, to 1913, and likely before.
We have some examples, starting around 1999, where the retention data numbers are attributed to Edgar Dale and his Cone of Experience. In the examples where a visual cone is presented, we include it in the third section "Corrupted Cone Timeline." If only the retention numbers were included, it is included in this section. What is particularly interesting about the attributions to Dale was that the percentages were applied to his cone, and then separated again, yet he maintains credit as the source, even without the cone present.
Note: If you have an example that you think should be included in this timeline, please contact me at betrusak@potsdam.edu