Interesting article written by PhD student Claire Browne from New Zealand. Dog training, do you get the timing right? A simple experiment shows a dramatic reduction in effective learning, in dogs whose rewards were delayed by just one second! This highlights the importance of using aids such as clickers, that improve the timing of rewards. It would have been interesting to see what effect a delay between the beep and the reward would have caused.
Re: New experiment highlights the importance of timing Nice piece of research evidence. I was kind of amazed (and kind of not) that in over 40% of instances dogs did nothing in response to their handlers' cues. I'd love to see a further round of studies where the impact of the timing of the marker and the timing of the reward are investigated separately. In that study the marker only was delayed or not and a 1 second marker delay was bad enough to have a significant negative effect on learning (fascinating result). I wonder how long after the marker you have to deliver the reward before the marker is less effective (relative to delivering the reward immediately). Do you have multiple seconds to deliver the reward, provided the marker is spot-on and you've got a strong association between the marker and the reward? When I used to teach puppy classes one of the exercises we did was getting the humans to pair up and teach each other something (e.g. pat yourself on the head) using a marker word only (with imaginary treats) and shaping. The idea was to demonstrate just what a hard task the dog has and also how important it is to be fast, accurate and consistent.
